Regulatory Impact Analysis

REGULATING
ADVICE
MARKETS
DEFINITION OF THE TERM
“FIDUCIARY”
CONFLICTS OF INTEREST RETIREMENT INVESTMENT
ADVICE
REGULATORY IMPACT
ANALYSIS FOR FINAL RULE AND
EXEMPTIONS
APRIL 2016

Table of Contents
Table of Contents .................................................................................................... i
List of Figures ....................................................................................................... vii
Regulatory Impact Analysis Glossary ..................................................................... ix
Executive Summary ................................................................................................ 1
1.

Introduction ................................................................................................ 16

2.

Legal Environment: ERISA and the IRC ........................................................ 20
2.1

Statutory Provisions ....................................................................................... 20

2.1.1 Provisions Relating to Plans ...................................................................... 20
2.1.2 Provisions Relating to IRAs ....................................................................... 21
2.1.3 Permissible Self-Dealing............................................................................ 21
2.2

Exemptions from the Prohibited Transactions Provisions ............................ 22

2.2.1 Statutory Investment Advice Exemption .................................................. 22
2.2.2 Administrative PTEs .................................................................................. 23
2.3

The 1975 Regulation ...................................................................................... 23

2.4

Relevant Advisory Opinions ........................................................................... 24

2.4.1 AO 97-15A (Frost Bank) and 2005-10A (Country Bank)............................ 24
2.4.2 AO 2001-09A (SunAmerica), Investment Advice Programs...................... 25
2.4.3 AO 2005-23A Regarding Rollovers ............................................................ 26
2.5

Interpretive Bulletin on Investment Education ............................................. 27

2.6

Intersections with Other Governing Authorities ........................................... 28

2.6.1 Regulation of BDs and RIAs ....................................................................... 30
2.6.2 Background on Regulation of Insurance Products and Producers ........... 36
2.6.3 Regulation of Annuity Products ................................................................ 39
2.6.4 SEC Staff Dodd-Frank Study ...................................................................... 43
2.6.5 Relevant Dodd-Frank Act Provisions......................................................... 43
2.6.6 FINRA Conflicts of Interest Report ............................................................ 45
2.7

2010 Proposal ................................................................................................ 47

2.8

2015 Proposal ................................................................................................ 48

2.8.1 Proposed PTEs........................................................................................... 51
2.9

2016 Final Rule and PTEs ............................................................................... 56

2.9.1 Final Rule ................................................................................................... 56
i

2.9.2 Final PTEs .................................................................................................. 68
2.10 Reform Abroad .............................................................................................. 75
2.10.1 The UK Retail Distribution Review ............................................................ 78
2.10.2 Australian Legislation Impacting Financial Advice .................................... 93
2.10.3 Is the RDR a Model for Wider European Regulation? .............................. 95
3.

IRA Market .................................................................................................. 96
3.1

Affected Universe .......................................................................................... 97

3.1.1 IRA Investors ............................................................................................. 97
3.1.2 Professional Advisers, BDs, RIAs, Insurance Agents ............................... 100
3.1.3 Product Providers ................................................................................... 103
3.2

Need for regulatory action .......................................................................... 105

3.2.1 IRAs Warrant Special Protection ............................................................. 106
3.2.2 Market Changes Since 1975.................................................................... 114
3.2.3 The IRA Advice Market............................................................................ 127
3.2.4 Magnitude of Harm................................................................................. 158
3.2.5 Conclusion ............................................................................................... 166
3.3

Gains to Investors ........................................................................................ 167

3.3.1 Quantified Gains to Investors ................................................................. 169
3.3.2 Qualitative Discussion ............................................................................. 178
4.

ERISA-Covered Plans ................................................................................. 184
4.1

Affected Entities........................................................................................... 185

4.1.1 Service Providers..................................................................................... 185
4.1.2 Plans and Participants ............................................................................. 186
4.2

Need for Regulatory Action ......................................................................... 187

4.2.1 Plan Level Advice .................................................................................... 188
4.2.2 Plan Participant Advice ........................................................................... 191
4.2.3 Department Enforcement Challenges .................................................... 197
4.3

The Final Rule and Exemptions’ Impact on Plan Participants ..................... 199

4.3.1 Promoting Good Advice and Education .................................................. 199
4.3.2 More Effective Enforcement................................................................... 204
5.

Cost ........................................................................................................... 206
5.1

Background .................................................................................................. 206

5.2

Affected Entities........................................................................................... 211
ii

5.2.1 Number of BDs ........................................................................................ 211
5.2.2 Number of RIAs ....................................................................................... 211
5.2.3 Number of ERISA Plan Service Providers ................................................ 212
5.2.4 Number of Insurers ................................................................................. 212
5.2.5 Banks ....................................................................................................... 213
5.2.6 Dividing Firms into Small, Medium, and Large Size Categories .............. 215
5.2.7 Determining the Share of Firms Servicing Plan or IRA Investors ............ 217
5.2.8 Accounting for Overlap Between Firm Types ......................................... 218
5.3

Methodology for Cost Estimates ................................................................. 219

5.3.1 Estimating BD Firm Costs for Complying With Final Rule and Exemptions223
5.3.2 RIA Firm Costs For Complying With Rule and Exemptions ..................... 233
5.3.3 Insurer Costs For Complying With Rule and Exemptions ....................... 237
5.3.4 Plan Service Provider Costs For Complying With Rule and Exemptions 238
5.4

Additional Costs of Assuming Fiduciary Status ............................................ 239

5.4.1 Increased Insurance Premiums/Litigation .............................................. 239
5.4.2 BD Conversion to RIA Status ................................................................... 242
5.4.3 Call Centers ............................................................................................. 243
5.5

Indirect Cost ................................................................................................. 244

5.5.1 Impact on Financial Sector Labor Markets ............................................. 244
5.5.2 Impact on Asset Providers ...................................................................... 244
5.6

Additional Costs Related to PTEs and Exceptions to Fiduciary Investment
Advice........................................................................................................... 245

5.7

Total Quantified Costs and Sensitivity Analysis ........................................... 246

6.

7.

Regulatory Flexibility Act........................................................................... 254
6.1

Need for and Objectives of the Rule............................................................ 254

6.2

Affected Small Entities ................................................................................. 254

6.3

Significant Issues Raised In Response to IRFA ............................................. 256

6.4

Response to Comments Filed by the Chief Counsel for Advocacy of the
Small Business Administration..................................................................... 256

6.5

Impact of the Final Rule and Exemptions .................................................... 257

6.6

Steps Agency Has Taken to Minimize Significant Economic Impact on Small
Entities ......................................................................................................... 260

6.7

Related Rules ............................................................................................... 260
Regulatory Alternatives ............................................................................. 262
iii

7.1

Populating Asset Allocation Models and Interactive Investment Materials
with Designated Investment Alternatives ................................................... 263

7.2

Extending Counterparty Exception to Include Smaller Plans, Participants
and Beneficiaries.......................................................................................... 264

7.3

Treating Appraisals, Fairness Opinions, or Similar Statements as Fiduciary
Investment Advice ....................................................................................... 267

7.4

Basing Exemptive Relief on Disclosure Alone .............................................. 268

7.5

Not Providing a Best Interest Contract Exemption ..................................... 271

7.6

ERISA-Covered Plans Have to Enter into the Best Interest Contract........... 272

7.7

Leave Best Interest Contract Exemption Unchanged from 2015 Proposal . 273

7.8

Arbitration.................................................................................................... 279

7.9

Assets Covered Under Best Interest Contract Exemption ........................... 281

7.10 Allowing Fixed-Indexed Annuities to be Covered under PTE 84-24 ............ 282
7.11 Wait for SEC Action ...................................................................................... 286
7.12 Alternative “Best Interest” Conduct Standard Formulations ...................... 288
7.13 Issue a Streamlined, “Low-Cost Safe Harbor” PTE ...................................... 290
7.14 Delaying Applicability Date .......................................................................... 291
7.15 Providing Streamlined Conditions in Best Interest Contract Exemption for
“Level-Fee Fiduciaries” ................................................................................ 294
7.16 Conclusion .................................................................................................... 296
8.

Uncertainty ............................................................................................... 298
8.1

Magnitude of Harm ..................................................................................... 298

8.1.1 Quantified Harm ..................................................................................... 299
8.1.2 Unquantified Harm ................................................................................. 302
8.2

Investor Gains .............................................................................................. 303

8.2.1 Quantified Investors Gains ..................................................................... 303
8.2.2 Unquantified Investor Gain..................................................................... 305
8.3

Compliance Costs ......................................................................................... 306

8.4

Secondary Market Effects ............................................................................ 307

8.4.1 Advisory Firms’ Responses...................................................................... 309
8.4.2 Product Manufacturers’ Responses........................................................ 311
8.4.3 Investors’ Responses............................................................................... 311
8.4.4 Impact on Small Plan and IRA Investors ................................................. 312
8.4.5 Innovation ............................................................................................... 318
iv

8.5
9.

Net Welfare Gains Considered Separately .................................................. 324
Conclusion ................................................................................................. 326

Appendix A: Analysis of Broker-Sold Mutual Fund Performance Using Data from
Morningstar ........................................................................................................ 330
A.1

Introduction ................................................................................................. 330

A.2

Data .............................................................................................................. 331

A.3

Broker-Sold Domestic Equity Mutual Fund Underperformance ................. 332

A.3.1 Hypothesis Tests ..................................................................................... 333
A.4

Aggregated Performance ............................................................................. 337

A.5

Conclusion .................................................................................................... 338

Appendix B: Bases for Estimates of Harm and Subsets of Gain to Investors ........ 340
B.1

Investment Performance ............................................................................. 340

B.2

Front-End-Load-Mutual-Fund-Gains-to-Investors Estimates ...................... 343

B.3

Assumptions and Uncertainty ..................................................................... 345

B.3.1 Effect of Loads on Returns ...................................................................... 346
B.3.2 Load and Performance Projections......................................................... 350
B.3.3 Front-End-Load Mutual Fund Assets ...................................................... 356
B.3.4 Aggregation of Yearly Gains to Investors................................................ 359
B.3.5 Effects of Proposed Regulatory Action on Market Trends ..................... 360
B.4

Estimates of the Harm Due to Conflicted Advice ........................................ 361

Appendix C: Small Saver Market......................................................................... 366
Appendix D: Partial Gains to Investors and Compliance Costs Accounting Table 374
Bibliography ....................................................................................................... 376

v

vi

List of Figures
Figure 2-1 ERISA and IRC Provisions Governing Fiduciary Advice on the Investments of Plan and IRA
Assets .............................................................................................................................................. 20
Figure 2-2 Intersection of Federal Laws ............................................................................................. 30
Figure 2-3 Reform Abroad .................................................................................................................. 76
Figure 2-4 Gross Retail Flows Through Highest-Charging Share Classes............................................ 83
Figure 2-5 Tracker Sales...................................................................................................................... 83
Figure 2-6 Decline in Investment Bonds ............................................................................................. 84
Figure 2-7 Number of Advisers per 10,000 Population ...................................................................... 92
Figure 3-1 Prevalence of IRAs ............................................................................................................. 98
Figure 3-2 Distribution of HHs and IRAs ............................................................................................. 98
Figure 3-3 IRAs by Financial Institution .............................................................................................. 99
Figure 3-4 Annuity Sales by Distribution Channel and Product Type in 2014 .................................. 102
Figure 3-5 Share of Sales by Top Provider by Type of Market in 2014 ............................................ 104
Figure 3-6 Taxable IRA Distributions 1990-2013 .............................................................................. 107
Figure 3-7 Retirement Assets ........................................................................................................... 114
Figure 3-8 Share of IRA Assets .......................................................................................................... 116
Figure 3-9 The Share (%) of Deferred Annuity Contracts Sold by Market and Type of Product
in 2014 ........................................................................................................................................... 117
Figure 3-10 Deferred Annuity Sales in the IRA Market by Product Type ......................................... 118
Figure 3-11 % Share of Deferred Annuity Sales in the IRA Market by Product Type ....................... 118
Figure 3-12 Year-End Deferred Annuity Assets in the IRA Market by Product Type ....................... 118
Figure 3-13 % Share of Year-End Deferred Annuity Assets in the IRA Market by Product Type ..... 118
Figure 3-14 Comparing Different Types of Deferred Annuities ........................................................ 124
Figure 3-15 Some Common Conflicts in Advice................................................................................ 127
Figure 3-16 Annuity Sales by Distribution Channel Within Each Product Type in 2014 .................. 131
Figure 3-17 Literature Investigating the Performance of Assets Held as a Result of Conflicted
Investment Advice ......................................................................................................................... 159
Figure 3-18 Front-load-mutual-fund-gains-to-investors Estimates by Time Horizon and Alternative
Scenario ......................................................................................................................................... 175
Figure 4-1 Service Providers to ERISA Plans ..................................................................................... 185
Figure 5-1 Affected Firms ................................................................................................................. 216
Figure 5-2 Start-up Costs Using Updated Number of Firms ............................................................. 221
Figure 5-3 Ongoing Costs Using Updated Number of Firms ............................................................ 221
Figure 5-4 FSI Categories of Costs Related to the Proposed Rules Requirements........................... 222
Figure 5-5 Calculations of Average Costs by Firm Size ..................................................................... 223
Figure 5-6 Percent Reduction of Costs of Final Rule Relative to the 2015 Proposal Estimates
by FSI ............................................................................................................................................. 224
Figure 5-7 Start-up Costs Before Adjusting for Data Discrepancy: Medium Reduction Scenario.... 232
Figure 5-8 Start-up Costs After Adjusting for Data Discrepancy: Medium Reduction Scenario ...... 232
Figure 5-9 Total Costs for BDs .......................................................................................................... 233
Figure 5-10 Total Costs for RIAs using an Exemption ....................................................................... 235
Figure 5-11 Total Costs for Insurers using an Exemption ................................................................ 238
Figure 5-12 Summary Figure of Costs............................................................................................... 248
Figure 5-13 Summary Figure of Quantified Costs ............................................................................ 249
Figure 5-14 Summary Figure of Quantified Costs, Complete Overlap of Service Providers to ERISA
Plans and IRAs ............................................................................................................................... 249
Figure 5-15 Summary Figure of Quantified Costs, Complete Overlap of Service Providers to ERISA
Plans and IRAs and RIAs Needing to use Best Interest Contract Exemption ................................ 250
Figure 5-16 Summary Figure of Costs by Provision .......................................................................... 250

vii

Figure 5-17 Comparison of Proposed and Final Best Interest Contract and Principal Transactions
Exemptions .................................................................................................................................... 251
Figure 6-1 Number of Entities Affected by Rulemaking Meeting SBA's Definition of a Small Entity256
Figure 6-2 Start-Up Costs Per Firm by DOL Firm Size Category in Medium Cost Reduction Scenario
....................................................................................................................................................... 259
Figure 6-3 Ongoing Costs Per Firm by DOL Firm Size Category in Medium Cost Reduction Scenario
....................................................................................................................................................... 260
Figure A-1 Risk-adjusted Domestic Equity Mutual Fund Performance by Distribution Channel, 19802015 ............................................................................................................................................... 332
Figure A-2 Broker-Sold Domestic Equity Mutual Fund Underperformance over Various Time Periods
....................................................................................................................................................... 333
Figure A-3 Results of Three Regressions of Domestic Equity Performance Difference (Direct-Sold
Performance Minus Broker-Sold Performance) on Year .............................................................. 334
Figure A-4 Broker-sold Domestic Equity Mutual Fund Underperformance, 1980-2015 ................. 334
Figure A-5 Results of Three Statistical t-tests on Average Broker-Sold Domestic Equity Mutual Fund
Underperformance Before and After January 1, 2008 ................................................................. 336
Figure A-6 Results of Three Statistical t-tests on Average Broker-Sold Domestic Equity Mutual Fund
Underperformance, 2008-2015 versus 1993-2007 ....................................................................... 337
Figure A-7 Risk-adjusted Equity Mutual Fund Performance by Investment Location and Distribution
Channel, 1980-2015 ...................................................................................................................... 338
Figure B-1 Calculation of Investment Performance Net of Loads for Baseline and Alternative
Scenarios ....................................................................................................................................... 342
Figure B-2 Distribution of Purchase Dates for Front-end-load Assets Owned in a Given Year........ 342
Figure B-3 Calculation of 10-year Front-load-mutual-fund-gains-to-investors Under Alternative
Scenario 1 ...................................................................................................................................... 344
Figure B-4 Assignments used in the Department's 10- and 20-year Front-load-mutual-fund-gains-toinvestors Estimates ....................................................................................................................... 345
Figure B-5 Alternative Assignments for the Distribution of Purchase Dates for Front-end-load Assets
Owned in a Given Year .................................................................................................................. 355
Figure B-6 Projected Total IRA Assets at Beginning of Year ............................................................. 356
Figure B-7 Projected Percentage of IRA Mutual Fund Assets Incurring a Front-end-load by Year .. 358
Figure B-8 Underperformance and Load Projections ....................................................................... 362
Figure B-9 Broker-sold Mutual Fund Underperformance ................................................................ 363
Figure B-10 Projected IRA Rollovers ................................................................................................. 364
Figure B-11 IRA Rollover Broker-sold Mutual Fund Underperformance ......................................... 364
Figure C-1 Few Small IRA Owners Are of Modest Means ................................................................ 366
Figure C-2 More Income = More Small IRAs..................................................................................... 366
Figure C-3 Job-based Plans Dominant (Except At the Top) .............................................................. 367
Figure C-4 Few with Modest Means Have IRAs ................................................................................ 367
Figure C-5 More $ Held at Work Than in IRAs .................................................................................. 368
Figure C-6 Take-Up ........................................................................................................................... 368
Figure C-7 Small Savers - and Debt ................................................................................................... 370
Figure C-8 More Small-Saver IRAs at Banks than at Brokerages ...................................................... 371
Figure C-9 Few Small Savers Advised by Brokers ............................................................................. 371
Figure D-1 Partial Gains to Investors and Compliance Costs Accounting Table............................... 374

viii

Regulatory Impact Analysis Glossary
1975 Regulation

29 C.F.R. 2510.3-21(c).

2010 Proposal

DOL “Proposed Definition of the Term ‘Fiduciary’” (Oct. 2010)

2015 Proposal

DOL “Definition of the Term ’Fiduciary'; Conflict of Interest Rule—
Retirement Investment Advice; Proposed Rule” (April 2015)

AACG

Advanced Analytical Consulting Group

Advisers Act

Investment Advisers Act of 1940

AO

Advisory Opinion

ASIC

Australian Securities and Investments Commission

BICE

Best Interest Contract Exemption

BD

Broker-Dealer

BGA

Brokerage General Agencies

CAP

DOL’s Consultant/Adviser Project

CEM

Christoffersen, Evans, and Musto (2013)

CFPB

Consumer Financial Protection Bureau

CFTC

Commodity Futures Trading Commission

DIA

Designated Investment Alternative

DB

Defined Benefit

DC

Defined Contribution

DOL

U.S. Department of Labor

Dodd-Frank

Dodd-Frank Wall Street Reform and Consumer Protection Act

DGR

Del Guercio and Reuter (2014)

ESOP

Employee Stock Ownership Plan

ERISA

Employee Retirement Income Security Act of 1974

ETF

Exchange-Traded Fund

EU

European Union

Exchange Act

Securities Exchange Act of 1934

FAMR

UK’s Financial Advice Market Review

FCA

UK’s Financial Conduct Authority

FMO

Field Marketing Organizations

FSA

UK’s Financial Services Authority
ix

FIO

Federal Insurance Office

FINRA

Financial Industry Regulatory Authority

FOCUS

Financial and Operational Combined Uniform Single Reports

FOFA

Australia’s Future of Financial Advice Act

FSI

Financial Services Institute

GAO

U.S. Government Accountability Office

GLBA

Gramm-Leach-Bliley Act

GMAB

Guaranteed Minimum Accumulation Benefits

GMWB

Guaranteed Minimum Withdrawal Benefits

IAC

SEC’s Investor Advisory Committee

IB

Interpretive Bulletin

ICI

Investment Company Institute

IMD2

Insurance Mediation Directive

IMO

Independent Marketing Organizations

IRA

Individual Retirement Account

IRC

Internal Revenue Code

IRS

Internal Revenue Service

MIFID II

Markets in Financial Instruments Directive

MSD

Massachusetts Securities Division

NAIC

National Association of Insurance Commissioners

NARAB

National Association of Registered Agents and Brokers

NARAB II

National Association of Registered Agents and Brokers Reform Act of
2015

NERA

National Economic Research Associates

NMO

National Marketing Organizations

NPRM

Notice of Proposed Rulemaking

OCIE

SEC’s Office of Compliance Inspections and Examinations

OMB

Office of Management and Budget

PJC

Parliamentary Joint Committee on Corporations and Financial Services

PPA

Pension Protection Act of 2006

PPGAs

Personal-Producing General Agents

PRA

Paperwork Reduction Act of 1995

PT

Prohibited Transaction

PTE

Prohibited Transaction Exemption
x

ReTIRE

SEC’s Retirement-Targeted Industry Reviews and Examinations

RDR

UK’s Retail Distribution Review

REITS

Real Estate Investment Trusts

RIA

Registered Investment Adviser

SBA

Small Business Administration

SEC

U.S. Securities and Exchange Commission

SCF

Survey of Consumer Finances

SIFMA

Securities Industry and Financial Markets Association

SRO

Self-Regulatory Organization

Treasury

U.S. Department of the Treasury

UK

United Kingdom

xi

Executive Summary
Tax-preferred retirement savings, in the form of private-sector, employer-sponsored
retirement plans, such as 401(k) plans, as well as Individual Retirement Accounts (IRAs), is
critical to the retirement security of most U.S. workers. Investment professionals play an
important role in guiding their investment decisions. However, these professional advisers1
often are compensated in ways that create conflicts of interest, which can bias the investment
advice that some render and erode plan and IRA investment results. In order to limit or mitigate
conflicts of interest and thereby improve retirement security, the Department of Labor (the
Department or DOL) is issuing a final rule that will attach fiduciary status to more of the advice
rendered to plan officials, plan participants, and beneficiaries (plan investors) and IRA investors.
The Department is also granting final exemptions from certain “prohibited transactions"
restrictions applicable to fiduciaries.
The Employee Retirement Income Security Act of 1974 (ERISA)2 and the Internal
Revenue Code (IRC or Code) together assign fiduciary status to any person who “renders
investment advice for a fee or other compensation, direct or indirect” with respect to plan or IRA
investments. The determination of who is a fiduciary is of central importance under this
statutory framework. One of the primary ways ERISA protects employee benefit plans and their
participants and beneficiaries is by requiring fiduciaries to comply with fundamental obligations
rooted in the law of trusts. In particular, ERISA requires fiduciary advisers to plan investors to
manage plan assets prudently and with undivided loyalty to the plan’s participants and
beneficiaries. In addition, ERISA and the IRC together forbid fiduciary advisers to both plan
and IRA investors from engaging in broadly-defined prohibited transactions in which the
advisers’ and investors’ interests might conflict. Under ERISA, plan fiduciaries are personally
liable for plan losses stemming from breach of these duties, and under the IRC, both plan and
IRA fiduciaries are liable for excise taxes when they engage in prohibited transactions.
While fiduciary advisers generally must avoid conflicts of interest, ERISA and the IRC
provide certain parallel statutory prohibited transaction exemptions (PTEs) that allow some
transactions that involve conflicts of interest to proceed provided that adequate consumer
protections are in place. For example, one statutory PTE allows fiduciary advisers to receive
indirect compensation from third parties in connection with investment products they
recommend as long as the compensation does not vary depending on the investments chosen or
the advice is generated by a computer model that is independently certified to be unbiased and
certain other conditions are met. The Department has the authority to issue additional individual
and class administrative PTEs if it finds the exemptions are administratively feasible, in the
interest of plan participants and IRA investors, and protective of their rights. PTE 86-128, an
existing class exemption3 allows fiduciary advisers to receive brokerage commissions for
executing transactions they recommend.

1

2
3

By using the term “adviser,” the Department does not intend to limit its use to investment advisers registered under the Investment Advisers
Act of 1940 (Advisers Act) or under state law. For example, as used herein, an adviser can be, among other things, a representative of a
registered investment adviser, a bank or similar financial institution, an insurance company, or a broker-dealer.
29 U.S.C. § 1001 et seq. (hereinafter cited as ERISA).
As discussed in Section 2.9 below, PTE 86-128 is being amended as part of the final rule and exemptions package.

1

The Department also has authority to issue rules under both ERISA and the IRC that
determine when persons rendering advice on the investment of plan or IRA assets must act as
fiduciaries.4 The prior rule, issued in 1975 (1975 regulation),5 narrowly limited fiduciary status;
it was written 40 years ago when IRAs had just been created and the vast majority of consumers
were not managing their own retirement savings or relying on investment advice to do so. The
1975 regulation provided a five-part test for determining whether an adviser was a fiduciary.
Under the test, the person must: (1) make recommendations on investing in, purchasing or
selling securities or other property, or give advice as to the investments’ value; (2) on a regular
basis; (3) pursuant to a mutual understanding that the advice; (4) would serve as a primary basis
for investment decisions; and (5) would be individualized to the particular needs of the plan. An
investment adviser was not treated as a fiduciary unless each element of the five-part test was
satisfied for each instance of advice. Subsequent Department interpretive guidance further
narrowed fiduciary status by ruling that advice to plan participants to roll over assets from a plan
to specific new investments in an IRA does not constitute fiduciary investment advice unless the
advice is provided by someone who already is a fiduciary.6
ERISA and IRC rules governing advice on the investment of plan and IRA assets are
separate from provisions of federal securities laws, such as the Securities Exchange Act of 1934
(Exchange Act) and the Investment Advisers Act of 1940 (Advisers Act), and rules issued by the
Securities and Exchange Commission (SEC or Commission) that govern the conduct of
Registered Investment Advisers (RIAs) and broker-dealers (BDs), who advise retail investors.
Congress, as part of the Dodd-Frank Wall Street Reform and Consumer Protection Act7 (DoddFrank Act), directed the SEC to consider a uniform fiduciary standard for RIAs and BDs who
advise retail customers. The SEC staff in January 2011 issued a report recommending that the
Commission pursue such reform (SEC Staff Dodd-Frank Study).8 As part of the analysis
supporting its recommendation, the report included a detailed discussion of the scope and limits
of current regulation of RIAs and BDs.9 The Commission in March 2013 issued a Request for
Information seeking data to further inform its consideration of these issues, and received
numerous responses.10 As further discussed in Section 2.6, in November 2013, an Investor
Advisory Committee established by Section 911 of the Dodd-Frank Act issued a recommended

4

5

6
7
8

9

10

Reorganization Plan No. 4 of 1978, 43 Fed. Reg. 47713 (Oct. 17, 1978), 92 Stat. 3790, 5 U.S.C. § App. (2010); available at:
http://www.gpo.gov/fdsys/pkg/USCODE-2010-title5/pdf/USCODE-2010-title5-app-reorganiz-other-dup102.pdf.
29 C.F.R. 2510.3-21(c), 40 Fed. Reg. 50843 (Oct. 1975); available at: http://www.gpo.gov/fdsys/pkg/C.F.R.--2011-title29-vol9/pdf/C.F.R.-2011-title29-vol9-sec2510-3-21.pdf.
DOL Advisory Opinion 2005-23A (Dec. 7, 2005).
Pub. Law No. 111-203, 124 Stat. 1376 (2010).
Staff of U.S. Securities and Exchange Commission, “Study on Investment Advisers and Broker-Dealers, As Required by Section 913 of the
Dodd-Frank Wall Street Reform and Consumer Protection Act” (hereinafter, the “The SEC Staff Dodd-Frank Study”) (Jan. 2011);
available at: http://www.sec.gov/news/studies/2011/913studyfinal.pdf.
This analysis briefly discusses the securities laws’ regulation of BDs and RIAs in Section 2.6. The SEC Staff Dodd-Frank Study includes
some additional information on the scope, terms, and limits of the securities laws in this regard, including a discussion of investor
confusion about financial service provider’s obligations and standards of conduct; the standards applicable to BDs and RIAs when
providing investment recommendations; regulation of compensation; licensing and registration requirements; the availability and
limitations on private rights of actions; requirements for proof of scienter in private actions; FINRA arbitration; and other matters. The
reader is generally referred to the SEC Staff Dodd-Frank Study for such additional discussion of the securities laws’ regulatory framework
for advice.
U.S. Securities and Exchange Commission, SEC Release No. 69013, IA 3558, “Duties of Brokers, Dealers, and Investment Advisers” (Mar.
1, 2013); available at: http://www.sec.gov/rules/other/2013/34-69013.pdf.

2

framework for a uniform fiduciary duty governing BDs and RIAs under the securities laws.11 To
date, the SEC has not issued any regulatory guidance regarding a uniform fiduciary standard,
although a proposed rule related to personalized investment advice for retail investments is listed
in the SEC’s Fall 2015 agenda published by the Office of Management and Budget (OMB) for
2016.12 Any new framework, if adopted, would not alter the obligation of BDs and RIAs to
comply with their separate obligations under ERISA and the IRC when giving advice on taxpreferred retirement investments. In addition, there are many transactions involving retirement
savings (such as advice to purchase some insurance annuity products, real estate and
commodities) to which federal securities laws do not apply, but ERISA and the IRC do.
Since the Department issued its 1975 regulation, the retirement savings market has
changed profoundly. Individuals, rather than large employers and professional money managers,
are increasingly responsible for their own investment decisions as IRAs and 401(k)-type defined
contribution (DC) plans have supplanted defined benefit pensions as the primary means of
providing retirement security. Financial products are increasingly varied and complex. Retail
investors are now confronted with myriad choices of how and where to invest, many of which
did not exist or were uncommon in 1975. These include, for example, market-tracking,
passively managed and so-called “target-date” mutual funds; exchange-traded funds (ETFs)
(which may be leveraged to multiply market exposure); hedge funds; private equity funds; real
estate investment trusts (both traded and non-traded); various structured debt instruments;
insurance products that offer menus of direct or formulaic market exposures and guarantees from
which consumers can choose; and an extensive array of derivatives and other alternative
investments. These choices vary widely with respect to return potential, risk characteristics,
liquidity, degree of diversification, contractual guarantees and/or restrictions, degree of
transparency, regulatory oversight, and available consumer protections. Many of these products
are marketed directly to retail investors via email, website pop-ups, mail, and telephone. All of
this variety creates the opportunity for retail investors to construct and pursue financial strategies
closely tailored to their unique circumstances – but also sows confusion and the potential for
very costly mistakes.
Plan and IRA investors often lack investment expertise and must rely on experts – but are
unable to assess the quality of the expert’s advice or guard against its conflicts of interest. Most
have no idea how advisers are compensated for selling them products. Many are bewildered by
complex choices that require substantial financial expertise and welcome advice that appears to
be free, without knowing that the adviser is compensated through indirect third-party payments
creating conflicts of interest or that hidden fees that go to the adviser over the life of the
investment will reduce their returns. The risks are growing as baby boomers retire and move
money from plans, where their employer has both the incentive and the fiduciary duty to
facilitate sound investment choices, to IRAs, where both good and bad investment choices are
more numerous and much advice is conflicted. These “rollovers” are expected to approach $2.4
trillion cumulatively from 2016 through 2020.13 Advice on rollovers typically was not covered

11

12

13

U.S. Securities and Exchange Commission Recommendation of the Investor Advisory Committee: Broker-Dealer Fiduciary Duty (Nov.
2013); available at: http://www.sec.gov/spotlight/investor-advisory-committee-2012/fiduciary-duty-recommendation-2013.pdf.
SEC Agency Rule List - Fall 2015, Personalized Investment Advice Standard of Conduct; available at:
http://www.reginfo.gov/public/do/eAgendaViewRule?pubId=201510&RIN=3235-AL27. Any proposal that emerges will then have to go
through a final rule stage before the four appointed SEC commissioners and Chair will vote on it.
Cerulli Associates, “Retirement Markets 2015: Growth Opportunities in Maturing Markets,” 2015.

3

by the 1975 regulation, even though decisions about rollovers are often the most important
financial decisions that consumers make in their lifetime. An ERISA plan investor who rolls her
retirement savings into an IRA could lose 6 to 12 and possibly as much as 23 percent of the
value of her savings over 30 years of retirement by accepting advice from a conflicted financial
adviser.14 Timely regulatory action to redress advisers’ conflicts is warranted to avert such
losses.
As IRAs have grown, so has the demand for personalized advice. Many professionals
offering such advice, such as brokers and insurance agents, traditionally have been paid by
commission and other means that can introduce conflicts and that are prohibited for fiduciary
advisers. However, many have been able to calibrate their business practices to steer around the
narrow 1975 regulation and thereby avoid fiduciary status and the prohibited transaction rules
for accepting conflicted compensation. Many advisers market retirement investment services in
ways that clearly suggest the provision of tailored or individualized advice, while at the same
time relying on the 1975 regulation to disclaim any fiduciary responsibility in the fine print of
contracts and marketing materials. Thus, at the same time that marketing materials characterize
the financial adviser’s relationship with the customer as one-on-one, personalized, and based on
the client’s best interest, footnotes and legal boilerplate disclaim the mutual agreement,
arrangement, or understanding that the advice is individualized or serves as a primary basis for
investment decisions that was requisite for fiduciary status. What is presented to an IRA
investor as trusted advice was often paid for by a financial product vendor in the form of a sales
commission or shelf-space fee, without adequate counter-balancing consumer protections
designed to ensure that the advice is in the investor’s best interest. In another variant of the
same problem, brokers and others who received conflicted compensation recommend specific
products to customers under the guise of general education to avoid triggering fiduciary status
and responsibility.
Likewise in the plan market, pension consultants and advisers that plan sponsors rely on
to guide their decisions often avoid, under the 1975 regulation, fiduciary status under the fivepart test, while receiving conflicted payments. For example, where a plan hires an investment
professional on a one-time basis for an investment recommendation on a large, complex
investment, the adviser has no fiduciary obligation to the plan under ERISA. Even if the plan
official, who lacks the specialized expertise necessary to evaluate the complex transaction on his
or her own, invests all or substantially all of the plan’s assets in reliance on the consultant’s
professional judgment, the consultant is not a fiduciary because he or she did not advise the plan
on a “regular basis” and therefore could stand to profit from the plan’s investment due to a
conflict of interest that affected the consultant’s best judgment. Too much has changed since
1975, and too many investment decisions are made based on one-time advice rather than advice
provided on a regular basis, for the five-part test to be a meaningful safeguard any longer.

14

For example, an ERISA plan investor who rolls $200,000 into an IRA, earns a 6 percent nominal rate of return with 2.3 percent inflation,
and aims to spend down her savings in 30 years, would be able to consume $11,034 per year for the 30-year period. A similar investor
whose assets underperform by 0.5, 1, or 2 percentage points per year would only be able to consume $10,359, $9,705, or $8,466,
respectively, in each of the 30 years. The 0.5 and 1 percentage point figures represent estimates of the underperformance of retail mutual
funds sold by potentially conflicted brokers. These figures are based on a large body of literature cited in the 2015 NPRM Regulatory
Impact Analysis, comments on the 2015 NPRM Regulatory Impact Analysis, and testimony at the DOL hearing on conflicts of interest in
investment advice in August 2015. The 2 percentage point figure illustrates a scenario for an individual where the impact of conflicts of
interest is more severe than average.

4

To be clear, many advisers do put their customers’ best interest first and there are many
good practices in the industry. But the balance of research and evidence indicates that the
aggregate harm from cases in which consumers receive bad advice based on conflicts of interest
is large.
To deal with these issues and update the 1975 regulation for application to the current
business environment, in October 2010, the Department proposed amendments to the 1975
regulation (the 2010 Proposal)15 that would have broadened the definition of fiduciary
investment advice under both ERISA and the IRC, making more advisory activities fiduciary in
nature. The proposal elicited extensive comments and prompted vigorous debate. While many
stakeholders championed the goals of the proposal and some feedback was positive, others
expressed concerns. Some commenters rejected the premise that conflicts pose any dangers to
plan or IRA investors, asserting that the Department had not provided adequate evidence of
tainted advice or adverse consequences. Recurrent themes from the comments were that the
Department should wait until the SEC completes its consideration of related reforms and that the
Department’s Regulatory Impact Analysis was inadequate, because it neglected to consider the
impact the rule would have on the IRA market. Some comments predicted that the 2010
Proposal would have highly negative impacts on IRA investors with small balances. Many
asked the Department to issue PTEs that would allow advisers to continue their current
compensation practices, which would otherwise be prohibited transactions if they engaged in
them as fiduciaries. Recognizing the need to study the issue in greater detail and to produce a
more robust and thorough economic impact analysis, the Department announced in September
2011 its intent to develop and issue a revised proposal in due course.
On April 20, 2015, the Department published a notice in the Federal Register
withdrawing the 2010 Proposal and issued a new proposal (the 2015 Proposal)16 that made many
revisions to the 2010 Proposal, although it also retained aspects of that proposal’s essential
framework.17 Under the 2015 Proposal, the definition of fiduciary investment advice generally
would have covered specific recommendations on investments, investment management, the
selection of persons to provide investment advice or management, and appraisals in connection
with investment decisions. Persons who provided such advice would fall within the proposed
regulation's ambit if they either (a) represented that they were acting as an ERISA fiduciary or
(b) made investment recommendations pursuant to an agreement, arrangement, or understanding
that the advice is individualized or specifically directed to the recipient for consideration in
making investment or investment management decisions regarding plan or IRA assets. The
2015 Proposal specifically included as fiduciary investment advice (under both ERISA and the
IRC) recommendations concerning the investment of assets that are rolled over or otherwise
distributed from a plan. This would have superseded guidance the Department provided in
Advisory Opinion 2005-23A that concluded that such recommendations did not constitute
fiduciary advice. The 2015 Proposal also provided that an adviser does not act as a fiduciary

15

16

17

DOL “Proposed Definition of the Term ‘Fiduciary’,” 75 Fed. Reg. 65263 (Oct. 2010); available at:
http://webapps.dol.gov/FederalRegister/PdfDisplay.aspx?DocId=24328.
DOL “Definition of the Term ’Fiduciary’; Conflict of Interest Rule—Retirement Investment Advice; Proposed Rule,” 80 Fed. Reg. 21928
(April 2015); available at: http://webapps.dol.gov/FederalRegister/PdfDisplay.aspx?DocId=28201.
80 Fed. Reg. 21928.

5

merely by providing plan or IRA investors with information about plan distribution options,
including the tax consequences associated with the available types of benefit distributions.
Critics of the 2010 Proposal had identified a number of activities and circumstances that
they believed would have been unjustifiably swept into fiduciary status. In response, the 2015
Proposal more clearly distinguished situations in which plans, plan participants, and IRA
investors should expect adherence to a fiduciary standard of impartiality and trust, from those
transactions that do not warrant such an expectation, by excluding the following:


Sales pitches involving large plan clients (refining the 2010 Proposal’s similar
exclusion);



Specified communications by counterparties in certain swap transactions;



Certain communications by parties known as “platform providers” who merely make
available a roster of investment options that plan officials can use to populate 401(k)
plan investment menus;



The provision of investment data or identification of investments that meet objective
criteria specified by plan officials;



Recommendations made to plan sponsors by their own employees;



Valuations provided for reporting and disclosure purposes rather than in connection
with transactions; and



Financial education that does not include specific investment recommendations.

Also in response to comments, the 2015 Proposal did not include the 2010 Proposal’s
provision that would have treated all RIAs as fiduciary advisers based upon their status even if
their communication would not otherwise have met the conditions of the regulation.
The Department also responded to the 2010 commenters’ requests for additional
exemptive relief. The 2015 Proposal proposed to narrow and attach new protective conditions to
some existing PTEs. At the same time, it included new flexible, more principles-based proposed
PTEs that apply to a broad range of compensation practices and that included strong protective
conditions. These elements of the proposal reflect the Department’s effort to ensure that advice
is in the best interest of consumers, while avoiding larger and costlier than necessary disruptions
to existing business arrangements or constraints on future innovation.
As part of the 2015 Proposal, the Department conducted an in-depth economic
assessment of current market conditions and the likely effects of reform and conducted and
published a detailed regulatory impact analysis18 pursuant to Executive Order 12866 and other
applicable authorities. That analysis examined a broad range of evidence, including public
comments on the 2010 Proposal; a growing body of empirical, peer-reviewed, academic research
into the effect of conflicts of interest in advisory relationships; a recent study conducted by the
White House Council of Economic Advisers;19 and some other countries’ early experience with

18

19

Department of Labor, “Fiduciary Investment Advice: Regulatory Impact Analysis” (April 14, 2015); available at:
http://www.dol.gov/ebsa/pdf/conflictsofinterestria.pdf.
White House Council of Economic Advisers, “The Effects of Conflicted Investment Advice on Retirement Savings” (Feb. 2015); available
at: https://www.whitehouse.gov/sites/default/files/docs/cea_coi_report_final.pdf.

6

related reform efforts, among other sources. Taken together, the evidence demonstrated that
advisory conflicts are costly to retail and plan investors. The Department’s analysis concluded
that its 2015 Proposal would produce gains for IRA and plan investors, comprising social
welfare improvements and transfers to investors from the financial industry that together would
easily justify associated compliance costs.
The Department took significant steps to give interested persons an opportunity to
comment on the 2015 Proposal and to participate in the rulemaking process. The Department
initially provided a 75-day comment period, ending on July 6, 2015. In response to stakeholder
requests, the Department extended the comment period until July 21, 2015. The Department
also held a four-day public hearing on the new regulatory package in Washington, D.C. on
August 10-13, 2015, at which over 75 speakers testified. A significant portion of the hearing on
August 11 was devoted expressly to testimony from stakeholders specifically regarding the
Department’s Regulatory Impact Analysis. The Department made the hearing transcript
available on EBSA’s website on September 8, 2015, and provided additional opportunity for
interested persons to comment on the proposed regulation and PTEs and hearing transcript until
September 24, 2015. In total, the Department received over 3,000 individual comment letters on
the proposal. The Department also received over 300,000 submissions made as part of 30
separate petitions submitted on the proposal. The comments and petitions came from consumer
groups, plan sponsors, financial services companies, academics, elected government officials,
trade and industry associations, and others, both in support and in opposition to the rule. The
Department also held numerous meetings with interested stakeholders at which the Regulatory
Impact Analysis was discussed. After careful consideration of the issues raised by stakeholders,
the Department is now issuing a final rule and related prohibited transaction exemptions.
The final rule clarifies and rationalizes the definition of fiduciary investment advice. The
rule covers: recommendations by a person who represents or acknowledges that they are acting
as a fiduciary within the meaning of the Act or the Code; advice rendered pursuant to a written
or verbal agreement, arrangement, or understanding that the advice is based on the particular
investment needs of the advice recipient; and advice directed to a specific advice recipient or
recipients regarding the advisability of a particular investment or management decision with
respect to securities or other investment property of the plan or IRA. Under the final rule, a
“recommendation” is a communication that, based on its content, context, and presentation,
would reasonably be viewed as a suggestion that the advice recipient engage in or refrain from
taking a particular course of action.
To further clarify the meaning of recommendation, the final rule provides examples of
services or materials that are not treated as a recommendation. These include services or
materials that provide general communications and commentary on investment products (such as
television, radio, and public media talk show commentary, remarks in widely attended speeches
and conferences, and financial newsletters), marketing or making available a menu of investment
alternatives that a plan fiduciary could choose from, identifying investment alternatives that
meet objective criteria specified by a plan fiduciary, and providing information and materials
that constitute investment education or retirement education. Similarly, the final rule does not
treat as fiduciary advice recommendations made to fiduciaries with financial expertise in an
arm’s length transaction where there is generally no expectation of fiduciary investment advice,
provided that the final rule’s specific conditions are met. In addition, the final rule does not treat
advice rendered to the employer by employees of the plan sponsor, and persons who offer or
enter into swaps or security-based swaps with plans as investment advice.

7

Finally, in addition to the final rule, the Department is simultaneously publishing a final
Best Interest Contract Exemption, Principal Transactions Exemption, and adopting final
amendments to existing PTEs that were proposed as part of the 2015 Proposal. These final PTEs
and amendments to existing PTEs parallel those included in the 2015 Proposal, but reflect some
important revisions that were prompted by public comments. Most of these revisions adjust
some protective conditions to ease compliance, and/or broaden exemptive relief. For example,
the Best Interest Contract Exemption enables fiduciary advisers to receive variable and indirect
compensation such as commissions, 12(b)-1 fees, and revenue sharing, subject to protective
conditions including an enforceable obligation to act in investors’ best interest and institutional
policies and procedures to mitigate adviser conflicts. In response to public comments, the
Department extended the scope of the Best Interest Contract Exemption to include advisers to
fiduciaries of small, participant-directed plans and streamlined conditions related to disclosure
and implementation of firms’ contractual obligations, among other revisions. Responding to
other comments, other revisions strengthen certain consumer protections. For example, an
amendment to a PTE generally available for commissions on insurance products, known as PTE
84-24, was narrowed further from the proposed amendment to exclude advice on insurance
products known as fixed-indexed annuities. These products, which blend limited financial
market exposures with minimum guaranteed values, can play a beneficial and important role in
retirement preparation. However, public comments and other evidence demonstrate that these
products are particularly complex, beset by adviser conflicts, and vulnerable to abuse.
Exemptive relief for commissions on these products remains available under the Best Interest
Contract Exemption, subject to its more protective conditions.
This document presents the Department’s regulatory impact analysis of the final rule and
exemptions. The analysis finds that conflicted advice is widespread, causing serious harm to
plan and IRA investors, and that disclosing conflicts alone would fail to adequately mitigate the
conflicts or remedy the harm. By extending fiduciary status to more advice and providing
flexible and protective PTEs that apply to a broad array of compensation arrangements, the final
rule and exemptions will mitigate conflicts, support consumer choice, and deliver substantial
gains for retirement investors and economic benefits that more than justify its costs.
Advisers’ conflicts of interest take a variety of forms and can bias their advice in a
variety of ways. For example, advisers and their affiliates often profit more when investors
select some mutual funds or insurance products rather than others, or engage in larger or more
frequent transactions. Advisers can capture varying price spreads from principal transactions
and product providers reap different amounts of revenue from the sale of different proprietary
products. Adviser compensation arrangements, which often are calibrated to align their interests
with those of their affiliates and product suppliers, often introduce serious conflicts of interest
between advisers and retirement investors. Advisers often are paid substantially more if they
recommend investments and transactions that are highly profitable to the financial industry, even
if they are not in investors’ best interests. These financial incentives sometimes bias the
advisers’ recommendations. Many advisers do not provide biased advice, but the harm to
investors from those that do can be large in many instances and is large on aggregate.
Following such biased advice can inflict losses on investors in several ways. They may
choose more expensive and/or poorer performing investments. They may trade too much and
thereby incur excessive transaction costs. They may chase returns and incur more costly timing
errors, which are a common consequence of chasing returns.
A wide body of economic evidence supports the Department’s finding that the impact of
these conflicts of interest on retirement investment outcomes is large and negative. The
8

supporting evidence includes, among other things, statistical comparisons of investment
performance in more and less conflicted investment channels, experimental and audit studies,
government reports documenting abuse, and economic theory on the dangers posed by conflicts
of interest and by the asymmetries of information and expertise that characterize interactions
between ordinary retirement investors and conflicted advisers. In addition, the Department
conducted its own analysis of mutual fund performance across investment channels, producing
results broadly consistent with the academic literature.
A careful review of the evidence, which consistently points to a substantial failure of the
market for retirement advice, suggests that IRA holders receiving conflicted investment advice
can expect their investments to underperform by an average of 50 to 100 basis points per year
over the next 20 years. The underperformance associated with conflicts of interest – in the
mutual funds segment alone – could cost IRA investors between $95 billion and $189 billion
over the next 10 years and between $202 billion and $404 billion over the next 20 years.
While these expected losses are large, they represent only a portion of what retirement
investors stand to lose as a result of adviser conflicts. The losses quantified immediately above
pertain only to IRA investors’ mutual fund investments, and with respect to these investments,
reflect only one of multiple types of losses that conflicted advice produces. The estimate does
not reflect expected losses from so-called timing errors, wherein investors invest and divest at
inopportune times and underperform pure buy-and-hold strategies. Such errors can be especially
costly. Good advice can help investors avoid such errors, for example, by reducing panic-selling
during large and abrupt downturns. But conflicted advisers often profit when investors choose
actively managed funds whose deviation from market results (i.e., positive and negative “alpha”)
can magnify investors’ natural tendency to trade more and “chase returns,” an activity that tends
to produce serious timing errors. There is some evidence that adviser conflicts do in fact
magnify timing errors.
The quantified losses also omit losses that adviser conflicts produce in connection with
IRA investments other than mutual funds. Many other products, including various annuity
products, among others, involve similar or larger adviser conflicts, and these conflicts are often
equally or more opaque. Many of these same products exhibit similar or greater degrees of
complexity, magnifying both investors’ need for good advice and their vulnerability to biased
advice. As with mutual funds, advisers may steer investors to products that are inferior to, or
costlier than, similar available products, or to excessively complex or costly product types when
simpler, more affordable product types would be appropriate. Finally, the quantified losses
reflect only those suffered by retail IRA investors and not those incurred by plan investors, when
there is evidence that the latter suffer losses as well. Data limitations impede quantification of
all of these losses, but there is ample qualitative and in some cases empirical evidence that they
occur and are large both in instance and on aggregate.
Disclosure alone has proven ineffective to mitigate conflicts in advice. Extensive
research has demonstrated that most investors have little understanding of their advisers’
conflicts of interest, and little awareness of what they are paying via indirect channels for the
conflicted advice. Even if they understand the scope of the advisers’ conflicts, many consumers
are not financial experts and therefore, cannot distinguish good advice or investments from bad.
The same gap in expertise that makes investment advice necessary and important frequently also
prevents investors from recognizing bad advice or understanding advisers’ disclosures. Some
research suggests that even if disclosure about conflicts could be made simple and clear, it could
be ineffective – or even harmful.

9

This final rule and exemptions aim to ensure that advice is in consumers’ best interest,
thereby rooting out excessive fees and substandard performance otherwise attributable to
advisers’ conflicts, producing gains for retirement investors. Delivering these gains will entail
some compliance costs, – mostly, the cost incurred by new fiduciary advisers to avoid prohibited
transactions and/or satisfy relevant PTE conditions – but the Department has attempted to
minimize compliance costs while maintaining an enforceable best interest standard.
The Department expects compliance with the final rule and exemptions to deliver large
gains for retirement investors by reducing, over time, the losses identified above. Because of
data limitations, as with the losses themselves, only a portion of the expected gains are
quantified in this analysis. The Department’s quantitative estimate of investor gains from the
final rule and exemptions takes into account only one type of adviser conflict: the conflict that
arises from variation in the share of front-end-loads that advisers receive when selling different
mutual funds that charge such loads to IRA investors. Published research provides evidence that
this conflict erodes investors’ returns. The Department estimates that the final rule and
exemptions, by mitigating this particular type of adviser conflict, will produce gains to IRA
investors worth between $33 billion and $36 billion over 10 years and between $66 and $76
billion over 20 years
These quantified potential gains do not include additional potentially large, expected
gains to IRA investors resulting from reducing or eliminating the effects of conflicts in IRA
advice on financial products other than front-end-load mutual funds or the effect of conflicts on
advice to plan investors on any financial products. Moreover, in addition to mitigating adviser
conflicts, the final rule and exemptions raise adviser conduct standards, potentially yielding
additional gains for both IRA and plan investors. The total gains to retirement investors thus are
likely to be substantially larger than these particular, quantified gains alone.
The final exemptions include strong protections calibrated to ensure that adviser conflicts
are fully mitigated such that advice is impartial. If, however, advisers’ impartiality is sometimes
compromised, gains to retirement investors consequently will be reduced correspondingly.
The Department estimates that the cost to comply with the final rule and exemptions will
be between $10.0 billion and $31.5 billion over 10 years with a primary estimate of $16.1
billion, mostly reflecting the cost incurred by affected fiduciary advisers to satisfy relevant
consumer-protective PTE conditions. Costs generally are estimated to be front-loaded,
reflecting a substantial amount of one-time, start-up costs. The Department’s primary 10-year
cost estimate of $16.1 billion reflects the present value of $5.0 billion in first-year costs and $1.5
billion in subsequent annual costs. These cost estimates may be overstated insofar as they
generally do not take into account potential cost savings from technological innovations and
market adjustments that favor lower-cost models. They may be understated insofar as they do
not account for frictions that may be associated with such innovations and adjustments.
Just as with IRAs, there is evidence that conflicts of interest in the investment advice
market also erode the retirement savings of plan participants and beneficiaries. For example, the
U.S. Government Accountability Office (GAO) found that defined benefit pension plans using
consultants with undisclosed conflicts of interest earned 1.3 percentage points per year less than

10

other plans.20 Other GAO reports have found that adviser conflicts may cause plan participants
to roll plan assets into IRAs that charge high fees or 401(k) plan officials to include expensive or
underperforming funds in investment menus.21 A number of academic studies find that 401(k)
plan investment options underperform the market, and at least one study attributes such
underperformance to excessive reliance on funds that are proprietary to plan service providers
who may be providing investment advice to plan officials that choose the investment options.
The final rule and exemptions’ positive effects are expected to extend well beyond
improved investment results for retirement investors. The IRA and plan markets for fiduciary
advice and other services may become more efficient as a result of more transparent pricing and
greater certainty about the fiduciary status of advisers and about the impartiality of their advice.
There may be benefits from the increased flexibility that the final rule and related exemptions
will provide with respect to fiduciary investment advice currently falling within the ambit of the
1975 regulation. The final rule’s defined boundaries between fiduciary advice, education, and
sales activity directed at independent fiduciaries with financial expertise may bring greater
clarity to the IRA and plan services markets. Innovation in new advice business models,
including technology-driven models, may be accelerated, and nudged away from conflicts and
toward transparency, thereby promoting healthy competition in the fiduciary advice market.
A major expected positive effect of the final rule and exemptions in the plan advice
market is improved compliance and the associated improved security of ERISA plan assets and
benefits. Clarity about advisers’ fiduciary status will strengthen the Department’s ability to
quickly and fully correct ERISA violations, while strengthening deterrence.
A portion of retirement investors’ gains from the final rule and exemptions represents
improvements in overall social welfare, as some resources heretofore consumed inefficiently in
the provision of financial products and services are freed for more valuable uses. The remainder
of the projected gains reflects transfers of existing economic surplus to retirement investors
primarily from the financial industry. Both the social welfare gains and the distributional effects
can promote retirement security, and the distributional effects more fairly (in the Department’s
view) allocate a larger portion of the returns on retirement investors’ capital to the investors
themselves. Because quantified and additional unquantified investor gains from the final rule
and exemptions comprise both welfare gains and transfers, they cannot be netted against
estimated compliance costs to produce an estimate of net social welfare gains. Rather, in this
case, the Department concludes that the final rule and exemptions’ positive social welfare and
distributional effects together justify their cost.
A number of comments on the Department’s 2015 Proposal, including those from
consumer advocates, some independent researchers, and some independent financial advisers,
largely endorsed its accompanying impact analysis, affirming that adviser conflicts cause
avoidable harm and that the proposal would deliver gains for retirement investors that more than
justify compliance costs, with minimal or no unintended adverse consequences. In contrast,
many other comments, including those from most of the financial industry (generally excepting
only comments from independent financial advisers), strongly criticized the Department’s

20

21

U.S. Government Accountability Office, GAO-09-503T, Private Pensions: Conflicts of Interest Can Affect Defined Benefit and Defined
Contribution Plans (2009) available at: http://www.gao.gov/new.items/d09503t.pdf.
GAO Publication No. GAO-11-119, 36.

11

analysis and conclusions. These comments collectively argued that the Department’s evidence
was weak, that its estimates of conflicts’ negative effects and the proposal’s benefits were
overstated, that its compliance cost estimates were understated, and that it failed to anticipate
predictable adverse consequences including increases in the cost of advice and reductions in its
availability to small investors, which the commenters said would depress saving and exacerbate
rather than reduce investment mistakes. Some of these comments took the form of or were
accompanied by research reports that collectively offered direct, sometimes technical critiques of
the Department’s analysis, or presented new data and analysis that challenged the Department’s
conclusions. The Department took these comments into account in developing this analysis of
its final rule and exemptions. Many of these comments were grounded in practical operational
concerns which the Department believes it has alleviated through revisions to the 2015 Proposal
reflected in this final rule and exemptions. At the same time, however, many of the reports
suffered from analytic weaknesses that undermined the credibility of some of their conclusions.
Many comments anticipating sharp increases in the cost of advice neglected the costs
currently attributable to conflicted advice including, for example, indirect fees. Many
exaggerated the negative impacts (and neglected the positive impacts) of recent overseas reforms
and/or the similarity of such reforms to the 2015 Proposal. Many implicitly and without support
assumed rigidity in existing business models, service levels, compensation structures, and/or
pricing levels, neglecting the demonstrated existence of low-cost solutions and potential for
investor-friendly market adjustments. Many that predicted that only wealthier investors would
be served appeared to neglect the possibility that once the fixed costs of serving wealthier
investors was defrayed, only the relatively small marginal cost of serving smaller investors
would remain for affected firms to bear in order to serve these consumers.22
The Department expects that, subject to some short-term frictions as markets adjust,
investment advice will continue to be readily available when the final rule and exemptions are
applicable, owing to both flexibilities built into the final rule and exemptions and to the
conditions and dynamics currently evident in relevant markets. Moreover, recent experience in
the United Kingdom suggests that potential gaps in markets for financial advice are driven
mostly by factors independent of reforms to mitigate adviser conflicts. Commenters’
conclusions that stem from an assumption that advice will be unavailable therefore are of limited
relevance to this analysis. Nonetheless, the Department notes that these commenters’ claims
about the consequences of the rule would still be overstated even if the availability of advice
were to decrease. Many commenters arguing that costlier advice will compromise saving
exaggerated their case by presenting mere correlation (wealth and advisory services are found
together) as evidence that advice causes large increases in saving. Some wrongly implied that
earlier Department estimates of the potential for fiduciary advice to reduce retirement investment
errors – when accompanied by very strong anti-conflict consumer protections – constituted an
acknowledgement that conflicted advice yields large net benefits.

22

Fixed costs include start-up costs for complying with the new exemptions, such as system changes, drafting contracts, putting in place
policies and procedures, and training staff. The marginal cost of servicing small accounts include providing investors with the contract
which is not investor specific (except to say whether and the extent to which they will monitor the account), providing information upon
request about each investment which is investment specific, but not customer specific, and for level fee fiduciaries, documenting why a
level fee account is good for the investor.

12

The negative comments that offered their own original analysis, and whose conclusions
contradicted the Department’s, also are generally unpersuasive on balance in the context of this
present analysis. For example, these comments collectively neglected important factors such as
indirect fees, made comparisons without adjusting for risk, relied on data that are likely to be
unrepresentative, failed to distinguish conflicted from independent advice, and/or presented as
evidence median results when the problems targeted by the 2015 Proposal and the proposal’s
expected benefits are likely to be concentrated on one side of the distribution’s median.
In light of the Department’s analysis, its careful consideration of the comments, and
responsive revisions made to the 2015 Proposal, the Department stands by its analysis and
conclusions that adviser conflicts are inflicting large, avoidable losses on retirement investors,
that appropriate, strong reforms are necessary, and that compliance with this final rule and
exemptions can be expected to deliver large net gains to retirement investors. The Department
does not anticipate the substantial, long-term unintended consequences predicted in the negative
comments.
In conclusion, the Department’s analysis indicates that the final rule and exemptions will
mitigate adviser conflicts and thereby improve plan and IRA investment results, while avoiding
greater than necessary disruption of existing business practices. The final rule and exemptions
have the potential to deliver large gains to retirement investors, reflecting a combination of
improvements in economic efficiency and worthwhile transfers to retirement investors from the
financial industry, and a variety of other positive economic effects, which, in the Department’s
view, will more than justify its costs.

13

14

15

1. Introduction
Tax-preferred retirement savings, in the form of private sector, employer-sponsored
retirement plans (plans) and IRAs, are critical to the retirement security of most U.S. workers. It
is therefore imperative that these savings are invested well. Investment professionals play a
major and largely beneficial role in guiding the investment decisions of plan officials, plan
participants, and IRA investors. But many of these professionals are compensated in ways that
may introduce conflicts of interest between them and the plan officials, plan participants, and the
IRA investors they advise. If the conflicts taint the investment advice they render,
underperformance could result and the retirement security of millions of America’s workers and
their families could be threatened. In economic terms, imperfect information may cause the
market for investment advice to fail: plan officials, plan participants, and IRA investors may
sometimes unknowingly pay for and follow tainted advice and consequently suffer large but
mostly hidden opportunity costs. The analysis that follows concludes that this is in fact
occurring today.
ERISA and the IRC together provide that anyone paid to provide advice on the
investment of plan or IRA assets is a fiduciary. As fiduciaries, they are subject to certain duties,
including the general avoidance of conflicts of interest. However, a 1975 regulation narrowly
construed these ERISA and IRC provisions, effectively relieving many advisers of these duties.
The Department is issuing a final rule that will revise the 1975 regulation to expand the
definition of fiduciary to include those who are not fiduciaries under the existing rule but should
be based on their conduct. The final rule extends fiduciary duties to more advisers to remedy
failures in the present day marketplace and thereby improve plan and IRA investing for the longterm retirement security of participants and IRA investors.
Executive Orders 1356323 and 1286624 direct agencies to assess all costs and benefits of
available regulatory alternatives and, if regulation is necessary, to select regulatory approaches
that maximize net benefits (including potential economic, environmental, public health and
safety effects, distributive impacts, and equity). Executive Order 13563 emphasizes the
importance of quantifying both costs and benefits, of reducing costs, of harmonizing and
streamlining rules, and of promoting flexibility. It also requires Federal agencies to develop a
plan under which the agencies will periodically review their existing significant regulations to
make the agency’s regulatory program more effective or less burdensome in achieving
regulatory objectives.
Under Executive Order 12866, “significant” regulatory actions are subject to the
requirements of the Executive Order and review by the Office of Management and Budget
(OMB). Section 3(f) of the Executive Order defines a “significant regulatory action” as an
action that is likely to result in a rule: (1) having an annual effect on the economy of $100
million or more, or adversely and materially affecting a sector of the economy, productivity,
competition, jobs, the environment, public health or safety, or State, local, or tribal governments
or communities (also referred to as “economically significant” regulatory actions); (2) creating
serious inconsistency or otherwise interfering with an action taken or planned by another

23
24

Exec. Order No. 13563, 3 C.F.R. 13563 (2011); available at: http://www.gpo.gov/fdsys/pkg/FR-2011-01-21/pdf/2011-1385.pdf.
Exec. Order No. 12866, 3 C.F.R. 638 (1993); available at: http://www.archives.gov/federal-register/executive-orders/pdf/12866.pdf.

16

agency; (3) materially altering the budgetary impacts of entitlement grants, user fees, or loan
programs or the rights and obligations of recipients thereof; or (4) raising novel legal or policy
issues arising out of legal mandates, the President's priorities, or the principles set forth in the
Executive Order. OMB has determined that this final rule and related prohibited transaction
exemptions are economically significant within the meaning of Section 3(f)(1) of the Executive
Order because they would be likely to have an effect on the economy of $100 million in at least
one year. Accordingly, OMB has reviewed the final rule and exemptions pursuant to the
Executive Order.
Executive Order 12866 and OMB Circular A-4 (Circular A-4)25 require agencies to
provide the public and OMB with a careful and transparent regulatory impact analysis of
economically significant regulatory actions. In particular, a regulatory impact analysis should
include: (1) an assessment and (to the extent feasible) a quantification and monetization of the
benefits and costs the Department anticipates will result from the regulation; (2) a clear
explanation of the need for regulatory action, including a description of the problem the agency
seeks to address; and (3) a range of potentially effective and reasonably feasible regulatory
alternatives.
The Department believes this Regulatory Impact Analysis has been conducted pursuant
to the requirements of Executive Order 12866 and Circular A-4. Chapters 3 and 4 provide a
detailed description of the need for regulatory action and establish a baseline for the magnitude
of harm that results from conflicts of interest in the IRA and ERISA plan markets. These
chapters also explain how the final rule and new and amended exemptions will meet that need by
ensuring that investment advisers provide advice that is in the clients’ best interest. The
estimates in the Regulatory Impact Analysis are based on reasonable, obtainable scientific,
technical, and economic information, and the Department has strived to present them in an
accurate, clear and unbiased manner. The data, sources, and methods used are cited and
described in this Regulatory Impact Analysis and its Technical Appendix, which allows the
regulated community, researchers, and other interested parties to replicate the results of the
Department’s analysis.
In Chapter 3, the Department quantifies and monetizes the anticipated gains to investors
the regulatory action may deliver, including economic efficiency gains and transfers from the
financial services industry, and qualitatively describes the benefits that will be derived in the
plan market in Chapter 4. The Department quantifies and monetizes the anticipated costs of the
final rule in Chapter 5 and concludes that the final rule and exemptions’ positive social welfare
and distributional effects together justify their costs. In calculating its estimates, the Department
uses appropriate discount rates specified in Circular A-4 for benefits and costs that are expected
to occur in the future.26
In Chapter 7, the Department assesses potentially effective and reasonably feasible
alternatives to the regulation and explains why the regulation is preferable to the identified
potential alternatives. Chapter 8 discusses uncertainties attendant to the Department’s analysis.
Finally, the Executive Summary, above, provides a plain-language explanation of the regulatory

25
26

OMB Circular A-4, available at: https://www.whitehouse.gov/omb/circulars_a004_a-4/.
The Department uses three and seven percent discount rates for benefits and costs as required by Circular A-4.

17

action and regulatory impact analysis, assessing the benefits, costs, and transfer impacts of the
regulatory action, including the qualitative and non-monetized benefits, costs, and transfer
impacts.

18

19

2. Legal Environment: ERISA and the IRC
When Congress enacted ERISA, it established special consumer protections for taxpreferred retirement savings. These include ERISA and IRC provisions designed to ensure
accountability and curb conflicts of interest among advisers to plan and IRA investors. The
Department is responsible for interpreting these ERISA and IRC provisions. This section
describes the current ERISA and IRC legal environment, major intersections between these
regimes and other laws, and the regulatory and subregulatory guidance the Department has
issued in the investment advice area to implement the relevant ERISA and IRC provisions.

2.1

Statutory Provisions

ERISA established several provisions governing advice on the investment of plan and
IRA assets. Some of these provisions were included in ERISA itself and made applicable to
advice on the investment of plan assets only, while others were added to the IRC and made
applicable to advice on the investment of both plan and IRA assets.

2.1.1 Provisions Relating to Plans
Under both ERISA and
Figure 2-1 ERISA and IRC Provisions Governing Fiduciary
the IRC, any person paid directly
Advice on the Investment of Plan and IRA Assets
or indirectly to provide plan
officials or participants with
Statute
ERISA
IRC
advice on the investment of plan
Be prudent and loyal to
Avoid conflicts.
Fiduciary
assets is a fiduciary.27 ERISA
participants’
interests.
Advisers’ Duties
requires fiduciaries to discharge
Avoid conflicts.
their duties prudently and solely
Personal liability for any
Excise tax
Sanctions
in the interest of plan
28
loss
arising
from
breach
of
participants. ERISA also
duty
generally requires fiduciaries to
Plans only
Plans and IRAs
Applies to
refrain from certain prohibited
transactions, which may involve
DOL
DOL
Rulemaking
conflicts of interest. Under
authority
ERISA’s prohibited transactions
DOL
IRS
Enforcement
provisions, fiduciaries generally
may not self-deal. In other
authority
words, they may not deal with
plan assets for their own interest or account, or be paid by a third party in connection with a
transaction involving plan assets.29 Fiduciaries may be held personally liable under ERISA for
any loss of plan assets arising from breaches of these duties.30 The IRC contains prohibited
transactions provisions parallel to ERISA.31 Under these IRC provisions, fiduciary advisers who

27
28
29
30
31

ERISA § 3(21)(A)(ii) and IRC § 4975(e)(3)(B).
ERISA § 404.
ERISA § 406.
ERISA § 409.
IRC § 4975.

20

self-deal are subject to an excise tax equal to 15 percent of the amount involved, or, if the
prohibited transactions is not corrected in a timely fashion, 100 percent of that amount.

2.1.2 Provisions Relating to IRAs
The regulatory structure governing IRAs is complicated and frequently misunderstood.
ERISA does not apply to retail IRAs; however, the relevant IRC provisions do apply to them.
Under the IRC, any person paid to provide advice on the investment of IRA assets is a fiduciary.
As with plan fiduciaries, the IRC prohibited transactions provisions generally prohibit such IRA
fiduciaries from self-dealing, enforced through an excise tax. ERISA’s duties of prudence and
loyalty do not apply to IRA fiduciaries nor are they liable under ERISA for losses arising from
breaches of such duties or from prohibited transactions.32 There is no private right of action for
prohibited transactions violations under the IRC. However, since 1978, the authority to define
who is a fiduciary and to interpret the IRC prohibited transactions provisions (including the
ability to draft exemptions from those provisions) is delegated to the Secretary of Labor under
Reorganization Plan No. 4 of 1978.33

2.1.3 Permissible Self-Dealing
The foregoing statutory provisions of ERISA and the IRC generally prohibit fiduciary
investment advisers to plan and IRA investors from accepting compensation that introduces
conflicts of interest. Thus, under ERISA and the IRC, the default rule is not that fiduciaries must
disclose their conflicts of interest, but rather that they must refrain from engaging in such
conflicted transactions in the first place. ERISA and the IRC presume that retirement investors
are best protected by strict prohibitions on conflicts of interest, subject to the Department’s
authority to grant prohibited transaction exemptions (PTEs) based on conditions that are
administratively feasible, in the interests of plans, participants and beneficiaries, and IRA
investors, and protective of their rights.
In practice, however, many advisers in the retirement marketplace are highly conflicted.
They receive a wide variety of forms of compensation that vary based on the investment
decisions made pursuant to their advice. These forms of compensation include, but are not
limited to: transaction-related commissions; mutual fund distribution fees known as 12b-1 fees;
revenue sharing from various third parties with an interest in the investment decisions, such as
RIAs managing mutual funds; and mark-ups on securities sold from (or mark-downs on
securities bought by) their own or their affiliates’ own accounts. Without the new and amended
PTEs granted as part of this regulatory package, all of these practices would be prohibited.
There generally are two different legal paths available to advisers who wish to accept
variable compensation in connection with advice provided to plan or IRA investors. They can
rely on a PTE from the otherwise applicable ERISA and IRC prohibited transactions provisions.
Alternatively, they can calibrate their business practices to avoid being classified as fiduciary
investment advisers under the 1975 regulation. These paths are discussed below.

32
33

See Section 2.6 for a discussion of remedies under securities laws.
Reorganization Plan No. 4 of 1978, 43 Fed. Reg. 47713 (Oct. 17, 1978), 92 Stat. 3790, 5 U.S.C. § App. (2010); available at:
http://www.gpo.gov/fdsys/pkg/USCODE-2010-title5/pdf/USCODE-2010-title5-app-reorganiz-other-dup102.pdf.

21

2.2

Exemptions from the Prohibited Transactions Provisions

As noted above, advisers wishing to accept variable compensation in connection with
advice on the investment of plan or IRA assets can take advantage of one or more PTEs. ERISA
and the IRC each provide a limited, parallel set of statutory PTEs. The Department has the
authority to issue additional class or individual PTEs.
From a fiduciary's point of view, a PTE is permissive: it allows the fiduciary to engage in
certain transactions, such as self-dealing, that would otherwise be prohibited. From the
Department’s perspective, a PTE must be protective. ERISA provides that before a class or
individual PTE can be granted by the Department, the Secretary must find that the exemption is
administratively feasible, in the participants’ interests, and protective of participants’ rights.34
Because prohibited transactions generally involve potential conflicts of interest, the Department
typically attaches conditions to PTEs that are intended to ensure transparency, impartiality,
accountability, and to protect plan participants, beneficiaries, and IRA investors. A fiduciary
adviser who wishes to take advantage of a PTE must satisfy its conditions. Failure to satisfy the
conditions can result in a non-exempt PT and associated sanctions, such as the prohibited
transactions excise tax.
Relevant PTEs are discussed below.

2.2.1 Statutory Investment Advice Exemption
The Pension Protection Act of 2006 (PPA)35 amended ERISA and the Code to establish a
new statutory PTE for fiduciary investment advisers to plan participants and IRA investors.36
This PTE permits advisers to receive indirect compensation from third parties in connection with
the investment products they recommend to plan participants and IRA investors.
Congress recognized that such compensation can pose conflicts of interest that could
taint advice, resulting in poor or suboptimal investment recommendations. Consequently, relief
under this PTE is subject to a number of protective conditions. For example, the advice must be
provided under one of two types of “eligible investment advice arrangements.” Under one
permissible type of arrangement, any fees (including any commission or other compensation)
received by the fiduciary adviser and the adviser’s firm may not vary based on the investment
products selected by the plan participant or IRA investor. Under the terms of the exemption,
however, compensation paid to the fiduciary adviser’s affiliates may vary. The other type of
arrangement requires the adviser to provide (and not alter) the investment advice
recommendation derived from a computer model meeting certain requirements, including a
requirement that the model be independently certified to be unbiased in favor of investment
options offered by the fiduciary adviser or related persons and for all investment options under

34
35
36

ERISA § 408(a).
Pub. Law. No. 109-280, 120 Stat. 780.
ERISA §§ 408(b)(14) and 408(g), Code sec. 4975(d)(17) and 4975(f)(8).

22

the plan to be taken into account in specifying how a participant’s account balance should be
invested.37

2.2.2 Administrative PTEs
Since 1978, the Department has been solely responsible for interpreting and issuing
exemptions from the prohibited transactions provisions of both ERISA and the IRC.38
A number of existing class PTEs permit fiduciary advisers to engage in several classes of
prohibited transactions in connection with the provision of fiduciary investment advice to plans,
plan participants, or IRA investors.39 These PTEs are named for the year and sequential order in
which they were granted. These PTEs are being amended along with the issuance of the final
rule in 2016, as discussed in Section 2.9. Among other amendments, these PTEs are amended to
condition relief on “impartial conduct standards,” including a best interest standard as discussed
in more detail below. The most important of these exemptions are PTE 84-24 (e.g. covering
specified transactions involving mutual fund shares or insurance or annuity contracts); PTE 77-4
(concerning the purchase or sale of mutual funds in specified conditions); PTE 75-1, Part IV
(specified relief for “market-makers” in certain securities transactions): and PTE 86-128 (relief
in connection with specified securities and cross-transactions).

2.3

The 1975 Regulation

Under ERISA and the Code, a person is a fiduciary to a plan or IRA to the extent that he
or she engages in specified plan activities, including rendering “investment advice for a fee or
other compensation, direct or indirect, with respect to any moneys or other property of such
plan.”40 In 1975, the Department and the IRS issued parallel rules that narrowed the statutory
ERISA and IRC definitions of fiduciary investment advice (“1975 regulation.”) The 1975
regulation established five conditions, all of which were required to be satisfied in connection
with each instance in which advice was rendered before the person rendering the advice would
be classified as having acted as a fiduciary in rendering that advice. As discussed above, the five
conditions required that the adviser:
(1) Make recommendations on investing in, purchasing or selling securities or other

property, or give advice as to their value;
(2) On a regular basis;
(3) Pursuant to a mutual understanding that the advice;
(4) Will serve as a primary basis for investment decisions; and
(5) Will be individualized to the particular needs of the plan.

37

38
39
40

ERISA § 408(g)(2), in relevant part, states that both types of arrangements must be expressly authorized by a plan fiduciary other than the
person offering the advice program and have an annual audit performed by an independent auditor who issues a written report to the
authorizing fiduciary presenting specific findings regarding compliance of the arrangement with the statutory exemption. In addition, the
fiduciary adviser must provide detailed disclosures to plan participants and IRA investors.
Reorganization Plan No. 4 of 1978, 43, 5 U.S.C. § App. (2010).
Some of these PTEs also provide relief to other fiduciaries in connection with other activities.
ERISA § 3(21)(A) and IRC §4975(e)(3).

23

Under the 1975 regulation, an investment adviser does not act as a fiduciary unless each
element of the five-part test is satisfied for each instance of advice. Therefore, if a plan official
hires an investment advice professional on a one-time basis to provide advice on a large complex
investment, the adviser is not acting as a fiduciary, because the advice is not given on a “regular
basis” as the regulation required. Similarly, if an adviser provides one-time individualized, paid
advice to a worker nearing retirement on the purchase of an annuity, the adviser is not acting as a
fiduciary, because the advice is not provided on a regular basis. This is the result even though
the advice involves the investment of a worker's entire IRA or 401(k) account balance, or
defined benefit plan balance.
If the adviser is not acting as a fiduciary, the self-dealing and conflict of interest
prohibited transactions provisions of ERISA and the IRC do not apply. Therefore, such an
adviser to a plan official, participant, or IRA investor is free to self-deal by accepting variable
compensation that introduces a conflict of interest into the advisory relationship. ERISA’s
additional fiduciary duties of prudence and loyalty likewise do not apply and the adviser faces
no liability for breaches of such duties.
Advisers often deliberately calibrate their business practices to avoid satisfying one or
more of the 1975 regulation’s conditions in the course of rendering advice on the investment of
plan or IRA assets. Materials describing their adviser services for current and prospective plan
and IRA clients, such as customer agreements or advertisements, often specifically disclaim
satisfaction of one or more elements of the 1975 regulation.41

2.4

Relevant Advisory Opinions

From time to time, the Department issues “Advisory Opinions” (AOs). These are written
interpretive statements issued by the Department to an applicant or his or her authorized
representative that interpret and apply Title I of ERISA to a specific factual situation presented
by the applicant. Some of these AOs have in effect further narrowed the scope of what is
considered to be fiduciary investment advice under ERISA.

2.4.1 AO 97-15A (Frost Bank) and 2005-10A (Country Bank)
In AO 97-15A (May 22, 1997) issued to Frost Bank, the Department opined that, where a
fiduciary advises a plan to invest in mutual funds that pay additional fees to the advising
fiduciary, the advising fiduciary generally would engage in transactions that violated ERISA
Section 406(b)(1). However, to the extent that the fiduciary uses every dollar of fees the mutual
funds pay the fiduciary to offset fees that the plan is otherwise legally obligated to pay the
fiduciary (e.g., for trustee services), Section 406(b)(1) will not be violated because the fiduciary
is not considered to be dealing with plan assets for his or her own account. The Department
noted that the bank would be an ERISA fiduciary to the extent it would advise plan sponsors on

41

For example, one large financial services company included the following language in the fine print of its print and television
advertisements specifically offering to provide one-on-one investment help to individuals: “[g]uidance provided … is educational in nature,
is not individualized, and is not intended to serve as the primary basis for your investment and tax-planning decisions.” Notwithstanding
such disclaimers, whether the conditions of the 1975 regulation are met depends on the facts and circumstances associated with each
instance where advice is rendered. For example, if an IRA investor and his or her financial adviser in fact mutually understand that certain
advice will serve as the primary basis for an IRA investment decision, then the third and fourth conditions are met, notwithstanding any
disclaimer to the contrary that might be included in a customer agreement.

24

which mutual funds to invest in or to make available to participants, or reserve the right to add or
remove mutual funds that it makes available to the plans.
In a subsequent opinion, 2005-10A (May 11, 2005) issued to Country Bank, the
Department confirmed that the "fee leveling or offset" approach may be applied where advisory
services are delivered to an IRA and where fees are paid from either affiliated or unaffiliated
mutual funds. In this case, the bank advised its clients on how to invest IRA assets in a manner
consistent with five model investment strategies.

2.4.2 AO 2001-09A (SunAmerica), Investment Advice Programs
In Advisory Opinion 2001-09A (Dec. 14, 2001) issued to SunAmerica Retirement
Markets, Inc., the Department concluded that a financial institution could offer an investment
advice program to plan investors under which it would pay an independent financial expert to
formulate investment recommendations using a computer model that the financial institution
would furnish to plan participants to allocate their account assets among collective investment
vehicles (funds) that would in turn pay varying, and therefore potentially greater, investment
advisory fees to the financial institution and its affiliates without violating ERISA’s prohibited
transaction rules.
As represented by the financial institution, the program worked as follows: a plan
fiduciary independent of the financial institution, and its affiliates, would determine whether the
plan should participate in the program and designate the investment alternatives to be offered
under the plan with respect to which the financial institution would furnish recommendations to
participants regarding allocations. The plan’s independent fiduciary would be provided detailed
information concerning, among other things, the program, and the role of the financial expert in
the development of the asset allocations under the program. In addition, the plan’s independent
fiduciary would be provided, on an ongoing basis, a number of disclosures concerning the
program and the designated investments under the plan, including information pertaining to
performance and rates of returns on designated investments, and with respect to funds advised
by the financial institution designated under the arrangement, the expenses and fees of the funds,
and any proposed increases in investment advisory or other fees charged.
The financial institution’s decisions regarding whether to retain the financial expert were
represented to be independent of the revenue generated by the asset allocations under the
program. The independent financial expert’s compensation would not be dependent on
allocations among investment alternatives under the plan. The annual gross income of the
financial expert from the financial institution and its affiliates would not exceed five percent of
its total income.
The independent expert would have sole control over development and maintenance of
the computer model that would formulate the recommendations for participants in the form of
model portfolio asset allocations. The recommended allocations would reflect solely the input of
participant information into computer programs utilizing methodologies and parameters
provided by the financial expert and neither the financial institution, nor its affiliates, would be
retained as computer programmers to formulate the model or be able to change or affect the
output of the computer programs.
The Department concluded, assuming the truth of all the representations above, that the
individual investment recommendations provided under the program would not be the result of
the financial institution’s exercise of authority, control, or responsibility for purposes of ERISA
Section 406(b) and the applicable regulations, based on: (1) the fully informed approval of
participation in the program by the plan’s independent fiduciary; (2) the financial expert’s sole
25

and independent discretion over the development, maintenance and oversight of the
methodologies producing the investment recommendations, and the financial institution’s lack of
discretion over the communication to, or implementation of, investment recommendations under
the program; and (3) the absence of any compensation or arrangements that were tied to the
recommendations made under the program.
After the Department issued AO 2001-09A, some investment providers told the
Department and Congress that they wanted to develop their own computer models and use them
to provide advice to participants and beneficiaries rather than employing an independent
financial expert to develop and apply the model. In response, Congress recognized the potential
conflict of interest that was involved when firms used their proprietary computer models and
enacted the investment advice statutory exemption with appropriate safeguards and conditions as
part of PPA.42 As discussed in Section 2.2.1, above, the PPA statutory exemption allows
fiduciary advisers to receive indirect compensation for advice generated by their own computer
models so long as certain requirements are met, including requiring an independent investment
expert to certify that a computer model operates in a manner that is not biased in favor of
investments offered by the investment advice provider before the computer model is used. An
independent auditor must perform an annual audit of the arrangement for compliance with the
conditions of the statutory exemption.

2.4.3 AO 2005-23A Regarding Rollovers
In AO 2005-23A, the Department addressed whether a recommendation that a participant
take a distribution from his or her DC plan and roll over the funds to an IRA was subject to
ERISA’s fiduciary standards and associated prohibited transactions provisions of ERISA and the
IRC. Specifically, the AO addressed whether a recommendation that a participant roll over an
account balance to an IRA to take advantage of investment options not available under the plan
would constitute “investment advice” with respect to the plan or the participant. AO 2005-23A
concluded that advising a plan participant to take an otherwise permissible plan distribution,
even when that advice is combined with a recommendation as to how the distribution should be
invested, does not by itself constitute “investment advice” within the meaning of the 1975
regulation. The Department opined that a recommendation to take a distribution is not advice or
a recommendation concerning a particular investment (i.e., purchasing or selling securities or
other property) as contemplated by the 1975 regulation, and that any investment
recommendation regarding the proceeds of such a distribution would be advice with respect to
funds that are no longer plan assets. However, in instances where a plan officer or someone who
is already a plan fiduciary responds to participant questions concerning the advisability of taking
a distribution or the investment of amounts withdrawn from the plan, the Department opined in
AO 2005-23A that the fiduciary is exercising discretionary authority respecting management of
the plan and must act prudently and solely in the interest of the participant. As discussed below,
the final regulation issued today reverses the Advisory Opinion and treats paid advice with
respect to distributions and rollovers as fiduciary advice.

42

ERISA §§ 408(b)(14) and 408(g) and IRC §§ 4975(d)(17) and 408(f)(8).

26

2.5

Interpretive Bulletin on Investment Education

With the increase in the number of participant-directed individual account plans and the
number of investment options available to participants covered by such plans, plan sponsors and
service providers have increasingly recognized the importance of providing participants and
beneficiaries with information designed to assist them in making investment and retirementrelated decisions appropriate to their particular situations. Concerns were expressed to the
Department, however, that providing educational information to participants and beneficiaries
may be viewed as rendering “investment advice for a fee or other compensation,” thereby giving
rise to fiduciary status and potential liability under ERISA for investment decisions of plan
participants and beneficiaries.
In response to these concerns, the Department issued Interpretive Bulletin 96-1 (IB 961),43 which identified the following four categories of investment-related educational materials
that can be provided to participants and beneficiaries without providing fiduciary investment
advice: (1) Plan Information; (2) General Financial and Investment Information; (3) Asset
Allocation Models; and (4) Interactive Investment Materials. Each category of information is
discussed below.

43



Plan Information is defined as information and materials that inform a participant or
beneficiary about the benefits of plan participation, the benefits of increasing plan
contributions, the impact of preretirement withdrawals on retirement income, the
terms of the plan, or the operation of the plan; or information regarding investment
alternatives under the plan (e.g., descriptions of investment objectives and
philosophies, risk and return characteristics, historical return information, or related
prospectuses).



General Financial and Investment Information is defined as information and
materials that inform a participant or beneficiary about: (i) general financial and
investment concepts, such as risk and return, diversification, dollar cost averaging,
compounded return, and tax deferred investment; (ii) historic differences in rates of
return between different asset classes (e.g., equities, bonds, or cash) based on
standard market indices; (iii) effects of inflation; (iv) estimating future retirement
income needs; (v) determining investment time horizons; and (vi) assessing risk
tolerance.



Asset Allocation Models is defined as information and materials (e.g., pie charts,
graphs, or case studies) that provide a participant or beneficiary with models,
available to all plan participants and beneficiaries, of asset allocation portfolios of
hypothetical individuals with different time horizons and risk profiles.



Interactive Investment Materials is defined as questionnaires, worksheets,
software, and similar materials that provide a participant or beneficiary with the
means to estimate future retirement income needs and assess the impact of different
asset allocations on their retirement income.

Interpretive Bulletin 96-1 - Interpretive Bulletin Relating to Participant Investment Education, 29 C.F.R. 2509.96-1 (June 11, 1996),
available at: http://www.dol.gov/ebsa/regs/fedreg/final/96_14093.pdf.

27

Interpretive Bulletin 96-1 has largely been incorporated in the final rule with
modifications, as discussed below.

2.6

Intersections with Other Governing Authorities

Many comments on the 2010 rulemaking emphasized the need to harmonize the
Department’s efforts with rulemaking activities under the Dodd-Frank Act, such as the SEC’s
standards of care for providing investment advice and the Commodity Futures Trading
Commission’s (CFTC) business conduct standards for swap dealers. In addition, commenters
questioned the adequacy of coordination with other agencies regarding IRA products and
services, specifically state insurance regulators. They argued that subjecting SEC-regulated
investment advisers and broker-dealers (BDs) and state-regulated insurance companies to a
special set of rules for IRAs could lead to additional costs and complexities for individuals who
may have several different types of accounts at the same financial institution.
In the course of developing the proposal, the final rule and the related prohibited
transaction exemptions, the Department has consulted with staff of the SEC, other securities,
banking and insurance regulators, including the U.S. Treasury Department’s Federal Office of
Insurance, the National Association of Insurance Commissioners (NAIC), and the Financial
Industry Regulatory Authority (FINRA), a self-regulatory organization of the broker-dealer
industry, to better understand whether the rule and exemptions would subject investment
advisers and broker-dealers who provide investment advice to requirements that create an undue
compliance burden or conflict with their obligations under other federal laws as well as related
rules and regulations. As part of this consultative process, SEC staff has provided technical
assistance and information with respect to the agencies’ separate regulatory provisions and
responsibilities, retail investors, and the marketplace for investment advice. Although the
Department and the SEC have different statutory responsibilities, the Department consulted with
the SEC on regulatory issues in which their interests and responsibilities overlap, particularly
where action by one agency may affect the parties regulated by the other agency. The technical
assistance that the SEC staff and others have provided has helped the Department in its efforts to
ensure that the rule strikes a balance between adding important additional protections for
individuals looking to build their savings and minimizing undue disruptions to the many
valuable services the financial services industry provides today.
In pursuing its consultations, the Department has not aimed to make the obligations of
fiduciary investment advisers under ERISA and the Code identical to the duties of advice
providers under the securities laws, nor could it. Even if each of the relevant agencies were to
adopt an express definition of “fiduciary” that was in all respects identical, the legal
consequences of the fiduciary designation likely would vary between agencies because of
differences in the specific duties imposed by the different federal laws at issue. ERISA and the
Code place special emphasis on the elimination or mitigation of conflicts of interest and
adherence to substantive standards of conduct, as reflected in the prohibited transaction rules and
ERISA’s stringent standards of fiduciary conduct. Additionally, ERISA and the Code apply to
all forms of assets that a plan or IRA may hold, including real estate and insurance products, not
just investment securities transactions or recommendations by financial institutions regulated by
the SEC.
The specific duties imposed on fiduciaries by ERISA and the Code stem from legislative
judgments on the best way to protect the public interest in tax-preferred benefit arrangements
that are critical to workers’ financial and physical health. ERISA fiduciaries are subject to
express statutory duties of prudence and loyalty rooted in the law of trusts. A fiduciary must
28

carry out its responsibilities with respect to the plan “solely in the interest of the participants and
beneficiaries” and “with the care, skill, prudence and diligence under the circumstances then
prevailing that a prudent man acting in a like capacity and familiar with such matters would use
in the conduct of an enterprise of a like character and with like aims.”44 These standards have
been characterized as the “highest known to the law.”45 The prudence standard is an objective
standard of care that requires advice fiduciaries to investigate and evaluate investments, make
recommendations, and exercise sound judgment in the same way that knowledgeable and
impartial professionals would. “[T]his is not a search for subjective good faith – a pure heart
and an empty head are not enough.”46 Whether or not the fiduciary is actually familiar with the
sound investment principals necessary to make particular recommendations, the fiduciary must
adhere to an objective professional standard. Additionally, when fiduciaries have conflicts of
interest, they must take special care to ensure the prudence and impartiality of their actions.47
The Internal Revenue Code does not directly impose these duties of prudence and loyalty
on IRA fiduciaries, but the Code and ERISA together require that plan and IRA fiduciaries
refrain from engaging in “prohibited transactions,” unless the transaction is covered by an
exemption in the statute, or granted by the Secretary of Labor. Of particular relevance here are
the prohibitions on self-dealing and receiving payments from third parties dealing with the plan
or IRA in connection with a transaction involving assets of the plan or IRA.48 In this manner,
ERISA and the Code focus on the elimination or mitigation of conflicts of interest. Thus, under
ERISA and the Code, fiduciary advisers are generally prohibited from making recommendations
with respect to which they have a financial conflict of interest unless the Department of Labor
first grants an exemption with conditions designed to protect the interests of plan participants and
IRA owners. This is true regardless of whether the fiduciary has disclosed his or her conflicts of
interest to their plan or IRA customer. The prohibited transaction provisions reflect profound
concern about the dangers that conflicts of interest pose for plan participants and IRA investors.
Rather than permit fiduciaries to “cure” conflicts of interest through disclosure, the statutory
default is simply to prohibit the fiduciary from engaging in the conflicted transaction in the first
place.
The federal securities laws, administered by the SEC, apply to transactions involving a
narrower category of investments, (i.e., transactions in securities), but involving a broader class
of investor, (i.e., all clients or customers, not just retirement or employee benefit investors). In
contrast to the trust law roots of ERISA and Code fiduciary duties, the specific duties imposed
on advisers by the SEC stem, in large part, from statutory antifraud provisions. As a result, and
as explained more fully in the next section, the duties imposed on broker-dealers and registered
investment advisers under the federal securities laws differ in significant respects.

44
45
46

47

48

29 U.S.C. §1104(a)(1).
See Chao v. Hall Holding Co., Inc., 285 F.3d 415, 426 (6th Cir.2002). .
Donovan v. Cunningham, 716 F.2d 1455, 1467 (5th Cir. 1983), cert. denied, 467 U.S. 1251 (1984); see also DiFelice v. U.S. Airways, Inc.,
497 F.3d 410, 418 (4th Cir. 2007) (“Good faith does not provide a defense to a claim of a breach of these fiduciary duties; ‘a pure heart and
an empty head are not enough’.”).
Donovan v. Bierwirth, 680 F.2d 263, 271 (2d Cir. 1982) (“the decisions [of the fiduciary] must be made with an eye single to the interests
of the participants and beneficiaries”); see also Bussian v. RJR Nabisco, Inc., 223 F.3d 286, 298 (5th Cir. 2000); and Leigh v. Engle, 727
F.2d 113 (7th Cir. 1984).
29 U.S.C. § 1106(b), 26 U.S.C. § 4975(C)(1)(E) and (F).

29

In pursuing this rulemaking, the Department has taken great care to honor ERISA and the
Code’s specific text and purposes. At the same time, however, the Department has endeavored
to understand the impact of the final rule and exemptions on firms subject to the securities laws
and other state and federal laws, and to take the effects into account by appropriately calibrating
the impact of the rule on those firms. The final regulation and exemptions reflect these efforts.
In the Department’s view, the final rule and exemptions neither undermine, nor contradict, the
provisions or purposes of securities or insurance laws. Instead, the Department has sought to
draft them to work in harmony with other state and federal laws. The Department has consulted - and will continue to consult -- with other state and federal agencies and has sought to ensure
that the various legal regimes are appropriately harmonized to the fullest extent possible.
The Department has also consulted with the Department of the Treasury, particularly on
the subject of IRAs. Although the Department has responsibility for issuing regulations and
PTEs under Section 4975 of the Code, which applies to IRAs, the IRS maintains general
responsibility for enforcing the tax laws. The IRS’ responsibilities extend to the imposition of
excise taxes on fiduciaries who participate in prohibited transactions.49 As a result, the
Department and the IRS share responsibility for combating self-dealing by fiduciary investment
advisers to tax-qualified plans and IRAs. Paragraph (f) of the final regulation, in particular,
recognizes this jurisdictional intersection.

2.6.1 Regulation of BDs and RIAs
Investment advice, and the institutions and individuals who render it, are subject to a
variety of other governing authorities. The accompanying diagram provides a simplified
illustration of how the Department’s authority overlaps with the SEC. Whether ERISA and the
IRC apply depends on whether the advised client is a plan, plan participant, or IRA investor.
Whether the Advisers Act and related SEC rules apply, and whether the Exchange Act and
related rules of FINRA apply, depends on the activities and business practices of the entities and
individuals rendering the advice. Moreover, other authorities govern the recommending and
selling of various bank and insurance products.
All of these authorities impose some standards of conduct (often including some limits
on, or requirements to, disclose certain conflicts of interest) and make available to consumers
mechanisms for remedying harms arising from violations of such standards. However, to the
Department’s knowledge, none include anti-conflict provisions approximating the prohibited
transactions provisions of ERISA and the IRC. Only the Advisers Act (as interpreted by courts)
establishes a fiduciary duty for RIAs roughly analogous to the fiduciary duties of care and
loyalty established by ERISA for investment advisers to plans and plan participants.50 It appears
that in enacting ERISA (and thereby establishing fiduciary duties under ERISA and prohibited
transactions provisions under both ERISA and the IRC) Congress established separate, and in
important respects, higher protections against conflicted advice for designated, tax-preferred
retirement savings in the form of plans and IRAs.

49
50

Reorganization Plan No. 4 of 1978, 5 U.S.C. § App. (2010).
“While broker-dealers are generally not subject to a fiduciary duty under the federal securities laws, courts have found broker-dealers to
have a fiduciary duty under certain circumstances.” SEC Staff Dodd-Frank Study, iv, 54. 106.

30

Figure 2-2 Intersection of Federal Laws
ERISA, IRC Intersect with Other Federal Laws Governing Financial
Anyone who gives “investment advice” for pay (direct or indirect) to a retirement
plan or IRA is a fiduciary under ERISA and/or IRC.
If you give
investment advice
for pay, you
generally must
register under the
Investment
Advisers Act of
1940 (with the
SEC or a state),
and assume
fiduciary status
toward your
clients.

Registered
Investment
Advisers

Broker
Dealers

IRA and ERISA
Plan Clients

But, a BD whose
advice is solely
incidental to its
brokerage
services and who
does not receive
special
compensation for
the advice is
exempt from the
IA Act, and is
subject to a
lesser “suitability”
standard under
the Securities and
Exchange Act of
1934 and FINRA
rules.

Others rendering financial advice
ERISA/IRC also intersect with rules governing bankers, insurance agents, etc.
Note: Diagram is schematic and does not convey size of jurisdictions or intersections.

A large proportion of the financial professionals that provide investment advice to plans,
plan participants, and IRA investors are either BDs or RIAs. The discussion below is not
intended to provide an exhaustive review of the federal securities laws relating to the provision
of investment advice or to fully explain BD or RIA regulation.

2.6.1.1

SEC Regulation of BDs and RIAs

BDs are regulated, among other things, under the Exchange Act, as well as rules and
regulations promulgated thereunder by the SEC. Section 10(b) of this Act makes it unlawful to
use or employ, in connection with the purchase or sale of any security or swap agreement, any
manipulative or deceptive devise or contrivance in contravention of such rules as the
Commission may prescribe.51 As a general matter, SEC Rule 10b-5 prohibits any person, from
directly or indirectly, (a) employing any device, scheme, or artifice to defraud; (b) making
untrue statements of material fact or omitting to state a material fact necessary in order to make
the statements made, in the light of the circumstances, not misleading; or (c) engaging in any act
or practice or course of business which operates or that would operate as a fraud or deceit upon
any person in connection with the purchase or sale of any security.

51

15 U.S.C. §78(b).

31

Scienter must be shown to establish a violation of some, but not all, of the general antifraud provisions of the federal securities laws, including Securities Act section 17(a)(1), 52
Exchange Act section 10(b) and Rule 10b-5 thereunder, and Advisers Act section 206(1).
Generally, investors do not have a private right of action, under either SEC or FINRA rules.
Rule 10b-5 is an exception; the Supreme Court has found an implied private right of action for
investors under Rule 10b-5, but such an action requires the investor to establish, inter alia, that
the unsuitable recommendation (or other misrepresentation or omission or fraudulent or
deceptive act or practice) was made with scienter.53 This is a much more difficult standard of
proof than required under ERISA and the IRC, which generally do not require proof that the
adviser acted with the intent to deceive or defraud the customer.
There are additional sources of liability for BDs beyond Rule 10b-5. For example, BDs
can face liability if they fail to supervise their agents and employees. Generally, BDs or their
associated persons may face liability for failure to supervise in the absence of policies and
procedures (and systems for implementing and monitoring compliance with such procedures)
that are reasonably designed to prevent and detect violations of the federal securities laws and
regulations, as well as applicable SRO rules.54
Like BDs, RIAs are subject to liability under Rule 10b-5. Unlike BDs, however, RIAs
are also regulated under the Advisers Act, which generally requires anyone who is paid to
provide investment advice to register with the SEC or a state and adhere to specified rules, and
some are regulated under State law. BDs that provide investment advice are exempted from the
term “investment adviser” as long as the provision of advice is “solely incidental” to their
business as a BD and they receive no special compensation. While the term “fiduciary” was not
used in the Advisers Act as originally enacted, the U.S. Supreme Court held that Section 206(2)
of the Advisers Act “reflects a congressional recognition ‘of the delicate fiduciary nature of an
investment advisory relationship’ as well as a congressional intent to eliminate, or at least to
expose, all conflicts of interest which might incline an investment adviser – consciously or
unconsciously – to render advice which was not disinterested.” The Court stated that the
purpose of the Federal securities laws “was to substitute a philosophy of full disclosure for the
philosophy of caveat emptor and thus to achieve a high standard of business ethics in the
securities industry.”55
According to SEC Staff, the Investment Advisers Act imposes on RIAs an—
“[A]ffirmative duty of utmost good faith, and full and fair disclosure of all material facts,
as well as an affirmative obligation to employ reasonable care to avoid misleading their
clients and prospective clients. Fundamental to the federal fiduciary standard are the
duties of loyalty and care. The duty of loyalty requires an adviser to serve the best
interests of its clients, which includes an obligation not to subordinate the clients’
interests to its own. An adviser’s duty of care requires it to make a reasonable

52
53
54

55

15 U.S.C. § 77q.
Superintendent of Ins. of N.Y. v. Bankers Life & Casualty Co., 404 U.S. 6, 13, n.9 (1971).
See SEC Staff Dodd-Frank Study at 74-76, 135 (citing Exchange Act Sections 15(b)(4)(E) and (b)(6)(A)). See also FINRA Rule 3110
(“Supervision,” which is based on NASD Rule 3010) and FINRA Rule 3120 (“Supervisory Control System,” which is based on NASD Rule
3012)].
Ibid., 186;. SEC v. Capital Gains Research Bureau, 375 U.S. 180, 186, 191 (1963).

32

investigation to determine that it is not basing its recommendations on materially
inaccurate or incomplete information.”56
The SEC may enforce Section 206(2) of the Advisers Act, but there is no private right of
action for damages beyond rescission of fees paid.57
As fiduciaries, RIAs also owe their clients a duty to provide only suitable investment
advice. This duty generally requires a RIA to determine that the investment advice it gives to a
client is suitable for the client, taking into consideration the client’s financial situation,
investment experience, and investment objectives.58 Under the Advisers Act, if RIAs do not
avoid a conflict of interest that could impact the impartiality of their advice, they must provide
full and fair disclosure of the conflict to their clients. The Advisers Act requires RIAs to fully
and fairly disclose conflicts, and as applicable, obtain client consent to such transactions to
occur. The SEC staff has taken the position that generally, they cannot use their clients’ assets
for their own benefit or the benefit of other clients, but a client can waive this protection and
consent to such activities. 59
One significant difference between the standards applicable to BDs and RIAs under the
federal securities laws involves principal trading. The Advisers Act prohibits RIAs from trading
securities with clients out of their own account unless the RIA provides advance written
disclosure to the client and obtains consent.60 The Exchange Act does not impose a similar
restriction on BDs.

2.6.1.2

FINRA Regulation of BDs

BDs are also generally required to become members of FINRA, a self-regulatory
organization (SRO) as defined in the Exchange Act. FINRA is a national securities association
that is registered with the SEC and operates under SEC oversight. FINRA is responsible for
regulating and examining securities firms that do business with the public, including with respect
to professional training, testing, and licensing of registered persons, as well as arbitration and
mediation of disputes.61 FINRA Rule 2111 establishes a “suitability” standard of conduct for
BDs, which requires them to “have a reasonable basis to believe that a recommended transaction
or investment strategy involving a security or securities is suitable for the customer, based on the
information obtained through the reasonable diligence of the [firm] or associated person to
ascertain the customer’s investment profile.”62

56
57
58
59

60
61

62

SEC Staff Dodd-Frank Study at 22 (internal quotation marks and citations omitted).
15 U.S.C. § 80b-9; Transamerica Mtg. Advrs, Inc. v. Lewis, 444 U.S. 111 (1979).
Investment Advisers Act Release No. 1406 (March 16, 1994).
Ibid. at iii. See “Information for Newly-Registered Investment Advisers,” prepared by the staff of the SEC’s Division of Investment
Management and Office of Compliance Inspections and Examinations; available at:
https://www.sec.gov/divisions/investment/advoverview.htm.
15 U.S.C. § 80b-6(3).
See Exchange Act Release, 34-56145 (July 26, 2007), 72 FR 42169, 42170 (Aug. 1, 2007) (SEC order approving the consolidation of
member firm regulatory functions of NASD and NYSE Regulation).
Financial Industry Regulatory Authority, FINRA Manual, Rule 2111, available at:
http://finra.complinet.com/en/display/display_main.thml?rbid=2403&element_id=9859. Under FINRA Rule 2111, a customer’s investment
profile would generally include the customer’s age, other investments, financial situation and needs, tax status, investment objectives,
investment experience, investment time horizon, liquidity needs, and risk tolerance. The rule also explicitly covers recommended
investment strategies involving securities, including recommendations to “hold” securities. The rule, moreover, identifies the three main
suitability obligations: reasonable-basis, customer-specific, and quantitative suitability. Activities such as excessive trading and churning
have been found to violate quantitative suitability obligations, but not the others (SEC Staff Dodd-Frank Study, 65).

33

In some circumstances FINRA may also impose a “heightened” suitability standard on
BDs. Rule 2330, for example, requires BDs to carefully determine suitability before selling a
variable annuity to a customer based on such factors as the customer’s age, the likelihood of the
customer being able to benefit from various features of the annuity, and potential account
surrender charges.63
The underlying customer agreement between a BD and a customer will generally require
that the customer seek redress for disputes through the FINRA arbitration process except in
cases when BDs are not FINRA members. FINRA can suspend or cancel the registration of a
brokerage firm or broker who does not comply with an arbitration award or settlement related to
an arbitration or mediation. Disputes that arise between a RIA and a customer are not required
to be heard through FINRA’s arbitration process. Many RIAs are also BDs and are FINRA
members, and as FINRA members, these RIAs may use the arbitration process for nearly all
claims related to their business practices. Accordingly, it is not uncommon for RIAs to use
arbitration as a means of resolving disputes with customers. The arbitrator can award monetary
relief for an investor, and arbitration may afford broader remedies than are available to investors
in federal court for careless investment advice.
In the IRA market, as discussed below, the new and amended final PTEs would be
conditioned on fiduciaries’ adherence to a “best interest” standard, which is an articulation of
ERISA’s fiduciary standards of prudence and loyalty as applicable to investment advice
fiduciaries - the standard is an objective standards of professional care and undivided loyalty,
which does not require proof of fraud, recklessness, or bad intent. In recent years, FINRA and
the SEC have suggested that the broker should have a reasonable basis to believe an investment
is in the customer’s best interest.64 However, absent fraud or bad intent, the Department believes
that the suitability standard often permit brokers to recommend investments that favor their own
financial interests or the financial interests of their firm in preference to better investments that
favor the customers’ interests, so long as the investment is suitable in terms of the customer’s
investment profile, which includes factors such as tax status, risk profile, and investment time
horizon.
FINRA is subject to oversight by the SEC. The SEC’s Office of Compliance Inspections
and Examinations (OCIE) conduct examinations of SROs, including FINRA, to ensure
compliance with federal securities laws and with standards of integrity, competence, and
financial soundness. The SEC may also discipline BDs and associated persons that fail to
comply with applicable requirements. The disciplinary procedures employed by the SROs are
subject to SEC oversight under the Exchange Act.65

63

64

65

See also FINRA Rule 2360 (requiring heightened account opening and suitability obligations regarding options) and FINRA Rule 2370
(requiring heightened account opening and suitability obligations regarding securities futures).
In its guidance on FINRA Rule 2111, FINRA explained that “[i]n interpreting FINRA’s suitability rule, numerous cases explicitly state that
‘a broker’s recommendations must be consistent with his customers’ best interests,’” and provides examples of conduct that would be
prohibited under this standard, including conduct that this exemption would not allow. (FINRA Regulatory Notice 12-25, p. 3 (2012)). The
guidance goes on to state that “[t]he suitability requirement that a broker make only those recommendations that are consistent with the
customer’s best interests prohibits a broker from placing his or her interests ahead of the customer’s interests.” This guidance, however, has
not been formalized as a rule and the cases involved egregious conduct and recommendations that, in the DOL’s view, would have been
clearly unsuitable under the language of FINRA rule 2111 without reference to a “best interest” standard. The scope of the guidance also
differs from the combined scope of the exemptions in this rulemaking. For example, fiduciary insurance providers who decide to accept
conflicted compensation will need to comply with the terms of an applicable exemption, but, may not be subject to FINRA's guidance.
See SEC Staff Dodd-Frank Study at A-6-7.

34

2.6.1.3

Additional Regulation

While BDs generally are not subject to a fiduciary duty under the federal securities laws,
courts have applied a fiduciary duty standard to certain actions by a BD. For example, BDs who
handles discretionary accounts for customers have often been held to a fiduciary duty standard.66
Such fiduciary duty may also arise in some circumstances under common law that varies by
state.67 However, BDs are required to deal fairly with their customers and to observe high
standards of commercial honor and just and equitable principles of trade.68 FINRA rules and
guidance also generally require a BD’s prices for securities and compensation for services to be
fair and reasonable based on all the relevant circumstances. Moreover, the courts and the
Commission have held that under the antifraud provisions of the federal securities laws, BDs
must charge prices reasonably related to the prevailing market price.69
As set forth above, the standards imposed on BDs and RIAs under the federal securities
laws for dealing with the recipients of their investment advice, while substantial, differ
significantly from what ERISA requires of fiduciary investment advisers to plans and to plan
participants and beneficiaries. The court-developed fiduciary standard for RIAs is largely based
on good faith and on making full and fair disclosures so as to avoid misleading the investor.
BDs do not typically have fiduciary duties. Instead, they look to the prohibitions on misleading
investors, the suitability standard, and to general principles of equity and fairness.70 Under
ERISA, subjective good faith is a necessary element of the ERISA fiduciary standard, but it is
not sufficient.71 ERISA’s trust law based duties encompass strict standards of prudence and
loyalty, together with more specific prohibitions on self-dealing and other conflicts. The
prudence standard set forth in both ERISA and the exemptions requires brokers and advisers to
adhere to an impartial expert standard of care in making investment recommendations that are in
the retirement investment customer’s best interests. Full disclosure is not a defense and the
fiduciary must put the customer first and act without regard to the broker’s or adviser’s own
interest.72 Remedies available under state securities laws would not generally afford the same
protection against conflicts of interest. As with Federal securities laws, they focus more on
issues of fraud, suitability, or careless execution of transactions. Additionally, the Best Interest
Contract and Principal Transactions Exemptions require an express acknowledgment of
fiduciary status; impose a requirement for stringent anti-conflict policies and procedures; and
prohibit quotas bonuses, incentives, performance measures, and similar measures that misalign
the interests of advisers and their customers by incentivizing advisers to favor the firms’ interests
at the customer’s expense.73

66

67
68
69
70

71
72
73

See Leib v. Merrill Lynch, Pierce, Fenner & Smith, Inc., 461 F.Supp. 951 (E.D. Mich. 1978). A discretionary account is one in which an
investor allows the broker-dealer to purchase and sell securities without having to give his or her consent for each transaction. In a
nondiscretionary account the broker-dealer buys and sells securities only as ordered by the investor.
SEC Staff Dodd-Frank Study at 51.
See, e.g., FINRA Rule 2010.
See FINRA Rule 2121; and discussion at Section 913 Study pages 66-67.
See, e.g., FINRA Rule 2010 (requiring FINRA members to "observe high standards of commercial honor and just and equitable principles
of trade”).
Donovan v. Cunningham, 716 F.2d 1455, 1467 (5th Cir. 1983).
However, BDs cannot disclaim any responsibilities under FINRA’s suitability rule (see Rule 2111.02).
These standards apply to all fiduciary investment advisers with respect to advice provided for a fee to retirement investors. In addition to
BDs and RIAs regulated by the SEC, the rule would also broadly cover insurance agents, and anyone else rendering investment advice to
retirement investors.

35

2.6.2 Background on Regulation of Insurance Products and
Producers
The insurance business primarily is regulated by the states. Insurance laws are enacted
by state legislators and governors and implemented and enforced by state regulators. The less
expansive role of the federal government as an insurance regulator was established in a Supreme
Court Case, Paul v. Virginia.74 The case involved several New York insurers that hired an agent
to sell insurance policies in Virginia. The insurers refused to obtain a licensing bond required by
Virginia law; therefore, Virginia denied the agent a license. The New York insurers filed a
lawsuit claiming that the Virginia law was unconstitutional, in part, because it violated the
Commerce Clause. The Supreme Court held against the insurers because it found the business
of insurance was not a transaction in interstate commerce. The Court in Paul v. Virginia stated
that “the Federal government can no more regulate the commerce of a State than a State can
regulate the commerce of the Federal government.”75 The decision effectively placed insurance
regulation solely under state purview.
The Supreme Court’s decision did not eliminate multistate insurance activity, and state
regulators sought ways to promote coordination among the states. In 1871, George W. Miller,
New York’s superintendent of insurance, invited the insurance commissioners from all 36 states
to participate in a meeting to discuss insurance regulation. Representatives from 19 states
attended the inaugural meeting of the association known today as the National Association of
Insurance Commissioners (NAIC).76 Today, the NAIC is a U.S. standard-setting and regulatory
support organization created and governed by the chief insurance regulators from the 50 states,
the District of Columbia and five U.S. territories.77 NAIC members, who are elected or
appointed state government officials, form the national system of state-based insurance
regulation for the conduct of insurance companies and agents in their respective state or
territory.78
As the insurance market grew and became a larger part of the economy, in a 1944
decision, the Supreme Court reversed its decision in Paul v. Virginia and held that the
Commerce Clause gives the Federal government the authority to regulate insurance transactions
across state lines.79 Shortly after the decision, Congress enacted the McCarran-Ferguson Act,80
which provides that state laws governing the business of insurance are not invalidated, impaired,
or superseded by any federal law unless the federal law specifically relates to the business of
insurance. Additionally, ERISA Section 514(b)(2)(A) specifically saves state insurance laws

74
75
76

77

78
79
80

75 U.S. 168 (1868).
Ibid, 183-184.
See FIO, “How to Modernize and Improve the System of Insurance Regulation in the United States,” 11 (2013); available at:
https://www.treasury.gov/initiatives/fio/reports; and
nottices/Documents/How%20to%20Modernize%20and%20Improve%20the%20System%20of%20Insurance%20Regulation%20in%20the
%20United%20States.pdf.
Through the NAIC, state insurance regulators establish standards and best practices, conduct peer review, and coordinate their regulatory
oversight. NAIC staff supports these efforts and represents the collective views of state regulators domestically and internationally. See
http://www.naic.org/index_about.htm#
See http://www.naic.org/index_about.htm#.
United States v. Southeastern Underwriters Association, 322 U.S. 533 (1944).
The McCarren-Ferguson Act, 15 U.S.C. §§ 1011-1015.

36

from ERISA preemption to the extent they do not conflict with ERISA’s provisions. See also
John Hancock Mutual Life Insurance Co. v. Harris Trust & Savings Bank, 510 U.S. 86 (1993).
Policymakers and industry leaders have debated the extent to which the federal
government should be involved in regulating insurance for many years. Proponents of federal
involvement maintain that the current state-based system does not impose the uniformity
necessary for the United States insurance market to operate efficiently. They argue that “state
regulation is often duplicative or inconsistent, that the multiplicity of jurisdictions makes state
regulators more prone to ‘capture,’ and that differences in standards between the states provide
opportunities for arbitrage, if not a race to the bottom.”81
There have been several federal legislative efforts to establish consistency and uniformity
in the regulation of insurance among the states. For example, state insurance agents who wish to
sell, solicit, or negotiate insurance in more than one state must complete separate applications,
pay multiple licensing fees, and meet different continuing education requirement established by
each state in which they seek to be licensed as a “Producer.”
On January 12, 2015, as part of the Terrorism Risk Insurance Program Reauthorization
Act of 2015, President Obama signed into law the National Association of Registered Agents
and Brokers Reform Act of 2015 (NARAB II). NARAB II reestablishes the National
Association of Registered Agents and Brokers (NARAB), which originally was authorized under
the Gramm-Leach-Bliley Act in 1999, but never established.82 The purpose of NARAB II is to
streamline the non-resident producer licensing process and preserve the states’ ability to protect
consumers and regulate producer conduct by establishing an independent non-profit corporation,
known as NARAB, controlled by its Board of Directors.83
In order to provide a federal role in the insurance market in response to the 2008
financial crisis, Title V of the Dodd-Frank Act, signed into law July 21, 2010 by President
Obama, established the Federal Insurance Office (FIO) within the U.S. Department of the
Treasury (Treasury).84 The statute provides FIO with the following authorities and
responsibilities:

81
82

83
84



Monitor all aspects of the insurance industry, including identifying issues or gaps in
the regulation of insurers that could contribute to a systemic crisis in the insurance
industry or the U.S. financial system;



Monitor the extent to which traditionally underserved communities and consumers,
minorities, and low- and moderate-income persons have access to affordable
insurance products regarding all lines of insurance, except health insurance;



Recommend to the Financial Stability Oversight Council (Council) that it designate
an insurer, including the affiliates of such insurer, as an entity subject to regulation as

See FIO, “How to Modernize and Improve the System of Insurance Regulation in the United States,” 11 (2013).
The Gramm-Leach-Bliley Act (GLBA) called for the establishment of the NARAB if a majority of the states and territories did not meet a
2002 deadline for reciprocity in producer licensing. The NARAB never was established, because in 2002 state regulators certified that 35
states and territories had satisfied the GLBA requirement, enough to constitute a majority and thereby avoid creation of the NARAB.
See http://www.naic.org/cipr_topics/topic_producer_licensing_narab_II.htm.
31 U.S.C. § 313.

37

a nonbank financial company supervised by the Board of Governors of the Federal
Reserve System (Federal Reserve);


Assist the Secretary of the Treasury (the Secretary) in administering the Terrorism
Insurance Program established in the Treasury under the Terrorism Risk Insurance
Act of 2002;



Coordinate federal efforts and develop federal policy on prudential aspects of
international insurance matters, including representing the United States, as
appropriate, in the International Association of Insurance Supervisors (IAIS) and
assisting the Secretary in negotiating covered agreements;



Determine whether state insurance measures are preempted by covered agreements;



Consult with the states (including state insurance regulators) regarding insurance
matters of national importance and prudential insurance matters of international
importance; and



Perform such other related duties and authorities as may be assigned to FIO by the
Secretary.

The Dodd-Frank Act required the FIO Director to “conduct a study and submit a report
to Congress on how to modernize and improve the system of insurance regulation in the United
States.”85 In a December 2013 report responding to this Congressional directive, FIO
acknowledged the limitations of state-based insurance regulation, such as market inefficiencies
created by the lack of regulatory uniformity and differences in regulatory standards among states
that provide opportunities for regulatory arbitrage, but did not recommend that the federal
government displace state regulation.86 The report concludes that the “debate is best reframed as
one in which the question is where federal involvement is warranted, not whether federal
regulation should completely displace state-based regulation.”87
Insurance agents may be either independent (who sell a variety of products for multiple
carriers), or may work exclusively for a single insurer under an employer-employee
relationship.88 In general, regardless of whether an agent is independent, insurance companies
retain responsibility with respect to the actions of agents who sell their products. If the
insurance agent is an independent agent, i.e., not an employee of the company, the insurance
agent is considered to be an agent of the carrier, and subject to the well-settled law of agency,
under which it is bound by the duties imposed upon its principals, the insurance companies.89
As the principal of independent agents, and as employer of its employees, insurance companies
are responsible for, and may be held liable for, acts and omissions of the agents.90
Unlike insurance agents, who are considered to be agents of the company, insurance
brokers act on behalf of the insured, helping the insured obtain an insurance or annuity contract.

85
86
87
88
89
90

31 U.S.C. § 313(p).
See FIO, “How to Modernize and Improve the System of Insurance Regulation in the United States,” 11 (2013).
Ibid., 65.
See, e.g., N.Y. Ins. Law § 2101(b); 215 Ill. Comp. Stat. Ann. 5/141.02(1); Cal. Ins. Code § 31.
See, e.g., Andrew J. Corse & Son, Inc. v. Harnett, 92 Misc. 2d 569m, 722-73, 400 N.Y.S.2d 1009, 1012 (Sup. Ct. 1977).
See, e.g., Loehr v. Great Republic Ins. Co., 226 Cal. App. 3d 727, 734, 276 Cal. Rptr. 667, 671-72 (Ct. App. 1990).

38

The divide between agent and broker can be somewhat ambiguous under state statutes91 and as
interpreted by courts.92 In general, when an insurer takes some action, or otherwise represents
that the broker has general authority to sell and market its products, the insured may infer that
the broker is the agent of the company.
Most states have adopted some form of the NAIC’s Suitability in Annuity Transactions
Model Regulation (Model Suitability Regulation) establishing suitability obligations with
respect to annuity transactions. The Model Suitability Regulation was adopted specifically to
establish a framework under which insurance companies, not just the agent or broker, are
“responsible for ensuring that the annuity transactions are suitable.93 The Model Suitability
Regulation requires insurance companies to develop a system of supervision reasonably
designed to achieve the insurers’ and their licensed producers’ (i.e., the individuals selling the
annuities) compliance with the Model Suitability Regulation and suitability obligations.94 This
supervisory system requires insurance companies to establish reasonable policies and procedures
to, inter alia, review each recommendation to ensure that it is suitable for the customer and to
detect recommendations that are not suitable.95
As of September 2015, 35 states and the District of Columbia have adopted some form of the
Model Suitability Regulation regarding suitability. For example, the California Code places
record-keeping and oversight duties on insurers, requiring them to “establish a supervision
system that is reasonably designed to achieve” compliance with the suitability rules by both the
insurer and the insurance producer.96 The insurer is also responsible for providing training and
establishing procedures to ensure that each recommendation is suitable. The insurance company
has oversight responsibility and can be held liable for actions or omissions of its agents and
employees.97 California also requires insurance companies to maintain records (including any
information used as the basis for a recommendation), and to make those records available for
five years after the insurance transaction is complete.98

2.6.3 Regulation of Annuity Products
An annuity is a contract between a customer and an insurance company that is designed
to meet retirement and other long-range goals. Generally, the customer makes a lump-sum
payment or series of payments. In return, the insurer agrees to make periodic payments
beginning immediately or at some future date (referred to as annuitization). If the payments are
made immediately, the annuity is referred to as an immediate annuity. If the payments are

91
92

93

94

95
96

97
98

See, e.g., Mass. Gen. Laws Ann. Ch. 175, § 162.
See Foisy v. Royal Maccabees Life Ins. Co., 356 F.3d 141, 150 (1st Cir. 2004) (finding the court must “look to the agent’s conduct in the
relevant transaction to determine the nature of the various relationships”); Bennion v. Allstate Ins. Co., 284 A.D.2d 924, 925, 727 N.Y.S.2d
222, 224 (2001).
NAIC, Suitability in Annuity Transactions Model Regulation, Executive Summary, available
at: http://www.naic.org/documents/committees_a_suitability_reg_guidance.pdf.
NAIC Model Suitability Regulation, § 6(F)(1) (2010): (“An insurer shall establish a supervision system that is reasonably designed to
achieve the insurer's and its insurance producers' compliance with this regulation including, but not limited to, the following: . . . (d) The
insurer shall maintain procedures for review of each recommendation prior to issuance of an annuity that are designed to ensure that there is
a reasonable basis to determine that a recommendation is suitable. . . .”)
NAIC Model Suitability Regulations, § 6(F)(1) (2010).
California defines an Insurance Producer as any person required to be licensed under California law to sell, solicit, or negotiate insurance,
including annuities. CA Ins. Code § 10509.913(e). The definition includes both independent agents and employees.
CA Ins. Code § 10509.916(a).
CA Ins. Code § 10509.917(a).

39

deferred until a future date, the annuity is referred to as a deferred annuity.99 With respect to
deferred annuities, annuitization is typically only one of the options offered to receive income
from the annuity. Owners also have the right to make partial withdrawals or surrender the
annuity subject to the terms of the contract.
Annuities typically offer tax-deferred growth of earnings and may include a death benefit
that will pay a designated beneficiary a specified minimum amount, such as the total purchase
payments. While tax is deferred on earnings growth, gains are taxed as ordinary income rates
and not capital gains rates, when withdrawals are taken from the annuity. If customers withdraw
their money early from an annuity, they may have to pay substantial surrender charges to the
insurance company, as well as tax penalties.100
There are generally three types of annuities — fixed, indexed, and variable. Each type of
annuity and the regulatory regimes to which each is subject are discussed below.
In a fixed annuity, the insurance company agrees to credit no less than a specified rate of
interest during the time that the account value is growing. The insurance company also agrees
that the periodic payments will be a specified amount per dollar in the account. These periodic
payments may last for a definite period, such as 20 years, or an indefinite period, such as the
later of the customer’s lifetime or the lifetime of his or her spouse.
For many years, fixed annuities dominated the annuities market. Over time, the annuities
market has become more complex. Some consumers perceived fixed-rate annuities as low-value
because their returns are based on interest rates, which often are significantly lower than stock
market returns, although the insurance guarantees can provide significant protection from market
downturns. To overcome this drawback, insurers introduced variable annuity products.
Generally, variable annuities are deferred annuity products for which the underlying assets are
held in separate accounts of the insurer that are segregated from the insurer’s operations, usually
with a variety of underlying investment options such as mutual funds, allowing for the
realization of market returns. However, customers make or lose money depending on the
performance of the chosen investment options and the contract value, and income payments are
variable and not guaranteed. Over time, variable annuities sales surpassed fixed annuities sales
and they now comprise the majority of sales in the annuity market. In 2012, total U.S.
individual annuity sales were $219 billion. Out of $219 billion, 67 percent ($147 billion) of total
sales were attributed to variable annuities.101
In a 1959 decision in SEC v. VALIC, the U.S. Supreme Court held that variable annuities
are subject to the Securities Act.102 The court determined that variable annuities do not fall
within the Securities Act exemption because “the variable annuity places all the investment risks
on the annuitant” by not guaranteeing any fixed return. Therefore, variable annuity products
generally must be registered with the SEC before they can be sold to the public. As variable
annuity contracts are securities under the Securities Act, a prospectus must be provided in
connection with the sale of variable annuities. This prospectus must follow requirements
specifically tailored to variable annuities, and the use of a prospectus subjects the insurance

99
100
101
102

SEC, “Annuities,” available at: http://www.sec.gov/answers/annuity.htm.
Ibid.
LIMRA (formerly “Life Insurance Market and Research Association,”), U.S. Individual Annuity Yearbook – 2014, (2015).
SEC v. VALIC, 359 U.S. 65 (1959).

40

company to various liability provisions under the Securities Act. Also, pursuant to the Exchange
Act, variable annuities must be distributed through registered broker-dealer firms and their
registered representatives who are members of FINRA. Variable annuities separate accounts
and underlying funds are subject to regulation under the Investment Company Act of 1940.
State insurance regulators also have jurisdiction over variable annuity products and sales.
Following several state-based market condition examinations that revealed unsupervised
sales of annuities that were not appropriate for the consumer’s profile, in 2010 NAIC adopted
the Suitability in Annuity Transactions Model Regulation (which created the national suitability
standard for annuity sales). As explained in Section 2.6.2, the Model Suitability Regulation sets
standards for “suitable annuity recommendations” and requires insurers to establish a system to
supervise annuity recommendations and provide both general and product-specific training to
producers. Specifically, the Model Suitability Regulation requires the insurance provider to
have “reasonable grounds for believing that the recommendation is suitable for the consumer on
the basis of the facts disclosed by the consumer as to his or her investments and other insurance
products and as to his or her financial situation and needs, including the consumer’s suitability
information.”103 Moreover, the insurance producer must have a “reasonable basis” for believing:
that the consumer has been “reasonably informed” of the annuity’s features; that the consumer
would benefit from certain of those features; that the annuity as a whole, the underlying
investment options selected by the consumer, and any riders or similar product enhancements are
suitable for the consumer; and in the case of an exchange or replacement of an annuity, the
exchange or replacement is suitable. The Model Suitability Regulation is similar in many
respects to FINRA Rule 2330, which imposes a wide range of requirements tailored specifically
to deferred variable annuity transactions, including suitability, principal review, supervision, and
training.
The annuity market has further evolved. Variable annuity sales have declined for three
consecutive years after they peaked in late 2011.104 In contrast, fixed-indexed annuity sales hit
record levels in 2014. It appears that the recent gains in the sales of fixed-indexed annuities
(alternatively referred to as “equity-indexed annuities”) have come at the expense of variable
annuities. While fixed-indexed annuities offer principal guarantees, the returns are based on
changes in an index such as the S&P 500 index or Dow Jones Industrial Average, during a set
period of time.105 Once the index’s returns are added to the annuity at the end of a set period,
those returns are locked in (fixed) and the changes in the index in the next period do not affect
them. Because fixed-indexed annuities’ returns are linked to indexes measuring overall market
performance, similar to variable annuities, fixed-indexed annuities can also provide investors
with higher returns than typical fixed-rate annuities during rising markets. The potential for
higher returns and fixed-indexed annuities’ principal guarantees have contributed to the recent
growth of fixed-indexed annuities.
In response to the growth in sales of equity-indexed annuities, the SEC sought to regulate
them when it proposed Rule 151A in 2008.106 The proposing release made clear that equity-

103

104
105
106

NAIC Suitability in Annuity Transactions Model Regulation, LH-275-1 (April 2010), [hereinafter NAIC Model Regulation] available at
http://www.naic.org/store/free/MDL-275.pdf.
LIMRA, “U.S. Individual Annuity Yearbook – 2014” (2015).
National Association of Insurance Commissioners (NAIC) (2013) “Buyer’s Guide for Deferred Annuities.
17 C.F.R. 230.151A.

41

indexed annuities did not fit within the safe harbor established in Rule 151 and were within the
SEC’s regulatory jurisdiction under the Securities Act. In fact, the SEC stated that the intent
behind Rule 151A was to register equity-indexed annuities as securities so purchasers receive
the benefits of federally-mandated disclosure, as well as anti-fraud, and sales practice
protections.
Shortly after the adoption of Rule 151A, in American Equity Inv. Life Ins. Co. v. SEC,107
the U.S. Court of Appeals for the D.C. Circuit struck down the SEC’s proposed rule. The court
found that the SEC reasonably concluded that the 1933 Securities Act’s exclusion for annuity
contracts did not include fixed indexed annuities, however, it vacated the rule because it found
that the SEC was arbitrary and capricious in failing to properly consider the effect of the rule on
efficiency, competition, and capital formation as required by Section 2(b) of the Securities Act.
As a result of the decision and the Harkin Amendment described below, equity-indexed
annuities remain subject to state regulation under current law.
After American Equity, in a provision referred to as the Harkin Amendment that was part
of the Dodd-Frank Act, Congress directed the SEC to treat an insurance policy or annuity
contract (collectively, any “insurance product” or “contract”) as exempt under the Securities Act
if (1) the value of the insurance product does not vary according to the performance of a separate
account; (2) the insurance product satisfies the NAIC Model Standard Nonforfeiture Law for
Life Insurance or the NAIC Model Standard Nonforfeiture Law for Individual Deferred
Annuities; and (3) the indexed annuity is issued in a state that has adopted the Model Suitability
Regulation or by an insurer that adopts and implements practices on a nationwide basis for the
sale of any insurance product that meets or exceeds the minimum requirements established by
Model Suitability Regulation.108
However, not all states have adopted the NAIC Model Suitability Regulation discussed
above and variation is found throughout many of the states. As mentioned above, according to
the Annual Report on the Insurance Industry by FIO published in September 2015, 35 states plus
the District of Columbia have adopted some version of the Suitability Model, but not
uniformly.109 According to the Annual Report, this is particularly concerning for complex
products, such as fixed-indexed annuities, because a uniform standard does not govern their sale.
Moreover, the SEC’s role with respect to regulating indexed annuities issued in the states after
American Equity is unclear, further complicating the regulatory landscape.110
In its Modernization Report and most recent Annual Reports, FIO urges that states adopt
the Model Suitability Regulation so that prospective annuity owners nationwide receive uniform
protection. In its September 2015 Annual Report, FIO states that “[a]s unprecedented numbers
of seniors reach retirement age with increased longevity, and as life insurers continue to
introduce more complex products tailored to consumer demand, the absence of national annuity
suitability standards is increasingly problematic.”111

107
108
109
110

111

613 F.3d.166 (D.C. Circuit 2010).
See § 989j, 15 U.S.C. § 77c(a)(8).
Annual Report on the Insurance Industry, FIO, U.S. Department of the Treasury, (Sept. 2015).
For example, certain indexed annuities are now being offered that include the potential for investment losses due to adverse market
performance and fall outside the Harkin Amendment. Generally, these types of indexed annuities are registered with the SEC.
Annual Report on the Insurance Industry, FIO, U.S. Department of the Treasury, (Sept. 2015), 54.

42

In fact, FINRA's Notice-to-Members 05-50 from August of 2005 addresses the
responsibility of firms to supervise the sale by their associated persons of equity-indexed
annuities that are not registered under the Federal securities laws and was intended to push firms
to adopt procedures to oversee that business.112 Specifically, FINRA was concerned with the
sales materials associated with unregistered equity-indexed annuities because they often did not
fully and accurately describe the products and the sales material could confuse or mislead
investors.113

2.6.4 SEC Staff Dodd-Frank Study
The Dodd-Frank Act required the SEC to conduct a study to evaluate the effectiveness of
existing legal or regulatory standards of care for providing investment advice to retail customers,
and whether there are gaps, shortcomings, or overlaps in the protection afforded retail customers
that should be addressed by rule or statute. In January 2011, the SEC staff published its study,
which included recommendations to the Commission.114
According to the study, the SEC staff’s recommendations to the Commission are
intended to make consistent the standards of conduct applying when retail customers receive
personalized investment advice about securities from BDs or RIAs. The SEC staff
recommended establishing a uniform fiduciary standard for RIAs and BDs when providing
investment advice about securities to retail customers that used the current RIA standard. The
recommendations also included suggestions to harmonize the BD and RIA regulatory regimes,
with a view toward enhancing their effectiveness in the retail marketplace.
On March 7, 2013, the SEC formally requested data and other information from the
public and interested parties about the benefits and costs of the current standards of conduct for
BDs and RIAs when providing advice to retail customers, as well as alternative approaches to
the standards of conduct.115 The SEC’s 2015 Fall Semi-annual Regulatory Agenda for 2016
included a new rule for the uniform fiduciary standard of conduct for broker-dealers and
investment advisers when providing personalized investment advice about securities to retail
customers.116 Any rule would have to be finalized after a notice and comment period before
being brought to a vote by the Commission.

2.6.5 Relevant Dodd-Frank Act Provisions
Section 911 of the Dodd-Frank Act established a new Investor Advisory Committee
(IAC) to advise the SEC on regulatory priorities, the regulation of securities products, trading
strategies, fee structures, the effectiveness of disclosures, and on initiatives to protect investor
interests and to promote investor confidence and the integrity of the securities marketplace. The

112

113
114

115

116

National Association of Security Dealers, Equity--Indexed Annuities, Notice to Members 05-50 (2005), available at:
https://www.finra.org/sites/default/files/NoticeDocument/p014821.pdf.
See Lazaro and Edwards (2015).
SEC Staff Dodd-Frank Study, (Jan. 2011). The study was not intended to provide an exhaustive discussion of federal securities laws relating
to the provision of investment advice or to fully explain investment adviser and BD regulation. Instead, as required by Congress by section
913 of the Dodd-Frank Act, the study evaluated the effectiveness of the existing standard of care and whether there were any regulatory
gaps, shortcomings, or overlaps, for example.
SEC Release No. 69103, “Duties of Brokers, Dealers, and Investment Advisers” (March 7, 2013); available at:
http://www.sec.gov/rules/other/2013/34-69013.pdf.
SEC Agency Rule List - Fall 2015.

43

Dodd-Frank Act authorizes the IAC to submit findings and recommendations for review and
consideration by the SEC.
In November 2013, the IAC recommended a framework for a uniform fiduciary duty
governing BDs and RIAs under the securities laws.117 The IAC’s favored approach was for the
SEC to use its rulemaking authority under the Advisers Act to propose rules that narrowed the
Broker-Dealer Exclusion from the Advisers Act, while providing a safe harbor for brokers who
did not engage in broader investment advisory services or hold themselves out as providing such
based either on the titles they used or the manner in which they marketed their services.
The IAC stated that one benefit of this approach is that it would provide a firm assurance
that the fiduciary standard for investment advice by BDs and RIAs would be the same and would
be no weaker than the existing standard. It also would “ensure that the existing legal precedent,
staff interpretations, and no-action positions developed under the Advisers Act and
accompanying rules would also apply to investment advice by brokers.”118 A BD that wished to
take advantage of the safe harbor could do so by limiting itself to transaction-specific
recommendations, avoiding holding itself out as an adviser or as providing advisory services,
and making an affirmative disclosure that the BD is acting solely as a salesperson and not as an
objective adviser.
The IAC also made an alternative recommendation for rulemaking pursuant to the
Exchange Act, as amended by Section 913(g) of the Dodd-Frank Act, to incorporate an
enforceable principles-based obligation to act in the best interest of the customer. The IAC
acknowledged that Section 913 of the Dodd-Frank Act posed some “significant implementation
challenges.”119 The IAC stated that the Dodd-Frank Act includes provisions specifying that
certain BD business practices — such as earning commissions, selling proprietary products, and
selling from a limited menu of products — should not automatically be deemed to constitute a
violation of the fiduciary standard. It intentionally avoided applying provisions of the Advisers
Act with regard to principal trades to brokers, but without specifying how principal trades by
brokers should be regulated under a fiduciary standard. Moreover, it specified that brokers
would not necessarily have an ongoing duty of care after the advice is rendered. The IAC
concluded that depending on how certain of these provisions are interpreted and enforced —
particularly those regarding selling from a limited menu of products and the ongoing duty of
care — such an approach could result in a significant weakening of the existing Advisers Act.120
Nonetheless, should the SEC choose to conduct rulemaking under the Exchange Act, the
IAC supported a three-pronged approach:


117

118
119
120

To ensure the standard is no weaker than the existing Advisers Act standard, any
fiduciary rule adopted must incorporate an enforceable, principles-based obligation to
act in the best interests of the customer.

SEC, “Recommendation of the Investor Advisory Committee: Broker-Dealer Fiduciary Duty” (Nov. 2013); available at:
http://www.sec.gov/spotlight/investor-advisory-committee-2012/fiduciary-duty-recommendation-2013.pdf.
Ibid., 5.
Ibid., 7.
Ibid.

44



To ensure the continued availability of transaction-based recommendations, any
standard adopted should be sufficiently flexible to permit the existence of certain
sales-related conflicts of interest, subject to a requirement that any such conflicts be
fully disclosed and appropriately managed.



While some forms of transaction-based payments would be acceptable under a
fiduciary standard, the SEC should fulfill the Dodd-Frank Act’s mandate to “examine
and, where appropriate, promulgate rules prohibiting or restricting certain sales
practices, conflicts of interest, and compensation schemes for brokers, dealers, and
investment advisers that the [SEC] deems contrary to the public interest and the
protection of investors.”

The IAC also recommended that the SEC adopt a uniform, plain English disclosure
document to be provided to customers and potential customers of BDs and RIAs at the start of
an engagement, and periodically thereafter, that covers basic information about the nature of
services offered, fees and compensation, conflicts of interest, and disciplinary record. The IAC
explained that disclosure alone is not sufficient to address the harm that can result when BDs act
on conflicts of interest, but stated that it believes improved disclosure should be included as part
any fiduciary rulemaking. The IAC suggested that the Form ADV provides a reasonable starting
point for designing a new disclosure document, and encouraged the SEC to work with disclosure
design experts to ensure that any document it develops is effective in conveying the relevant
information to investors in a way that enables them to act on the information.
A BD’s or RIA’s status under the federal securities laws is not directly relevant to the
determination of fiduciary status under ERISA and the IRC. Rather, fiduciary status under
ERISA and the IRC is determined by the functions that BDs and RIAs perform with respect to
plan and IRA investors. A BD generally is not a fiduciary under federal securities laws.
However, if the BD engages in activity defined in Department of Labor regulations, the BD is an
investment advice fiduciary under ERISA. RIAs generally are fiduciaries under federal
securities laws, but they are investment advice fiduciaries under ERISA or the IRC only if they
advise plan participants or IRA investors as defined under Department of Labor regulations.
The intersections between ERISA and the IRC on the one hand and federal securities
laws on the other follow from terms in the statutes. Because the statutes differ in material ways,
and reflect a deliberate congressional choice to apply different standards, agency rules, and other
guidance, DOL and SEC rules will necessarily vary in substance, even as the agencies work to
ensure that, to the extent possible, the various legal regimes are appropriately harmonized.
Many RIAs and some BDs that provide services to plan officials currently understand that they
are subject to both ERISA and relevant SEC rules and structure their practices to comply with
both, often taking advantage of one or more available PTEs.

2.6.6 FINRA Conflicts of Interest Report
FINRA began a conflicts of interest initiative in 2012 to review BDs’ approaches to
conflicts management and to identify effective practices. FINRA used firms’ responses to a
FINRA conflicts of interest letter, in-person meetings, and a follow-up compensation

45

questionnaire to develop observations detailed in an October 2013 report.121 One area of focus
in the FINRA report is firms’ approaches to identifying and managing conflicts with respect to
compensating those acting as brokers to private clients. In response to FINRA’s letter, firms
summarized the most significant conflicts they face in their businesses. The firms identified
potential conflicts of interest related to their retail and private wealth business that relate mostly
to the pursuit of revenue by the firm or its registered representatives at a client’s expense
including the following:


Firms offering or promoting particular products or product providers because of their
revenue or profit potential to the firm, such as through revenue sharing;



Registered representatives offering or giving preference to certain products or
services because of their income potential for the firm;



Registered representatives recommending transactions in order to generate revenue
for the firm without due regard to suitability;



Firms offering incentive programs to employees; and



Firms or employees giving preference to proprietary products.

The report highlights the following as examples of effective practices used by firms to
mitigate instances where the compensation structure may potentially affect the behavior of
registered representatives:

121

122
123



Avoiding creating compensation thresholds that enable a registered representative to
increase his or her compensation disproportionately through an incremental increase
in sales;



Monitoring activity of representatives approaching compensation thresholds such as
higher payout percentages, back-end bonuses, or participation in a recognition club,
such as a President’s Club;



Refraining from providing higher compensation or other rewards for the sale of
proprietary products or products for which the firm has entered into revenue sharing
arrangements;



Monitoring the suitability of recommendations around key liquidity events in the
investor’s lifecycle where the recommendation is particularly significant (e.g., when
an investor rolls over his or her pension or 401(k) account); and



Developing metrics for good and bad behavior (red flag processes) and using
clawbacks122 to adjust compensation for employees who do not properly manage
conflicts of interest.123

FINRA, “Report on Conflicts of Interest” (Oct. 2013); available at:
http://www.finra.org/web/groups/industry/@ip/@reg/@guide/documents/industry/p359971.pdf.
A “clawback” generally refers to a contractual clause allowing a firm to revoke some or all of an employee’s deferred compensation.
The report also suggested that firms could use “neutral compensation grids.” Under the terms of the Department’s Best Interest Contract
Exemption and Principal Transactions Exemption, in constructing such grids, however, the firm would need to be careful to ensure that it
was not simply transmitting firm-level conflicts to the adviser by tying the adviser’s compensation directly to the profitability of a
recommendation to the firm. The Best Interest Contract Exemption and the Principal Transactions Exemption do not permit compensation
practices that a reasonable person would view as encouraging persons to violate the best interest standard by, for example, favoring the
firm’s financial interest at the customers’ expense.

46

The report states that conflicts also may arise in recommending the type of account a
client should open with a firm. For example, firms that are dually registered as a BD and a RIA
should consider whether a commission-based or fee-based account is more appropriate for a
customer. The report notes that depending on the circumstances, fee-based accounts may be
preferable for customers with a fair amount of trading activity or the desire for active account
monitoring and ongoing advice, while commission-based accounts may be more cost-effective
or appropriate for customers with low trading activity. The report recommends that firms
examine their procedures to ensure that they are reasonably designed to monitor inappropriate
behavior.

2.7

2010 Proposal

Since 1978, the Department has been solely responsible for issuing rules and other
interpretations of the prohibited transactions provisions of both ERISA and the IRC.124 In
October 2010, the Department proposed amendments to the 1975 regulation that would have
broadened the definition of fiduciary investment advice under both ERISA and the IRC, making
more investment advisory activities fiduciary in nature (“2010 Proposal”).125 Under the 2010
Proposal, advice could be fiduciary if it consisted of a single recommendation given once
(relaxing the 1975 regulation’s requirement that the advice be given on a regular basis). Advice
would be fiduciary if it was agreed that the advice may be considered in investment decisions, or
if the adviser was otherwise a fiduciary to the plan or IRA, represented that he or she was a
fiduciary, or was a RIA (relaxing the 1975 regulation’s requirement of a mutual agreement that
the advice would
serve as a primary basis for investment decisions). The 2010 Proposal
also generally would have treated the valuation of plan or IRA assets (including employer
securities held by ESOPs) as fiduciary advice, superseding AO 76-65A. Recommendations
made as part of certain pure sales activities, however, would not have constituted fiduciary
investment advice under the 2010 Proposal.
The 2010 Proposal was motivated by the Department’s concern that conflicts of interest
often compromised advice rendered to plan officials, participants, and IRA investors. In
addition, the Department’s experience enforcing the fiduciary provisions of ERISA had shown
that abuses by plan advisers were numerous and difficult to remedy. By broadening the
fiduciary definition, the 2010 Proposal would have extended the ERISA and IRC prohibited
transactions provisions to cover more advice rendered to plan officials, plan participants, and
IRA investors, thereby limiting the self-dealing that can compromise that advice. It also would
have extended ERISA’s statutory fiduciary duties and liability for any breaches of such duties to
more advice rendered to plan investors, thereby raising the standards of conduct applicable to the
professionals rendering advice and holding those professionals accountable for adhering to the
standards. In issuing the 2010 Proposal, the Department presented a Regulatory Impact Analysis
pursuant to Executive Order 12866, which concluded that the 2010 Proposal’s benefits would
justify its cost.

124
125

Reorganization Plan No. 4 of 1978, 5 U.S.C. § App. (2010).
EBSA Proposed Rule, “Definition of the Term ‘Fiduciary’,” 75 Fed. Reg. 65263 (Oct. 22, 2010); available at:
http://webapps.dol.gov/FederalRegister/PdfDisplay.aspx?DocId=24328. For an analysis of the proposed rule, see Munnell, Webb, and
Vitagliano (Working Paper 2013-4).

47

The 2010 Proposal elicited extensive comments and prompted vigorous debate. The
Department heard from a very wide range of stakeholders in a variety of forums. Some
feedback was positive, but financial services industry feedback was largely negative. Some of
the negative feedback was specific, accepting at least some of the Department’s premises and
aims, but stating that particular proposed provisions were poorly calibrated or targeted to achieve
the Department’s stated aims. Some stakeholders requested the Department issue PTEs to
permit certain existing business practices that would involve prohibited fiduciary self-dealing
under the 2010 Proposal. Some of the negative feedback was broader. For example, some
comments rejected the premise that conflicts of interest sometimes compromise advice,
maintaining that the Department had not provided adequate evidence of the harm resulting from
conflicted advice. Some commenters argued that the Department should take no regulatory
action in connection with investment advice until the SEC completes its consideration of its
staff’s recommendations on a uniform fiduciary standard for BDs and RIAs under the securities
laws. Some argued that the fiduciary duty of loyalty might conflict with an appraiser’s duty to
impartially assess value, and that treating appraisals as fiduciary advice would make valuations
more costly for ESOPs and other plans. Some commenters complained that the Department’s
Regulatory Impact Analysis was inadequate, and that it neglected to consider certain potential
major, negative impacts on the retail IRA market. Two formal written comments provided
alternative analysis predicting that the 2010 Proposal would have highly negative impacts on the
IRA market and small investors.126
To obtain additional feedback on the 2010 Proposal and associated policy questions, the
Department held two full days of open public hearings on March 1 and 2, 2011, taking testimony
from 38 witnesses and receiving more than 60 post-hearing written comments. The Department
also met individually with many stakeholder groups that sought additional opportunities to
explain their views. Along the way the Department heard from various members of Congress,
representatives of many segments of the financial services industry, as well as plan sponsors,
advocates for small investors, plan participants, service providers, and academics who study the
roles of financial intermediaries and the effects of conflicts of interest between consumers and
their expert advisers.
In response to this feedback the Department announced in September 2011 that it
intended to withdraw the 2010 Proposal and develop and issue a revised proposal in due course.
The Department also expressed its intention to provide a more thorough and robust regulatory
impact analysis with the re-proposal than was provided with the 2010 Proposal.

2.8

2015 Proposal

On April 20, 2015, the Department published in the Federal Register a Notice
withdrawing the 2010 Proposal, and issuing a new proposed amendment to the 1975 regulation
(2015 Proposal).127 On the same date, the Department published proposed new and amended

126

127

See Oliver Wyman report, “Assessment of the Impact of the Department of Labor’s Proposed `Fiduciary’ Definition Rule on IRA
Consumers” (Apr. 12, 2011) (data for Department use submitted Jan. 2012); available at:
http://www.dol.gov/ebsa/pdf/WymanStudy041211.pdf; and Daniel R. Fischel and Todd D. Kendall, “Comment To The Department Of
Labor On A Proposed Rule Regarding Fiduciary Status Under ERISA” (Apr. 12, 2011); available at: http://www.dol.gov/ebsa/pdf/1210AB32-PH056.pdf. A detailed response to these comments is provided in Chapter 4.
80 Fed. Reg. 21927.

48

exemptions from ERISA’s and the Code’s prohibited transactions rules designed to allow certain
BDs, insurance agents, and others that act as investment advice fiduciaries to nevertheless
continue to receive common forms of compensation that would otherwise be prohibited, subject
to appropriate safeguards. The 2015 Proposal made many revisions to the 2010 Proposal,
although it also retained aspects of that proposal’s essential framework. The proposal set forth
the following types of advice, which, when provided in exchange for a fee or other
compensation, whether directly or indirectly, would be “investment advice” unless one of the
carve-outs in the proposal applied. The listed types of advice were—(i) A recommendation as to
the advisability of acquiring, holding, disposing of, or exchanging securities or other property,
including a recommendation to take a distribution of benefits or a recommendation as to the
investment of securities or other property to be rolled over or otherwise distributed from the plan
or IRA; (ii) A recommendation as to the management of securities or other property, including
recommendations as to the management of securities or other property to be rolled over or
otherwise distributed from the plan or IRA; (iii) An appraisal, fairness opinion, or similar
statement whether verbal or written concerning the value of securities or other property if
provided in connection with a specific transaction or transactions involving the acquisition,
disposition, or exchange of such securities or other property by the plan or IRA; or (iv) A
recommendation of a person who is also going to receive a fee or other compensation to provide
any of the types of advice described in paragraphs (i) through (iii) above.
The 2015 Proposal provided that unless a carve-out applied, a category of advice listed in
the proposal would constitute “investment advice” if the person providing the advice, either
directly or indirectly (e.g., through or together with any affiliate)—(i) Represented or
acknowledged that it is acting as a fiduciary within the meaning of the Act or Code with respect
to the advice described in paragraph (a)(1); or (ii) Rendered the advice pursuant to a written or
verbal agreement, arrangement, or understanding that the advice is individualized to, or that such
advice is specifically directed to, the advice recipient for consideration in making investment or
management decisions with respect to securities or other property of the plan or IRA.
The proposal included several carve-outs for persons who do not represent that they are
acting as ERISA fiduciaries, some of which were included in some form in the 2010 Proposal
but many of which were not. Subject to specified conditions, these carve-outs covered ––
(1) Statements or recommendations made to a “large plan investor with financial
expertise” by a counterparty acting in an arm’s length transaction;
(2) Offers or recommendations to plan fiduciaries of ERISA plans to enter into a swap
or security-based swap that is regulated under the Securities Exchange Act or the
Commodity Exchange Act;
(3) Statements or recommendations provided to a plan fiduciary of an ERISA plan by
an employee of the plan sponsor if the employee receives no fee beyond his or her
normal compensation;
(4) Marketing or making available a platform of investment alternatives to be selected
by a plan fiduciary for an ERISA participant-directed individual account plan;
(5) The identification of investment alternatives that meet objective criteria specified
by a plan fiduciary of an ERISA plan or the provision of objective financial data to
such fiduciary;
(6) The provision of an appraisal, fairness opinion or a statement of value to an ESOP
regarding employer securities, to a collective investment vehicle holding plan
assets, or to a plan for meeting reporting and disclosure requirements; and
49

(7) Information and materials that constitute “investment education” or “retirement
education.”
The 2015 Proposal applied the same definition of “investment advice” to the definition of
“fiduciary” in IRC Section 4975(e)(3) and thus applied to investment advice rendered to IRAs.
“Plan” was defined in the proposal to mean any employee benefit plan described in ERISA
Section 3(3) and any plan described in IRC Section 4975(e)(1)(A).128 Under the 2015 Proposal,
a recommendation was defined as a communication that, based on its content, context, and
presentation, would reasonably be viewed as a suggestion that the advice recipient engage in or
refrain from taking a particular course of action. The Department specifically requested
comments on whether the Department should adopt the standards that FINRA uses to define
“recommendation” for purposes of the suitability rules applicable to BDs.
Many of the differences between the 2015 Proposal and the 2010 Proposal reflect the
input of commenters on the 2010 Proposal as part of the public notice and comment process.
For example, some commenters argued that the 2010 Proposal swept too broadly by making
investment recommendations fiduciary in nature simply because the adviser was a plan fiduciary
for purposes unconnected with the advice or an investment adviser under the Advisers Act. In
their view, such status-based criteria were in tension with the Act’s functional approach to
fiduciary status and would have resulted in unwarranted and unintended compliance issues and
costs. Other commenters objected to the lack of a requirement for these status-based categories
that the advice be individualized to the needs of the advice recipient. Under the 2015 Proposal,
an adviser’s status as an investment adviser under the Advisers Act or as an ERISA fiduciary for
reasons unrelated to advice were not explicit factors in the definition. In addition, the proposal
provided that unless the adviser represented that he or she is a fiduciary with respect to advice,
the advice must be provided pursuant to a written or verbal agreement, arrangement, or
understanding that the advice is individualized to, or that such advice is specifically directed to,
the recipient for consideration in making investment or management decisions with respect to
securities or other property of the plan or IRA.
Furthermore, under the proposal the carve-outs that treat certain conduct as non-fiduciary
in nature were modified, clarified, and expanded in response to comments on the 2010 Proposal.
For example, the carve-out for certain valuations from the definition of fiduciary investment
advice was modified and expanded. Under the 2015 Proposal, appraisals and valuations for
compliance with certain reporting and disclosure requirements were not treated as fiduciary
investment advice. The proposal additionally provided a carve-out from fiduciary treatment for
appraisal and fairness opinions for ESOPs regarding employer securities. Although, the
Department remained concerned about valuation advice concerning an ESOP’s purchase of
employer stock and about a plan’s reliance on that advice, the Department concluded, at the
time, that the concerns regarding valuations of closely held employer stock in ESOP transactions
raised unique issues that were more appropriately addressed in a separate regulatory initiative.
Additionally, the carve-out for valuations conducted for reporting and disclosure purposes was
expanded to include reporting and disclosure obligations outside of ERISA and the Code, and
was applicable to both ERISA plans and IRAs.

128

For ease of reference the proposal defined the term “IRA” inclusively to mean any account described in IRC § 4975(e)(1)(B) through (F),
such as an individual retirement account described under Code § 408(a) and a health savings account described in Code § 223(d).

50

The Department took significant steps to give interested persons an opportunity to
comment on the 2015 Proposal and proposed related exemptions. The proposal and proposed
related exemptions initially provided for 75-day comment periods, ending on July 6, 2015, but
the Department extended the comment periods to July 21, 2015. The Department also held a
public hearing in Washington, DC from August 10 through 13, 2015, at which over 75 speakers
testified. The transcript of the hearing was made available on September 8, 2015, and the
Department provided additional opportunity for interested persons to submit comments on the
proposal and proposed related exemptions or transcript until September 24, 2015. The
Department received a total of over 3,000 comment letters on the 2015 Proposal. There were
also over 300,000 submissions made as part of 30 separate petitions submitted on the proposal.
These comments and petitions came from consumer groups, plan sponsors, financial services
companies, academics, elected government officials, trade and industry associations, and others,
both in support of and in opposition to the proposed rule and proposed related exemptions.

2.8.1 Proposed PTEs
The 2015 Proposal included several proposed new and amended class PTEs, which
together would permit fiduciary investment advisers to plan and IRA investors to engage in
certain specified types of transactions that would otherwise be prohibited subject to a number of
protective conditions.
As discussed above, under the 2015 Proposal, a person would be an investment advice
fiduciary if he or she provided a recommendation to a plan, plan fiduciary, plan participant or
beneficiary or IRA investor regarding the advisability of acquiring, holding, disposing or
exchanging securities or other property pursuant to a written or verbal agreement, arrangement
or understanding that the advice was specifically directed to the advice recipient for
consideration in making investment decisions with respect to securities or other property. Once
a person was an investment advice fiduciary, the person was prohibited by the PT provisions
from engaging in certain kinds of transactions involving the plan or IRA, including transactions
in which the fiduciary affected or increased his or her own compensation or that of a person in
which such fiduciary has an interest which may affect the exercise of the fiduciary’s best
judgment. Receipt by a fiduciary of certain common types of fees and compensation, such as
brokerage or insurance commissions, in connection with investment transactions entered into by
the plan or IRA, fell within the prohibition.
The Department recognized the concerns expressed in the comments from
representatives of BDs and other IRA advisers regarding the potential disruption to current fee
arrangements that would arise by applying the IRC prohibited transactions rules more broadly in
the retail IRA market. Therefore, simultaneous with the publication of these proposed
regulations, the Department proposed several new and amended PTEs that would allow certain
currently common fee practices to persist subject to conditions provided in the exemption that
protect plans, plan participants, and IRA investors from advisers’ conflicts of interest.

2.8.1.1

Proposed Best Interest Contract Exemption

Many comments on the 2010 Proposal requested relief for the receipt by investment
advice fiduciaries of a variety of fees and compensation resulting from agency transactions
involving plans and IRAs. These transactions involve, according to the commenters,
investments in mutual fund shares, collective trusts, insurance products, commodities, futures
and private funds. The Department was urged to propose an exemption that would permit
investment advice fiduciaries to continue to recommend investments historically used by plans
and IRA investors.
51

In response to these comments, the Department proposed the Best Interest Contract
Exemption. As proposed, the exemption would have permitted investment advice fiduciaries
and certain related entities to receive compensation for services provided in connection with the
purchase, sale, or holding by plan participants, beneficiaries, IRAs and small employee benefit
plans of certain assets as a result of the investment advice. The proposed exemption permitted
fiduciary advisers and their firms to receive fees such as commissions, 12b-1 fees, and revenue
sharing in connection with investment transactions by the plan participants, beneficiaries, IRAs
and small plans, thus preserving many current fee practices.
In order to ensure compliance with its broad protective standards and purposes, the
exemption required fiduciary advisers and their firms to enter into a written contract with the
plan/IRA investor. The existence of enforceable rights and remedies were intended to give firms
and their advisers a powerful incentive to comply with the exemption’s standards, implement
policies and procedures that are more than window-dressing, and carefully police conflicts of
interest to ensure that the conflicts of interest do not taint the advice. The contract could not
contain exculpatory provisions disclaiming or otherwise limiting liability of the fiduciary adviser
and firm for violation of the contract’s terms. Some of the main conditions of the exemption
provided that:


The contract must state that the adviser and firm are fiduciaries to the extent they
make investment recommendations.



The contract also must provide that the adviser and firm will adhere to impartial
conduct standards including: acting in the “best interest” of the plan/IRA investor,
charging no more than reasonable compensation and not making misleading
statements.



The adviser and firm must warrant in the contract that they would comply with
applicable federal and state law related to the provision of advice and the investment
transaction.



The adviser’s firm must warrant in the contract that it has put in place policies and
procedures to mitigate material conflicts of interest and to ensure compliance with the
impartial conduct standards. This includes a warranty that the firm does not allow
employment incentives that would encourage advisers to violate the best interest
standard.



Under the best interest standard, advice must reflect the care, skill, prudence, and
diligence under the circumstances then prevailing that a prudent person would
exercise based on the investment objectives, risk tolerance, financial circumstances
and needs of the plan or IRA Investor without regard to the financial or other interests
of the adviser, firm, or any affiliate, related entity or other parties.



The Best Interest Contract Exemption would require that if firms limit
recommendations based on proprietary products or receipt of third-party payments or
for other reasons they must disclose those limitations, and make a specific
determination that the limitation does not prevent the adviser from providing
investment advice that is in the best interest of the firm’s plan and IRA clients or
otherwise adhering to the impartial conduct standards. The adviser must further
notify the plan or IRA investor if the adviser does not recommend a sufficiently broad
range of assets to meet the plan’s or IRA investor’s needs. Payments received by
such firms must be reasonable in relation to a specific service rendered in exchange
for the payment.
52



The proposed Best Interest Contract Exemption would have required firms to provide
customers with a chart with respect to the recommended investment before execution
of the purchase. Among other things, the chart would have been required to show the
total cost of the investment at the point of sale over 1-, 5-, and 10-year periods,
including the acquisition cost (such as commissions), ongoing costs, (such as revenue
sharing), disposition costs and other costs that reduce the investment’s return. On an
annual basis, the customer would have been required to receive a summary of the
investments purchased or sold, and the adviser’s and financial institution’s total
compensation as a result of the listed investments over the period.129 Firms also
would have been required to create a public webpage disclosing their compensation
arrangements with the third parties whose products they recommend. Firms also
were required to retain specified data on investments and returns for six years to
enable later analysis by the Department.

2.8.1.2

Proposed Principal Transactions PTE

Commenters responding to the 2010 Proposal also indicated that if the current regulation
is amended, the entities that would be newly defined as investment advice fiduciaries would
need exemptive relief for principal transactions between a plan or IRA and a fiduciary adviser.
In this regard, both ERISA and the IRC prohibit a fiduciary from dealing with the assets of the
plan or IRA investor in his or her own interest or for his or her own account. ERISA further
prohibits a fiduciary from, in his or her individual or any other capacity, acting in any transaction
involving the plan on behalf of a party (or representing a party) whose interests are adverse to
the interests of the plan or the interests of its participants or beneficiaries. As a result, the
purchase or sale of a security in a principal transaction between a plan or IRA investor and an
investment advice fiduciary, resulting from the fiduciary’s provision of investment advice within
the meaning of 29 C.F.R. 2510.3-21 to the plan is generally prohibited under both ERISA and
the IRC.
Therefore, as part of the 2015 proposed regulatory package, the Department proposed
relief for principal transactions in certain debt securities between a plan or IRA and an
investment advice fiduciary where the principal transaction was a result of the provision of
investment advice to a plan or IRA by the investment advice fiduciary. While commenters
requested relief with respect to a broad range of principal transactions (e.g., those involving
equities, debt securities, futures, currencies, etc.), the Department elected to propose relief solely
with respect to debt securities. The Department believed that debt securities uniquely represent
a category of investments that are widely and deeply held, yet are still reliant on principal
transactions for the majority of executions. Like the proposed Best Interest Contract Exemption,
the proposed Principal Transactions Exemption would have required firms and the advisers to

129

For securities, this disclosure regime was designed to supplement current SEC and FINRA disclosure requirements. For example, RIAs
disclose who they are compensated by in general with the Form ADV (Part 2A). SEC Staff Dodd Frank Study (2011), 40. In addition, the
fees and expenses related to a mutual fund are disclosed in the fund’s prospectus. FINRA rules also require disclosure of certain obvious
conflicts such as if the BD is trading as a principal or acting as a market maker for the recommended security. Case law has determined that
BDs should provide additional disclosures necessary for customers to evaluate a recommendation. Ibid., 56. The chart described here,
however, would have provided information on the dollar amount of costs that flow from the particular transaction recommended by the
provider. This customized information, together with the timing requirement would have given retirement plan customers significant
assistance in evaluating the cost of an investment and the adviser’s and financial institution’s potential conflicts.

53

contractually commit to adhere to the impartial conduct standards, warrant as to compliance with
applicable federal and state law, and warrant that the firm has adopted policies and procedures
designed to mitigate the impact of material conflicts of interest and ensure that the individual
advisers adhere to the impartial conduct standards. Certain disclosures would have been
required and the plan or IRA investor would have been required to consent to the principal
transaction.

2.8.1.3

Proposed Amendment to PTE 75-1, Part V

An existing class exemption, PTE 75-1, Part V, provides relief for extensions of credit to
plans by BDs. Under the exemption, BDs who possess or exercise any discretionary authority or
control (except as a directed trustee) with respect to the investment of the plan assets involved in
the transaction, or render investment advice with respect to those assets, may not receive
compensation in return for the extension of credit. Commenters responding to the 2010 Proposal
requested that the Department provide exemptive relief for compensation for extensions of credit
to a plan or IRA investor by investment advice fiduciaries, because many BDs that have
historically relied upon the relief provided by PTE 75-1, Part V, would not be able to rely on
such relief if they became investment advice fiduciaries under the 2015 Proposal.
Therefore, as part of the 2015 Proposal, the Department proposed to amend PTE 75-1,
Part V to add a new section that would have provided an exception to the requirement that
fiduciaries not receive compensation under the exemption. The amendment would have
provided that an investment advice fiduciary may receive reasonable compensation for
extending credit to a plan or IRA to avoid a failed purchase or sale of securities involving the
plan or IRA subject to several conditions. Under the proposal, relief would not be available if
the potential failure of the purchase or sale of the securities was the result of the action or
inaction by the broker-dealer or any affiliate.
Additionally, the terms of the extension of credit would have been required to be at least
as favorable to the plan or IRA as the terms available in an arm’s length transaction between
unaffiliated parties. Finally, the plan or IRA investor would have been required to receive
written disclosure of certain terms prior to the extension of credit. This disclosure did not need
to be made on a transaction-by-transaction basis, and could be part of an account opening
agreement. As proposed, the disclosure had to include the rate of interest or other fees that
would be charged on such extension of credit, and the method of determining the balance upon
which interest will be charged. The plan or IRA also would have been required to be provided
with prior written disclosure of any changes to these terms.

2.8.1.4

Proposed Amendment to PTE 86-128

Another existing class exemption, PTE 86-128, provides relief for an investment advice
fiduciary’s use of its authority to cause a plan to pay a fee to such fiduciary or its affiliate for
effecting or executing securities transactions. The exemption also provides relief for an
investment advice fiduciary to act as the agent in an agency cross transaction for both the plan
and one or more other parties to the transaction, and to receive reasonable compensation therefor
from one or more other parties to the transaction.
The Department proposed to amend PTE 86-128 to require all fiduciaries relying on the
exemption to adhere to the same impartial conduct standards required in the Best Interest
Contract Exemption and the Principal Transactions Exemption. The proposed amendment also
would have added a new covered transaction that would have permitted certain fiduciaries that
are BDs (and who are not the principal underwriter for or affiliated with a mutual fund) to use
54

their authority to cause plans or IRAs to purchase mutual fund shares in riskless principal
transactions from the fiduciary and receive a commission in connection with the transaction.
Relief for this transaction was available in a different class exemption, PTE 75-1, Part II (2). As
the Department believes that this transaction should be engaged in pursuant to conditions set
forth in PTE 86-128, the 2015 regulatory package proposed to move relief for this transaction to
PTE 86-128 and to revoke PTE 75-1, Part II(2).
The Department also proposed an amendment to PTE 86-128 that eliminated relief
provided by PTE 86-128 to fiduciary investment advisers to IRAs. The proposal reflected the
Department’s view that the provisions of the Best Interest Contract Exemption better protect the
interests of IRAs with respect to investment advice regarding securities transactions.

2.8.1.5

Proposed Amendment to PTE 84-24

The Department also proposed in 2015 to amend PTE 84-24 to require all fiduciaries
relying on the exemption to adhere to the same impartial conduct standards required in the Best
Interest Contract Exemption and the Principal Transactions Exemption. At the same time, the
proposed amendment would have revoked PTE 84-24 in part so that investment advice
fiduciaries to IRA owners would not be able to rely on PTE 84-24 with respect to (1)
transactions involving variable annuity contracts and other annuity contracts that constitute
securities under federal securities laws, and (2) transactions involving the purchase of mutual
fund shares. Investment advice fiduciaries instead would have been permitted to rely on the Best
Interest Contract Exemption for compensation received in connection with these transactions.
The 2015 proposed amendment to PTE 84-24 also would have required the fiduciary
engaging in a transaction covered by the exemption to maintain records necessary to enable the
Department and IRA owners (and certain persons described in the proposed amendment) to
determine whether the conditions of this exemption have been met. This requirement replaced
the more limited existing recordkeeping requirement in the original version of PTE 84-24.

2.8.1.6

Proposed Amendments to PTEs 75-1, Part III, 75-1,
Part IV, 77-4, 80-83, and 83-1

The 2015 Proposal included proposed amendments to prohibited transaction exemptions
75-1, Part III, 75-1, Part IV, 77-4, 80-83, and 83-1 to require all fiduciaries relying on the
exemption to adhere to the impartial conduct standards. These exemptions provide the following
relief:


PTE 75-1, Part III permits a fiduciary to use its authority to cause a plan or IRA to
purchase securities from a member of an underwriting syndicate other than the
fiduciary, when the fiduciary is also a member of the syndicate;



PTE 75-1, Part IV permits a plan or IRA to purchase securities in a principal
transaction from a fiduciary that is a market maker with respect to such securities;



PTE 77-4 provides relief for a plan’s or IRA’s purchase or sale of open-end
investment company shares where the investment adviser for the open-end investment
company is also a fiduciary to the plan or IRA;



PTE 80-83 provides relief for a fiduciary’s use of its authority to cause a plan or IRA
to purchase a security when the proceeds of the securities issuance may be used by
the issuer to retire or reduce indebtedness to the fiduciary or an affiliate;



PTE 83-1 provides relief for the sale of certificates in an initial issuance of
certificates, by the sponsor of a mortgage pool to a plan or IRA, when the sponsor,
55

trustee or insurer of the mortgage pool is a fiduciary with respect to the plan or IRA
assets invested in such certificates.

2.9

2016 Final Rule and PTEs
2.9.1 Final Rule

After carefully evaluating the full range of public comments and extensive record
developed on the 2015 Proposal, the 2016 final rule replaces the restrictive five-part test in the
1975 regulation with a new definition that better comports with the statutory language in ERISA
and the Code. The final rule largely adopts the general structure of the proposal but with
modifications in response to commenters seeking clarification of certain provisions in the
proposal.
The final rule provides that a person shall be deemed to be rendering investment advice if
they provide for a fee or other compensation certain categories or types of advice. The listed
types of advice are—
(i)

(ii)

A recommendation as to the advisability of acquiring, holding, disposing of
or exchanging securities or other investment property, or a recommendation
as to how securities or other investment property should be invested after
the securities or other investment property are rolled over, transferred, or
otherwise distributed from the plan or IRA;
A recommendation as to the management of securities or other investment
property, including, among other things, recommendations on investment
policies or strategies, portfolio composition, selection of other persons to
provide investment management services, selection of investment account
arrangements (e.g., brokerage versus advisory); or recommendations with
respect to rollovers, transfers or distributions from a plan or IRA, including
whether, in what amounts, in what form, and to what destination such a
rollover, transfer or distribution should be made.

The 2015 Proposal contained a separate provision covering recommendations to hire
investment advisers and investment managers, which commenters indicated created uncertainty
about the breadth of the proposal. Some commenters expressed concern about their perceived
breadth of this prong as encompassing a service or investment provider’s solicitation efforts on
its own (or an affiliate’s) behalf to potential clients, including routine sales or promotion activity,
such as the marketing or sale of one’s own products or services to plans, participants, or IRA
owners. These commenters argued that this provision could be interpreted broadly enough to
capture as investment advice nearly all marketing activity that occurs during initial conversations
with plan fiduciaries or other potential retirement investors associated with hiring a person who
would either manage or advise as to plan assets. Some service providers argued that the
proposal could preclude them from being able to provide information and data on their services
to plans, participants and IRA owners, during the sales process in a non-fiduciary capacity. For
example, commenters questioned whether the mere provision of a brochure or a sales
presentation, especially if targeted to a specific market segment, plan size, or group of
individuals, could be fiduciary investment advice under the 2015 Proposal. Commenters stated
that a similar issue exists in the distribution and rollover context regarding a sales pitch to
participants about potential retention of an adviser to provide retirement services outside of the
plan.

56

Many commenters were also concerned that the provision would treat responses to
requests for proposal (RFP) as investment advice, especially in cases where the RFP requires
some degree of individualization in the response or where specific representations were included
about the quality of services being offered. For example, they noted that a service provider may
include a sample fund line up or discuss specific products or services. Commenters argued that
this or similar individualization should not trigger fiduciary status in an RFP context.
Consistent with prior guidance, the Department continues to believe that the
recommendation of another person to be entrusted with investment advice or investment
management authority over retirement assets is often critical to the proper management and
investment of those assets and should be fiduciary in nature. The Department did not intend,
however, for the definition of investment advice to capture as fiduciary in nature the normal
activity of marketing oneself or an affiliate as a potential fiduciary to be selected by a plan
fiduciary or IRA owner. Thus, the Department revised the final rule as described above to more
clearly and simply set forth the scope of the subject matter covered by the rule.
The 2015 Proposal, like the 1975 regulation, included advice as to “the value of
securities or other property,” and covered certain appraisals and valuation reports as fiduciary in
nature. However, it was considerably more focused than the 2010 Proposal. As discussed
above, responding to comments to the 2010 Proposal, the 2015 Proposal covered only appraisals,
fairness opinions, or similar statements that relate to a particular investment transaction. The
2015 Proposal also expanded the 2010 Proposal’s carve-out for general reports or statements of
value provided to satisfy required reporting and disclosure rules under ERISA or the Code. In
this manner, the 2015 Proposal focused on instances where the plan or IRA owner is looking to
the appraiser for advice on the market value of an asset that the investor is considering to
acquire, dispose, or exchange. Nonetheless, the Department received many comments raising
questions about the scope and application of these provisions in the 2015 Proposal. It continues
to be the Department’s opinion that, in many transactions, a proper appraisal of hard-to-value
assets is the single most important factor in determining the prudence of the transaction.
Accordingly, the Department believes that employers and participants could benefit from the
imposition of fiduciary standards on appraisers when they value assets in connection with
investment transactions. However, after carefully reviewing the comments on the 2015
Proposal, the Department has concluded that the issues related to valuations are more
appropriately addressed in a separate regulatory initiative. Therefore, unlike the proposal, the
final rule does not address appraisals, fairness opinions, or similar statements concerning the
value of securities or other property in any way. Consequently, in the absence of regulations or
other guidance by the Department, appraisals, fairness opinions and other similar statements will
not be considered fiduciary investment advice for purposes of the final rule. A person would be
considered a fiduciary investment adviser in connection with a recommendation of a type
discussed above if the recommendation is made either directly or indirectly (e.g., through or
together with any affiliate) by a person who:
(i)

Represents or acknowledges that it is acting as a fiduciary within the
meaning of the Act or Code with respect to the advice;

(ii)

Renders the advice pursuant to a written or verbal agreement, arrangement
or understanding that the advice is based on the particular investment needs
of the advice recipient; or

(iii) Directs the advice to a specific advice recipient or recipients regarding the
advisability of a particular investment or management decision with respect
to securities or other investment property of the plan or IRA.
57

As in the 2015 Proposal, under the final rule, advisers who claim fiduciary status under
ERISA or the Code in providing advice are required to honor their words. They may not say
they are acting as fiduciaries and later argue that the advice was not fiduciary in nature.
The proposal alternatively required that “the advice be rendered pursuant to a written or
verbal agreement, arrangement or understanding that the advice is individualized to or that such
advice is specifically directed to, the advice recipient for consideration in making investment or
management decisions with respect to the plan or IRA.” Commenters focused on several aspects
of this provision. First, they argued that the “specifically directed” and “individualized” prongs
were unclear, overly broad, and duplicative, because any advice that was individualized would
also be specifically directed at the recipient. Second, they said it was not clear whether there had
to be an agreement, arrangement or understanding that advice was specifically directed to a
recipient, and, if so, what would be required for such an agreement, arrangement or
understanding to exist. They expressed concern about fiduciary status possibly arising from a
subjective belief of a participant or IRA investor. And third, they requested modification of the
phrase “for consideration,” believing the phrase was overly broad and set the threshold too low
for requiring that recommendations be made for the purpose of making investment decisions. A
number of other commenters explicitly endorsed the phrases “specifically directed,” and
“individualized to,” believing that these are appropriate and straightforward thresholds to attach
fiduciary status. After reviewing the comments, the Department concluded that the provision in
the proposal could be improved and clarified. Therefore, the Department revised the provision
in the final rule in two respects. First, the phrase “for consideration” has been removed from the
provision. After reviewing the comments, the Department believes that that clause as drafted
was largely redundant to the provisions in paragraph (a)(1) of the proposal and that the final rule
sets forth the subject matter areas to which a recommendation must relate to constitute
investment advice. The final rule revises the condition to require that advice be “directed to” a
specific advice recipient or recipients regarding the advisability of a particular investment or
management decision. Second, although the preamble to the 2015 Proposal stated that the
“specifically directed to” provision, like the individualized advice prong, required that there be
an agreement, arrangement or understanding that advice was specifically directed to the
recipient, the Department agrees that using that terminology for both the individualized advice
prong and the specifically directed to prong serves no useful purpose for defining fiduciary
investment advice. The point of the proposal’s language concerning advice specifically directed
to an individual was to distinguish specific investment recommendations to an individual from
“recommendations made to the general public or no one in particular.” Examples included
general circulation newsletters, television talk show commentary, and remarks in speeches and
presentations at conferences. As discussed below, the final rule now includes a new provision to
make clear that such general communications are not advice because they are not
recommendations within the meaning of the final rule.
In the final rule, the initial threshold of whether a person is a fiduciary by virtue of
providing investment advice continues to be whether that person makes a recommendation as to
the various activities described in paragraph (a)(1) of the rule. The final rule continues to define
“recommendation” as a communication that, based on its content, context, and presentation,
would reasonably be viewed as a suggestion that the advice recipient engage in or refrain from
taking a particular course of action. Thus, communications that require the adviser to comply
with suitability requirements under applicable securities or insurance laws will be viewed as a
recommendation. The final rule also includes additional text intended to clarify the nature of
communications that would constitute recommendations. The final rule makes it clear that the

58

determination of whether a “recommendation” has been made is an objective rather than
subjective inquiry.
To further clarify the meaning of recommendation, the final rule describes certain
services or materials that do not constitute recommendations, such as general communications
and commentary on investment products such as financial newsletters, marketing or making
available a menu of investment alternatives that a plan fiduciary could choose from, identifying
investment alternatives that meet objective criteria specified by a plan fiduciary, and providing
information and materials that constitute investment education or retirement education.
These provisions were described as “carve-outs” in the 2015 Proposal. As the
Department described in the proposal, the purpose of the carve-outs was to highlight that, in
many circumstances, plan fiduciaries, participants, beneficiaries, and IRA owners may receive
recommendations that, notwithstanding the general definition set forth in the proposal, should
not be treated as fiduciary investment advice. However, the Department agrees with many of the
commenters that much of the conduct and information described in the proposal for certain of
the carve-outs did not meet the technical definition of investment advice under the proposal such
that they should be excluded from that definition. Some were more in the nature of examples of
education or other information which would not rise to the level of a recommendation to begin
with. Thus, the final rule retains these provisions, with changes made in response to comments,
but presents them as examples to clarify the definition of recommendation and does not
characterize them as carve-outs. The final rule also incorporates, with modifications, “carveouts” from the proposal that addressed counterparty transactions, swaps transactions, and certain
employee communications. Here again, the final rule does not use the term “carve-outs,” but
these provisions still recognize circumstances in which plans, plan fiduciaries, plan participants
and beneficiaries, IRAs, and IRA owners may receive recommendations the Department does
not believe should be treated as fiduciary investment advice notwithstanding the general
definition set forth in the rule. Each of the provisions has been modified from the proposal to
address public comments and refine the provision.
Platform Providers and Selection and Monitoring Assistance -- Similar to the 2015
Proposal, this provision of the final rule is directed to service providers, such as recordkeepers
and third-party administrators, that offer a “platform” or selection of investment alternatives to
participant-directed individual account plans under ERISA and plan fiduciaries of these plans
who choose the specific investment alternatives that will be made available to participants for
investing their individual accounts. The provision makes clear that such persons would not
make recommendations simply by making available, without regard to the individualized needs
of the plan or its participants and beneficiaries, a platform of investment vehicles from which
plan participants or beneficiaries may direct the investment of assets held in, or contributed to,
their individual accounts, as long as the plan fiduciary is independent of the person who markets
or makes available the investment alternatives and the person discloses in writing to the plan
fiduciary that they are not undertaking to provide impartial investment advice or to give advice
in a fiduciary capacity. The provision also makes clear that certain common activities that
platform providers may carry out to assist plan fiduciaries in selecting and monitoring the
investment alternatives that they make available to plan participants are not recommendations.
Thus, identifying offered investment alternatives meeting objective criteria specified by the plan
fiduciary or providing objective financial data regarding available investment alternatives to the
plan fiduciary would not cause a platform provider to be a fiduciary investment adviser.
Investment Education -- The 2015 Proposal carved out investment education from the
definition of investment advice. In doing so, the Department incorporated much of the
59

Department’s earlier 1996 Interpretive Bulletin (IB 96-1),130 but with the important exceptions
that asset allocation models and interactive investment materials could not include or identify
any specific investment product or specific investment alternative available under the plan or
IRA. The Department understood that not incorporating these provisions of IB 96–1 into the
proposal represented a significant change in the information and materials that may constitute
investment education. Accordingly, the Department specifically invited comments on whether
the change was appropriate.
A few commenters supported this change. They argued that participants are highly
vulnerable to subtle, but powerful, influences by advisers when they receive asset allocation
information. They believe that ordinary participants may view these models, including specific
investments included therein, as specific investment recommendations even if the provider does
not intend it as advice and even if the provider includes caveats or statements about the
availability of other products. In contrast, other commenters argued – particularly with respect
to ERISA-covered plans – that it is a mistake to prohibit the use of specific investment options in
asset allocation models used for educational purposes. They claimed that the inability to
reference specific investment options in asset allocation models and interactive materials would
greatly undermine the effectiveness of these models and materials as educational tools. They
said that without the ability to include specific investment products, participants could have a
hard time understanding how the educational materials relate to specific investment options.
After evaluating the comments and considerations above, the Department has determined
that asset allocation models and interactive investment materials can identify a specific
investment product or specific alternative available under ERISA-covered plans and fall within
the education provision in the final rule if (1) the alternative is a designated investment
alternative (DIA) under an employee benefit plan; (2) the alternative is subject to fiduciary
oversight by a plan fiduciary independent of the person who developed or markets the
investment alternative or distribution option; (3) the asset allocation models and interactive
investment materials identify all the other designated investment alternatives available under the
plan that have similar risk and return characteristics, if any; and (4) the asset allocation models
and interactive investment materials are accompanied by a statement that identifies where
information on those investment alternatives may be obtained. When these conditions are
satisfied with respect to asset allocation or interactive investment materials, the communications
can be appropriately treated as non-fiduciary “education” rather than fiduciary investment
recommendations, and the interests of plan participants are protected by fiduciary oversight and
monitoring of the DIAs.
The Department does not agree that the same conclusion applies in the case of
presentations of specific investments to IRA owners because of the lack of review and prudent
selection of the presented options by an independent plan fiduciary, and because of the
likelihood that such “guidance” or “education” would often amount to specific investment
recommendations in the IRA context. The Department was not able to reach the conclusion that
it should create a broad safe harbor from fiduciary status for circumstances in which the IRA
provider effectively narrows the entire universe of investment alternatives available to IRA

130

29 C.F.R. 2509.96-1 (IB 96-1),

60

owners to just a few coupled with asset allocation models or interactive materials. Nor could the
Department readily import the conditions applicable to such plan communications to IRA
communications.
Similarly, because the provision is limited to DIAs available under employee benefit
plans, the use of asset allocation models and interactive materials with specific investment
alternatives available through a self-directed brokerage account would not be covered by the
“education” provision in the final rule. Such communications lack the safeguards associated
with DIAs, and pose many of the same problems and dangers as identified with respect to IRAs.
These tools and models can be important in the IRA and self-directed brokerage account
context, just as in the plan context more generally. An asset allocation model for an IRA could
still qualify as “education” under the final rule, for example, if it described a hypothetical
customer’s portfolio as having certain percentages of investments in equity securities, fixedincome securities and cash equivalents. The asset allocation could also continue to be
“education” under the final rule if it described a hypothetical portfolio based on broad-based
market sectors (e.g., agriculture, construction, finance, manufacturing, mining, retail, services,
transportation and public utilities, and wholesale trade). The asset allocation model would have
to meet the other criteria in the final rule and could not include particular securities. In the
Department’s view, as an allocation becomes narrower or more specific, the presentation of the
portfolio gets closer to becoming a recommendation of particular securities. Although the
Department is open to continuing a dialog on possible approaches for additional regulatory or
other guidance in this area, when advisers use such tools and models to effectively recommend
particular investments, they should be prepared to adhere to fiduciary norms and to make sure
their investment recommendations are in the investors’ best interest.
General Communications -- As discussed above, many commenters expressed concern
about the phrase “specifically directed” in the 2015 Proposal and asked the Department to clarify
the application of the final rule to certain communications including casual conversations with
clients about an investment, distribution or rollovers, responding to participant inquiries about
their investment options, ordinary sales activities, providing research reports, sample fund menus
and other similar support activities. For example, they were concerned about communications
made in newsletters, media commentary, or remarks directed to no one in particular. Although
the Department believed that the definition of “recommendation” in the proposal sufficiently
distinguished such communications from investment advice, the Department has concluded that
it would be helpful if the final rule more expressly addressed these types of communications to
alleviate commenters’ concerns. Thus, the final rule includes a new “general communications”
paragraph as an example of communications that are not considered recommendations under the
definition. This paragraph affirmatively excludes from investment advice the furnishing of
communications that a reasonable person would not view as an investment recommendation,
including general circulation newsletters, television, radio, and public media talk show
commentary, remarks in widely attended speeches and conferences, research reports prepared for
general circulation, general marketing materials, general market data, including data on market
performance, market indices, or trading volumes, price quotes, performance reports, or
prospectuses.
Transactions with Independent Plan Fiduciaries with Financial Expertise -- The
proposed rule provided a carve-out (referred to as the “seller’s” or “counterparty” carve-out)
from the general definition for incidental advice provided in connection with an arm’s length
sale, purchase, loan, or bilateral contract between an expert plan investor and the adviser. The
exclusion also applied in connection with an offer to enter into such an arm’s length transaction,
61

and when the person providing the advice acts as a representative, such as an agent, for the
plan’s counterparty. In particular, the proposal provided a carve-out for incidental advice
provided in connection with counterparty transactions with a plan fiduciary with financial
expertise. As a proxy for financial expertise the rule required that the advice recipient be a
fiduciary of a plan with 100 or more participants or have responsibility for managing at least
$100 million in plan assets. Additional conditions applied to each of these two categories of
sophisticated investors that were intended to ensure the parties understood the non-fiduciary
nature of the relationship.
Some commenters on the proposal offered threshold views on whether the Department
should include a seller’s carve-out as a general matter or whether, for example, an alternative
approach such as requiring specific disclosures would be preferable. Others strongly supported
the inclusion of a seller’s carve-out, believing it to be a critical component of the proposal. As
explained in the proposal, the purpose of the proposed carve-out was to avoid imposing ERISA
fiduciary obligations on sales pitches that are part of arm’s length transactions where neither side
assumes that the counterparty to the plan is acting as an impartial trusted adviser. The premise
of the proposed carve-out was that both sides of such transactions understand that they are acting
at arm’s length, and neither party expects that recommendations will necessarily be based on the
buyer’s best interests, or that the buyer will rely on them as such.
Consumer advocates generally agreed with the Department’s views expressed in the
preamble that it was appropriate to limit the carve-out to large plans and sophisticated asset
managers. These commenters encouraged the Department to retain a very narrow and stringent
carve-out. They argued that the communications to participants and retail investors are generally
presented as advice and understood to be advice. Indeed, both FINRA and state insurance law
commonly require that recommendations reflect proper consideration of the investment’s
suitability in light of the investor’s particular circumstances, regardless of whether the
transaction could be characterized as involving a “sale.” Additionally, commenters noted that
participants and IRA owners cannot readily ascertain the nuanced differences among different
types of financial professionals (including differences in legal standards that apply to different
professionals) or easily determine whether advice is impartial or potentially conflicted, or assess
the significance of the conflict. Similar points were made concerning advice in the small plan
marketplace.
These commenters expressed concern, shared by the Department, that allowing
investment advisers to claim non-fiduciary status as “sellers” across the entire retail market
would effectively open a large loophole by allowing brokers and other advisers to use
disclosures in account opening agreements, investor communications, advertisements, and
marketing materials to evade fiduciary responsibility and accountability for investment
recommendations that investors rely upon to make important investment decisions. Just as
financial service companies currently seek to disclaim fiduciary status under the five-part test
through standardized statements disclaiming the investor’s right to rely upon communications as
individualized advice, an overbroad seller’s exception could invite similar statements that
recommendations are made purely in a sales capacity, even as oral communications and
marketing materials suggest expert financial assistance upon which the investor can and should
rely.
On the other hand, many commenters representing financial services providers argued for
extending the “seller’s” carve-out to include transactions in the market composed of smaller
plans and individual participants, beneficiaries and IRA owners. These commenters contended
that the lines drawn in the proposal were based on a flawed assumption that representatives of
62

small plans and individual investors cannot understand the difference between a sales pitch and
advice. They argued that failure to extend the carve-out to these markets will limit the ability of
small plans and individual investors to obtain advice and to choose among a variety of services
and products that are best suited to their needs. They also argued that there is no statutory basis
for distinguishing the scope of the fiduciary responsibility based on plan size. Some
commenters suggested that the Department could extend the carve-out to individuals that meet
financial or net worth thresholds or to “accredited investors,” “qualified purchasers,” or
“qualified clients” under federal securities laws. Some commenters also requested that the
Department expand the persons and entities that would be considered “sophisticated” fiduciaries
for purposes of the carve-out, for example asking that banks, savings and loan associations, and
insurance companies be explicitly covered. Others alternatively argued that the carve-out should
be expanded to fiduciaries of participant-directed plans regardless of plan size, which they said
is not a reliable predictor for financial sophistication, or if the plan is represented by a financial
expert such as an ERISA section 3(38) investment manager or an ERISA qualified professional
asset manager. Other commenters asked that the carve-out be expanded to all proprietary
products on the theory that investors generally understand that a person selling proprietary
products is going to be making recommendations that are biased in favor of the proprietary
product. Others suggested that the Department could address its concern about retail investor
confusion by requiring specified disclosures, warranties, or representations to investors or small
plan fiduciaries.
The Department does not believe it would be consistent with the language or purposes of
ERISA Section 3(21) to extend this exclusion to small retail employee benefit plan investors or
IRA owners. The Department explained its rationale in the preamble to the 2015 Proposal. In
summary, retail investors were not included in this carve-out because (1) the Department did not
believe the relationships fit the arm’s length characteristics that the seller’s carve-out was
designed to preserve; (2) the Department did not believe disclaimers of adviser status were
effective in alerting retail investors to nature and consequences of the conflicting financial
interests; (3) IRA owners in particular do not have the benefit of a menu selected or monitored
by an independent plan fiduciary; (4) small business sponsors of small plans are more like retail
investors compared to large companies that often have financial departments and staff dedicated
to running the company’s employee benefit plans; (5) it would be inconsistent with
congressional intent under ERISA section 408(b)(14) to create such a broad carve-out, as most
recently reflected in enactment of a statutory provision that placed substantial conditions on the
provision of investment advice to individual participants and IRA owners; and (6) there were
other more appropriate ways to ensure such retail investors had access to investment advice,
such as prohibited transaction exemptions, and investment education. In addition, and perhaps
more fundamentally, the Department rejects the purported dichotomy between a mere “sales”
recommendation, on the one hand, and advice, on the other in the context of the retail market for
investment products. As reflected in financial service industry marketing materials, the
industry’s comment letters reciting the guidance they provide to investors, and the obligation to
ensure that recommended products are at least suitable to the individual investor, sales and
advice go hand in hand in the retail market. When plan participants, IRA owners, and small
businesses talk to financial service professionals about the investments they should make, they
typically pay for, and receive, advice.
The Department continues to believe for all of those reasons that it would be an error to
provide a broad “seller’s” exemption for investment advice in the retail market.
Recommendations to retail investors and small plan providers are routinely presented as advice,
consulting, or financial planning services. In fact, in the securities markets, brokers’ suitability
63

obligations generally require a significant degree of individualization. Most retail investors and
many small plan sponsors are not financial experts, are unaware of the magnitude and impact of
conflicts of interest, and are unable effectively to assess the quality of the advice they receive.
IRA owners are especially at risk because they lack the protection of having a menu of
investment options chosen by an independent plan fiduciary that is charged to protect their
interests. Similarly, small plan sponsors are typically experts in the day-to-day business of
operating a company, not in managing financial investments for others. In this retail market,
such an exclusion would run the risk of creating a loophole that would result in the rule failing to
make any real improvement in consumer protections because it could be used by financial
service providers to evade fiduciary responsibility for their advice through the same type of
boilerplate disclaimers that some advisers use to avoid fiduciary status under the current “fivepart test” regulation. The Department also is not prepared to conclude that written disclosures,
including models developed by the Department, are sufficient to address investor confusion
about financial conflicts of interest. As discussed below in this Regulatory Impact Analysis,
research has shown that disclaimers are ineffective in alerting retail investors to the potential
costs imposed by conflicts of interest, or the fact that advice is not necessarily in their best
interest, and may even exacerbate these costs. In addition to problems with the effectiveness of
such disclosures, the possibility of inconsistent oral representations raise questions about
whether any boilerplate written disclosure could ensure that the person’s financial interest in the
transaction is effectively communicated as being in conflict with the interests of the advice
recipient.
Further, the Department is not prepared to adopt the approach suggested by some
commenters that the provision be expanded to include individual retail investors through an
accredited or sophisticated investor test that uses wealth as a proxy for the type of investor
sophistication that was the basis for the Department proposing some relationships as nonfiduciary. The Department agrees with the commenters that argued that merely concluding
someone may be wealthy enough to be able to afford to lose money by reason of bad advice
should not be a reason for treating advice given to that person as non-fiduciary. Nor is wealth
necessarily correlated with financial sophistication. Individual investors may have considerable
savings as a result of numerous factors unrelated to financial sophistication, such as a lifetime of
thrift and hard work, inheritance, marriage, business successes unrelated to investment
management, or simple good fortune.
Nonetheless, the Department agrees with the commenters that criticized the proposal
with arguments that the criteria in the proposal were not good proxies for appropriately
distinguishing non-fiduciary communications taking place in an arm’s length transaction from
instances where customers should reasonably be able to expect investment recommendations to
be unbiased advice that is in their best interest. The Department notes that the definition of
investment advice in the proposal expressly required a recommendation directly to a plan, plan
fiduciary, plan participant or IRA owner. The use of the term “plan fiduciary” in the proposal
was not intended to suggest that ordinary business activities among financial institutions and
licensed financial professionals should become fiduciary investment advice relationships merely
because the institution or professional was acting on behalf of an ERISA plan or IRA. The “100
participant plan” threshold was borrowed from annual reporting provisions in ERISA that were
designed to serve different purposes related to simplifying reporting for small plans and reducing
administrative burdens on small businesses that sponsor employee benefit plans. The “$100
million in assets under management” threshold was a better proxy for the type of financial
capabilities the carve-out was intended to capture, but it failed to include a range of financial

64

services providers that fairly could be said to have the financial capabilities and understanding
that was the focus of the carve-out.
Thus, after carefully evaluating the comments, the Department has concluded that the
exclusion is better tailored to the Department’s stated objective by requiring the communications
to take place with plan or IRA fiduciaries who are independent from the person providing the
advice and are either licensed and regulated providers of financial services or employee benefit
plan fiduciaries with responsibility for the management of $50 million in assets.131 This
provision does not require that the $50 million be attributable to only one plan, but rather allows
all the plan and non-plan assets under management to be included in determining whether the
threshold is met. Such parties should have a high degree of financial sophistication and may
often engage in arm’s length transactions in which neither party has an expectation of reliance
on the counterparty’s recommendations. The final rule revises and re-labels the carve-out in a
new paragraph that provides that a person shall not be deemed to be a fiduciary within the
meaning of Section 3(21)(ii) of the Act solely because of the provision of any advice (including
the provision of asset allocation models or other financial analysis tools) to an independent
person who is a fiduciary of the plan or IRA132 with respect to an arm’s length sale, purchase,
loan, exchange, or other transaction involving the investment of securities or other property, if
the person knows or reasonably believes that they are dealing with a fiduciary of the plan or IRA
who is independent from the person providing the advice and who is (1) a bank;133 (2) an
insurance carrier which is qualified under the laws of more than one State to perform the
services of managing, acquiring, or disposing of assets of a plan134; (3) a registered investment
adviser;135 (4) a broker-dealer registered under the Exchange Act; or (5) any other person acting
as an independent fiduciary that holds, or has under management or control, total assets of at
least $50 million.136 Additional conditions are intended to ensure that this provision is limited to
circumstances that involve true arm’s length transactions between investment professionals or
large asset managers who do not have a legitimate expectation that they are in a relationship of
trust and loyalty where they fairly can rely on the other person for impartial advice.

131

132

133

134

135

136

This provision does not require that the $50 million be attributable to only one plan, but rather allows all the plan and non-plan assets under
management to be included in determining whether the threshold is met.
Including a fiduciary to an investment contract, product, or entity that holds plan assets as determined pursuant to Sections 3(42) and 401 of
the Act and 29 C.F.R. 2510.3-101.
As defined in Section 202 of the Advisers Act or similar institution that is regulated and supervised and subject to periodic examination by a
State or Federal agency.
Prohibited Transaction Exemption 84-14 (PTE 84-14) permits transactions between parties in interest to a plan and an investment fund in
which the plan has an interest provided the fund is managed by a qualified professional plan asset manager (QPAM) that satisfies certain
conditions. Among the entities that can qualify as a QPAM is “an insurance company which is qualified under the laws of more than one
state to manage, acquire or dispose of any assets of a plan…” 49 FR 9494.
Registered under the Advisers Act or, if not registered as an investment adviser under such Act by reason of paragraph (1) of Section 203A
of such Act, is registered as an investment adviser under the laws of the State (referred to in such paragraph (1)) in which it maintains its
principal office and place of business.
The $50 million threshold in the final rule for “other plan fiduciaries” is similarly based upon the definition of “institutional account” from
FINRA rule 4512(c)(3) to which the suitability rules of FINRA rule 2111 apply and responds to the requests of commenters that the test for
sophistication be based on market concepts that are well understood by brokers and advisers. Specifically, FINRA Rule 2111(b) on
suitability and FINRA’s “books and records” Rule 4512(c) both use a definition of “institutional account,” which means the account of a
bank, savings and loan association, insurance company, registered investment company, registered investment adviser or any other person
(whether a natural person, corporation, partnership, trust or otherwise) with total assets of at least $50 million. Ibid. at Q&A 8.1. In regard
to the “other person” category, FINRA’s rule had used a standard of at least $10 million invested in securities and/or under management,
but revised it to the current $50 million standard. Ibid. at footnote 80. In addition, the FINRA rule requires: (1) that the broker have “a
reasonable basis to believe the institutional customer is capable of evaluating investment risks independently, both in general and with
regard to particular transactions and investment strategies involving a security or securities” and (2) that “the institutional customer
affirmatively indicates that it is exercising independent judgment.”

65

Swap and security-based swap transaction -- The proposal included a “carve-out”
intended to make it clear that communications and activities engaged in by counterparties to
ERISA-covered employee benefit plans in swap and security-based swap transactions did not
result in the counterparties becoming investment advice fiduciaries to the plan. As explained in
the preamble to the 2015 Proposal, swaps and security-based swaps are a broad class of financial
transactions defined and regulated under amendments to the Commodity Exchange Act and the
Securities Exchange Act by the Dodd-Frank Act.137 Special rules apply for swap and securitybased swap transactions involving “special entities,” a term that includes employee benefit plans
under ERISA. Under the business conduct standards in the Commodity Exchange Act as added
by the Dodd-Frank Act, swap dealers or major swap participants that act as counterparties to
ERISA plans, must, among other conditions, have a reasonable basis to believe that the plans
have independent representatives who are fiduciaries under ERISA.138 Similar requirements
apply for security-based swap transactions.139 The CFTC has issued a final rule to implement
these requirements and the SEC has issued a proposed rule that would cover security-based
swaps.140 In the Department’s view, when Congress enacted the swap and security-based swap
provisions in the Dodd-Frank Act, including those expressly applicable to ERISA-covered plans,
Congress did not intend that engaging in regulated conduct as part of a swap or security-based
swap transaction would automatically give rise to additional fiduciary obligations or restrictions
under Title I of ERISA.
In this regard, the disclosures required under the business conduct standards, do not in the
Department’s view compel counterparties to ERISA-covered employee benefit plans to make a
recommendation for purposes of paragraph (a) of the final rule or otherwise compel them to act
as fiduciaries in swap and security-based swap transactions conducted pursuant to section 54s of
the Commodity Exchange Act or section 15 F of the Exchange Act. The final rule provision on
swap and security-based swap transactions is intended to address this issue and includes
conditions are intended to ensure that this provision is limited to such swap and security-based
swap transactions.
Some commenters argued that IRA owners should be able to engage in a swap and
security-based swap transaction under appropriate circumstances, assuming the account owner is
an “eligible contract participant.” The Department notes that IRAs and IRA owners would not
appear to be “special entities” under the Dodd-Frank Act provisions and transactions with IRAs
would not be subject to the business conduct standards that apply to cleared and uncleared swap
and security-based swap transactions with employee benefit plans. Moreover, for the same
reasons discussed elsewhere that the Department declined to adopt a broad “seller’s” exception
for retail retirement investors, the Department does not believe extending the swap and securitybased swap provisions to IRA investors is appropriate. Rather, the Department concluded that it
was more appropriate to address this issue in the context of the “independent plan fiduciary with
financial expertise” provision.

137

138
139
140

Section 4s(h) of the Commodity Exchange Act (7 U.S.C. § 6s(h)) and Section 15F of the Securities Exchange Act of 1934 (15 U.S.C. §
78o-10(h) establish similar business conduct standards for dealers and major participants in swaps or security-based swaps.
7 U.S.C. § 6s(h)(5).
15 U.S.C § 78o-10(h)(4) and (5).
17 C.F.R. 23.400 to 23.451 (2012); 70 Fed. Reg. 42396 (July 18, 2011).

66

Employees of Plan Sponsors, Employee Benefit Plans, or Plan Fiduciaries -- The
final rule provides that a person is not an investment advice fiduciary if, in his or her capacity as
an employee of the plan sponsor of a plan, as an employee of an affiliate of such plan sponsor, as
an employee of an employee benefit plan, as an employee of an employee organization, or as an
employee of a plan fiduciary, the person provides advice to a plan fiduciary, or to an employee
(other than in his or her capacity as a participant or beneficiary of an employee benefit plan) or
independent contractor of such plan sponsor, affiliate or employee benefit plan, provided the
person receives no fee or other compensation, direct or indirect, in connection with the advice
beyond the employee’s normal compensation for work performed for the employer.
This exclusion from the scope of the fiduciary investment advice definition addresses
concerns raised by public comments seeking confirmation that the rule does not include as
investment advice the communication of fiduciaries’ employees working in a company’s payroll,
accounting, human resources, and financial departments, who routinely develop reports and
recommendations for the company and other named fiduciaries of the sponsors’ plans. The
exclusion was revised to make it clear that it covers employees even if they are not the persons
ultimately communicating directly with the plan fiduciary (e.g., employees in financial
departments that prepare reports for the Chief Financial Officer who then communicates directly
with a named fiduciary of the plan). The Department agrees that such personnel of the employer
should not be treated as investment advice fiduciaries based on communications that are part of
their normal employment duties if they receive no compensation for these advice-related
functions above and beyond their normal salary.
Similarly, and as requested by commenters, the exclusion covers communications
between employees, such as human resources department staff communicating information to
other employees about the plan and distribution options in the plan subject to certain conditions
designed to prevent the exclusion from covering employees who are in fact employed to provide
investment recommendations to plan participants or otherwise becoming a possible loophole for
financial services providers seeking to avoid fiduciary status under the rule. Specifically, the
exclusion covers circumstances where an employee of the plan sponsor of a plan, or as an
employee of an affiliate of such plan sponsor, provides advice to another employee of the plan
sponsor in his or her capacity as a participant or beneficiary of the employee benefit plan,
provided (1) the person’s job responsibilities do not involve the provision of investment advice
or investment recommendations, (2) the person is not registered or licensed under federal or state
securities or insurance law, (3) the advice they provide does not require the person to be
registered or licensed under federal or state securities or insurance laws, and (4) the person
receives no fee or other compensation, direct or indirect, in connection with the advice beyond
the employee’s normal compensation for work performed for the employer. The Department
established these conditions to address circumstances where an HR employee, for example, may
inadvertently make an investment recommendation within the meaning of the final rule. It also
is designed so that it does not cover situations designed to evade the standards and purposes of
the final rule. For example, the Department wanted to ensure that the exclusion did not create a
loophole through which a person could be detailed from an investment firm, or “hired” under a
dual employment structure, as part of an arrangement designed to avoid fiduciary obligations in
connection with investment advice to participants or insulate recommendations designed to
benefit the investment firm.
Execution of Securities Transactions --The final rule provides that a broker or dealer
registered under the Exchange Act that executes transactions for the purchase of securities on
behalf of a plan or IRA will not be a fiduciary with respect to an employee benefit plan or IRA
solely because such person executes transactions for the purchase or sale of securities on behalf
67

of such plan. This provision is unchanged from the current 1975 regulation and the 2015
Proposal.
The final rule is effective 60 days after publication in the Federal Register and will
become applicable one year after the date of publication.

2.9.2 Final PTEs
The 2015 Proposal included several proposed new and amended class PTEs, which
together would permit fiduciary investment advisers to plan and IRA investors to engage in
certain specified types of transactions that would otherwise be prohibited, subject to a number of
protective conditions.
As discussed above, under the 2015 Proposal, a person would be an investment advice
fiduciary if he or she provides a recommendation to a plan, plan fiduciary, plan participant or
beneficiary or IRA investor regarding the advisability of acquiring, holding, disposing or
exchanging of securities or other property pursuant to a written or verbal agreement,
arrangement or understanding that the advice is specifically directed to the advice recipient for
consideration in making investment decisions with respect to securities or other property. Once
a person is an investment advice fiduciary, the person is prohibited by the prohibited transactions
provisions from engaging in certain kinds of transactions involving the plan or IRA, including
transactions in which the fiduciary affects or increases his or her own compensation or that of a
person in which such fiduciary has an interest which may affect the exercise of the fiduciary’s
best judgment. Receipt by a fiduciary of certain common types of fees and compensation, such
as brokerage or insurance commissions, in connection with investment transactions entered into
by the plan or IRA, fall within the prohibition.
The Department recognized the concerns expressed in the comments received from
representatives of BDs and other IRA advisers regarding the potential disruption to current fee
arrangements that would arise by applying the IRC prohibited transactions rules more broadly in
the retail IRA market. Therefore, simultaneous with the publication of the 2015 proposed
regulation, the Department proposed several new and amended PTEs that would allow certain
currently common fee practices to persist subject to conditions provided in the exemptions that
protect plans, plan participants, and IRA investors from investment advice fiduciaries’ conflicts
of interest, which are discussed in Section 2.8. The Department has finalized these exemptions
and amended exemptions as part of the package. The discussion below provides a summary of
the PTEs with a focus on differences between the proposed and final PTEs.

2.9.2.1

Best Interest Contract Exemption

As discussed above, in response to the 2010 Proposal, the Department proposed the Best
Interest Contract Exemption as part of the 2015 Proposal. The final exemption will permit
investment advice fiduciaries and certain related entities to receive compensation as a result of
the investment advice. The Best Interest Contract Exemption will permit investment advice
fiduciaries to receive fees such as commissions, 12b-1 fees, and revenue sharing in connection
with investment transactions by the plan participants, beneficiaries, IRAs and small plans, thus
preserving many current fee practices.
The Best Interest Contract Exemption and other new and amended exemptions follow a
lengthy public notice and comment process, which gave interested persons an extensive
opportunity to comment on the proposed Regulation and exemption proposals. The proposals
initially provided for 75-day comment periods, ending on July 6, 2015, but the Department
extended the comment periods to July 21, 2015. The Department then held four days of public
68

hearings on the new regulatory package, including the proposed exemptions, in Washington, DC
from August 10 to 13, 2015, at which over 75 speakers testified. The transcript of the hearing
was made available on September 8, 2015, and the Department provided additional opportunity
for interested persons to provide comment on the proposals or hearing transcript until September
24, 2015. A total of over 3000 comment letters were received on the 2015 Proposal. There were
also over 300,000 submissions made as part of 30 separate petitions submitted on the proposal.
These comments and petitions came from consumer groups, plan sponsors, financial services
companies, academics, elected government officials, trade and industry associations, and others,
both in support and in opposition to the rule. After careful consideration of comments received,
the Department determined to grant the final exemption.
As finalized, the Best Interest Contract Exemption retains the core protections of the
proposed exemption, but with revisions designed to facilitate implementation and compliance
with the exemption’s terms. In broadest outline, the exemption permits investment advice
fiduciaries – both individual advisers and the financial institutions that employ them -- to receive
many common forms of compensation that ERISA and the Code would otherwise prohibit,
provided that they give advice that is in their customers’ best interest and they implement basic
protections against the dangers posed by conflicts of interest. In particular, to rely on the
exemption, financial institutions must:


Acknowledge fiduciary status with respect to investment transactions to the
retirement investor;



Adhere to “impartial conduct standards” requiring them to:


Give advice that is in the retirement investor’s best interest (i.e., prudent advice
that is based on the investment objectives, risk tolerance, financial circumstances,
and needs of the retirement investor, without regard to financial or other interests
of the adviser or financial institution);



Charge no more than reasonable compensation; and



Make no misleading statements about investment transactions, compensation, and
conflicts of interest;



Implement policies and procedures designed to ensure compliance with the impartial
conduct standards;



Refrain from giving or using incentives for advisers to act contrary to the customer’s
best interest; and



Fairly disclose the fees, compensation, and material conflicts of interest, associated
with their recommendations.

Individual advisers relying on the exemption must comply with the impartial conduct
standards when making investment recommendations.
More streamlined conditions apply to “level fee fiduciaries” that, with their affiliates,
will receive only a “level fee,” as defined in the exemption, that is disclosed in advance to the
retirement investor, for the provision of advisory or investment management services regarding
the plan or IRA assets. Level fee fiduciaries must provide written acknowledgment of fiduciary
status and comply with the impartial conduct standards when providing advice. When level fee
fiduciaries recommend a rollover from an ERISA plan to an IRA, a rollover from another IRA,
or a switch from a commission-based account to a fee-based account, they must document the

69

specific reason or reasons why the rollover was considered to be in the best interest of the
retirement investor.
The exemption neither bans all conflicted compensation, nor permits financial
institutions and advisers to act on their conflicts of interest to the detriment of the retirement
investors they serve as fiduciaries. Instead, it holds financial institutions and their advisers
responsible for adhering to fundamental standards of fiduciary conduct and fair dealing, while
leaving them the flexibility and discretion necessary to determine how best to satisfy these basic
standards in light of the unique attributes of their particular businesses. The exemption’s
principles-based conditions, which are rooted in the law of trust and agency, have the breadth
and flexibility necessary to apply to a large range of investment and compensation practices,
while ensuring that advisers put the interests of retirement investors first. When advisers choose
to give advice to retail retirement investors pursuant to conflicted compensation structures, they
must protect their customers from the dangers posed by conflicts of interest.
The following changes were made in the final PTE, among others:

141



The description of covered transactions was revised to make clear that prohibited
transactions arising from rollover or distribution advice, or advice regarding services,
is covered by the exemption if the conditions are satisfied;



The exemption was expanded to include advice to small participant-directed plans
and to include advice to all “retail fiduciaries,”141 not just plan sponsors and their
employees, officers, and directors;



As noted above, streamlined conditions have been provided for “level fee
fiduciaries.”



The definition of a limited category of covered “assets” has been eliminated in the
final exemption; thus, advisers and financial institutions using the exemption may
provide advice more broadly with respect to all securities and other investment
property if they comply with other safeguards adopted in the final exemption.



The conditions applicable to insurance companies and distributors of insurance
products are revised to make the exemption more usable and less costly with respect
to sales of annuities;



The exemption provides specific guidance on satisfaction of the Best Interest
standard by financial institutions and advisers that restrict recommendations, in
whole or part, to proprietary products or to investments that generate third-party
payments such as revenue sharing.



A written contract between an investor and financial institution only is required with
respect to advice regarding IRA investments and plans that are not covered by Title I
of ERISA, such as Keogh Plans.

A “retail fiduciary” is a fiduciary with respect to a plan or IRA (whether a natural person, corporation, partnership, trust, investment
committee, or otherwise), if such fiduciary is not a financial institution (including banks, insurance companies, registered investment
advisers and broker dealers) or a person that otherwise holds or has under management or control, total assets of $50 million or more.

70



The contract must be entered into before or at the same time as the execution of
the recommended transaction, not before advice is provided;



The contract for these investors must be maintained on the financial institution’s
website and be accessible by the investor;



The contract may be a master contract covering multiple recommendations and
can cover advice rendered before execution of the contract; the contract terms can
appear in a standalone document or in an investment advisory agreement,
investment program agreement, account opening agreement, insurance or annuity
contract or an application;



A contract is not required for ERISA-covered plans: however, the financial institution
must acknowledge fiduciary status for itself and its advisers in writing and comply
with the remaining conditions of the exemption;



For existing customers, the exemption permits customer assent to be evidenced by
either affirmative consent or a negative consent procedure;



The exemption does not require a chart illustrating the total cost of the recommended
investment for 1-, 5-, and 10- year periods expressed as a dollar amount; instead it
requires a disclosure focusing on the financial institution’s material conflicts of
interest, with more specific information to be provided upon request;



The website disclosure is also more general to reduce cost and burden and makes
clear that disclosure of compensation arrangements, not specific amounts of
compensation received, is required; a written description of the financial institution’s
policies and procedures, a model contract and disclosures must be maintained and
freely accessible to the public on the financial institution’s website;



Good faith provisions have been included to avoid loss of the exemption if the
financial institution, acting in good faith and with reasonable diligence, makes an
error or omission with respect to the required disclosures;



The annual disclosure requirement is eliminated;



The data request provision has been eliminated; and



Broader relief is provided for pre-existing investments so that additional investment
advice can be provided on all investments held prior to the applicability date.

The Best Interest Contract Exemption is effective 60 days after publication in the Federal
Register and will become applicable one year after the date of publication. The exemption
provides relief from the ERISA and IRC prohibited transactions rules prohibiting plan and IRA
fiduciary advisers from receiving compensation that varies based on their investment advice and
from third parties in connection with their advice. During the period between the Applicability
Date and January 1, 2018 (the “transition period”), full relief under the exemption will be
available for financial Institutions and advisers subject to more limited conditions than apply
after the transition period. The transition period is intended to provide financial institutions and
advisers with time to prepare for compliance with the conditions of the Best Interest Contract
Exemption, while safeguarding the interests of retirement investors.
The transition period conditions require the financial institution and its advisers to
comply with the impartial conduct standards when making recommendations to retirement
investors. The financial institution must additionally provide a written notice to the retirement
investor acknowledging its and its adviser(s) fiduciary status with respect to the recommended
71

transaction before or contemporaneously with the execution of the recommended transaction.
The financial institution also must state in writing that it and its advisers will comply with the
impartial conduct standards and describe its material conflicts of interest. Further, the financial
institution’s notice must disclose whether it recommends proprietary products or investments
that generate third-party payments. To the extent the financial institution or adviser limits
investment recommendations, in whole or part, to proprietary products or investments that
generate third-party payments, the financial institution must notify the retirement investor of the
limitations placed on the universe of investment recommendations. The disclosure may be
provided in person, electronically or by mail. It does not have to be repeated for any subsequent
recommendations during the transition period.
In addition, the financial institution must designate a person or persons, identified by
name, title or function, responsible for addressing material conflicts of interest and monitoring
advisers’ adherence to the impartial conduct standards. Finally, the financial institution must
comply with the recordkeeping provision of the exemption regarding the transactions entered
into during the transition period.
Similar to the disclosure provisions the transition conditions provide for exemptive relief
to continue despite errors and omissions with respect to the disclosures, if the financial
institution acts in good faith and with reasonable diligence.

2.9.2.2

Principal Transactions Exemption

Commenters responding to the 2010 Proposal also indicated that if the current regulation
is amended, the entities that would be newly defined as investment advice fiduciaries would
need exemptive relief for principal transactions between a plan or IRA and the investment advice
fiduciary. In this regard, both ERISA and the IRC prohibit a fiduciary from dealing with the
assets of the plan or IRA in his or her own interest or for his or her own account. ERISA further
prohibits a fiduciary from, in his or her individual or any other capacity, acting in any transaction
involving the plan on behalf of a party (or representing a party) whose interests are adverse to
the interests of the plan or the interests of its participants or beneficiaries. As a result, the
purchase or sale of a security in a principal transaction between a plan or IRA investor and an
investment advice fiduciary, resulting from the fiduciary’s provision of investment advice within
the meaning of 29 C.F.R. 2510.3-21 to the plan or IRA involves prohibited transactions under
ERISA and the IRC.
As part of the 2015 proposed regulatory package, the Department proposed relief for
principal transactions in certain debt securities between a plan or IRA and an investment advice
fiduciary where the principal transaction is a result of the provision of investment advice to a
plan or IRA by the investment advice fiduciary. The Principal Transactions Exemption, as
adopted, includes many of the safeguards included in the Best Interest Contract Exemption. The
exemption requires financial institutions to acknowledge in writing their fiduciary status and that
of their individual advisers with respect to the advice; adhere to impartial conduct standards,
including providing advice in retirement investors’ best interest; seeking to obtain the best
execution reasonably available under the circumstances; and making no misleading statements
about the transaction, compensation, and conflicts of interest; implement policies and procedures
designed to ensure compliance with the impartial conduct standards; refrain from giving or using
incentives for advisers to act contrary to the customer’s best interest; and make disclosures about
material conflicts of interest and the investment that is traded in the principal transaction.
Individual advisers relying on the exemption must comply with the impartial conduct standards
when making investment recommendations regarding principal transactions.
72

The following changes were made in the final Principal Transactions Exemption, among
others:


The final exemption covers two additional types of investments that can be sold to
plans or IRAs, certificates of deposit and unit investment trusts. For purchases from
plans or IRAs, the exemption would apply to all securities or other property.



The exemption does not require disclosure of the mark-up or mark-down on the
transaction, or of two comparable price quotes, as proposed. Instead, the financial
institution must seek to obtain the best execution reasonably available under the
circumstances with respect to the transaction.

The exemption is effective 60 days after publication in the Federal Register and will
become applicable one year after the date of publication. The exemption provides a transition
period under which relief from these provisions is available for financial institutions and
advisers during the period between the applicability date and January 1, 2018 (the “transition
period”). For the transition period, full relief under the exemption will be available for financial
institutions and advisers subject to more limited conditions. This period is intended to provide
financial institutions and advisers with time to prepare for compliance with the conditions of the
exemption while safeguarding the interests of retirement investors. The transition period
conditions are subject to the same exclusions for advice from fiduciaries with discretionary
authority over the customer’s investments and specified advice concerning in-house plans.
The transition period conditions require the financial institution and its advisers to
comply with the impartial conduct standards when making recommendations regarding principal
transactions to retirement investors. The financial institution must provide a written notice to the
retirement investor before the execution of the principal transaction acknowledging its and its
adviser(s) fiduciary status with respect to the recommended transaction. The financial institution
must also state in writing that it and its advisers will comply with the impartial conduct
standards, and disclose the circumstances under which the adviser and financial institution may
engage in principal transactions with the plan, participant or beneficiary account or IRA, and its
material conflicts of interest. The financial institution must comply with the recordkeeping
provision of the exemption for transactions entered into during the transition period.

2.9.2.3

PTE 75-1, Part V Amendment

An existing class exemption, PTE 75-1, Part V, provides relief for extensions of credit to
plans and IRAs by BDs. Under the exemption as originally granted, BDs who possessed or
exercised any discretionary authority or control (except as a directed trustee) with respect to the
investment of the plan assets involved in the transaction, or rendered investment advice with
respect to those assets, were not permitted to receive compensation in return for the extension of
credit. Commenters responding to the 2010 Proposal requested that the Department provide
exemptive relief for compensation for extensions of credit to a plan or IRA investor by
investment advice fiduciaries, because many BDs that have historically relied upon the relief
provided by PTE 75-1, Part V, would not be able to rely on such relief if they became
investment advice fiduciaries under the 2015 Proposal.
The Department amended PTE 75-1, Part V, by adding a new section that permits an
investment advice fiduciary to receive reasonable compensation for extending credit to a plan or
IRA to avoid a failed purchase or sale of securities involving the plan or IRA, subject to several
conditions. The potential failure of the purchase or sale of the securities may not be caused by
the broker-dealer or any affiliate.
73

Additionally, the terms of the extension of credit must be at least as favorable to the plan
or IRA as the terms available in an arm’s length transaction between unaffiliated parties.
Finally, the plan or IRA investor must receive written disclosure of certain terms prior to the
extension of credit. This disclosure does not need to be made on a transaction-by-transaction
basis, and can be part of an account opening agreement. The disclosure must include the rate of
interest or other fees that will be charged on such extension of credit, and the method of
determining the balance upon which interest will be charged. The plan or IRA must additionally
be provided with prior written disclosure of any changes to these terms.
The amended exemption is effective 60 days after publication in the Federal Register and
will become applicable one year after the date of publication.

2.9.2.4

PTE 86-128 Amendment

Another existing class exemption, PTE 86-128, provided relief for, among other things,
an investment advice fiduciary’s use of its authority to cause a plan or IRA to pay a fee to such
fiduciary or its affiliate for effecting or executing securities transactions. The exemption also
provided relief for an investment advice fiduciary to act as the agent in an agency cross
transaction for both the plan or IRA and one or more other parties to the transaction, and to
receive reasonable compensation therefor from one or more other parties to the transaction.
The Department amended PTE 86-128 to require that fiduciaries relying on the
exemption comply with the impartial conduct standards. As amended, PTE 86-128 also includes
a new covered transaction that permits certain fiduciaries that are BDs (and who are not the
principal underwriter for or affiliated with a mutual fund) to use their authority to cause plans to
purchase mutual fund shares from the fiduciary and receive a commission. Relief for this
transaction is currently available in a different class exemption, PTE 75-1, Part II (2), which the
Department is revoking as of the applicability date.
The Department also amended PTE 86-128 to eliminate relief provided by PTE 86-128
for investment advice fiduciaries to IRAs. The amendment reflects the Department’s view that
the conditions of the Best Interest Contract Exemption provide more appropriate safeguards of
these investors in connection with the transactions.
The amended exemption is effective 60 days after it is published in the Federal Register
and will become applicable one year after the date of publication.

2.9.2.5

PTE 84-24 Amendment

PTE 84-24 permitted insurance agents, insurance brokers and pension consultants to
receive, directly or indirectly, a commission for selling insurance or annuity contracts to plans
and IRAs. PTE 84-24 also permitted an investment company’s principal underwriter to receive
commissions in connection with a plan’s or IRA’s purchase of investment company securities.
The Department is finalizing its proposal to amend PTE 84-24 to require all fiduciaries
relying on the exemption to adhere to the same impartial conduct standards required in the Best
Interest Contract Exemption. At the same time, the amendment revokes PTE 84-24 in part so
that investment advice fiduciaries will not be able to rely on PTE 84-24 with respect to (1)
transactions involving annuities other than “fixed rate annuity contracts” as defined in the
exemption, and (2) transactions involving the purchase by IRAs of investment company shares.
Investment advice fiduciaries will be able to rely instead on the Best Interest Contract
Exemption for compensation received in connection with these transactions. Fixed rate annuity
contracts, as defined in the exemptions, do not include variable annuities, indexed annuities and
74

similar annuities. The Department believes that investment advice transactions involving
variable annuities, indexed annuities, and other similar annuities, as well as transactions
involving the purchase of investment company shares by IRAs, should occur under the
conditions of the Best Interest Contract Exemption because that exemption provides more
appropriate safeguards in connection with the transactions. Investment advice fiduciaries can
continue to rely on the exemption for receipt of commissions with respect to transactions
involving all insurance contracts and fixed-rate annuity contracts, and the receipt of
commissions with respect to ERISA plan purchases of investment company shares, but they
would be required to comply with all of the protective conditions described above.
The final amendment to PTE 84-24 also requires the fiduciary engaging in a transaction
covered by the exemption to maintain records necessary to enable the Department (and certain
persons described in the amended exemption) to determine whether the conditions of the
exemption have been met. This requirement would replace the more limited existing
recordkeeping requirement that existed prior to the amendment.
The amended exemption is effective 60 days after it is published in the Federal Register
and will become applicable one year after the date of publication.

2.9.2.6

Final Amendments to PTEs 75-1, Part III, 75-1, Part
IV, 77-4, 80-83, and 83-1

The Department is finalizing the amendments to PTEs 75-1, Part III, 75-1, Part IV, 77-4,
80-83, and 83-1 as proposed. These exemptions provide the following relief:


PTE 75-1, Part III permits a fiduciary to use its authority to cause a plan or IRA to
purchase securities from a member of an underwriting syndicate other than the
fiduciary, when the fiduciary is also a member of the syndicate;



PTE 75-1, Part IV permits a plan or IRA to purchase securities in a principal
transaction from a fiduciary that is a market maker with respect to such securities;



PTE 77-4 provides relief for a plan’s or IRA’s purchase or sale of open-end
investment company shares where the investment adviser for the open-end investment
company is also a fiduciary to the plan or IRA;



PTE 80-83 provides relief for a fiduciary’s use of its authority to cause a plan or IRA
to purchase a security when the proceeds of the securities issuance may be used by
the issuer to retire or reduce indebtedness to the fiduciary or an affiliate;



PTE 83-1 provides relief for the sale of certificates in an initial issuance of
certificates, by the sponsor of a mortgage pool to a plan or IRA, when the sponsor,
trustee or insurer of the mortgage pool is a fiduciary with respect to the plan or IRA
assets invested in such certificates.

Each of these exemptions is being amended to incorporate the Impartial Conduct
Standards set forth in the Best Interest Contract Exemption.
The amended exemptions are effective 60 days after they are published in the Federal
Register and will become applicable one year after the date of publication.

2.10

Reform Abroad

Since the global financial crisis of 2007-08 there has been an international regulatory
trend focusing on eliminating or mitigating the conflicts of interest inherent in the compensation
75

paid to advisers to improve the quality of advice and the long-term success of the financial
services market. Many countries first attempted to use disclosure-based regulatory regimes to
provide transparency to actual, potential, or perceived conflict of interest risks to clients but
concluded that disclosure, although a necessary part of mitigating these risks is not sufficient. A
number of countries from around the globe have gone much further than the
Department’s final rule and exemptions by banning commission payments, increasing
professional standards form advisers, and adopting a new best interest standard that more
broadly embraces the direct fee-for-service model in financial advice.142
As Regulators in several countries identified failures in their investment advice markets,
they have undertaken a range of regulatory and legislative initiatives that directly address
conflicted investment advice. The data show that the traditional commission model is in decline
in many countries although the trajectory and extent varies by country to country.143 Some
countries introduced or finalized regulations with measures that ban or strictly limit certain thirdparty payments and increase transparency in the system to improve consumer protection. The
table below, from the White House Council of Economic Advisers report issued in February
2015, portrays recent international regulatory changes addressing conflicted advice across the
globe:
Figure 2-3 Reform Abroad
Country

Description

Australia

Banned payments from product providers and conflicted remuneration payments for retail investments
and created a statutory duty for advisors to act in the best interest of their clients.

Canada

New regulations, implementation of which began in 2014, require much greater transparency about the
direct and indirect costs to the client for each account and details on advisor compensation by clients and
product providers.

India

Banned all front loads for mutual fund products beginning in 2009. Implemented heightened
requirements to disclose the value and justification for any commission payments to advisors.

142

143

In 2015, the Department commissioned RAND to examine the existing market practices for the provision of financial advice and the
regulatory frameworks which address conflicts of interest for financial advisers in order to identify effective approaches to limit biased
financial advice and its negative effects in several countries. Consequently, in August, 2015, RAND published "Financial Advice MarketsA Cross Country Comparison" which compares the financial advice markets in the United States, the United Kingdom, Australia, Germany,
Singapore, and the European Union that recently made regulatory changes aimed at improving financial advice and how the regulatory tools
used have affected their respective financial advice markets.
In sum, the report depicts considerable variation in the financial regulatory environment around the world following the financial crisis of
2007–2008 and in efforts aimed at mitigating conflicts of interest to improve the quality and suitability of advice provided to retail
investors. In contrast to the United States, many countries like the UK and Australia have taken a more stringent approach to adviser
remuneration by placing outright bans on certain commissions to help align incentives between advisers and their clients. In other countries
across the European Union and in Germany in particular, recent and impending legislation has sought to promote improved advice by
creating classes of advisers that are to be compensated solely on a fee basis. Early research into the RDR “provides suggestive evidence
that the regulation has reduced the amount of bias present in advice—fund flows into high-charging share classes have decreased
substantially, while flows into low-cost index funds have grown.” There is also suggestive evidence indicating that the cost of financial
advice may have increased modestly. In addition there is conflicting evidence on whether the RDR has led to an “advice gap,” as in some
cases lower-wealth clients may now find it more difficult to receive advice. However, the RAND report concludes that there is evidence
suggesting that the number of low-wealth clients who lost access to advice may be small. Due to the recent regulatory changes in these
countries there is only preliminary evidence about the impact of those changes on consumers in the long run.
For a more detailed review, the RAND report can be found on the Department’s website at:
http://www.dol.gov/ebsa/regs/conflictsofinterest.html#additionalresearchpapers.
“The Effects of Conflicted Investment Advice on Retirement Savings,” Council of Economic Advisers, (Feb. 2015).

76

Italy

Banned commissions for discretionary portfolio management services beginning in 2007.

Germany

Increased disclosures about the cost of advice and whether the advisors are compensated solely through
client fees or by payments from service providers.

The Netherlands

Banned all payments by a product issuer to an advisor relating to advice beginning in 2013. The ban
applies to investment, insurance, and mortgage protection (annuity) products.

United Kingdom

Banned conflicted payments, increased education and credential standards, and required advisors to
disclose whether they make recommendations from a restricted menu of products or across all products
beginning in 2013.

Source: Table from Council of Economic Advisers, "The Effects of Conflicted Investment Advice on Retirement Savings" (Feb. 2015)
(available at: https://www.whitehouse.gov/sites/default/files/docs/cea_coi_report_final.pdf), which cites as sources: BlackRock
Viewpoints, "The Changing Face of European Distribution: A Better Financial Future for Savers?" (2014); Australian Securities &
Investment Commission, "FOFA - Background and implementation" (2015); and Financial Conduct Authority, “Retail Distribution
Review (RDR)” (London, 2014).

The Department’s regulatory initiative represents a middle ground between no reform
and the outright bans on conflicted payments implemented in many countries as outlined above.
The Department’s approach allows businesses to continue to use a wide range of compensation
practices while minimizing the harmful impact of conflicts of interest on the quality of advice.
Advisers and financial institutions that opt to continue to receive compensation that would
otherwise be prohibited must adopt a new best interest standard and enact policies and
procedures to manage and mitigate the harmful impact of conflicted investment advice.
Two of the most far-reaching initiatives have occurred in the United Kingdom (UK),
where the Financial Conduct Authority (FCA) (formerly, the Financial Services Authority)
issued new regulations that were effective on January 1, 2013, called the Retail Distribution
Review (RDR) and in Australia, where the government adopted significant regulatory changes
under the Future of Financial Advice (FOFA) Act. These are examples of new regulatory
regimes with more transparent fee-for-service compensation structures.
The Department’s regulatory efforts and the RDR differ in scope. The UK enacted a
much more aggressive reform with an outright ban on commissions affecting all retail
investment products, not just those related to retirement savings.144 In addition, the RDR
required that customers in the UK be charged directly for advice and raised qualification
standards for advisers.
Early evidence from the UK indicates that regulatory changes that ban commissions
entirely have not resulted in consumers being abandoned by their financial advisers. Much of
the evidence presented in this report shows that despite a small reduction in adviser numbers,
firms have adapted successfully to the post-RDR world. Looking at all of the evidence and
results, the RDR has achieved lower costs for customers, hasn’t resulted in the sizable advice
gap that the advisers feared, and did not cause a large exit of advisers. While some advisers left
the market, overall availability of advice does not appear to have been significantly reduced, and

144

"Non–advised" services, or execution-only sales, where no advice or recommendation is given, falls outside of the RDR. Thus, a
commission is still permitted for non-advised annuity sales. The FCA is currently examining the risks that exist with the purchase of ‘nonadvised’ annuities. Please see: http://www.fca.org.uk/static/documents/consultation-papers/cp15-30.pdf.

77

some of those that left have since returned to the market.145 On balance, the UK’s experience
lends support of the Department’s conclusion that its reforms, which do not ban commissions or
increase adviser qualifications, are unlikely to result in a significant diminution of advice. The
Department does anticipate some transitional issues as firms compete in a new environment and
as the cost of advice becomes clearer to customers. As in the UK, the Department will continue
to watch these transitional issues carefully to see if additional guidance, assistance, or other
further action is needed.
Early evidence on Australia’s FOFA also suggests an improvement in transparency and
fairness in the financial advice industry. Under FOFA, advice given to Australians must be in
the best interests of the client and clients are given the opportunity to choose and agree on fees
up front.146

2.10.1

The UK Retail Distribution Review

The RDR is aimed at introducing more transparency and fairness into the investment
industry in the UK, reducing conflicts of interest and allowing clients to see how much advice is
costing them and, in turn, understand what benefit they derive from it. 147 The most significant
change is that financial advisers are no longer permitted to earn commissions in return for selling
recommending investment products. Instead, investors now have to agree on the fees148 for the
advice up-front. In addition, financial advisers now have to offer either "independent” or
“restricted” advice and explain the difference between the two – essentially making clear
whether their recommendations are limited to certain products or product providers.149 The
RDR eliminated commissions broadly for both retirement and non-retirement accounts.

2.10.1.1 The Problem
The FCA began working on the RDR in June of 2006 to address persistent problems that
emerged in the UK retail investment market. These include a series of commission-based misselling scandals by UK banks over a period of more than 20 years regarding sales of unsuitable
products, including pensions, as well as other problems concerning product and provider bias,
churning of products, and lack of access to financial advice.
The FCA was also concerned that (1) the commission-based compensation model
incentivized advisers to sell products whose providers paid them the largest commissions rather
than products that were in their clients’ best interests, and (2) the lack of fee transparency hid the
true cost of advice from consumers.

145

146
147

148

149

Andrew Hogg, “FSB to Financial Advisors: Retail Distribution Review Will Help, Not Hurt You, Biznews (Jan. 2015); available at:
http://www.biznews.com/interviews/2015/01/30/fsb-financial-advisers-retail-distribution-review-will-help-not-hurt/
Senate Inquiry into the Scrutiny of Financial Advice, Submission by the Australian Securities and Investments Commission, (Dec. 2014).
The Department consulted with staff of the UK’s FCA in drafting this section. As part of this consultative process, FCA staff provided
technical assistance to the Department in support of its efforts to accurately describe the RDR provisions and their market impact to date.
The RDR requires firms to work out an appropriate charging structure for calculating the adviser charge and provide a copy of this to the
client in writing before providing advice, rather than calculating a tailor-made charge for each client (Conduct of Business Source Book
(COBS) 6.1A.17R; available at: http://fshandbook.info/FS/html/handbook/COBS/6/1A). Whether the charging structure is based on a fixed
fee, an hourly rate or a percentage of funds invested will be up to the firm, provided it always bears in mind its duty to act in the client’s
best interests. When adviser charges vary inappropriately by the provider or product the best interest rule is not being met. Thus, firms are
not able to charge more for recommending one particular product instead of another substitutable product. Firms must base their charges on
services they provide rather than on the type of products they sell.
Financial Services Authority (FSA), “Retail Distribution Review: Independent and Restricted Advice” (June 2012).

78

2.10.1.2 The Rule
The FCA worked extensively with the financial services industry and other stakeholders
to identify areas that should be addressed by the RDR. After these consultations, the FCA
developed three broad objectives for the RDR: (1) provide a clear definition of independent
advice; (2) address the potential for remuneration bias; and (3) increase professional standards.
The RDR achieves these objectives by requiring “Independent Advisers” to: (1) consider
a broad range of products and (2) provide unbiased and unrestricted advice based on a
comprehensive and fair analysis of the relevant market. “Restricted Advisers” - those who
provide advice with respect to a limited range of products or providers - are required to meet the
suitability requirements150 for the advice that Independent Advisers must follow. Therefore,
Restricted Advisers cannot recommend a product that most closely meets a client’s needs from a
restricted range of products if that product is not suitable for the client. If advice is not
independent, then it must be described as restricted. This label covers firms that advise on their
own products or on a limited range of products, such as bank advisers and other single-tied and
multi-tied adviser firms.
The RDR requires Independent and Restricted Advisers to: (1) explicitly disclose and
separately charge clients for their services (this means that commission payments to advisers
will cease); (2) disclose to their clients whether they are providing independent or restricted
advice; (3) subscribe to a code of ethics; (4) have appropriate qualifications; (5) carry out at least
35 hours of continuing professional development annually; and (6) hold a Statement of
Professional Standing from an accredited body.
As stated above, the RDR prohibits financial advisers from receiving commissions when
they advise clients to invest in a product. Instead, they must charge a fee, expressed either as a
percentage of the amount invested, a fixed fee, or an hourly rate. Whether the charging structure
is based on a fixed fee, an hourly rate, or a percentage of funds invested will be up to the firm,
provided it always bears in mind its duty to act in the client’s best interests. The client should
only pay ongoing charges if the firm is providing an ongoing, value-added service, the details of
which have been properly disclosed to the client.151 The fee can be paid directly by the client or
can be taken from a product that they invest in, provided that the client knows exactly what the
charges are up front. The rules provide exceptions, however, in situations where a client
purchased a retail investment product before January 1, 2013. In such cases, the adviser can
continue to receive ongoing “trail commissions” in relation to the pre-RDR advice until the

150

151

Suitability is a well-established regulatory concept for the UK financial services industry. Principle 9 of the FCA’s Principles for Businesses
(PRIN 2.1) requires firms to take reasonable care to ensure the suitability of their advice and discretionary decisions for any customer who
is entitled to rely upon their judgment. The Conduct of Business Sourcebook defines the FCA’s rules and guidance on suitability. The
suitability requirements seek to ensure that, where firms provide investment advisory or portfolio management services, they obtain enough
information about their customers to be able to act properly for them, and that the business conducted for them, or on their behalf, is
appropriate to their circumstances. Failure to obtain all the relevant information, or evaluate it properly, can lead to the recommended
transaction or decision to trade being unsuitable. PRIN 2.1 is available at: http://fshandbook.info/FS/html/FCA/PRIN/2/1. This appears to
be a similar standard to FINRA Rule 2111, which establishes a “suitability” standard of conduct for BDs, which requires them to “have a
reasonable basis to believe that a recommended transaction or investment strategy involving a security or securities is suitable for the
customer, based on the information obtained through the reasonable diligence of the [firm] or associated person to ascertain the customer’s
investment profile.” See FINRA Manual, Rule 2111.
FCA “Conduct of Business Sourcebook” (COBS) 6.1A.22R; 6.1A.26G; available at: http://fshandbook.info/FS/html/handbook/COBS.

79

product is changed, matures or is terminated.152 Additionally, execution-only sales (where no
advice or recommendation is given) also fall outside the adviser charging regime.

2.10.1.3 The RDR’s Effects
The RDR appears to have achieved much of what it was designed to accomplish: adviser
recommendations are no longer influenced by commissions paid by product providers; advisers
are now better qualified; product prices have fallen in some areas as a result of more effective
competition in the market; and the costs to firms complying with the RDR have been in line with
or lower than expectations.153 According to one major final report from Europe Economics:154


The RDR has initiated a move towards increased professionalism among advisers.



The ban on third-party commissions has reduced product bias.



Consumers are increasingly shopping around for financial services.



Charges for retail investment products have been falling.



Initial evidence indicates that advisory firms appear slightly better placed to meet
their long-term commitments.



Costs of complying with the RDR have been in line with or lower than expectations.



The market is adjusting to offer advice which is more tailored to consumers’
demands.



Any advice gaps are questionable, not attributable to the RDR, or small and will be
resolved by the market.



The RDR has created an opportunity for innovation, and there are signs innovation is
coming, though [as of December 2014] actual innovation has been limited.



Those consumers receiving full advice are now receiving better advice due to
improved adviser qualifications and reduced adviser bias.



Disclosure has improved, but more improvement is needed to help consumers
understand whether advice is independent.

The Europe Economics report also states that advisers have capacity and have been
taking on new clients. According to the report, it appears that in the year ending March 31,
2014, while advisers dropped about 310,000 clients whom they no longer found profitable to
serve, they picked up a total of 820,000 clients. According to the authors, the net increase in
customers served suggests that dropped clients who looked for replacement advisers were
largely successful. Thus, the industry appears to have adjusted to the RDR’s commission ban
and as a result clients were essentially moved and served by other advisers.
The Europe Economics report also points out that by revealing the true cost of advice, the
RDR has led some consumers to consider the extent to which the advice they receive represents

152

153
154

A definition of “trail commission” can be found on the FCA’s website at: http://www.fca.org.uk/consumers/financial-servicesproducts/investments/financial-advice/trail-commission.
“The Financial Advice Market Review” (Oct. 2015), Financial Conduct Authority and HM Treasury.
Europe Economics, “Retail Distribution Review,” Post Implementation Review, (Dec. 16, 2014).

80

value for money, and in some cases conclude it does not. To the extent that there is demand
from some consumers for lower-cost advice not currently offered by the market, this demand
also existed pre-RDR.155 This group includes consumers who would be likely to pay for a
cheaper form of advice, for example, that which may be provided by a simpler advice model.
According to the report, there is little evidence that consumers perceive themselves to have been
abandoned by advisers.
The FCA engaged in a three-stage thematic review to assess investment advisory firms’
approaches to implementing the RDR. The first stage was completed in July 2013, and
concluded that the majority of firms had made progress and there was a willingness among them
to adapt to the new rules.156 The second stage of the FCA’s thematic review assessed how firms
had implemented the RDR in terms of whether firms that were describing themselves as
independent were acting independently in practice and whether firms were complying with the
disclosure requirements.157 The FCA found a high proportion of firms were failing to correctly
disclose the cost of their advice to clients or the type of services they offered, and many were not
disclosing the ongoing services they provided as required by the RDR. The FCA found the level
of noncompliance with the disclosure requirements “disappointing” and stated that the failure of
firms to meet their regulatory requirements was “unacceptable” and could lead to poor outcomes
for consumers, as some consumers could be left unaware of the true cost of advice (both initial
and ongoing) which would undermine their ability to make informed choices.158
The FCA third cycle of the thematic review, completed in December 2014, focused on
an assessment of firms’ adviser fees and disclosures and how firms were delivering these
services to clients in the UK in practice. The FCA sampled 110 firms to provide a representative
sample of firms across the financial advice sector and to ensure the results were robust. Almost
all of the 110 firms it reviewed offered their clients a type of ongoing service in exchange for an
ongoing adviser charge. In around half of firms the regulator reviewed, over 90 percent of their
clients were paying to receive an ongoing service.
Overall, the results of the third thematic review were positive and show material
improvements in how firms are complying with the RDR, including how they disclose the cost
of their advice, their scope of service, and the nature of their services to clients. The findings
demonstrate that the sector has responded to the two previous thematic reviews which found
significant issues with the quality of the information given to those seeking advice. The
improvements point to increasing professional standards and should mean those seeking advice
are better placed to understand the nature of a firm’s services and how much they will cost.
Specifically, the FCA found in the December 2014 review that due to the RDR’s higher level
qualification standards, the professionalism of advisers is increasing in the financial sector and
there was a material improvement in the way firms disclose the cost of their advice to clients.159

155
156

157

158
159

The Financial Advice Market Review, Oct. 2015, Financial Conduct Authority and HM Treasury.
Thematic Review (TR) 13/5, “Supervising Retail Investment Advice: How Firms are Implementing the RDR” (July 2013); available at:
https://www.fca.org.uk/static/documents/thematic-reviews/tr13-05.pdf.
TR 14/5, “Supervising Retail Investment Firms: Delivering Investment Advice” (March 2014); available at:
https://www.fca.org.uk/static/documents/thematic-reviews/tr14-05.pdf; and TR 14/6, “Supervising Retail Investment Firms: Being Clear
about Adviser Charges and Services” (April 2014); available at: https://www.fca.org.uk/static/documents/thematic-reviews/tr14-06.pdf.
TR 14/6, “Supervising Retail Investment Firms: Being Clear about Adviser Charges and Services” (April 2014).
TR 14/21, Retail Investment Advice: Adviser Charging and Services” (Dec. 2014); available at:
http://www.fca.org.uk/static/documents/thematic-reviews/tr14-21.pdf.

81

However, the review did show that some further improvements are needed, particularly in the
way that costs, in cash terms, of ongoing services are disclosed.

2.10.1.4 Cost of Investing
There have been positive impacts on the types of investments individuals are making
(e.g., index funds), and there are at least some indications that the demand for new lower-cost
models is increasing. Several commenters, however, stated that the RDR may have increased
the cost of financial advice in the UK. One commenter specifically noted from the RAND
Cross-Country Comparison Report that there is suggestive evidence that some investors now pay
0.5 percent to 1 percent in ongoing charges compared to pre-RDR trail commissions typically in
the range of 0.5 percent to 0.75 percent. The Europe Economics report also found there is still
confusion among consumers regarding how advisers charge for advice, stating that the
difference in charging structures has continued to confuse customers. However, the report also
found that the charges for retail investments have been falling post-RDR, although adviser
charges have not. The report speculated that higher adviser charges are likely due to limited
competition in the advice market and limited consumer awareness and understanding of adviser
charging, which limits consumers’ ability to shop around and exert downward pressure on
prices. Europe Economics notes that product prices have fallen by at least the amounts paid in
commission’s pre-RDR - and there is evidence some could have fallen even further. Despite
falling product prices, the report states that adviser charges have increased post –RDR at least
for some consumers. The markets are adjusting and more time may be needed for a complete
evaluation of the full impact of the RDR on adviser charges.
To illustrate how product prices are falling, the chart below shows how the introduction
of the RDR at the end of December 2012 corresponded with a move towards the use of less
costly share classes.160 The adoption of lower-charging share classes has gathered pace ever
since. By the end of May 2014, over 80 percent of flows were being directed into lowercharging share classes as opposed to the highest.

160

Investment Management Association, “Asset Management in the UK 2013–2014: The IMA Annual Survey,” London: Investment
Management Association, (Sept. 2014).

82

Figure 2-4 Gross Retail Flows Through Highest-Charging Share Classes
and Other Share Classes

Source: IMA Asset Management Survey Report 2013

The FCA has also recorded positive results on fund prices. As the graph below indicates,
sales of low-cost tracker funds (an index fund that tracks a broad market index or a market
segment) and investment trusts, which did not pay commission before the RDR, increased as of
quarter two in 2014. During that same period sales of high-commission paying bonds fell.

1,200

12%

1,000

10%

800

8%

600
400

6%
4%

200

2%

0

0%

Tracker fund net sales (£m)

Percentage of industry total

Source: IMA data

83

Percentage of industry total

Net sales (£m)

Figure 2-5 Tracker Sales

Figure 2-6 Decline in Investment Bonds

Source: Association of British Insurers (2014). Note data refer to ABI members only.

Evidence shows that from 2011 through the first half of 2015, passive-investment tracker
fund assets increased approximately 140 percent, and market share significantly increased from
about 7.4 percent to over 12 percent.161 This report states that “the increased transparency in
investment product fees, diminished influence of trail commissions from actively managed
products, and the wider adoption of investment platforms” could explain the increased
investment in passively-managed funds.162

2.10.1.5 Access to Advice
Several comments on the Department’s 2015 Proposal cite reports that the number of
advisers in the UK has declined as evidence of a UK advice gap.163 The numbers cited in these
reports, however, neglect other reports that find that numbers have rebounded, and obscure
evidence that there is sufficient advisory capacity and evidence that advisers are available to
serve even small investors.164

161

162
163

164

Morningstar, Financial Services Observer, “The US Department of Labor’s Fiduciary Rule for Advisors Could Reshape the Financial
Sector” (Oct. 2015); available at: http://www.advisor.ca/wp-content/uploads/2015/11/FinancialServicesObserver_DOL-Oct2015.pdf.
Ibid.
For example, a study entitled “Challenge and Opportunity: The Impact of the RDR on the UK’s Market for Financial Advice” by the Cass
Business School; (available at: http://www.cassknowledge.com/sites/default/files/article-attachments/bny-mellon-rdr-cass-knowledge.pdf)
reports that the number of UK financial advisers fell by 25 percent during the first year following adoption of the RDR. Some comments
point to more recent reports that the number of advisers declined from 40,000 to 31,000 between 2011 and January 2014 (FCA Professional
Standards data, as reported in Association of Professional Financial Advisers (APFA), 2014) while the estimated number of advisers
working at banks dropped from about 8,600 to 3,600 over the period (APFA, 2014. See also CFA Institute, “Restricting Sales Inducements:
Perspectives on the Availability and Quality of Financial Advice for Individual Investors,” Code, Standards and Position Papers, Vol. 2013,
No. 15, (Dec. 2013); available at: http://www.cfapubs.org/doi/pdf/10.2469/ccb.v2013.n15.1). Other comments cite a report that the number
of advisers offering professional advice fell from around 26,000 in 2011 to 24,000 in 2014 (Financial Advice Market Review: Call For
Input, October 2015, HM Treasury, UK). (This number, however, is not fully representative of the total of adviser numbers offering
professional advice because it has excluded certain entities, such as banks.)
For example, see Europe Economics, “Retail Distribution Review,” Post Implementation Review, (Dec. 2014), Towers Watson, “Advice
Gap Analysis: Report to the FCA” (Dec. 2014) and The Financial Advice Market Review, Call for Input, Financial Conduct Authority and
HM Treasury (October 2015).

84

Although there was a decrease in the number of financial advisers in the UK, the FCA
had previously predicted this result based on their expectations of the impact of the RDR due to
a number of significant factors. According to a letter from the UK’s FCA to the Department, a
substantial decrease in the number of financial advisers did not occur and the decline was in line
with the FCA’s expectations.165 These types of frictional and transitional issues (e.g., some
people who are unwilling to pay the now transparently-priced fee for advice) and issues of
matching people up with those advisers that are available in the market were predictable due to
the stringency of the reform. In addition, the new professional qualifications standard required
of advisers has likely caused some advisers to leave the industry. When Cass Business School
surveyed financial advisers and asked what factors had led to the declining number of advisers,
the most frequent response indicated that “an inability to meet the minimum standard for
professionalism, the QCF Level 4 qualification, would be the main reason why advisers might
choose to leave the industry.”166 Those advisers near retirement were particularly likely to exit.
In fact, as several commenters point out, the Cass Report also states that “even without the RDR,
the landscape for the advisory sector would have begun to change” as “[t]echnological advances
have been making the creation and delivery of investment products more accessible and cheaper
to a wider audience, whether guided by an advisor or not.” The report found that “[t]he industry
was already shrinking pre-RDR.”167 Moreover, an October 2015 report issued by Morningstar
states that when looking at the subgroup of advisers who indicated that they planned on leaving
the industry after the RDR was implemented, a survey done on behalf of the U.K. Financial
Services Authority found that 40 percent cited the professionalism requirement as a material
factor in their leaving.168
A few comments noted that UK investors were able to access advice through large banks
but that after the RDR was passed, several large UK banks that provide investment advice and
products exited the market because they determined it was too costly to service small investors,
while several banks which remained significantly increased their minimum investment amount
thresholds before offering investment advice. However, as the FCA indicated to the Department
in a letter dated February 10, 2014, banks have not been major players in the UK advice market
(contrary to many other European countries) and the majority of advice is provided by
independent financial advisers.169 According to the FCA, several of the banks had been fined in
recent years for failings in their advice arms (problems were exposed in the FSA’s mystery
shopping exercise in 2012)170 and at least one had publicly-cited issues with commercial
viability in this market. The FCA stated that it is considered likely that the closure of banks’
advice arms were largely strategic, rather than as a result of the RDR changes. For example, at

165
166

167
168

169
170

Letter from David Geale, FCA, UK to Joseph Piacentini, U.S. Department of Labor (2014).
Cass Business School, “Challenge and Opportunity: The Impact of the RDR on the UK’s Market for Financial Advice” (June 2013). The
report states on page 11 that 47 percent of survey participants gave this response. On the other hand, Figure 1.4 on page 12 shows that 16
percent of respondents cited this reason. Both discussions indicate that this reason is the most common reason, a finding which is
emphasized in the descriptive statements summarizing the survey’s findings.
Ibid.
Morningstar, Financial Services Observer, “The US Department of Labor’s Fiduciary Rule for Advisers Could Reshape the Financial
Sector” (Oct. 2015).
Letter from David Geale, FCA, UK to Joseph Piacentini, U.S. Department of Labor (2014).
The FSA has published the results of the Mystery Shopping Review, carried out between March and September 2012, looking into the
quality of investment advice given by banks and building societies by focusing on the quality of advice given to customers looking to invest
a lump-sum. According to the findings, one-quarter of investment advice given by banks and building societies is of questionable quality,
with customer suitability not being properly assessed and evidence unsuitable advice being routinely given, according to a FSA mystery
shopping review; available at: http://www.fsa.gov.uk/static/pubs/other/thematic_assessing_retail_banking.pdf.

85

least one large bank exited the advice market even before the legislation was passed “citing ‘a
decline in commercial viability for such services over recent years.’”171 The report issued by
Europe Economics in December of 2014 notes that there are some indications that a number of
banks are looking to re-enter the market, perhaps with more technology-supported
applications.172
Some commenters point to reports that advice has become more expensive and/or
minimum account balance requirements have increased in the UK.173 But this evidence appears
to be mostly anecdotal. Other evidence demonstrates that advice remains broadly available.
Recent data suggest that a number of UK banks are returning to the market but also with
revamped client services with lower minimum investible assets requirements.174 One large UK
bank is planning to offer a stand-alone investment service later in 2016, which will allow its
customers - including those with less than £50k - to meet with an advisor to discuss specific
investment needs.175
FCA-commissioned research found that most retail investment advisers continue to serve
clients with savings and investments between £20,000 and £75,000 and that a third of advisers
continue to serve clients with less than £20,000. The FCA noted that the emergence of new
ways to access advice using online technology has the potential to offer those with small
amounts to invest an efficient and cost-effective means to receive advice.176 A 2014 Towers
Watson report indicated demand for around 25,000 individual advisers, compared with estimates
of around 30,000 financial advisers currently active in the market, although supply and demand
may not be perfectly aligned across the market. According to the report, adviser business
models are likely beginning to adapt to meet and service the transactional demand that exists;
otherwise, a much faster reduction in the number of advisers would have been visible due to
declining prices and profitability resulting from excess supply.177 A 2014 NMG Consulting
research report stated that 83 percent of surveyed advisers indicated they had capacity to advise
additional clients seeking guidance on pension decumulation and only 19 percent claimed they
would not advise on accounts below a certain threshold, while 50 percent stated it would depend
on the particular case.178 Another NMG report found that the RDR had little impact on

171
172
173

174
175
176
177
178

Letter from David Geale, FCA, UK to Joseph Piacentini, U.S. Department of Labor (2014).
Europe Economics, “Retail Distribution Review,” Post Implementation Review, (Dec. 16, 2014).
For example, a number of commenters cite data from an August 2014 Morningstar article reporting that “[e]leven million investors consider
financial advice too expensive and have fallen through the advice gap following the industry regulation” and that “some investors prefer not
to have an upfront cost for financial advice –as this prices them out of the advice market” (Morningstar, “10 Million Find Advice Too
Expensive.” (Aug. 28, 2014)). Other commenters point to a study by Fidelity Worldwide and Cass Business School that suggests that the
“average level of investible assets needed to make a consumer commercially viable to an adviser is approximately £61,000” (approximately
$110,000 USD). See Professor Andrew Clare, “An Investigation of the UK’s Post-RDR Savings and Investment Landscape”, (Jan. 2013);
available at: https://www.cass.city.ac.uk/__data/assets/pdf_file/0014/202316/The-guidance-gap-report-Cass-version.pdf. Moreover, some
commenters argue that the minimum account size might even need to be higher than this amount in order to receive access to financial
advice and quote the former CEO of the FCA as stating that “people who have… below £50,000 or £100,000 [approximately $78,000$156,000 USD] are not getting the same service they were getting” prior to the RDR. See: Michelle Abrego, “FCA Chief Wheatley Admits
Concerns Over Advice Gap,” Citywire, (Sept. 2013); available at: http://citywire.co.uk/new-model-adviser/news/fca-chief-wheatley-admitsconcerns-over-advice-gap/a701957.
Valentina Romeo and Natalie Holt, “The Return of Bank Advice: Will Things be Different This Time,” Money Marketing, (Jan. 14, 2016).
Santander UK to Re-enter Investment Advice Market, Emma Dunkley, (Jan. 4, 2016).
The Financial Advice Market Review, Call for Input, Oct. 2015, Financial Conduct Authority and HM Treasury.
Towers Watson, “Advice Gap Analysis: Report to the FCA” (Dec. 5, 2014).
APFA, “The Advice Market Post RDR Review,” June 2014, (citing NMG Consulting, Financial Adviser Census for APFA, “The Guidance
Guarantee” May 2014); available at: http://www.apfa.net/documents/publications/APFA-report-the-advice-market-post-RDR-June2014.pdf.

86

consumers’ desires to use advice and among consumers it appears that the RDR and adviser
charging do not have a direct impact on the likelihood to seek advice.179
Moreover, it is likely that any advice gap could substantially be addressed by the market
through online platforms, advances in technology, and firms developing new simplified advice
models which are more cost-effective. These include self-directed platforms that allow
customers to make their own investments, and so-called “robo-advisers” which use new
technology to target clients using automated advice solutions and human advisers through the
process of setting up portfolios. The FCA has been looking with great interest at the emergence
of new models allowing consumers to access guidance and investment advice. The FCA has
conducted some thematic supervisory work and research looking at firms' models from across
different sectors, some with quite innovative technologies. The FCA published the preliminary
results of that work, along with a consultation on how they, as regulators, view the different
models.180 In addition, the FCA in October of 2014 launched Project Innovate, an initiative to
foster innovation in financial services for new and established businesses, to promote the
introduction of innovative financial products and services to the market.181
In conclusion, based on the available data from post-RDR reports since 2013, the
Department believes that the RDR has not significantly reduced availability of advice, and any
RDR-related advice gap is likely minor and temporary. Simple, affordable advice, which is
mostly likely to benefit many small investors, was scarce before the RDR, but indications are the
market is evolving to meet these needs under the RDR.

2.10.1.6 The Financial Advice Market Review
Several industry comments cited the UK’s recently initiated Financial Advice Market
Review (FAMR) as evidence that the UK is suffering an advice gap as a consequence of the
RDR. Other comments characterize FAMR as a comprehensive review of the RDR. One stated
the Department’s failure to discuss the negative implications of the FAMR undermined the
integrity of its 2015 NPRM regulatory impact analysis. These comments misapprehend the
nature of FAMR itself and the information FAMR has provided to date. The FAMR is best
understood as a general examination of the financial advice market, not a reconsideration of the
RDR.
FAMR examined the current regulatory and legal framework governing the provision of
financial advice to consumers and its effectiveness in ensuring that all consumers have access to
the information, advice and guidance necessary to empower them to make effective decisions
about their finances. It was not a commentary that the RDR is not working or should be scaled
back, but rather an examination of how financial advice markets and regulations could work
better for consumers. It was motivated in part by recent pension reforms which, among other
things, relaxed annuitization requirements, and gave consumers more access to lump sum
distributions of retirement savings. FAMR had a wide scope and aimed to look across the
financial services market to improve the availability of advice, with a focus on investment and

179

180

181

NMG Consulting, Impact of the Retail Distribution Review on Consumer Interaction with the Retail Investments Market: A Quantitative
Research Report, London: NMG Consulting, September 2014c.
See GC14/3, “Retail Investment Advice: Clarifying the boundaries and exploring the barriers to market development” (Nov. 2014);
available at: http://www.fca.org.uk/news/guidance-consultations/gc14-03.
See UK FCA website “Innovator business: Project Innovate” available at: https://innovate.fca.org.uk/.

87

pension advice. Specifically, FAMR solicited views on: advice gaps for consumers without
significant wealth or income; barriers to people seeking advice and barriers to firms providing
advice; opportunities for online services in offering advice; and how to encourage demand for
financial advice.
In October 2015, the United Kingdom’s Financial Conduct Authority and Treasury
jointly published a "Call for Input" on the first stage of FAMR. The FAMR Call for Input paper
did not draw new conclusions on the UK financial advice market, but rather solicited input from
all stakeholders regarding financial advice. The paper posited that “people who have some
existing savings but not significant wealth are less well served at present,” and that “[r]etirement
income is one area where there is an obvious need in the light of the pension reforms [referring
to reforms described below, not to the RDR], and where some people may be facing a complex
financial decision without being able to access appropriate professional advice or without
recognizing the benefit of seeking such advice.” However, to the extent that there is unmet
demand from some consumers for lower cost simplified advice, not currently offered by the
market, the paper cites the FCA’s post-implementation review of the RDR which clearly states
this same demand existed pre-RDR and is not a result of the RDR.182 As is the case in the U.S.,
the market works better for some consumers segments, particularly the wealthy, to receive
individualized professional advice than for less wealthy people. The UK is looking into means
of addressing this problem.
On March 14, 2016, the UK's HM Treasury and the FCA jointly published the final
FAMR report (the “Report”) setting forth findings from the FAMR.183 The Report considered
input provided by a wide range of stakeholders, including advisers, consumer groups, banks,
insurers, and individuals who responsed to the Call for Input. One of the most significant
objectives of the RDR was to eliminate conflicts of interests that were causing harm to UK
consumers in the financial advice market by banning commissions for advised investment sales.
According to the Report, a majority of respondents to the Call for Input who commented on the
RDR commission ban have found the reforms to be effective and beneficial to consumers and
and do not recommend that the UK should return to pre-RDR rules. The Report states that
“[g]iven the strong arguments against a commission-based system, such as the lack of
transparency and distortion of incentives, FAMR does not believe there is a case to consider this,
and is therefore not recommending a return to commission-based financial advice.”184
Some respondents to the Call for Input suggested that, despite the benefits of removing
commission bias, the RDR requirements for advisers to move to a fee-based compensaton model
contributed to an “advice gap” where many people are not able to get the advice they want and
need at a price they are willing to pay. As the Report states, respondents to the Call For Input
expressed a wide range of views with regard to the nature and size of the advice gap, and how to
define it. According to the the Report “[t]he vast majority of respondents believed that one or

182
183

184

Financial Conduct Authority and HM Treasury, “The Financial Advice Market Review, Call for Input” (Oct.2015).
FAMR Final Report pg. 6, HM Treasury and FCA, (March 2016); available at: https://www.the-fca.org.uk/financial-advice-market-reviewfamr#sthash.rJPhzihT.dpuf.
Ibid at 46. Moreover, the report discusses concerns that were raised during the review regarding the potential detriment to consumers that
arise from the receipt of commissions from non-advised sales, particularly the sale of annuity products that are exempt from the RDR
requirements. Ibid. See also “Pension Reforms- Proposed Changes to Our Rules and Guidance,” FCA (Oct. 2015); available at:
http://www.fca.org.uk/static/documents/consultation-papers/cp15-30.pdf.

88

more advice gaps exist, most believing that a gap exists for people on lower incomes or with
lower levels of assets who cannot afford to pay the fee for advice or find it harder to access.”185
Respondents cited a number of reasons for this gap, including supply-side and demand side
issues, which are addressed in the Report.
On the supply-side, the report states that adviser numbers in the UK have declined over
the period 2011-2014 for a range of reasons including the more stringent qualification
requirements imposed by the RDR. However, the report indicates that there are reasons for the
decline in the numbers that are not associated with the RDR, particularily in the banking industry
where the majority of advisers exited the market during this period, such as “declining
profitability of branch-based distribution models, a lesser role for branch-based activity, … and
the consequences of episodes of mass mis-selling (in terms of redress and reputational
damage).”186
With respect to the cost of providing advice, the Report states that a number of firms
focused their efforts on clientswith an increased focus on larger investors. The Report indicates
that there is some quantitative evidence supporting this, citing a 2016 survey finding that
although only 52% of advisers ask for a minimum portfolio, of those 52% the proportion that ask
for a minimum portfolio of over £100,000 has gone up from 13% in 2013 to 32% in 2015 (so for
the overall population it has risen to 16%).187 The Report states that “[a] consistent theme
emerging from the Call for Input was that there are significant minimum costs per customer
associated with supplying face-to-face advice.”188 This means that it may be less cost-effective
for individuals with small amounts to invest to obtain advice regardless of the whether their
advisers are compensated on a commission or fee basis.
The Report also states that firms might not be providing advice due to concerns about the
complexity of the regulatory requirements and liability concerns.
The Report provides positive data on the supply-side of the financial advice market in
stating that “[some] larger firms have recently signalled a return to the advice market. In some
cases this is being facilitated by effective and creative use of new technologies. A number of
firms currently in the advice market are also planning to increase the number of customers they
serve. The FCA’s recent survey of advisers found that around 30% of firms surveyed expect to
grow the number of advisers over the next year.”189
With respect to the demand side, the Report states that responses to the Call for Input
suggest that a lack of consumer demand is an important factor underlying the advice gap. The
Report concludes that the low levels of consumer demand are attributable to several factors
including high costs (especially relative to small amounts available to invest), a trend toward
consumers making and executing their own financial decisions, and limited confidence in
seeking out and acting on financial advice. There is also evidence that lack of take up of
investment advice is due to mistrust of advisers among the general population, particularly

185

186
187
188
189

Ibid at 24. A large number of respondents stated that advice gaps exits with respect to “saving into a pension, taking an income in
retirement, and savings and investment.”
Ibid. at 18.
Blue and Green Tomorrow, Voice of the Adviser Survey, 2016.
FAMR Final Report at 19.
Ibid. at 20, citing a forthcoming FCA survey of 233 firms on the provision of financial advice which included questions relating to FAMR.
This survey is due to be published April 2016.

89

following the past “mis-selling” scandals that led to the passage of the RDR. According to the
report, “although the RDR has made significant progress in professionalizing the advice
industry, and levels of trust are high amongst those who already receive advice, there is evidence
that trust in advisers remains low among the general population.” 190
The report states that more technological, automated advice models need to be developed
to make available more cost-effective and potentially more engaging forms of advice.
Therefore, the Report recommends that the FCA issue guidance clearly setting out how
‘streamlined advice’ models can be designed to comply with regulatory requirements. In
addition, the FCA is planning to set up an ‘Advice Unit’ (building on Project Innovate, which
supports innovative financial services businesses) to help firms develop automated advice
models. The review has recognized previous attempts to design a system which allows
consumers to access cheaper forms of advice have not been as successful as hoped. In
particular, many firms have been deterred from offering “basic” and “simplified” advice due to
concerns that complaints relating to such advice may ultimately be judged by the regulator and
the Financial Ombudsman Service (which resolves disputes between consumers and firms)
against the standard for “full advice”. According to the report, “at present, a number of firms do
not have the confidence to develop advice services to meet simple consumer needs.” As a result,
many consumers who want to receive this kind of support are either left without it, or are
required to pay for full advice. The review recommends developing a clear framework to give
firms the confidence to deliver advice on simple consumer needs in a proportionate way.
The review also notes that firms have also reported a lack of demand for these services.
FAMR recommends drawing a clearer line between “advice” and “guidance”, by changing the
definition of ‘regulated advice’ to mean a personal recommendation. This would enable more
firms to give more ‘guidance’ to consumers on their options, without needing to meet the same
suitability requirements as full advice (although this guidance would still need to be clear, fair
and not misleading; and firms would still need to act honestly, fairly and professionally in
accordance with the best interests of their clients. Also, it seems that firms in the UK advice
market remain unclear regarding where the boundary exists between providing non-advisory and
providing advisory services. Thus, the Report recommends changing the definition of regulated
advice, and issuing additional guidance to help firms distinguish between advice and other forms
of assistance to help clients make their own investment decisions. This relates to the
Department’s initiative to help clarify the distinction between investment education and advice,
as discussed previously in Section 2.9.
In summary, the Report concludes that the RDR has brought about a positive change in
the quality of advice available but it also suggests that more can and should be done to make the
provision of advice and guidance to the mass market more accessible and cost-effective. In
order to address these issues, the Report provides 28 recommendations in the following three
key areas:


190

Affordability – These include proposals to make the provision of advice and
guidance to the mass market more cost-effective. FAMR makes a number of
recommendations intended to allow firms to develop more streamlined services and

Ibid. at 23.

90

engage with customers in a more engaging and effective way. These include a
proposal that the FCA should set up a dedicated team to assist firms that are seeking
to develop large scale automated advice models to bring those to market more
quickly.


Accessibility – These are aimed at increasing consumer engagement and confidence
in dealing with financial advice. FAMR proposes a number of measures to help
consumers engage more effectively with advice. These include making their own
information more easily available to them and those that advise them; the
development of 'rules of thumb' and the use of nudges to encourage customers to seek
support at key life stages and recommendations intended to help employers give more
support to their staff in financial matters.



Redress – Some industry stakeholders suggested that concerns about future liability
are preventing them giving advice. FAMR has made a number of recommendations to
address these concerns while ensuring consumers have adequate protection. FAMR
has made recommendations to increase the transparency of the Financial Ombudsman
Service, and to consider making the Financial Service Compensation Scheme levy
more manageable for advisory firms in order to adequately manage consumer
complaints.

FCA and HM Treasury will jointly report on progess in implementing the
recommendations in twelve months and review the outcome of the recommendations in three
years.
Thus, FAMR is part of an ongoing effort to closely monitor of market developments to
see how firms have responded to the challenges presented by the reforms, how they have met
changing consumer demands and to assess how the protections put in place are effective in
delivering good consumer outcomes. Moreover, FAMR cannot be viewed in isolation without
also examining the other reforms in the UK outside the RDR that are currently and will further
impact the pension and retirement market. In 2014, the UK Government announced reforms
giving people more freedom to access their pension savings from age 55, further increasing the
attractiveness of pension saving. Where previously people were persuaded to save and then
defaulted to an annuity purchase, now many are automatically enrolled into pension saving but
given the freedom to decide how and when to access their savings from age 55.

2.10.1.7 Implications for the Department’s Final Rule and
Exemptions
The Department has devoted significant time to studying the effects of the RDR.
Department leadership and staff have consulted with the UK’s FCA and Treasury counterparts,
including staff involved in the FAMR, to engage in a meaningful dialogue in order to set forth a
proper assessment in this analysis. The UK reforms reflect the premise that strong reforms were
needed to protect investors from advisory conflicts, and the UK’s experience suggests that
advisory companies can and do adapt to serve investors under new, stronger rules, and that retail
investing aligns better with investors’ interests once strong protections are in place.
Many industry comments on the Department’s 2015 Proposal have wrongly analogized
the UK reforms to those of the Department’s proposal by drawing straight comparisons between
the two countries, ignoring or understating the substantial differences between the two which
make it impossible for an apples-to-apples comparison to be made. The Department’s final rule
and exemptions reflect its effort to mitigate advisory conflicts effectively while preserving
sufficient flexibility to minimize even minor and temporary negative consequences.
91

Unlike the RDR, the Department’s rule does not ban commissions. The Best Interest
Contract Exemption that accompanies the final regulation provides conditional relief for
common compensation arrangements, such as commissions and revenue sharing, that an adviser
and the adviser’s employing firm might receive in connection with investment advice to retail
retirement investors. The RDR, on the other hand, specifically bans payments of commissions
from product providers with respect to advised investment sales.
In addition to the full commission ban, the RDR rules imposed new extensive and
rigorous professional certification standards on advisers, which some have found burdensome.
The Department’s regulation does not include any analogous qualification standards for advisers.
The Department’s regulation is focused solely on retirement plans and accounts, while
the RDR’s ban on commissions applies very broadly to both retirement and non-retirement
accounts in the UK. The Department’s rule protects ERISA plan participants and IRA investors
who, as discussed later in this analysis, merit special protection, from advisory conflicts of
interest and related self-dealing.
Several industry comments assert that the RDR has caused an “advice gap” in the UK,
and that Department’s 2015 Proposal would cause an advice gap here in the US. This assertion
is belied by the facts, however. In fact, advice is amply available in the UK under the RDR.
While some UK advisers reduced services to small investors, this trend existed independent of
the RDR, and RDR provisions that may have contributed to the trend (mainly higher
qualification standards, and possibly the ban on commissions) are absent from the Department’s
final rule and exemptions. Moreover, the U.S. has five times as many advisers per person as the
UK, and almost 4 times what the UK had even before the RDR was passed.

Figure 2-7 Number of Advisers per 10,000 Population
UK Advisers Pre-RDR number

6.2

UK Advisers Post-RDR number

4.8
20.3

US Broker Dealer Representatives

8.7

US RIA Representatives

25.3

US Total

The Department believes that the UK experience supports a finding that strong
protections against advisory conflicts are warranted and can produce substantial benefits for
consumers. The Department anticipates that its final rule and exemptions will deliver strong
protections, but viewed against the RDR, it provides greater flexibility to ensure that possible
transitional negative effects are minimized.

92

2.10.2

Australian Legislation Impacting Financial Advice

In a similar development to the UK, the Australian government enacted the Future of
Financial Advice (FOFA) legislation on June 25, 2012.191 The FOFA was passed to improve the
integrity of the financial advice market, making dramatic changes to the delivery and receipt of
financial advice with the goal of mitigating conflicts of interest. Prior to FOFA, poor advice had
been strongly linked to the presence of commissions and advisers failing to act in a client’s best
interests.192 The legislation was initiated as a government response to the Parliamentary Joint
Committee on Corporations and Financial Services' Inquiry (PJC) into financial products and
services due to the collapse of several financial services companies during the financial crisis of
2007-08 and mis-selling scandals in which financial advisers switched clients out of deposit
accounts into funds which increased their compensation.193 A PJC report on the inquiry was
issued in early 2012. The FOFA became effective on July 1, 2012. Compliance with the new
measures was voluntary until July 1, 2013, and became mandatory thereafter.194
Most notably, FOFA imposes the following standards on financial advisers:


Bans conflicted remuneration structures including commissions with respect to the
distribution of advice on retail investment products, including managed
investments;195



Requires financial advisers who charge ongoing fees to retail clients to provide a
renewal notice every two years, in addition to an annual fee disclosure statement;



Prohibits an ongoing fee from being charged to clients if they do not renew by optingin every two years (clients are presumed to have opted out if they do not opt-in);



Prohibits licensees or representatives who provide financial product advice (personal
and general) to retail clients, which could reasonably be expected to influence the
choice of financial product recommended or the financial advice given, from
accepting soft-dollar benefits over $300 where it could be expected to have influence
over the choice of financial product recommendation or the advice given to retail
clients (limited exceptions apply for general insurance, execution-only services and
other prescribed benefits); and



Provides a new statutory duty for financial advisers to act in the best interest of their
clients.

On December 20, 2013, the Australian Government announced a package of regulatory
changes to FOFA through the Corporations Amendment Regulation of 2014 (Streamlining
Future of Financial Advice) to reduce compliance costs and regulatory burden on the financial

191

192

193

194

195

Australian Securities and Investments Commission, FOFA Background and Implementation; available at: http://asic.gov.au/regulatoryresources/financial-services/future-of-financial-advice-reforms/fofa-background-and-implementation/.
For example, see ASIC (2003) Report 18 Survey on the quality of financial planning advice, 5-6, ASIC (2012) Report 279 Shadow
shopping study of retirement advice, 8.
A. Ferguson and C. Vedelago, “Targets, Bonuses, Trips- Inside the CBA Boiler Room,” Sydney Morning Herald, June 22, 2013; and
Australian Government: The Treasury, media release, “Delivering Affordable and Accessible Financial Advice” (Dec. 20, 2013).
The Treasury of Australia, “The Future of Financial Advice” (2012); available at:
http://futureofadvice.treasury.gov.au/Content/Content.aspx?doc=home.htm. As stated above, FOFA became mandatory on July 1, 2013.
The ban does not apply to some products and advice services—for example, general insurance products, some life insurance products and
basic banking products.

93

services sector under FOFA. The ban on commissions and conflicted remuneration for financial
advisers was not amended to re-introduce commissions or conflicted remuneration for financial
advisers. These regulatory changes commenced on July 1, 2014. However, the regulations were
repealed by a Motion of Disallowance passed by the Senate on November 19, 2014. Therefore,
the law reverted back to the original FOFA legislation and future modifications remains unclear.
According to the Australian Government, it is considering further legislative refinements and in
which FOFA will be given time to work.196
Because of the debate on whether to amend certain provisions of FOFA, there is a lack of
detailed evidence as to the post implementation effectiveness and impacts on the Australian
market in comparison with the UK. For example, in March of 2014, the Australian Treasury
published a Regulation Impact Statement (RIS) that stated that the number of financial advisers
had declined and compliance costs had increased following the economic crisis and
implementation of FOFA.197 However, as the RAND Research Report, “Financial Advice
Markets: A Cross Country Comparison” notes, the RIS did not directly cite any publically
available evidence to support these statements, did not try to distinguish what trends might be
attributable to the impacts of the economic downturn versus what might be attributable to FOFA,
and did not make clear if changes observed in the industry had actually led to consumers seeing
cost increases or declines in their ability to access advice.198
On September 17, 2014, the Australian Securities and Investments Commission (ASIC)
reported on the results of two financial advice industry engagement projects conducted regarding
the implementation of the FOFA reforms by interviewing 60 licensees who accounted for close
to 10,000 advisers and 4.6 million retail clients. ASIC sought to assess how the advice industry
had adapted to the new requirements and to obtain a greater understanding of key issues facing
licensees. Some key findings of the report included:


Over 90 percent of licensees indicated that there had been no change in the number of
their advisers or authorized representatives as a result of the reforms;



The majority of respondents indicated there was no change to the type of advice
services they offered;



Ongoing commissions decreased as a proportion of revenue but revenues from fees
increased; and



The biggest challenge identified by the licensees related to the requirement to provide
fee disclosure statements and the changes they needed to make to their systems.199

Although there is not as much direct evidence and data as there is in the UK market, the
Australian financial sector is one of the largest and most sophisticated in the world and the
pensions market is the fifth largest in terms of size of assets, mainly accumulating in defined

196
197

198

199

The Treasury of Australia, “The Future of Financial Advice” (2012).
Australian Government, Department of the Prime Minister and Cabinet, “Future of Financial Advice Amendments — Details-Stage
Regulation Impact Statement,” Canberra: Office of Best Practice Regulation, Department of the Treasury, (March 19, 2014).
See RAND Research Report, “Financial Advice Markets: A Cross Country Comparison” (April 21, 2015). RAND cites on page 29 that the
RIS notes “much of the evidence in this RIS has been provided to the Treasury under commercial-in-confidence arrangements and cannot
be directly quoted. Where this is the case, the evidence is paraphrased and no source is referenced.”
Australian Securities and Investments Commission, Review of the Financial Advice Industry’s Implementation of the FOFA Reforms,
Report 407, (Sept. 2014).

94

contribution plans.200 The Department has carefully studied, and will continue to monitor the
consequences, both expected and unintended, that result from the application of a ban on
commissions and a fiduciary duty for advisers. Australia, similar to the UK, was one of the first
countries to dramatically alter the landscape of financial services to address the conflicts of
interest inherent in current commission practices with the goal of providing advice in the client's
best interest.

2.10.3

Is the RDR a Model for Wider European Regulation?

The RDR appears to be influencing the future of distribution of investment advice both
within the regulatory bodies of the European Union (EU) and within several member states. The
European Parliament addressed the payment of commission to both retail and professional
clients in the new version of the Markets in Financial Instruments Directive (MiFID II). On
October 26, 2012, the European Parliament approved a revised version of MiFID II which
includes a ban on the acceptance of commissions in relation to advice, but only where the firm
has informed the client that the advice is given on an independent basis- and also in relation to
portfolio management services. MiFID II covers the sale and distribution of investment products
such as investment funds in and structured bank-based products. However, this revised draft
expressly permits Member States to adopt more restrictive measures. MiFID II is the subject of
discussions between the European parliament, European Commission and European Council.
Political agreement on the MiFID II proposals was reached on January 14, 2014, after several
months of negotiations between the Commission, Parliament, and Council. Parliament endorsed
the MiFID II on April 15, 2014, and the Council adopted the legislation on May 13, 2014 and it
is currently expected to become effective in 2018.201 Under the agreement, firms providing
independent advice or portfolio management may not accept any fees, commissions, or monetary
or non-monetary benefits from third parties in relation to the advice or service.202
Additionally, the Insurance Distribution Directive (IDD) of the EU aims to improve the
regulation of retail insurance sales and distribution practices across the single European market
by adding measures requiring greater transparency and restricting certain commission payments
for all intermediaries where a conflict of interest arises. One of the goals of IDD is to improve
consumer protection in the insurance sector through requirements for increased information
provision and advice and by creating common standards for insurance sales. EU member
countries would be allowed to impose higher standards if they wish. IDD will likely come into
force in member states in 2018, two years after its adoption in 2016.

200

201

202

Willis Towers Watson, “Global Pension Assets Study” (Feb. 2016); available at:
https://www.willistowerswatson.com/en/insights/2016/02/global-pensions-asset-study-2016.
MiFID II was published in the Official Journal on June 12, 2014 and entered into force on July 2, 2014, 20 days after publication. As a
directive, MiFID II must be transposed into national law by Member States by July 3, 2016, and must generally apply within Member States
by Jan. 3, 2017.
Minor non-monetary benefits that could enhance the quality of service may be permitted, provided they are disclosed and do not impair
compliance with the firm’s duty to act in its client’s best interest.

95

3. IRA Market
The current market for investment advice to IRA investors is replete with conflicts of
interest between advisers and investors. Well qualified advisers compete vigorously for
investors’ business, but investors’ high information costs – i.e., the fact that most investors lack
the information and expertise necessary to evaluate the price and quality of advice – prevent this
competition from producing efficient results. Many investors do not know how much they are
paying for advice or whether the advice is of high quality. They cannot effectively discourage
advisers from acting on their conflicts by, for example, taking their business to non-conflicted
advisers. Academic studies and comments on the NPRM provide persuasive evidence that
conflicts of interest sometimes bias advisers’ recommendations in ways that harm investors, and
that as a result, investors overall pay more and earn lower returns than they would in the absence
of harmful conflicts.
A careful review of this data, which consistently point to a substantial failure of the
market for retirement advice, suggests that IRA holders receiving conflicted investment advice
can expect their investments to underperform by an average of 50 to 100 basis points per year
over the next 20 years. The underperformance associated with conflicts of interest – in the
mutual funds segment alone – could cost IRA investors between $95 billion and $189 billion
over the next 10 years and between $202 billion and $404 billion over the next 20 years. While
these expected losses are large, they represent only a portion of what IRA investors stand to lose
as a result of adviser conflicts. Data limitations impede quantification of all of these losses, but
there is ample qualitative and in some cases empirical evidence that they occur and are large
both in instance and on aggregate.
This final rule and exemptions strengthen consumer protections in the retail IRA market,
in order to deliver better results for IRA investors. Existing protections, including relevant
securities and insurance regulations, have proven inadequate to prevent adviser conflicts from
inflicting excessive losses on investors. Such existing rules generally do too little to mitigate
advisers’ conflicts. Adviser compensation arrangements permissible under existing rules
sometimes create strong incentives for advisers to make recommendations that are not in their
customers’ best interest. Moreover, existing requirements that recommendations be “suitable”
leave some room for advisers to subordinate their customers’ interests to their own. The
rulemaking is designed to prevent conflicts of interest from compromising the quality and
inflating the price of investment advice provided to IRA investors, and to ensure that that advice
honors IRA investors’ best interest. The regulation broadens the IRC definition of fiduciary
investment advice rendered to retail IRAs. This would limit or mitigate conflicts of interest in
such advice by subjecting more of it to the IRC prohibited transactions provisions. Some
conflicts would remain permissible, subject to protective conditions, pursuant to the prohibited
transaction exemptions established and amended as part of this rulemaking. If the regulatory
aims are achieved, the result will be lower fees, more appropriate risks, and higher risk-adjusted
returns for many IRA investors.
The Department estimates that the final rule and exemptions will deliver large gains for
retirement investors by reducing, over time, the losses identified above. Because of data
limitations, as with the losses themselves, only a portion of the potential gains are quantified in
this analysis. Available empirical research isolates the effect of front-end-load-shares paid to
brokers recommending mutual funds on the performance these funds deliver, distinguishing the
effect of conflicts of interest from other factors that impact performance. While consistent with
the broader literature, one study provides a picture of the benefit to investors of mitigating one
particular type of conflict in a subset of the IRA market. The Department estimates that the
96

gains to IRA front-end-load mutual fund investors alone will be worth between $33 billion and
$36 billion over 10 years and between $66 billion and $76 billion over 20 years.203 These gains
estimates may include both cost savings and efficiency benefits, as well as transfers from the
financial industry to IRA investors. These estimates, being limited only to investor gains from
mitigation of adviser conflicts attributable to variation in mutual fund front-end load sharing,
omit a broad array of potential additional gains to investors and social benefits from the final
rule and exemptions.
The Department’s 2015 NPRM Regulatory Impact Analysis similarly concluded that
adviser conflicts are costly for IRA investors, and predicted that the 2015 proposed rulemaking
would deliver financial benefits that justify its costs. Some comments disputed these
conclusions. Some offered alternative analyses concluding that the proposal’s costs would
exceed its benefits. Many of these comments argued that the proposal would make advice more
expensive or less available, particularly for less wealthy IRA investors. These investors would
get less advice and their consequent losses would exceed any gain they might derive from
mitigation of adviser conflicts, these comments said. After close review, much of the analysis
set forth in these comments proved to be flawed or otherwise unpersuasive. Nonetheless, as
detailed below, the Department took these comments into careful consideration in developing its
final rule and exemptions and this associated Regulatory Impact Analysis. The final rule and
exemptions reflect responsive changes to the 2015 NPRM that reduce compliance costs and
disruption to existing arrangements and practices. This Regulatory Impact Analysis reflects a
fresh review of the evidence, assumptions and methods that underlie the 2015 NPRM Regulatory
Impact Analysis, consideration of all material public comments, and targeted analyses
examining disagreements between the two. It concludes that adviser conflicts harm IRA
investors and that this final rule and exemptions will deliver financial benefits and worthwhile
distributional impacts that justify its costs.

3.1

Affected Universe

This rulemaking, as applied to the retail IRA marketplace, will directly affect two major
groups: IRA investors, and the professionals who render investment advice to them. It will
indirectly affect others, such as vendors of financial products that IRA investors choose pursuant
to advice.

3.1.1 IRA Investors
Tax-preferred retirement savings, in the form of plans and IRAs, are critical to the
retirement security of most U.S. workers. These savings totaled nearly $15.3 trillion in the third
quarter of 2015.204 Workers and retirees themselves are responsible in whole or part for
directing the investment of the vast majority of these savings. Individual IRA investors

203

204

The gains estimates reflect the Department’s assumption that, as required under the final rule and exemptions, advisers’ recommendations
will not be influenced by variation in the share of mutual fund front-end loads paid to them.
Board of Governors of the Federal Reserve System, “Financial Accounts of the United States: Flow of Funds, Balance Sheets, and
Integrated Macroeconomic Accounts: Third Quarter 2015” (Dec. 2015) Federal Reserve Statistical Release Z.1; available at:
http://www.federalreserve.gov/releases/Z1/current/z1.pdf. DB assets do not include claims of pension fund on sponsor. Also see
Investment Company Institute (ICI), “U.S. Retirement Market, Third Quarter 2015,” 2015.

97

currently direct the investment of approximately $7.3 trillion in IRA assets, and can choose from
a near endless variety of financial products, securities, or other property in the marketplace.205
IRAs play a major role in U.S. households’ retirement security. In contrast to plans,
which are available to less than two-thirds of private-sector employees,206 IRAs are the only taxadvantaged retirement savings vehicle available to virtually all of America’s workers.207
In 2013, 34 million households (28 percent of all U.S. households) had an IRA,
according to tabulations of the Survey of Consumer Finances (SCF) prepared for the Department
(see Panis and Brien 2016). The median value of these accounts for such families was $50,000
and the median household income for these families was $93,000. Higher-income households
are more likely to have IRAs, but middle- and upper middle-income households on aggregate
hold a larger share of their financial assets in IRAs. Viewed another way, IRA-owning
households tend to have higher incomes than households overall. IRA assets are concentrated at
still higher income levels, but are not nearly as skewed toward higher incomes as are financial
assets overall (see Figure 3-1 and Figure 3-2).
Significant shares of IRA
investors belong to demographic
groups that tend to be more susceptible
to financial exploitation. As elaborated
in Section 3.2.1.2 below, older and less
financially literate investors generally
are less able to distinguish good
investment advice from bad. More
than half of all IRA investors are age
55 or older, and nine percent are 75 or
older. It is likely that over time IRA
ownership will become more skewed
toward more advanced ages, for two
reasons: the DC retirement plan
system is maturing, and the population
is aging.

% of Total within Income Group

Figure 3-1 Prevalence of IRAs
80%
70%
60%

% Share of HHs with IRAs
% Share of Financial
assets in IRAs

50%
40%
30%
20%
10%
0%
$<50K

$100K-<$200K

$500K+

HH Income

205
206

207

ICI, “U.S. Retirement Market, Third Quarter 2015,” 2015.
See U.S. Bureau of Labor Statistics, “Employee Benefits in the United States – March 2015” available at:
http://www.bls.gov/ncs/ebs/sp/ebnr0021.pdf.
IRA tax advantages, however, vary depending on income and plan participation. Taxpayers above certain income thresholds are not
eligible to contribute directly to Roth IRAs. Taxpayers above other income thresholds cannot deduct IRA contributions if they are also plan
participants, but can defer taxation of earnings on IRA investments until money is withdrawn.

98

% of US Total

Some comments on the 2010 Proposal suggested that a large majority of IRAs, especially
small IRAs, are held in brokerage accounts.208 This claim seems to be based on what may be a
misleading comparison based on selected information of just two types of financial investment
accounts: brokerage and
advisory. However, in one
Figure 3-2 Distribution of HHs and IRAs
40%
major survey, more IRAAll HHs
35%
owning households report
HHs w/IRA
holding IRAs at commercial
30%
All $Fin
banks (50 percent) than at
25%
$IRA
brokerages (41 percent).
20%
Among IRA-owning
15%
households with less than
10%
$10,000 in their IRAs, 47
5%
percent held IRAs at
0%
commercial banks, 32 percent
at brokerages, and 16 percent
at credit unions.
Commission-based brokerage
HH Income
does not appear to dominate
the small IRA market (see
Figure 3-3).
Figure 3-3 IRAs by Financial Institution

% of IRA-owning HHS

60%
Households with IRAs
report receiving financial
advice from many sources.
50%
The most popular sources are
internet/online services (44
40%
percent), financial planners (42
percent), friends and relatives
30%
(37 percent), and bankers (33
percent). Just 17 percent
obtain financial advice from
20%
brokers, slightly fewer than
from magazines, newspapers,
10%
and books (18 percent). Even
households that hold IRAs with
0%
brokerages appear to rely less
Top 1% by wealth
With IRA <$10,000 All IRA-owning HHs
on brokers than on other
Commercial Bank
Brokerage
Credit union
S&L
Insurance co.
Finance or loan co.
sources for financial advice:
52 percent use internet/online
services, 44 percent financial planners, 41 percent friends and relatives, and 25 percent bankers,
compared with 23 percent consulting brokers.209

208

209

See Oliver Wyman report, “Assessment of the Impact of the Department of Labor’s Proposed `Fiduciary’ Definition Rule on IRA
Consumers” (Apr. 12, 2011).
Percentages do not add to 100 percent because multiple answers were allowed.

99

Other data sources, in addition to the SCF, paint a similar picture. The Investment
Company Institute (ICI) reports that the median IRA investor in their database is 54 years old,
has a household income of $87,500, and an IRA balance of $50,000. IRAs comprise 32 percent
of household financial assets for households with IRAs.210 These assets are invested in a variety
of investment vehicles: 64 percent of IRAs include investments in mutual funds, 40 percent of
IRAs include investments in individual equities, and 31 percent of IRAs include investments in
annuities. Smaller numbers of IRAs invest in bank accounts and bonds, as well as ETFs and
other investment options.211 According to the U.S. Federal Reserve Flow of Funds Report, in
2014, 83 percent of IRA assets were invested in mutual funds or other self-directed
investments.212
Rollovers from employment-based plans account for most IRA funding. Almost half of
all IRAs include at least some rollover funds. Rollovers may be due to job change, layoffs, or
termination (69 percent of rollovers), retirement (34 percent), as well as other reasons (13
percent).213 In 2014, new IRA rollover contributions amounted to $376.5 billion. Cerulli
Associates projects that by 2020, new IRA rollover contributions will total $517 billion per
year.214 According to the ICI IRA Owners Survey, 49 percent of IRA investors with rollovers
consulted a professional financial adviser as their primary source of information, and 63 percent
of IRA investors with rollovers consulted a professional financial adviser in some capacity
regarding their rollover decision.215

3.1.2 Professional Advisers, BDs, RIAs, Insurance Agents
Registered Investment Advisers (RIAs) and broker-dealers (BDs) are the two main types
of advisory firms affected by the rule. In addition to RIAs and BDs, insurance agents will be
affected by the rule. As discussed previously, brokers and financial planners are the two biggest
named professional sources of financial advice for IRA investors. Insurance agents provide
financial advice for their clients as well. Within the financial industry, many BDs market
themselves with titles that give the impression of specialized advisory expertise, such as wealth
adviser, wealth planner, financial planner, financial adviser, retirement planner, or investment
adviser. In some cases, outside professional groups govern the terms and circumstances under
which an individual can claim a designation, as in the case of the title “Certified Financial
Planner.” In other cases, anyone can use the title, as in the case of “Financial Adviser.” For

210

211

212

213

214
215

ICI Research Perspective, “The Role of IRAs in U.S. Households’ Saving for Retirement, 2015” (Feb. 2016), 8; available at:
http://www.ici.org/pdf/per22-01.pdf.
ICI Research Perspective, “Appendix: Additional Data on IRA Ownership in 2014” (Jan. 2015), 10; available at:
http://www.ici.org/pdf/per21-01a.pdf.
Board of Governors of the Federal Reserve System, “Financial Accounts of the United States” Federal Reserve Statistical Release Z.1,
Washington, D.C. (Dec. 2015), available at: http://www.federalreserve.gov/releases/z1/current/data.htm. The Federal Reserve Board’s
Flow of Funds Report defines “other self-directed investments” to include securities held in brokerage accounts excluding money market
fund and other mutual fund assets held by households through brokerage accounts (e.g., ETFs, equities, or bonds held at Fidelity or
Vanguard).
ICI Research Perspective, “The Role of IRAs, 2015.” 13. Because multiple responses were allowed, they added up to more than 100
percent.
Cerulli Associates, “Retirement Markets 2015: Growth Opportunities in Maturing Markets” 2015.
ICI Research Perspective, “The Role of IRAs, 2015,” 16-17.

100

many of these titles, both BDs and RIAs (and people who are neither) can use them, which can
confuse consumers.216

3.1.2.1

RIAs

Over 10,500 RIA firms are registered with the SEC. These SEC-registered RIAs
managed more than $38 trillion. In addition, there are more than 15,000 state-registered RIAs
and there are more than 275,000 state-registered RIA representatives. Most RIAs charge their
clients fees based on the percentage of assets under management, while others may charge
hourly or fixed rates.217 Depending on an RIA’s particular customer base affiliations and
business and compensation model, fees may be materially affected by this rule.

3.1.2.2

Broker-Dealers

The SEC and FINRA oversee approximately 4,000 BD firms.218 As of the end of 2009,
FINRA-registered BDs held over 109 million retail and institutional accounts. Approximately
18 percent of FINRA-registered BDs also are registered as RIAs with the Commission or a
state.219
Most BDs receive transaction-based compensation. An industry survey conducted by the
Financial Services Institute (FSI) found that 60 percent of all revenue received by BDs is
commissions received from financial entities. An additional 31 percent of revenue is received in
the form of asset management and advisory fees. About 13 percent of assets held by BDs are in
securities held for resale. Most BD representatives service small books of business. Forty-five
percent of representatives produce less than $50,000 in revenue for their BDs annually,220 while
only two percent of representatives produce more than $1 million. Additionally, 41 percent of
BDs offer production bonuses and 68 percent of BDs have minimum production requirements
for representatives.221

3.1.2.3

Insurance Agents

Insurance products that are exempt from registration as securities (in this context,
generally, fixed-rate and fixed-indexed annuities) generally are distributed by state-licensed
insurance agents. There generally are two types of insurance agents – career agents and
independent agents – that sell annuity products. As commonly defined, career agents devote
more than three-quarters of their time to one insurance company’s products. Insurance
companies often provide career agents with financing, training, and office space. Independent
agents are salespersons who earn commissions by selling insurance products from multiple

216

217

218
219
220

221

Consumer Financial Protection Bureau, “Senior Designations for Financial Advisers: Reducing Consumer Confusion and Risks” (Apr. 18,
2013); available at: http://files.consumerfinance.gov/f/201304_CFPB_OlderAmericans Report.pdf; and GAO Report, “Consumer
Finance: Regulatory Coverage Generally Exists for Financial Planner but Consumer Protection Issues Remain” ( 2011), GAO-11-235;
available at: http://www.gao.gov/new.items/d11235.pdf.
SEC “SEC Staff Dodd-Frank Study,” 2011, iii and 2012-2013 Report of North American Securities Administrators Association, “The
Pillars of Protection;” available at:
http://www.nasaa.org/wp-content/uploads/2011/08/NASAA-2012-2013-Report.pdf.
FINRA Newsroom at: http://www.finra.org/newsroom/statistics.
SEC “SEC Staff Dodd-Frank Study,” 2011, iii.
While some of these BD representatives may derive their incomes entirely from the limited revenue they generate, others may earn
additional fee income as RIAs or financial planners. Some may work part-time as BD representatives, possibly in addition to other paid
work.
Financial Services Institute (FSI), “2013 Broker-Dealer Financial Performance Study,” 32-83.

101

companies. Independent agents include Personal-Producing General Agents (PPGAs). Bankaffiliated insurers or insurance agents or licensed BDs can sell insurance products to bank
customers. This practice was authorized by the Gramm-Leach-Bliley Act in 1999. Salaried
employees of insurance companies sometimes engage in sales of annuity products through direct
mail or telemarketing. These employees may receive bonuses, but no commissions are paid. In
the distribution channel, this group is sometimes called “direct response.”
Figure 3-4 Annuity Sales by Distribution Channel and Product Type in 2014
Variable
Fixed-rate
Fixed Indexed Total
23%
1%
3%
27%
Independent BD
15%
3%
1%
19%
Career Agents
10%
1%
0%*
12%
Full Service National BD
8%
6%
3%
17%
Banks
7%
0%*
0%
8%
Direct Response
0%*
3%
15%
18%
Independent Agents
0%
0%
0%
0%
Other
64%
14%
22% 100%
Total
Source: DOL’s own calculation using LIMRA Annuity Yearbook-2014
* Sales were positive. They appear as 0% due to rounding.

Figure 3-4 summarizes the share of annuity sales by distribution channel and product
type in 2014. According to this figure, the main product sold by independent BDs is variable
annuities (23 percent of all annuity sales in the market), as compared to fixed-rate (1 percent)
and fixed-indexed annuities (3 percent). The figure also illustrates that the main product sold by
independent agents is fixed-indexed annuities (15 percent), as compared to variable (0 percent)
and fixed-rate annuities (3 percent).
A variety of intermediaries between insurers and independent agents exist in the annuity
market. Such insurance intermediaries are commonly referred to independent marketing
organizations (IMOs), field marketing organizations (FMOs), national marketing organizations
(NMOs), or brokerage general agencies (BGAs). These intermediaries play a role as middlemen
in the distribution system. The main function of these intermediaries is to market, distribute and
wholesale various insurance products.222 The terms IMO, FMO, NMO, and BGA often are used
interchangeably, although they carry slightly distinct meanings.223 In this section, we will use
the term IMO to represent all different types of insurance intermediaries. This intermediary
structure can be appealing to both insurance carriers (insurers) and independent insurance
producers (insurance agents) because it allows insurance carriers to cut their overhead costs and
at the same time it can allow independent insurance producers to make twice the commission of
their captive counterparts.224 An IMO can provide independent producers with support that

222

223

224

Joe Simonds, “Warning: Why FMOs will be extinct soon,” Lifehealthpro (Dec. 5, 2012); available at:
http://www.lifehealthpro.com/2012/12/05/warning-why-fmos-will-be-extinct-soon.
Ben Nevejans, “Know your IMO: Finding the Right Support System,” Lifehealthpro (Oct. 4, 2007); available at:
http://www.lifehealthpro.com/2007/10/04/knowyourimofindingtherightsupportsystem.
Ibid.

102

career agents can get from their large insurance carriers. An IMO often provides independent
producers with on-hand sales support, product recommendations, and training. Although less
common, it sometimes provides coaching and mentoring programs. Furthermore, many IMOs
offer leads that can boost sales. In exchange for these leads and services, independent producers
often share the commissions with the IMOs. For example, an IMO can provide independent
producers with various types of sales support and receive some portion of commissions once a
sale is closed. Various factors determine how the commissions are split between IMOs and
independent producers. One main factor is the marketing allowance that insurance carriers
provide to IMOs. An IMO with a good relationship with the insurance carrier often receives a
marketing allowance (funds to cover marketing costs) from the carrier, which the IMO can retain
or, if it so chooses, share with producers.
Many insurance intermediaries have adapted to digital environments, and use digital
marketing organizations (DMOs) to assist with marketing efforts. In turn, DMOs heavily rely on
the internet to increase their visibility, brand themselves, attract new clients and close sales.
Many broker-dealers identified automated transactions and electronic capabilities as important
aspects when they choose whom they do business with.225 Therefore, other insurance
intermediaries also need to utilize and incorporate technologies into their business models.
Overall high utilization of technology in this market would lead to lower costs as it would
greatly reduce printing and mailing costs and make transactions easier in general.

3.1.2.4

Overlaps Among Professional Advisers

Approximately 5 percent of SEC-registered RIAs are also registered as BDs, and 22
percent have a related person that is a BD.226 According to one survey, 63 percent of licensed
insurance agents are also BDs and/or RIA representatives.227 They offer not only variable life
and annuities but also mutual funds, stocks and bonds. Additionally, approximately 88 percent
of RIA representatives are also registered representatives of BDs. A majority of SEC-registered
RIAs reported that over half of their assets under management related to the accounts of
individual clients.

3.1.3 Product Providers
3.1.3.1

Mutual Fund Companies

In 2014, more than 9,000 U.S. registered mutual fund companies held approximately $16
trillion in assets. Investment companies as a whole, a majority of which provide mutual funds,
and their service providers, employed approximately 166,000 individuals in 2013.228
The Department expects some mutual funds companies, insurance companies, and other
product providers to be significantly affected by the proposal. This is because many incentivize

225

226
227
228

Joe Simonds, “Warning: Why FMOs will be extinct soon,” Lifehealthpro (Dec. 5, 2012); available at:
http://www.lifehealthpro.com/2012/12/05/warning-why-fmos-will-be-extinct-soon.
See SEC Staff Dodd-Frank Study at 12.
Jim Mitchel & Shannon O’Keefe, November 2010, NAIFA “2010 Membership Survey”.
ICI, “2015 Investment Company Fact Book: A Review of Trends and Activities in the U.S. Investment Company Industry,” 55th Edition
(2015); available at: http://www.ici.org/research/stats/factbook.

103

advisers to recommend particular mutual funds or insurance products to their clients.229 To the
extent that the proposal is effective in mitigating conflicts of interest with respect to the advice
given by advisers, mutual fund companies, insurance companies and other product providers that
currently sell their products by making payments to brokers and other advisers may sometimes
find it more profitable to employ different methods to promote their products.

3.1.3.2

Insurers

The annuity insurance market is concentrated. The top 20 life insurance companies in
deferred annuity assets hold about $2.2 trillion, or 79 percent, of the total of $2.7 trillion in
industry assets.230 In terms of sales, the top 10 companies in variable annuity sales in 2014 have
a 77 percent market share.231 In the fixed-indexed annuity market, the top 10 companies in sales
in 2014 have a 74 percent market share.232
The annuity insurance market is not only concentrated but also specialized. For
example, only two companies made the lists of top ten carriers in sales for both fixed and
variable annuities in 2014.233 In general, the top ten carriers in terms of sales of fixed annuities
do not overlap with the top ten carriers in terms of sales of variable annuities.234 Out of the top
twenty insurance companies in terms of fixed-rate annuity sales, all but three companies sold
fixed-indexed annuities in 2014. In contrast, among companies ranked between 51st and 75th in
sales of fixed-rate annuities, only four companies sold fixed-indexed annuities. The
development of new designs of fixed-indexed annuity products may have increased the
disparities between large and smaller companies.
Concentration in the annuity market varies by market segment. The employer plan space
is the most concentrated, whereas the non-qualified annuity market is the least concentrated.
About 45 percent of total sales in employer plan annuities are through one provider. About 12
percent of total sales in the IRA market are through one provider. The dominating providers in
the IRA market and in the employer plan market are different (Figure 3-5).
Figure 3-5 Share of Sales by Top Provider by Type of Market in 2014
IRA
Non-Qualified
Employer Plan
Share (%) in Sales
12%
10%
45%

Annuity product providers develop and rely on particular distribution channels to deliver
their newly designed products to potential customers. For example, according to one industry
report, in the indexed annuity market insurers heavily rely on independent insurance agents. In
the fourth quarter of 2014, 80 percent of fixed-indexed annuity sales were attributed to
independent insurance agents, whereas only 10 percent of fixed-indexed annuity sales were

229
230
231
232
233
234

See Section 3.2.3.1 for details on these practices.
LIMRA, “U.S. Individual Annuity Yearbook – 2014” (2015).
DOL’s own calculation based on LIMRA, “U.S. Individual Annuity Yearbook – 2014” (2015).
Wink’s Sales & Market Report, 4th Quarter, 2014.
LIMRA (2015) U.S. Individual Annuity Yearbook-2014 Data; LifeHealth Pro, March 12, 2015.
LIMRA (2015) “U.S. Individual Annuity Yearbook – 2014.”

104

completed through banks, 2014.235 The type of products and the distribution channels are
intertwined partly due to legal and regulatory regimes.

3.1.3.3

Other Product Providers

There are other product providers whose products can be held in IRAs. In addition to
investing in mutual funds and annuities, retirement investors invest in other products, e.g.
certificates of deposits (CDs), hedge funds, stocks, bonds, real estate investment trusts (REITs)
and exchange-traded funds (ETFs).

3.2

Need for regulatory action

The Department collected and studied a wide range of evidence in order to determine
with confidence whether there is a harmful failure in the market for IRA advice, and if so, what
if any regulatory solution would be most beneficial. This evidence included public comments on
the 2015 and 2010 NPRMs (including the 2015 public hearing record); academic research papers
related to conflicts of interest in the market for financial advice and the effects of disclosure,
among other relevant topics; and government and industry statistics on the IRA marketplace,
including information on financial products and services, vendors and intermediaries, and
consumers. The Department also consulted with analysts at the SEC, FINRA, the Council of
Economic Advisers, the National Economic Council, the Domestic Policy Council, the OMB,
the Department of the Treasury, the Consumer Financial Protection Bureau, and the GAO, as
well as with academic researchers in the field, the Financial Conduct Authority (previously, the
Financial Services Authority) of the United Kingdom, and the Australian Securities and
Investments Commission, among others. As elaborated below, the evidence supports the
following conclusions:

235



The IRA market warrants consumer protections beyond those applicable to other
retail investment accounts.



Material changes in the marketplace since 1975 have rendered the 1975 regulation
obsolete and ineffective.



The IRA advice marketplace exhibits characteristics that economic theory
suggests would lead to market failures harmful to advice recipients. That is,
because of agency conflicts between advisers and investors that reflect the way
advisers are compensated and IRA investors’ high information costs, IRA
investors will sometimes receive and follow advice that subordinates their
financial interests to their advisers’, and consequently their net investment results
will suffer.



Such harm that exists in the IRA marketplace can be expected to amount to at
least tens and probably hundreds of billions of dollars over the next 10 years.
This harm persists despite existing consumer protections, including those
provided under securities and insurance regulations.

Wink’s Sales & Market Report 4th Quarter, 2014, Part 1.

105



Regulatory action that effectively mitigates advisers’ conflicts and ensures that
advice puts IRA investors’ interests first can deliver large, welfare-improving
financial benefits for IRA investors that justify associated costs.

3.2.1 IRAs Warrant Special Protection
A number of comments on the 2015 NPRM argued that special protections for IRAs are
not warranted. These comments generally maintained that protections applicable to all retail
investors are adequate for IRA investors and that adding special protections for the latter would
increase costs and cause confusion.
IRAs warrant special protections in addition to those applicable to other retail accounts
because of their importance to retirement security, their preferential tax treatment, and IRA
investors’ vulnerability to abuse. Congress recognized this when, in 1974, it amended the IRC
to give fiduciary status to advice on the investment of IRA assets under the IRC’s new
prohibited transactions provisions. Under the narrow 1975 regulation, however, IRA advisers
generally can and do avoid fiduciary status, thereby stripping IRA investors of the protections
the IRC’s prohibited transactions provisions were enacted to provide. There is convincing
evidence that, notwithstanding other existing protections (see Section 2.6 above), advice
conflicts inflict losses on IRA investors.
A number of comments on the 2015 NPRM questioned the need for Department action to
regulate investment advice rendered to IRA investors. These comments argued that various
other federal and state rules governing retail investment advice already provide sufficient
consumer protections, and that subjecting such advice to the prohibited transaction provisions of
the IRC was therefore unnecessary. Some questioned whether the Department had any
legitimate role in regulating advice to retail IRAs because they are not ERISA-covered
retirement plans, and argued that another agency, primarily the SEC, is the proper regulator of
retail investment advice.
The Department understands the roles of the SEC and other federal and state agencies in
the regulation of financial advice provided to retail investors. At the same time, however, the
IRC prohibited transaction provisions, as enacted by Congress as part of ERISA in 1974,
specifically apply to IRA investment advice, and the Department is solely responsible for
interpreting these provisions.236 It is thus incumbent on the Department to protect IRA investors
from harmful adviser conflicts. An examination of trends and evidence accumulated since 1974
suggests that such special protections, if anything, are even more critical today than when
Congress first enacted ERISA more than 40 years ago. The Department’s role in applying these
protections is well established under law and in practice.

3.2.1.1

Importance of IRA Investments

Comments on the 2015 NPRM Regulatory Impact Analysis strongly supported its
conclusion that IRAs are important to retirement security. In the Department’s view, that
importance underscores the need for the protections provided by its final rule and exemptions.

236

See Reorganization Plan No. 4 of 1978, 5 U.S.C. § App. (2010).

106

IRAs were established in 1974 as a vehicle to promote retirement savings. In supporting
IRAs, lawmakers pointed to the need to provide tax preferences similar to those applicable to
job-based pensions to workers who did not have access to such pensions. They also pointed to
rollover IRAs’ potential to make job-based pensions more portable.
The special protections for IRAs embodied in the IRC prohibited transactions provisions
are mirrored by the large tax subsidies IRAs enjoy under other IRC provisions. These direct
subsidies are estimated to amount to $17 billion in 2016 alone.237 This figure dramatically
understates the degree to which current IRA savings
have been subsidized by taxpayers, however. Most of Figure 3-6
Taxable IRA Distributions 1990-2013
the savings flowing into IRAs comes not from direct
15
contributions but from rollovers primarily from jobNumber of Taxpayers
based retirement plans, mostly from DC plans
(millions)
including 401(k)s238 – and much of the savings
10
currently in these plans may eventually be rolled over
into IRAs. The tax preference for DC plans is
5
estimated to amount to $65 billion in 2016.239
IRA’s importance to retirement security in the
United States is widely documented.240 In aggregate
dollar terms, IRAs now represent the single largest and
fastest growing form of retirement saving, outstripping
both private-sector DC and DB plans (see Section
3.2.2.1 below). Almost $2.4 trillion is projected to be
rolled over from plans to IRAs between 2016 and
2020.241 As the baby boom generation begins to retire,
IRA distributions represent a large and growing source
of retirement income (Anguelov, Iams and Purcell
2012). In 2013, 13 million taxpayers reported $120
billion in constant 1990 dollars from taxable IRA
distributions, up from 4 million reporting just $18
billion in 1990. Taxable IRA distributions averaged
$8,990 per taxpayer in 2013 in constant 1990 dollars,
up from $4,951 in 1990 (see Figure 3-6).
All of this suggests that IRAs have become
critical to the retirement security of a large proportion
of America’s middle class, and therefore merit special
protections beyond those applicable to other retail
savings and investment accounts.

237

238
239
240

241

0
150
100

Aggregate Amount
($billions)

50
0
15,000
10,000

Mean Amount
($)

5,000
0

1990 1995 2000 2005 2010
Note: In 1990 dollars
Source: Internal Revenue Service
Statistics of Income

Office of Management and Budget, “Fiscal Year 2017 Analytical Perspectives” (2016), 231. Available
at: https://www.whitehouse.gov/sites/default/files/omb/budget/fy2017/assets/spec.pdf.
ICI, “U.S. Retirement Market, Third Quarter 2015,” 2015.
Office of Management and Budget, “Fiscal Year 2017 Analytical Perspectives” (2016), 231.
ICI Research Perspective, “The Role of IRAs, 2015;” Employee Benefit Research Institute (EBRI) Issue Brief, Number 399, “Individual
Retirement Account Balances, Contributions, and Rollovers, 2012; With Longitudinal Results 2010–2012: The EBRI IRA Database” (May
2014); available at: http://www.ebri.org/pdf/briefspdf/EBRI_IB_399_May14.IRAs.pdf.
Cerulli Associates, “Retirement Markets 2015,” Exhibit 9.08.

107

3.2.1.2

IRA Investors’ Vulnerability

Many comments on the 2015 NPRM Regulatory Impact Analysis supported its
conclusion that many IRA investors lack financial sophistication and, absent adequate
protections, are vulnerable to abuse. There is ample evidence that retail investors generally and
IRA owners in particular are vulnerable to abuse. They face challenges in successfully
navigating today’s complex financial markets. Many cannot effectively assess the quality of the
investment advice they receive or even the investment results they achieve. Disclosures often
lack salience or suffer from complexity, so IRA investors often overlook or misunderstand them
and often gloss over the information presented to them. Research also documents that most
individuals cannot distinguish between the different types of advisers or the different standards
of conduct to which different advisers must adhere, and this confusion is exacerbated by industry
marketing and other practices, especially if the adviser is dually registered as a BD and an RIA.
In addition, IRA investors, in particular, face unique circumstances that easily lend
themselves to abuse. Because most IRAs are retirement income vehicles fed by job-based
pension plans, balances tend to be highest at advanced ages, close to before and after retirement.
Households headed by individuals over age 55 held 79 percent of IRA assets in 2013. This
contrasts with just 45 percent of job-based DC plan assets (Panis and Brien 2016). Yet under
current rules the former – the group more susceptible to abuse – typically lack the protection
associated with a fiduciary standard of conduct, while the latter generally enjoys such
protection.242 Vulnerability is often particularly acute at the moment of retirement, as investors
roll over large balances from more protected, job-based DC (or even DB) plans to less-protected
IRAs. As noted in Section 2.4.3 above, under current Department guidance, advice on such
rollovers need not adhere to ERISA and IRC fiduciary standards. If such advice is tainted by
conflicts, the participant may suffer serious negative consequences. For example, conflicts may
lead an adviser to recommend that a plan participant retire earlier than planned in order to roll
his or her balance into an IRA, offering unwarranted assurances that investment opportunities
available in the IRA will adequately provide for the participant’s retirement income needs.
In a January 2015 letter announcing its regulatory and examination priorities for 2015,
FINRA stated that “a central failing [it] has observed is firms not putting customer’s interests
first. The harm caused by this may be compounded when it involves vulnerable investors (e.g.,
senior investors) or a major liquidity or wealth event in an investor’s life (e.g., an inheritance or
Individual Retirement Account rollover). Poor advice and investments in these situations can
have especially devastating and lasting consequences for the investor.”243 On January 5, 2016,
FINRA released its 2016 examination priorities letter stating its intention to formalize its
assessment of firm culture while continuing its focus on conflicts of interest and ethics.244
There is evidence that, as investors age, they become more vulnerable to and targeted for
abuse. By several measures, according to academic research, financial capability begins to
decline around age 53 (Agarwal, et al. 2009). Individuals over the age of 55 often “lack even a

242

243
244

FINRA Regulatory and Examinations Priorities Letter (Jan. 6, 2015); available at:
http://www.finra.org/web/groups/industry/@ip/@reg/@guide/documents/industry/p602239.pdf.
Ibid.
FINRA Regulatory and Examinations Priority Letter (Jan. 5, 2016), available at: https://www.finra.org/industry/2016-regulatory-andexamination-priorities-letter.

108

rudimentary understanding of stock and bond prices, risk diversification, portfolio choice, and
investment fees” (Lusardi, Mitchell and Curto 2009). While financial literacy falls at advanced
ages, confidence in financial capability may actually increase, leading to poor investment
decisions (Finke, Howe and Huston 2011) and vulnerability to fraud (Gamble, et al. 2015).
SEC examinations of “free lunch” sales seminars found that these events often target
older investors, offering attractive inducements to attend. The seminars commonly employ a
variety of misleading and abusive sales practices. They are often promoted as educational
workshops led by expert financial advisers. Attendees “may not understand that the seminar is
sponsored by an undisclosed company with a financial interest in product sales.”245 Financial
advisers often use “senior designations” – titles that denote special expertise in financial advice
for older individuals – in these and other forums. The report by the SEC, NASAA, and FINRA
indicates that the hosts of free lunch sales seminars often target seniors and approach senior
citizens using titles that suggest they have special credentials or certifications, such as “Certified
Senior Adviser” or “Elder Care Asset Protection Specialist,” when there is in fact no regulatory
qualification that recognizes such expertise.246 The Consumer Financial Protection Bureau
(CFPB) has found that these designations are confusing to consumers. A recent CFPB report
documents older investors’ vulnerability to abuse, and explains how some advisers use senior
designations to create an impression of unbiased expertise when their true aim is to sell products
in which they have a financial interest. CFPB recommends improving standards for acquisition
of senior designations and for the conduct of individuals holding such designations.247 FINRA,
noting that some BDs misleadingly purport to offer free, “no-fee” IRAs, recently opined that
materials making such claims violate applicable advertising rules.248 In addition, the
Massachusetts Securities Division (MSD) finds that the use of various designations and
credentials targeting seniors has increased, leading it to adopt regulations addressing this type of
misconduct targeting senior citizens in 2007.249
IRA annuity purchasers may be particularly vulnerable insofar as they tend to be at or
near retirement age, when individuals are older and have the most assets at stake. Annuities
have been primarily attracting individuals at or near retirement age (Poterba, 1997). This appeal
to individuals at or near retirement age remains similar even in the current time period. Among
individuals holding variable annuities with Guaranteed Minimum Withdrawal Benefits
(GMWBs), 45 percent are age 70 or older. Individuals who hold variable annuities with
Guaranteed Minimum Accumulation Benefits (GMABs) tend to be younger than policy-holders

245

246

247
248

249

SEC, “Protecting Senior Investors: Report of Examinations of Securities Firms Providing “Free Lunch” Sales Seminar” (Sept. 2006);
available at: http://www.sec.gov/spotlight/seniors/freelunchreport.pdf.
SEC, NASAA, FINRA, “Protecting Senior Investors: Report of Examinations of Securities Firms Providing “Free Lunch” Sales Seminars”
(Sept. 2007); available at: http://www.sec.gov/spotlight/seniors/freelunchreport.pdf.
See CFPB, “Senior Designations for Financial Advisers,” 2013 and GAO Publication No. GAO-11-235.
See FINRA Regulatory Notice 13-23, “Brokerage and Individual Retirement Account Fees” (July 2013); available at:
http://www.finra.org/web/groups/industry/@ip/@reg/@notice/documents/notices/p304670.pdf. In the notice, FINRA stated that BDs
marketing campaigns often emphasize that fees are not charged in connection with their retail brokerage accounts and IRAs. Nevertheless,
while certain types of fees may not be charged, others will be. For example, accounts offered by broker-dealers may be subject to fees for
opening, maintaining or closing accounts. FINRA concluded that referring to an IRA account as a “free IRA” or “no-fee IRA” where costs
exist would fail to comply with Rule 2210’s prohibition of false, exaggerated, unwarranted, promissory or misleading statements or claims.
950 Mass. Code Regs. 12.204(2)(i) (2007): 950 Mass. Code Regs. 12. 205(9) (c) (15)(2007); accessed at:
http://www.sec.state.ma.us/sct/sctpropreg/propreg.htm.

109

of other Guaranteed Living Benefits. The average age of GMABs buyers was about 53 year old
in 2012.250
Recently there have been reports that some financial advisers were targeting older
consumers and selling them inappropriate and/or fraudulent products.251 Financial decisions are
most consequential for individuals at or near retirement age when assets have been accumulated.
The population of those 65 or older in 2050 is projected to be 88 million, almost doubling from
46 million in 2014.252 An increased retirement age population puts even more individuals at risk
from inappropriate sales. As previously discussed, although older people make many financial
decisions over their lifetimes, they are still not well versed in financial matters (Lusardi, Mitchell
and Curto, 2012). This lack of sophistication in financial knowledge is further exacerbated as
older individuals are likely to experience cognitive decline.
All of this suggests that IRAs not only merit but also need special protections. By
broadening the application of fiduciary provisions to more financial advice rendered to IRA
investors, this final rule and exemptions will reduce or mitigate the adviser conflicts that can
otherwise motivate abuse.

3.2.1.3

Current Protections

A number of comments on the 2015 NPRM Regulatory Impact Analysis argued that
there are already adequate consumer protections in the IRA market. The Department
understands that a variety of protections currently exist, but has concluded that despite these
protections adviser conflicts inflict excessive losses on IRA investors. The evidence supporting
this conclusion reflects actual experience of investors under the regulatory regimes applicable at
the time the underlying data were collected. As elaborated in Section 3.2.4 below, this
Regulatory Impact Analysis examines more recent evidence and reaches similar conclusions.
Some comments on the 2015 NPRM Regulatory Impact Analysis argued that many of
the proposal’s protections overlap with existing protections, and that overlapping protections
would impose costs but would not add any benefits. The Department agrees that measures that
merely duplicate existing protections are not likely to yield meaningful benefits for IRA
investors. The Department’s final rule and exemptions amend the 2015 Proposal in a number of
ways to minimize such duplication, and these amendments are reflected in the cost estimates
presented in Chapter 5 below. The benefits of this final rule and exemptions, like the benefits
previously attributed to the 2015 Proposal, will be attributable to provisions establishing
different or stronger protections than currently exist. These include provisions that hold all
potentially conflicted IRA advisers to a best interest, rather than suitability, standard, and that
call for policies and procedures to mitigate adviser conflicts more than is currently required.
Existing protections do not always limit or mitigate potentially harmful adviser conflicts as
robustly as would the combination of these protections with those contained in the IRC
prohibited transactions provisions. As elaborated in Section 3.2.4 below, notwithstanding

250

251

252

LIMRA, 2014, “Variable Annuity Guaranteed Living Benefits Utilization. 2012 Experience. A Joint Study Sponsored by the Society of
Actuaries and LIMRA.”
Consumer Financial Protection Bureau, “Senior Designations for Financial Advisers: Reducing Consumer Confusion and Risks” (April 18,
2013).
U.S. Census Bureau, “Projections of the Size and Composition of the U.S. Population: 2014 to 2060,” March 2015; available at:
https://www.census.gov/library/publications/2015/demo/p25-1143.html.

110

existing protections, there is convincing evidence that advice conflicts are inflicting losses on
IRA investors. Therefore, IRA investors will gain from this final rule’s and exemptions’
extension of fiduciary standards and other contract provisions designed to mitigate the effects of
adviser conflicts on advice and investment decisions.
As noted in Chapter 2, the rules governing retail investment advice can vary depending
on the nature of the advice, the financial products that are being recommended, and whether the
assets are held in an IRA. Under the 1975 regulation certain advice rendered to IRA investors is
already subject to the prohibited transactions provisions of the IRC. Retail advice on securities
investing generally is governed by the Advisers Act, pursuant to which advisers generally must
register with the SEC or a state and adhere to fiduciary standards of care and loyalty to client
interests. However BDs who render investment advice about securities to their clients are
excluded from the Advisers Act if the advice is “solely incidental” to brokerage services, and the
broker receives no special compensation for providing the advice. Instead such BDs and their
representatives must register under the Securities Exchange Act of 1934, deal fairly with clients,
recommend only suitable investments, and seek best execution of trades. The suitability
standard is widely understood to be less exacting than the fiduciary duty to act in a customer’s
best interest.253 In a January 2015 letter announcing its regulatory and examination priorities,
FINRA stated that “[i]rrespective of whether a firm must meet a suitability or fiduciary standard,
FINRA believes that firms best serve their customers – and reduce their regulatory risk – by
putting customer’s interest first. This requires the firm to align its interests with those of its
customers.”254 In practice, however, as detailed in this Regulatory Impact Analysis, BD’s
interests are not so aligned. Moreover, BDs generally are not subject to a fiduciary duty under
the federal securities law, and are subject only to the lower suitability standard.
Advice on relevant insurance products generally is required to be suitable for the
consumer. Variable annuities are regulated as securities and therefore subject to FINRA
standards applicable to BDs. Insurance products that are not regulated as securities – in this
context, fixed and fixed-indexed annuities – are sold by state-licensed insurance agents.
Insurance agents rarely are held to a fiduciary standard, and historically have typically been held
to a negligence standard instead (Beh and Willis 2009). Insurance agents’ advice on annuities
generally is required to be suitable, under state rules that often resemble an NAIC model, which
in turn resembles the FINRA standard. However, state standards are not uniform (nor uniformly
administered) across all states – just 35 states and the District of Columbia have adopted some
version of the NAIC model 255– and one state currently lacks any suitability requirement.256 Still
other federal or state rules may apply where bank representatives recommend bank products.
The type and level of disclosure advisers must provide about their duties to their
customers (and their potential conflicts of interest) likewise vary depending on whether the
adviser is acting as an RIA, BD, insurance agent or other professional.

253
254

255

256

See e.g., Laby (2012, 707, 710, 725-744).
FINRA Regulatory and Examinations Priorities Letter, 2015. FINRA also has stated that suitability also requires consistency with a best
interest standard. Also see FINRA Regulatory Notice 12-25, “Additional Guidance on FINRA’s New Suitability Rule” (May 18, 2012);
available at: http://www.finra.org/web/groups/industry/@ip/@reg/@notice /documents/notices/p126431.pdf.
U.S. Department of the Treasury, Federal Insurance Office, “Annual Report on the Insurance Industry” (Sept. 2015); available at:
https://www.treasury.gov/initiatives/fio/reports-and-notices/Documents/2015%20FIO%20Annual%20Report_Final.pdf.
ACLI 2015 NPRM Comment Letter; (July 21, 2015); available at: http://www.dol.gov/ebsa/pdf/1210-AB32-2-00621.pdf; and Lazaro and
Edwards (2015).

111

Overall, many retail customers do not understand the difference between the different
regimes or know which regime their adviser is subject to.
The protections provided under these different regimes vary substantially. Generally all
but the IRC prohibited transactions provisions permit advisers to provide advice where their own
interests conflict with those of their clients. These regimes tend to rely heavily on disclosure to
mitigate conflicts, but the degree to which and manner in which such conflicts must be disclosed
to clients vary. The general fiduciary duties imposed on investment advisers under the Advisers
Act are enforced under its antifraud provisions. Accordingly, certain conflicts of interest do not
result in violations if the investment advisers provide full and fair disclosures. Other conflicts of
interest, however, may require such disclosure and client consent; and yet other conflicts of
interest may be prohibited under SEC rules that cannot be satisfied by disclosure alone. In
contrast, ERISA and the Code place special emphasis on the elimination or mitigation of
conflicts. Absent an exemption designed to protect the interests of plan participants and IRA
owners, an investment adviser subject to the prohibited transaction rules is forbidden from
giving conflicted advice, regardless of whether he or she has fully disclosed the conflict of
interest.
As elaborated below, conflicts of interest are widespread in retail investment advice
services, disclosure appears to be largely ineffective in mitigating potential harm from such
conflicts, and there is evidence that existing conflicts are associated with large costs in the
aggregate to investors. Broader application of the IRC prohibited transactions provisions would
reduce and/or more effectively mitigate conflicts in advice rendered to IRA investors, and
thereby prevent some harm that other regimes alone fail to prevent.
The wide variety of advisers’ titles and business models and practices sows confusion
among investors and thereby leaves them more vulnerable to harm and/or prone to expensive
errors. The SEC has “expressed concern when specific regulatory obligations depend on the
statute under which a financial intermediary is registered instead of the services provided.”257
SEC staff in 2011 concluded that investors “should not have to parse through legal distinctions,”
but instead should be “protected uniformly when receiving personalized investment advice.”258
Laby (2012) argues that because brokers routinely market their services as advisory, investors’
reasonably expect advice loyal to their interests, and their expectations justify application of a
fiduciary standard of conduct to their advisory activities. Broader application of the IRC
prohibited transactions provisions will provide strong, complementary protections for all
investment advice regarding IRAs.

3.2.1.4

The Department’s Role

A number of comments on the 2015 NPRM argued that the Department is not well
positioned to establish its proposed protections for IRA investors, and suggested that other
agencies, such as the SEC, are better equipped. Some questioned whether certain aspects of the
proposal might fall outside the Department’s authority. A formal examination of the scope of
the Department’s legal authority is outside the scope of this Regulatory Impact Analysis
(although an overview is included in Chapter 2). However, review of relevant legislative and

257
258

SEC Release No. 69013, IA-3558, “Duties of Brokers, Dealers and Investment Advisers” (2013), 5.
SEC Staff Dodd-Frank Study (2011), 101.

112

regulatory history plainly illustrates the Department’s responsibility to regulate advice regarding
IRAs, which was established in 1978259 and underscored in 2006 by the PPA’s addition to
ERISA and the IRC of a statutory investment advice exemption.260
As noted above, since 1978 the Department has been solely responsible for interpreting
and issuing exemptions from the prohibited transactions provisions of both ERISA and the IRC.
As discussed in the Legal Environment section above, since that time the Department has issued
a number of regulations related to the IRC prohibited transactions provisions, as well as a
number of PTEs that grant fiduciary investment advisers certain relief from those provisions.
Notably, pursuant to certain provisions of the PPA,261 the Department issued a number of
regulations and exemptions related to fiduciary investment advice to IRA investors,262
culminating in the 2011 promulgation of a final regulation implementing a statutory PTE for
fiduciary investment advisers to plan participants and IRA investors.263 The regulation includes
strong safeguards to ensure that advice is not tainted by conflicts of interest. Generally, either
the adviser’s compensation must not vary depending on the IRA investor’s investment choices,
or the recommendations must be generated by a computer model that was independently
certified to be unbiased, among other protections. In developing and issuing the regulation, the
Department provided regulatory impact analyses that pointed to research on the potential for
harm from conflicted financial advice as a reason why such strong safeguards were necessary
and why the Department elected not to provide additional, administrative exemptive relief.264
The Department also held a public hearing, in which several witnesses’ testimony addressed the
implications of the statutory PTE, the implementing regulation, and potential additional
exemptive relief for investment advice regarding IRAs.265
Also of note, the PPA specifically charged the Secretary of Labor with determining
whether relief under the statutory PTE could be used by fiduciary advisers in connection with
IRAs.266 To reach its determination, the Department obtained public input via a Request for
Information published in the Federal Register,267 direct outreach to major IRA custodians, and a
public hearing.268 In a 2008 Report to Congress, the Department issued its determination,269
thereby making the aforementioned relief available to fiduciary advisers in connection with
IRAs.

259
260
261
262

263
264
265
266
267
268
269

See Reorganization Plan No. 4 of 1978, 5 U.S.C. § App. (2010).
ERISA §§ 408(b)(14) and 408(g) and IRC §§ 4975(d)(17) and 408(f)(8) as added by the PPA.
Ibid.
See IB 96-1, in which the Department identified categories of investment-related information and materials that do not constitute investment
advice; AOs 97-15A and 2005-10A, in which the Department explained that a fiduciary investment adviser could provide investment advice
with respect to investment funds that pay it or an affiliate additional fees without engaging in a prohibited transaction if those fees are offset
against fees that the plan otherwise is obligated to pay to the fiduciary; and AO 2001-09A in which the Department concluded that the
provision of fiduciary investment advice, under circumstances where the advice provided by the fiduciary with respect to investment funds
that pay additional fees to the fiduciary is the result of the application of methodologies developed, maintained and overseen by a party
independent of the fiduciary, would not result in prohibited transactions.
29 C.F.R. 2550.408g-1 and 408g-2.
74 Fed. Reg. 60156 (Nov. 20, 2009); 76 Fed. Reg. 66136 (Oct. 25, 2011).
A transcript of the hearing is available at: http://www.dol.gov/ebsa/pdf/investmentadvicetranscript102108.pdf.
PPA § 601(b)(3)(B), Pub. .L. No. 109-280, 120 Stat. 965.
71 Fed. Reg. 70427 (Dec. 4, 2006).
72 Fed. Reg. 34043 (June 20, 2007).
DOL Report to Congress (Aug. 21, 2008); available at: http://www.dol.gov/ebsa/publications/reporttocongress.html.

113

As this history demonstrates, the Department’s role regulating fiduciary investment
advice to IRAs long predates the 2010 Proposal – it was established 35 years prior and was
recently explicitly recognized and expanded by the PPA in 2006. The new rule and exemptions
fit squarely within the Department’s scope of responsibility to interpret the IRC prohibited
transactions provisions and issue PTEs in connection with investment advice regarding IRAs.

3.2.2 Market Changes Since 1975
Comments on the 2015 NPRM Regulatory Impact Analysis generally did not dispute its
characterization of changes that have affected the IRA advice marketplace since 1975. Rather
the comments described in rich detail a current market profoundly different from that which
existed then.
In the Department’s view, market changes have rendered the 1975 regulation ineffective,
exposing IRA investors to excessive losses attributable to adviser conflicts.
Retirement savings in 1975 existed mostly in the form of DB pensions270 and DC plans
in which investment choices were made mostly by plan managers and not participants.271 IRAs
had just been enacted. In the private sector, ERISA in 1974 established fiduciary duties for the
individuals who chose plan investments, and for individuals who advised with respect to such
choices. The 1975 regulation was drafted in an environment where its application was mostly to
advice rendered to plan managers; that is, to institutional investors, not to consumers.
Today’s retirement savings marketplace is dramatically different from that which existed
when the 1975 regulation was issued. Compared with 1975, America’s workers and retirees
today are far more responsible for providing for their own retirement security. At the same time,
the investments available to them have grown in variety and complexity. Their need for
investment advice or other effective support is great and growing.
The market for investment advice and other support is likewise changing rapidly. The
types of help available are multiplying. Distinctions between the functions of different types of
professionals have blurred. The web of relationships and revenue streams between product
manufacturers, distributors, and advisers has become more intricate and less transparent,
multiplying opportunities for conflicts of interest to taint advice. This growing complexity
breeds confusion among consumers, making them more vulnerable to abuse.

3.2.2.1

Changes in Retirement Savings

Comments on the 2015 NPRM Regulatory Impact Analysis largely confirmed its
characterization of changes in retirement savings since 1975. The extent of an individual’s
responsibility for providing for his or her own retirement security depends on the type of
retirement savings or benefit program he or she relies on to achieve that security. DB plans
typically provide participants with a specified benefit – the worker or retiree has no
responsibility for investment decisions. DC plan participants usually are responsible for

270

271

See Private Pension Plan Bulletin Historical Tables and Graphs, U.S. Department of Labor, Employee Benefits Security Administration
(Sept. 2015); available at: http://www.dol.gov/ebsa/pdf/historicaltables.pdf.
The law creating 401(k) plans was not effective until Jan. 1, 1980.

114

investing their own accounts (although this was less common in 1975 than it is today272).
However, their choice is usually limited to a menu of options pre-selected for them by a
responsible plan fiduciary.273 The menu often features a default option,274 chosen by the
fiduciary to be well suited to the needs of many participants. Investment advice provided to
participants often is understood by the advisers to be fiduciary advice under the 1975 regulation,
comments on the 2015 NPRM suggest.
IRA investors, in contrast, are fully responsible for choosing investments (or hiring a
professional to choose for them) from among a near endless variety of securities, financial
products, and other property in which they are permitted by law to invest their IRAs. There is no
fiduciary responsible for constructing a menu or identifying an appropriate default option. And
advisers generally do not consider the advice they render to IRA investors to be fiduciary advice
under the 1975 regulation, according to comments on the 2015 NPRM.
American workers and retirees today are far more responsible for providing for their own
retirement security than they were in 1975, due to a major decline in the role of DB plans, a
corresponding increase in the role of DC plans (and a shift toward more participant direction of
investment in these plans), and an even larger increase in the role of IRAs. In 1975, IRAs had
just been established (when ERISA was enacted in 1974). By 1984, IRAs still held just $159
billion in assets, compared with $589 billion in private-sector DB plans and $279 billion in
private-sector DC plans (See Figure 3-7). By the end of the third quarter of 2015, in contrast,
IRAs held $7.3 trillion, far surpassing both DB plans ($2.8 trillion) and DC plans ($5.2
trillion).275 If current trends continue, DB plans’ role will decline further, and IRA growth will
continue to outstrip that of DC plans as the workforce ages, the baby boom generation retires,
and more DC accounts (and sometimes lump sum payouts of DB benefits) are rolled into IRAs.
Almost $2.4 trillion is projected to be rolled over from plans to IRAs between 2016 and 2020.276

272

273
274
275

276

This is due to the fact that participants became more responsible for managing the investments in their accounts when 401(k) plans were
created. The law creating them did not become effective until Jan. 1, 1980.
Plan Sponsor Council of America’s (PSCA), “58th Annual Survey Reflecting 2014 Plan Experience,” Tables 64, 70 and 71.
Ibid., Table 116.
Federal Reserve Board, “Financial Accounts of the United States, 1945–2014,” Federal Reserve Statistical Release Z.1, (Dec. 2015).
DB assets do not include claims of pension fund on sponsor. Also see ICI, “U.S. Retirement Market, Third Quarter 2015,” 2015.
Households and non-profit organizations sector refers to brokerage accounts held by households.
Cerulli Associates, “Retirement Markets 2015,” Exhibit 9.08.

115

Figure 3-7 Retirement Assets
8
7

ERISA DB

6
$Trillions

5
4
3

ERISA DC
IRA

2
1
0
1981 1985 1989 1993 1997 2001 2005 2009 2013

Source: Federal Reserve Board, Financial Accounts of the
United States

IRAs’ growth has made more middle- and lower-income families into investors, and
sound investing more critical to such families’ retirement security. As a result the pool of
consumers needing expert financial advice or other support is growing to include more modest
income families, who often lack financial expertise.
As more families have invested, investing has become more complicated. As IRAs grew
during the 1980s and 1990s, their investment pattern changed, shifting away from bank products
and toward mutual funds (see Figure 3-8).277 Bank products typically provide a specified
investment return, and perhaps charge an explicit fee. Single issue securities lack diversification
and have uncertain returns, but the expenses associated with acquiring and holding them
typically take the form of explicit up-front commissions and perhaps some ongoing account
fees.278 Mutual funds are more diversified (and in this respect can simplify investing), but also
have uncertain returns, and their fee arrangements can be more complex, and can include a
variety of revenue sharing and other arrangements that can introduce conflicts into investment
advice and that usually are not fully transparent to investors.

277
278

Federal Reserve Board, “Financial Accounts of the United States, 1945–2014,” Federal Reserve Statistical Release Z.1 (Dec. 2015).
The transparency of fees associated with single issue securities should not be taken to suggest that conflicts of interest are not a concern in
this area. As discussed later, conflicts can be harmful even when the presence and magnitude of the conflict is known, and disclosure alone
is rarely a sufficient remedy.

116

Figure 3-8 Share of IRA Assets
100%
90%

Households and nonprofit organizations

80%

Mutual funds

70%
60%

Money market

50%
40%

Life insurance companies

30%
Credit unions

20%
10%

Bank products

2014

2011

2008

2005

2002

1999

1996

1993

1990

1987

1984

1981

0%

Source: Federal Reserve Board, Financial Accounts of the United States

Insurance products’ share of IRA assets has remained relatively steady of late. At the
same time, however, the annuities market has changed significantly. Fixed-rate annuities, which
were once dominant, have yielded to variable, and more recently fixed-indexed, annuities. In
2012, total U.S. individual annuity sales were $219 billion. Out of $219 billion, 67 percent
($147 billion) of total sales were attributed to variable annuity sales. Variable annuity sales
declined for three consecutive years after peaking in late 2011.279 In contrast, fixed-indexed
annuity sales hit record levels in 2014. While this evolution has brought more choice and
flexibility for retirement investors, it has also brought increasing complexity, as detailed later in
this chapter. The constant development and introduction of new products also can present
challenges to regulators in promoting fair competition for companies and ensuring appropriate
safeguards for consumers.
In 2014, the IRA market accounted for nearly two-thirds of fixed-indexed annuity sales,
compared with less than one-half of variable annuity sales and just more than one-third of fixedrate annuity sales (Figure 3-9).
Figure 3-9 The Share (%) of Deferred Annuity Contracts Sold by Market and
Type of Product in 2014
Variable
Fixed-rate
Indexed
Total

IRA

Non-Qualified

Employer Plan

Total

44%
36%
62%
47%

28%
58%
36%
36%

27%
6%
2%
17%

100%
100%
100%
100%

Source: DOL’s own calculation using LIMRA Annuity Yearbook-2014

279

LIMRA, U.S. Individual Annuity Yearbook - 2014 (2015).

117

Sales of variable annuities in the IRA market declined for three consecutive years (Figure
3-10). In contrast, sales of fixed-indexed annuities in the IRA market have steadily increased
and accounted for about 27 percent of total sales of the annuities in the IRA market in 2014
(Figure 3-11).
Figure 3-10 Deferred Annuity Sales in the IRA Market by Product Type ($ in Billions)
2009
2010
2011
2012
2013
Variable
60.3
71.0
81.7
74.4
71.4
Fixed-rate
21.5
11.0
9.2
8.0
10.7
Fixed-Indexed
16.4
17.8
18.6
20.0
23.2
IRA Total
98.2
99.8
109.5
102.4
105.3

2014
69.6
11.0
29.2
109.8

Source: LIMRA Annuity Yearbook 2010, 2011, 2012, 2013, 2014.

Figure 3-11 % Share of Deferred Annuity Sales in the IRA Market by Product Type
2009
2010
2011
2012
2013
Variable
61%
71%
75%
73%
68%
Fixed-rate
22%
11%
8%
8%
10%
Fixed-Indexed
17%
18%
17%
20%
22%
IRA Total
100%
100%
100%
100%
100%

2014
63%
10%
27%
100%

Source: DOL’s own Calculation based on LIMRA Annuity Yearbook 2010, 2011, 2012, 2013, 2014.

Figure 3-12 Year-End Deferred Annuity Assets in the IRA Market by Product Type ($ in Billions)
2009
2010
2011
2012
2013
2014
Variable
451
522
546
667
724
735
Fixed-rate
140
145
148
155
152
152
Fixed-Indexed
78
94
105
117
144
167
IRA Total
669
761
799
939
1020
1054
Source: LIMRA Annuity Yearbook 2010, 2011, 2012, 2013, 2014.

While fixed-indexed annuity sales are rapidly gaining market share compared to variable
annuity sales, in terms of assets, variable annuities still account for 70 percent of IRA annuity
investments and fixed-rate and fixed-indexed annuities have nearly equal shares of the remainder
(Figure 3-13).
Figure 3-13 % Share of Year-End Deferred Annuity Assets in the IRA Market by Product Type
2009
2010
2011
2012
2013
2014
Variable
67%
69%
68%
71%
71%
70%
Fixed-rate
21%
19%
19%
17%
15%
14%
Fixed-Indexed
12%
12%
13%
12%
14%
16%
IRA Total
100%
100%
100%
100%
100%
100%
Source: DOL’s own Calculation based on LIMRA Annuity Yearbook 2010, 2011, 2012, 2013, 2014.

Even though fixed-rate, fixed-indexed and variable annuity products differ in important
ways, they also have similarities that impact the investment component, insurance component
and fees associated with the respective products. Variable annuities are regulated as securities,
while fixed-indexed annuities and fixed-rate annuities are regulated through state insurance
regulation. Fixed-rate annuities have specified interest rates set by the insurance company,
118

subject to minimum requirements under state insurance laws. In contrast, fixed-indexed annuity
contracts provide crediting for interest based on changes in a market index.280 These
investment-oriented features differentiate fixed-indexed annuities from fixed-rate annuities
although both products are treated as exempt securities under current federal law.
According to one industry survey report, there are 317 indexed annuities as segmented by
product and 1,648 index annuity strategies as segmented by index-crediting method. The
selection of the crediting index or indices is an important, and often complex, decision. Today
many indexed annuity carriers offer bond or equity indices beyond the S&P 500, which in earlier
years dominated fixed-indexed product designs. In the 4th quarter of 2014, approximately 55
percent of indexed annuity sales involved products linked to the S&P 500 followed by 27.8
percent of annuity sales that were based on hybrid indices (i.e., indices that are derived from one
or more other indices). However, there are also products in the market that use gold or
international emerging equity market indices as their index.281 In fixed-indexed annuities,
indexed-linked gains are generally not fully credited. How much of the index gain that is
credited depends on the particular features of the annuity such as participation rates, interest rate
caps, and spread/margin asset fees.282 In contrast, in a fixed-rate annuity the interest rate for any
crediting period is set by the insurance company and is not tied to a market or other index.
These annuity products may offer insurance features such as death benefits and
guaranteed living benefits. There are three types of guaranteed living benefits - guaranteed
minimum income, guaranteed minimum accumulation, and guaranteed minimum withdrawal
(including lifetime withdrawal benefits).283 But these benefits may come at an extra cost and,
because of their variability and complexity, may not be fully understood by the consumer.

3.2.2.2

Changes in Advice Provision

As more American workers have become IRA investors, the types of investment advice
services available to them have changed and multiplied.284 Compared with 1975, today’s
services are more likely to involve a wider variety of conflicts of interest, operate under a wider
variety of rules, and saddle consumers with more confusion and more varied risks of abuse.
Before 1975, brokerage and advisory services were relatively distinct. Brokerage mostly
involved execution of trades. These trades involved substantial labor input, commissions were
fixed in law, and BDs and their representatives could and did derive their revenue mostly from
commission payments for execution. BD representatives’ advice was limited and mostly
incidental to transactions, and therefore comfortably excluded from regulation under the
Advisers Act. Advisory services were understood to be different and separate from brokerage,
and regulated under the Advisers Act. Advisers were compensated mostly by means of assetbased advisory fees, and generally were subject to a fiduciary standard of conduct toward their
retail customers. Also at that time the investments on which advice was rendered were less
likely to involve complex fee arrangements that can introduce a variety of less transparent
conflicts into advice. For example, in 1975 there were just 426 U.S. mutual funds holding $46

280
281
282
283
284

NAIC Buyer’s Guide for Deferred Annuities, 2013.
Wink’s Sales & Market Report, 4th Quarter, 2014, Part 1.
FINRA Investor Alert “Equity-Indexed Annuities: A Complex Choice” 2012.
LIMRA, “Variable Annuity Guaranteed Living Benefits Utilization: 2012 Experience.”
For a fuller discussion of some of these changes, see Laby 2012 (726-731).

119

billion in assets. In 2014, more than 7,900 mutual funds held nearly $16 trillion.285 Almost
contemporaneous with Congress’ passage of ERISA, changes under the securities laws created
competitive pressures that motivated BDs and their representatives to provide services in
addition to transaction execution, including research and fuller financial advice. In 1975, the
SEC and Congress deregulated fixed commissions. Discount brokers entered the business. As
years passed, technological advances facilitated deeper discounts. Two-tier pricing emerged
consisting of high-priced, full-service brokerage bundled with personalized financial advice, and
low-priced, discount brokerage with no or limited ancillary services.
In 1983, the FDIC made clear that banks are permitted to provide discount brokerage
services. From 1980 to 1992, discount brokers’ market share of retail commissions grew from
1.3 percent to 12.9 percent.286 The available commission rates for retail customers fell
substantially. During the mid-1990’s commissions for a 100-share trade with a full-service BD
ranged from $75 to $150. By 1996, discount brokers introduced online trading. Soon, online
brokers were offering commissions as low as $7 per trade (Bakos et al. 1999, 4).
As noted earlier, BDs who receive a special fee for investment advice generally must
register with the SEC or a state pursuant to the Advisers Act and assume fiduciary duties.287 The
higher commissions associated with full service brokerage might appear to be (and arguably
often function as) a special fee for advice. The SEC recognized this tension. It also recognized
that BD representatives who give fuller financial advice and are compensated by transactionbased commissions have an incentive to recommend higher trading volumes than would be
optimal for their customers. To address both the legal tension and the conflict, the SEC
proposed in 1999 to essentially waive the special-fee condition to avoid registration under the
Adviser Act, by allowing certain non-discretionary fee-based brokerage accounts provided that
BDs include prominent statements that the account is a brokerage account and not an advisory
account and that the BDs’ interests may not always be the same as the customer’s in
advertisements, contracts and other documents.288 However, a group representing RIAs who
objected to this policy successfully challenged it in court, and the rule was vacated.289
As advice services evolved, so did the means by which they were compensated290
particularly for BD representatives recommending and selling mutual funds. In 1980, the SEC
issued rule 12b-1, which permitted mutual funds to pay “distribution fees” to BDs to promote
and sell the funds.291 So-called 12b-1 fees largely precipitated the development of the different
mutual fund share classes available today. Different classes generally carry different investor
costs to buy, sell, or hold what is otherwise the same fund, and entail different compensation
streams from the mutual funds to the BDs that distribute them. Mutual fund asset managers also
frequently share revenue with BDs who distribute the funds they manage. BDs in turn can share
this compensation in various ways with their representatives who recommend the funds.

285
286

287
288

289
290
291

ICI, “2015 Investment Company Fact Book” 2015.
SEC, “Market 2000, An Examination of Current Equity Market Developments, Structure of the U.S. Equity Markets” (Jan. 27, 1994) at p.
II-5; available at: https://www.sec.gov/divisions/marketreg/market2000.pdf.
15 U.S.C. § 80b-2(a)(11)(C).
SEC Proposed Rule, “Certain Broker-Dealers Deemed Not to be Investment Advisers,” 64 Fed. Reg. 61226, Release Nos. IA-1845, 3442099; File No. S7-25-99 (1999).
See Financial Planning Association v. SEC, 482 F.3d 481 (D.C. Cir. 2007).
For a fuller discussion of some changes in adviser compensation, see Howat and Reid (2007).
17 C.F.R. 270.12b-1.

120

Because of these various compensation practices, BD representative compensation can vary
depending on what fund and what share class their customers select. This creates a conflict that
can bias their recommendations. These conflicts often are not transparent to investors, even if
they are financially sophisticated.
Compensation arrangements that create conflicts in advice extend beyond BD-sold
mutual funds, however. For example, BDs often sell securities, such as corporate bonds, to retail
customers out of their own accounts at mark-ups that are not transparent.292 Nor are such
conflicts limited to BD representatives. For example, many RIAs receive variable compensation
other than asset-based fees or are affiliated with other entities (generally, BDs or insurance
agents or brokers) that do, and while this is disclosed in general terms to their customers, the
disclosures generally do not quantify the conflict that pertains to a particular recommendation
and often are not understood or even read by investors (Hung et al. 2008). Insurance agents and
brokers who distribute and recommend products that are not registered as securities typically are
compensated by commission and may be otherwise rewarded for achieving various sales goals.
The conflicts facing a particular adviser can become more numerous and complicated if that
adviser is authorized to act in more than one capacity, as a BD representative, RIA, and/or
insurance agent or broker, a practice sometimes referred to as “hat-switching,” or if the adviser
is affiliated with BDs or insurance agents or brokers, or with other advisers who wear different
hats. This poses a particular problem to retail customers, many of whom are not aware of the
differences in regulatory approaches for these entities and the differing duties that flow from
them.
Many of the trends in retail investing since 1975 have been favorable to consumers.
Discount brokerage in particular has reduced many investors’ trading costs. This, together with
competition and growth in the mutual fund industry, has contributed to substantial declines in
mutual fund loads and expense ratios293 (although the total net effect on mutual fund investor
results is less certain294). In recent years, new technologies and innovations in financial products
appear to be making advice and other potentially effective investment support more affordable
and available to many consumers. Some of these newer business models lean toward
independence in advice (but absent policy changes such as those included in the Department’s
final rule and exemptions, they likely would face the same competitive pressures that led more
conflicted models to prevail in the past).
Notwithstanding these positive developments, however, the major changes in advice and
compensation arrangements and associated conflicts of interest since 1975 compelled the

292

293

294

Ferrell (2011) reports that, in the market for lower-priced, less liquid equities, mark-ups and mark-downs have decreased in size over the
last 40 years. However, he also finds that a BD’s principal status and solicitation of trades are associated with larger mark-ups. It is not
clear whether his finding would hold in the very different market for investment grade corporate bonds, where IRA investors are more
likely to be active. The BD’s financial incentive to maximize mark-ups is facially the same in both markets, however, which raises concern
that, because of BD conflicts, IRA investors may sometimes pay more than fair prices for corporate bonds. In corporate bond markets,
Trade Reporting and Compliance Engine (TRACE), lowered costs for retails investors and squeezed dealer revenue and service
(Bessenbinder and Maxwell, JEP Spring 2008). Yet spread arguably is high at 1.24 percent for a $20,000 trade (Edwards, Harris and
Piwouar, 2007).
ICI,” 2015 Investment Company Fact Book, Chapter 5: Mutual Fund Expenses and Fees,” 2015; available at:
http://www.icifactbook.org/fb_ch5.html.
Investor results are further affected by fund performance and timing of trades, and are generally known to lag the performance of funds
themselves. Dalbar, “2014 Quantitative Analysis of Investor Behavior;” available at:
http://www.dalbar.com/ProductsServices/AdvisorSolutions/QAIB/tabid/214/Default.aspx.

121

Department to reexamine the 1975 regulation. All of the trends discussed directly affect IRAs
and therefore retirement security. The increasing complexity and variety in advisory services,
and related compensation arrangements and consumer protections, cause confusion among
consumers – a conclusion reached by GAO295 and the CFPB,296 and supported by a carefully
researched study by RAND for the SEC (Hung et al. 2008). Palaveev (2008) describes how BD
representatives have adopted a new role as advisers who control client relationships, and “the
center of the relationship has shifted from the product to the skills of the advisor.” Conflicts of
interest associated with many of these relationships raise serious concerns that advice will
sometimes be biased and IRAs will be vulnerable to abuse. Palaveev recommends that advisers
who produce revenue for BDs should be aware of BDs’ “hidden profit centers,” that stem from
“marketing fees from mutual fund[s] and investment management funds,” which can “represent
a conflict of interest, because [BD]s have [an] incentive to promote such funds and programs
even if they aren’t in the long-term interest of clients.” Palaveev’s article reveals how BDs and
their producing advisers compete with each other for revenue and profit, often at investors’
expense.
A contingent commission is an arrangement in which an insurance agent or broker
receives a percentage of the premiums realized by the insurer, if the agent of broker meets
certain goals in terms of volume, persistency, and profitability in the business it places with the
insurer (Cheng et al. 2010). Contingent commissions are one of the ways that the insurer makes
sure the insurance agent’s incentive is aligned with the insurer’s interest. This function of
contingent commissions – aligning the incentives between insurance agents and insurance
companies for their mutual profit – has been reported in several empirical studies (Ghosh and
Hillard 2012; Cheng et al. 2010). When a lawsuit targeting large insurance brokers for
inappropriate uses of contingent commissions was filed in New York in 2004, stock prices of
both the insurance brokers and the insurance companies heavily relying on contingent
commissions plummeted. (Ghosh and Hillard, 2012; Cheng et al. 2010). Another empirical
study found that contingent commissions distort sales by insurance agents and tilt sales toward
the insurers with such market service agreements (Wilder 2002). These studies all empirically
show that contingent commissions align the insurance agent or broker’s incentive with the
insurance company, not with the consumer. These studies examine the commercial propertycasualty insurance market, not the annuity insurance market. However, the conflicts of interest
between insurance agents and consumers are relevant and applicable in the annuity market as
well. If anything, the potential harm from conflicts of interest would be larger in the annuity
market because purchasers of annuities are often older individuals who are less sophisticated in
financial matters than the purchasers of commercial property-casualty insurance.
Figure 3-14 compares three different types of deferred annuities. A deferred annuity has
an accumulation phase and a payout phase. In the accumulation phase, the owner pays a
premium or premiums into the contract and accumulates an account value. In the payout phase,
the owner receives payouts following his or her election to convert or “annuitize” the account
value into a stream of income. Figure 3-14 focuses on the accumulation phase.

295
296

GAO Publication No. GAO-11-235.
See CFPB “Senior Designations for Financial Advisers,” 2013.

122

In Figure 3-14, features of the three types of annuities are compared in three categories:
“allocation of investment risk,” and “guaranteed optional benefits,” and “fees.” All deferred
annuities have investment components as well as insurance components. In examining
investment components, Figure 3-14 shows that the insurance company bears the investment risk
in a fixed-rate annuity, because the insurer guarantees a minimum interest rate at the beginning
of crediting period. In contrast, in a variable annuity, the investment risk is borne by the contract
owner because the account value fluctuates based on the performance of underlying funds.
Fixed-indexed annuities fall between fixed-rate annuities and variable annuities in terms of the
extent to which insurers bear investment risks. In fixed-indexed annuities, insurers generally
guarantee at least a zero return. However, as long as the return is above the minimum guarantee,
the actual return on a fixed-indexed annuity is not determined until the end of the crediting
period and is based on the performance of a specified index or other external reference. Similar
to variable annuities, the returns of fixed-indexed annuities can vary widely, which results in a
risk to investors. Furthermore, insurers generally reserve rights to change participation rates,
interest caps, and fees, which can limit the investor’s exposure to the upside of the market and
effectively transfer investment risks from insurers to investors.
Figure 3-14 also shows that fixed-indexed annuities are as complex as variable annuities,
if not more complex. Traditionally, common indexes used in fixed-indexed annuities were
equity indexes such as the S&P 500 or Dow Jones Industrial Average. Although annuities using
the S&P 500 index still represent the majority of fixed-indexed annuity sales in 2014, various
alternative indexes - including gold and a hybrid derived from one or more other indexes – have
gained market share.297 In addition, there are several methods for determining changes in the
index such as point-to-point, annual reset, high-water-mark, and low-water-mark.298 Because
different indexing methods can result in varying rates of return, investors need to understand the
trade-offs that they make by choosing a particular indexing method. The rate of return is further
affected by participation rates, cap rates, and the rules regarding interest compounding.
Understanding all these different options and their impacts on returns requires significant
time and expertise from investors. In this regard, investors in fixed-indexed annuities are acutely
dependent on financial advice they receive from broker-dealers and insurance agents. As shown
in Figure 3-14, fixed-indexed annuities are distinguished from fixed-rate annuities by their
complex designs and the exposure to investment risks, and have many similarities with variable
annuities. Unbiased and sound advice is important to all investors but it is even more crucial in
guarding the best interests of investors in fixed-indexed annuities and variable annuities.

297
298

Wink’s Sales & Market Report, 4th Quarter, 2014.
The point-to-point method compares the index values at the beginning of the term to the end of the term. The annual reset method compares
the index value at the beginning of each contract year to the end of that year. The high-water-mark method measures the difference between
the highest index value at various points during the term and the index value at the start of the term. The low-water-mark method measures
the difference between the index value at the end of the term and the lowest index value at various points during the term. In all four
methods, interest is added to the annuity at the end of the term. (NAIC Buyers’ Guide to Fixed Deferred Annuities with Appendix for
Equity-Indexed Annuities, 1999).

123

Overview

Figure 3-14 Comparing Different Types of Deferred Annuities
Fixed-Rate

Fixed-Index

Variable

• A contract providing a
guaranteed, specified rate of
interest on premiums paid.

• A contract providing for the
crediting of interest based on
changes in a market index.

• Premiums are guaranteed to
earn at least a minimum
specified interest rate. The
insurance company may in its
discretion credit interest at
rates higher than the
minimum.
• Under most current states
laws, upon surrender of the
contract the buyer is
guaranteed to receive at least
87.5% of premiums paid,
credited with a minimum
interest rate such as 1%. This
is known as the Nonforfeiture
Amount.

• Returns are less predictable
because the interest credited at
the end of each index period
depends on changes in a market
index.

• Returns are variable based on the
performance of underlying funds in
the subaccounts.1

• The surrender value must
always equal at least the
Nonforfeiture Amount and the
interest rate is guaranteed to
never be less than zero during
each index period.

• The insurance company does not
guarantee investment performance.
Investment risk is borne by the
contract owner.

• In general, returns depend on
what index is linked and how
the index-linked gains are
calculated.3 Many current
product designs offer
alternatives to traditional
indexes such as the S&P 500
and allow owners to allocate
premiums among different
indexes. These alternative
indexes may include precious
commodities, international and
emerging markets, and
proprietary indexes developed
by insurance companies.
• Changes in the index can be
determined by several methods
such as annual reset, high water
mark, low water mark, point-topoint, and index averaging.3

• A variable annuity contract can
offer hundreds of subaccounts and
generally allows owners to transfer
or reallocate their account values
among the various subaccounts.

• A contract with an account value
that rises or falls based on the
performance of investment options,
known as “subaccounts,” chosen by
the contract owner.

Allocation of Investment Risk

Returns

Returns
• Index-linked gains are not
always fully credited. How
much of the gain in the index
will be credited depends on the
particular features of the
annuity such as participation
rates, interest rate caps, and
spread/margin/asset fees.3

124

• The insurer generally reserves
the right to change participation
rates, interest rate caps, and
spread/margin/asset fees,
subject to minimums and
maximums specified in the
contract.3

Fees

Surrender Charges & Surrender Period
• If the owner withdraws all
or part of the value out of the
annuity within a specified
period, surrender charge will
be applied.1
•The buyer can often receive
a partial withdrawal (usually
up to 10%) without paying
surrender charges1 and the
charge may be waived in
certain circumstances, such as
confinement in a nursing
home.
• State laws generally require
“free-look” provisions under
which the owner can return
the contract free of charge
within a stated number of
days after purchase.2
•Some annuities have a
market value adjustment
(MVA). If at the time of
surrender interest rates are
higher than at the time of
purchase, the MVA could
reduce the amount paid on
surrender; conversely, if
interest rates have fallen, the
MVA could increase the
surrender value.1,2

• Same as fixed-rate.

• Same as fixed-rate.

• Same as fixed-rate.

• Same as fixed-rate.

• Same as fixed-rate.

• Same as fixed-rate.

•Generally no express fees6

• Generally no express fees6
• Often sold with a guaranteed
lifetime withdrawal benefit,
which requires a rider fee.

• Same as fixed-rate.

Other Fees & Charges
• Contract Fee2
• Transaction Fee
• Mortality and Expense risk fee
• Underlying fund fees
• Additional fees or charges for
certain product features (often
contained in “riders” to the base
contract) such as stepped-up death
benefits, guaranteed minimum
income benefits, and principal
protection.4

125

Guaranteed Optional Benefits

Guaranteed Living Benefit Riders7
•Seldom offered.

• The most popular benefit, the
guaranteed lifetime withdrawal
benefit, is offered with 84% of
all new fixed indexed annuity
sales in 2014.5

• Annuities pay a death
benefit to the beneficiary
upon death of the owner or
annuitant during the
accumulation phase.2 Benefit
is typically the greater of the
accumulated account value or
the Nonforfeiture Amount.
Different rules govern death
benefits during the payout
phase.

• Same as fixed-rate.

• Contracts constituting 83% of all
new variable annuity sales in 2014
offered guaranteed living benefit
riders.5

Death Benefit
• If the owner dies during the
accumulation period, the
beneficiary generally receives the
greater of (a) the accumulated
account value or (b) premium
payments less prior withdrawals.
An enhanced guaranteed minimum
death benefit may be available for
an additional fee.8

Sources:
1: NAIC Buyer’s Guide for Deferred Annuities, 2013
2: NAIC Buyers’ Guide to Fixed Deferred Annuities with Appendix for Equity-Indexed Annuities, 1999
3: FINRA Investor Alert “Equity-Indexed Annuities: A Complex Choice,” 2012
4: FINRA Investor Alert “Variable Annuities: Beyond the Hard Sell,” 2012
5: LIMRA “U.S. Individual Annuity Yearbook 2014”
6: The insurer covers its expenses via the margin of premiums received over the cost of the annuity benefits, commonly referred to a
“spread.”
7: Guaranteed living benefits are available for additional fees and generally protect against investment risks by guaranteeing the level of
account values or annuity payments, regardless of market performance. There are three types of guaranteed living benefits—
guaranteed minimum income, guaranteed minimum accumulation, and guaranteed minimum withdrawal (including lifetime withdrawal
benefits).
8: Some fixed-indexed annuities also offer this benefit for an additional fee.

126

3.2.3 The IRA Advice Market
A number of comments on the 2015 NPRM Regulatory Impact Analysis argued that the
IRA advice market currently functions well, and that advisers’ potential conflicts do not bias
their advice in ways that harm IRA investors. This Regulatory Impact Analysis examines and
rejects those arguments in this chapter and in Chapter 8. These and other comments, however,
generally affirmed the Regulatory Impact Analysis’s characterization of adviser compensation
arrangements and market practices.
In the Department’s view, economic theory suggests that IRA advisers’ conflicts are
likely to harm IRA investors. According to academic literature, it is likely that advisers’
conflicts often bias their advice, and IRA investors often follow biased advice. The nature,
theory, and evidence of this market failure are investigated in detail throughout Section 3.2.3.
The proliferation and adherence to biased advice results in social welfare losses – IRA investors
make suboptimal decisions about their purchases of advice and, following biased advice, about
their investments. Suboptimal investment decisions may allocate capital inefficiently in the
national economy. It also results in transfers, as advisers and producers of the products they
recommend capture surplus from IRA investors. Both of these effects erode IRA investors’
retirement security.
The market for IRA advice exhibits at least three noteworthy characteristics, which
together may render IRA investors vulnerable to harm from advisers’ conflicts. First, conflicts
are widespread in the market even in spite of the existing regulatory framework (see Section
3.2.3.1 below). Second, advisers incur substantial costs pursuing IRA customers, and IRA
investors ultimately bear such cost (see Section 3.2.3.2 below). Third, and almost certainly
underlying the other two, IRA investors face high “information costs” – i.e., they face barriers in
evaluating the quality of advice (see Section 3.2.3.3 below).

3.2.3.1

Existence of Conflicts

Conflicts of interest are widespread and often acute in the market for IRA investment
advice. Many IRA advisers, including many BDs, RIAs, insurance agents, and bank
representatives, are conflicted. Figure 3-15 illustrates conflicts present when BDs distribute
mutual funds.299 Advisers often have an interest in recommending products that are proprietary
to their employers or their employers’ affiliates, or that generate greater revenue for themselves,
their employers, or affiliates.300

299
300

See Section 3.2.3.3.3 for an explanation of the color scheme used in Figure 3-15.
This discussion is not intended to be exhaustive with respect to compensation arrangements that may introduce conflicts into investment
advice. For some additional discussion of the types of conflicts affecting such advice, see Howat and Reid (2007), Hung et al. (2008),
Turner and Muir (2013), and Robinson (2007).

127

Figure 3-15
Some Common Conflicts in Advice:
Full Service Brokerage IRAs
Brokerage Commissions

Mutual Fund

Distribution (12b1)

Shares, Returns Net of Fees

IRA Accounts Loads

Deducts Fees
from Assets

Represented on MF Board

Selects Broker

IRA

Might
be the
same
BD

Investment
Adviser to MF

Load sharing, Other fees (e.g. subaccounting)
Mark-Ups
Corp. Bonds

Broker-Dealer
Firm for MF

Broker-Dealer
Firm for IRA
(Sells MFs as
agent, corp.
bonds as
principal)

Cash, Variable commissions

BD
Representative
Variable compensation

BDs and their representatives often have a financial stake in the investment decisions that
IRA investors make pursuant to the representatives’ advice. BDs and their representatives often
stand to gain if IRA investors trade more, buy or hold certain mutual funds or other products, or
buy securities out of the BD’s own inventory. The attendant conflicts often play out at two
levels: variation in the revenue received by the BD, and variable compensation paid by the BD
to its representatives who render IRA advice. Figure 3-15 provides a simplified representation
of some of the common payments and relationships that can give rise to such conflicts.
Mutual funds compensate the BDs that distribute them in various ways, and RIAs acting
as mutual fund asset managers also often share revenue with BDs who distribute the funds they
manage. BDs share this compensation in various ways with their representatives who
recommend funds to IRA investors and other retail clients.
Many of the mutual fund shares distributed through BDs are so-called “class A” shares,
which charge a front-end-sales-load. The mutual fund’s principal underwriter typically shares
this load with the BD who distributed the shares. Many mutual funds also deduct so-called 12b1 fees from fund assets to pay distribution costs. Some of this fee often is paid to the distributing
BD, perhaps as compensation for selling the shares, sometimes called a “trailing commission,” or
for promoting the fund to customers, sometimes called a payment for “shelf space.” The mutual
fund might pay the distributing BD to perform services, such as “sub-accounting,” where the BD
aggregates many customer accounts to act as one large shareholder, relieving the mutual fund
from administering many small accounts. The mutual fund also pays an RIA to manage the
fund’s assets, and that adviser may share some of the revenue earned with BDs who distribute

128

the fund. Different mutual funds provide different combinations of these payments, in different
amounts, to distributing BDs, so the BDs’ revenue will be increased if IRA investors select
mutual funds that provide more and larger payments.
Additional conflicts can arise if the distributing broker also executes trades for the mutual
fund.301 The mutual fund’s adviser may arrange for the mutual fund to pay the BD more than the
lowest available commissions in nominal exchange for providing the adviser with research or
other services that help the adviser manage the fund’s assets, in a practice known as “soft
dollars” (because there is no explicit or “hard dollar” fee paid for the service).302
BD conflicts are not limited to those associated with the distribution of mutual funds.
BDs’ revenue can likewise vary in connection with their distribution of other financial products,
such as variable annuities. Variable annuities often carry larger commissions than mutual funds,
and therefore may sometimes introduce more acute advisory conflicts. Conflicts also can arise
where advisers recommend variable annuities that are proprietary to their employers. Unlike
mutual funds, variable annuity prices reflect spreads captured by insurers that are not transparent
to consumers. Such spreads compensate insurers for assuming risk, but also can introduce
conflicts, and with conflicts, the risk that recommended insurance protections are insufficiently
valuable to the consumer to justify the associated fees.
BD revenue is also affected by so-called “principal transactions,” where the firm acts as a
dealer, or “principal,” rather than as a broker or agent, and executes the transaction between the
customer and its own account. In one common transaction, a BD sells corporate bonds to an
IRA investor from its own inventory, charging some mark-up over the bonds’ market value as
compensation for its dealer service. Of course, executing securities transactions as an agent, for
example buying equity shares on a stock exchange for a customer’s account, also generates
revenue, in the form of commissions, for a BD.
Importantly, many of the aforementioned types of BD revenue increase with their
customers’ trading volume. More trades can generate more load sharing, more mark-ups, and
more commissions.
BDs typically pass much of their variable revenue on to their representatives who
recommend the mutual funds, as different types of variable compensation. One common type of
compensation known as payout generally amounts to a specified fraction of the revenue that the
representative produces for the BD. The fraction often increases with the representative’s
production, and may be different for different asset classes, different products, and products from
different vendors.303 Depending on the payout formula, BD representatives, like BDs, often
stand to gain if IRA investors trade more, buy or hold certain mutual funds or other products, or
buy securities out of the BD’s own inventory. Some BD representatives receive higher

301

302

303

Section 17(e)(2) of the Investment Company Act of 1940 generally limits the remuneration that an affiliated person of a fund, acting as
broker, may receive for effecting purchases and sales of securities on a securities exchange on behalf of the fund, or a company the fund
controls, to the “usual and customary broker's commission.” Rule 17e-1 under the Act describes the circumstances in which remuneration
received by an affiliated person of a fund qualifies as the “usual and customary broker's commission.”
Section 28(e) of the Exchange Act provides a safe harbor for advisers who charge higher commissions as long as the amount is reasonable
in relation to the research or other services provided.
Hung et al. (2008) reports that “[a] common source of compensation is payout, the amount that a broker receives from total revenue that he
or she generated for the firm.” The payout percentage depends on the type of relationship between the firm and the broker, the level of
production, the products involved, and the broker’s rank in the firm… In general, payouts are structured to increase incrementally as
production increases” (29-30).

129

compensation for distributing the BD’s proprietary or affiliated mutual funds rather than a
competitor’s funds.304
Prentice (2011) lists common conflicts by which financial advisers can profit at investors’
expense, including churning, reverse churning, excessive mark-ups and commissions, failing best
execution, failing to disclose market-maker status, price manipulation, unauthorized trading,
selling unsuitable securities, and operating boiler rooms.305
Conflicts of interest likewise often arise in connection with compensation arrangements
common to RIAs, insurance agents and brokers, and bank representatives who advise IRA
investors.
A RAND study for the SEC found that RIAs who provide investment advice to retail
clients are often highly conflicted. The study notes that RIAs often face “conflicts” arising from
“various practices in which an adviser may have pecuniary interest (through, e.g., fees or profits
generated in another commercial relationship, finder’s fees, outside commissions or bonuses) in
recommending a transaction to a client” (Hung, et.al. 2008). According to the study, 13 percent
of SEC-registered RIAs with individuals as clients received commissions. Many engaged in socalled “hat switching”: 7 percent were BDs, 12 percent were registered representatives of a BD,
and 16 percent were insurance agents or brokers. Thirty percent sold products or provided
services other than investment advice to advisory clients. Twenty-two percent were affiliated
with a BD, 11 percent with an investment company, 9 percent with a bank, and 17 percent with
an insurance company or agency. An even larger fraction conducted discretionary business with
BDs: 61 percent determined and 78 percent recommended the BD for some client account
transactions. Sixty percent received products or services other than execution from a BD (Hung
et al. 2008).
Nearly all RIAs with individuals as clients – 97 percent – received some compensation in
the form of a fee tied to assets under management. This form of compensation is free of many of
the types of conflicts described above but may still introduce other potential conflicts. Reliance
on asset-based fees might discourage an RIA from recommending the purchase of annuities and
other instruments that have the effect of removing assets from the account under management.
Asset-based fees also have sometimes raised concerns about the potential for “reversechurning,” or charging an ongoing fee that is excessive because the account investor rarely
trades and the adviser provides little ongoing service to the investor. (RIAs, however, generally
are fiduciaries under securities law and acting on such conflicts could breach their fiduciary
duty.)
Commissions are a common practice in the insurance market and reflect how distributors
– insurance agents or broker-dealers – get compensated after a transaction is completed.

304

305

Hung et al. (2008) also document complex webs of affiliations (41 and 59) and revenue streams (25-26) among financial products and
services firms. For example, “[f]und companies pay the broker-dealers a certain percentage of the sales that brokers bring in, on top of the
commissions that investors pay the broker” (25). These affiliations and revenue streams create myriad potential conflicts. The authors were
unable to fully examine such affiliations and revenue streams, however. Although the authors “had access to extensive databases based on
regulatory filings,” gaps, “inaccuracies” and “inconsistencies” in such filings make it “difficult to make systematic and conclusive
comparisons between the different types of firms.” (59-61).
According to the SEC, “Dishonest brokers set up ‘boiler rooms’ where a small army of high-pressure salespeople use banks of telephones to
make cold calls to as many potential investors as possible. These strangers push investors to buy ‘house stocks’—stocks that the firm buys
or sells as a market maker or has in its inventory.” (See http://www.sec.gov/answers/boiler.htm.)

130

Annuities are sold through different types of distributors. Independent BDs, full service national
BDs, independent agents, career agents, and banks collectively account for over 90 percent of
annuity sales and all are paid by commissions – see Figure 3-16. These commissions create
conflicts of interests between salespersons and consumers as a salesperson may have an
incentive to sell an annuity product paying higher commissions even though buying that
particular annuity product may not be in the best interest of the consumer. The conflicts of
interest in the annuity market can be even more detrimental than the mutual fund market because
the decision to purchase an annuity product can be costly to reverse due to contractual surrender
charges. Commissions are also associated with product features that may be detrimental to
customers. For example, annuities sold by an intermediary who receives a commission more
often include surrender charges than annuities sold directly to customers.306
Figure 3-16 Annuity Sales by Distribution Channel Within Each Product Type in 2014
Variable
Fixed-rate
Fixed Indexed Total
Independent BD
36%
5%
13%
27%
Career Agents
24%
22%
5%
19%
Full Service National BD
16%
10%
2%
12%
Banks
12%
42%
14%
17%
Direct Response
11%
3%
0%
8%
Independent Agents
1%
18%
66%
18%
Other
0%
0%
0%
0%
Total
100%
100%
100% 100%
Source: LIMRA U.S. Individual Annuity Yearbook-2014

Insurance product commissions are often substantially higher than BDs’ mutual fund
load-shares or securities commissions. James and Song (2001) reported that U.S. sales
commissions on annuities were about 4% of premiums. Commissions on indexed annuities
average 6.3 percent of the principal payment, according to one observer.307 U.S. life insurers’
aggregate commission payments accounted for 7 percent of aggregate total expenses and
amounted to 9 percent of total premiums in 2013. 308 Moreover, insurance product commissions
can vary widely across both products and insurers. Such high and variable commissions can
encourage agents and brokers to recommend products that are not suitable for their customers
and/or to favor one suitable product over others that would better serve their customers’ interests
(Schwarcz 2009).
Scholars and regulators recently have singled out so called “contingent commissions” as
concerning and worthy of special scrutiny for the acute conflicts of interest they introduce
between insurance agents and their customers. Contingent commissions are cash or in-kind
bonuses awarded to independent insurance agents or brokers by insurers for meeting specified

306
307
308

LIMRA, “U.S. Annuity Persistency, Second Quarter, 2015” (2015).
Scism, Leslie. "Insurance Fees, Revealed," Wall Street Journal, March 30, 2012.
Department calculations based on “American Council of Life Insurers: Life Insurers Fact Book 2014,” available at:
https://www.acli.com/Tools/Industry%20Facts/Life%20Insurers%20Fact%20Book/Documents/FB14All%20ChaptersFinal.pdf. These
figures include all life insurers’ product lines. Commissions are not reported separately for individual annuity products.

131

volume or profitability goals. Their size and structure vary widely, introducing a complex
variety of potential conflicts. For example, an insurance broker could be rewarded for steering
customers toward insurers whose production goals they are approaching, or for steering higher
risk customers away from insurers who pay bonuses contingent on profitability (net of claims)
(Schwarcz 2007; Schwarcz and Siegelman 2015; Beh and Willis 2009). Contingent
commissions and the attendant potential conflicts generally are not transparent to retail
customers. Although the studies that closely examined contingent commissions mostly focused
on commercial property-casualty insurance, the incentive structures that were the focus of the
studies parallel practices in the annuity market.
Concerns about the conflicts of interest in annuity sales are underscored by various
media reports reflecting how conflicts of interest influence agents’ recommendations.309 While
regulators and industry participants debated how to regulate fixed-indexed annuities, fixedindexed annuities grew substantially in the fragmented regulatory environment. In 2014, fixedindexed annuity sales reached a record-high of $48.2 billion, up 23 percent from 2013. These
increased sales of fixed-indexed annuities have been followed by complaints that the products
were being sold to customers who did not need them. Some attribute these increased sales to
unusually high commissions on fixed-indexed annuities, which provide insurance agents with a
strong incentive to sell the products even if they are not right for customers.310 Some reports
suggest the commission that insurance companies pay their agents for selling fixed-indexed
annuities is on average 6 percent311 but ranges up to 12 percent.312 In a study on senior financial
exploitation, the Financial Planning Coalition found that “over half of the [Certified Financial
Planner] professional respondents…personally had worked with an older client who previously
had been subjected to unfair, deceptive or abusive practices. Of these, 76 percent reported
financial exploitation that involved equity-indexed or variable annuities.”313 Several high profile
class action lawsuits involving variable annuity and fixed-indexed annuity sales have been
documented.314 These media reports and lawsuits illustrate that pervasive conflicts of interest
are embedded in current industry practices and demonstrate the clear need for regulatory action
in the annuity market.
Potential conflicts of interest in advisers’ recommendations concerning insurance
products are not limited to those associated with insurance product commissions. Insurance
brokers, like BD representatives and RIAs, often engage in hat-switching, and/or are affiliated
with vendors or distributors of products other than insurance products. Moreover, because
variable annuities, likely the insurance product most widely marketed to retail investors, are
regulated as securities, the advisers who distribute them are BD representatives, whose potential
conflicts are documented immediately above in this section.

309

310
311
312
313

314

Kathy Kristof, April 25, 2006, “Unions’ Advice is Failing Teachers” The Los Angeles Times; Leslie Scism, June 7, 2015, “A New Warning
on Indexed Annuities” The Wall Street Journal; Chris Serres, October 13, 2009, “A split decision in Allianz Life annuity lawsuit.” Star
Tribune; Zeke Faux and Margaret Collins, January 20, 2011, “Indexed Annuities Obscure Fees as Sellers Earn Trip to Disney” Bloomberg
Business.
Leslie Scism, Fixed Indexed Annuities Merit Caution, Wall Street Journal, August 16, 2013
Ibid.
Zeke Faux and Margaret Collins, January 20, 2011, “Indexed Annuities Obscure Fees as Sellers Earn Trip to Disney” Bloomberg Business
Financial Planning Coalition’s comment letter to the Department dated July 21, 2015;’ available at: http://www.dol.gov/ebsa/pdf/1210AB32-2-00702.pdf.
Andrea Robinson, 2009, “Annuity Class Action Litigation-Trends and Strategies for Effective Class Action Defense.”

132

Bank representatives who distribute bank products, such as certificates of deposit, to IRA
investors, generally are bank employees who distribute only proprietary products. Many banks,
however, have affiliates that provide or distribute investment products that are not bank products,
and bank employees may be encouraged to direct customers to such distributors and products.
The U.S. financial services industry itself widely acknowledges that compensation
arrangements that the Department believes pose potential conflicts of interest are pervasive
among professionals who provide investment advice to IRA investors. This is borne out in
public comments on the 2015 NPRM. Many of the comments specifically reference
compensation arrangements such as commissions and revenue sharing that can pose conflicts.
The major role such compensation arrangements play in the current market for IRA investment
advice appears to be a primary motivation for many of the industry’s objections to the 2015
Proposal. Many comments question whether various conflicts impact advice, arguing that
countervailing market forces, business practices designed to make advice impartial, and/or
various rules governing advice effectively prevent existing conflicts from tainting advice. Some
argue that compensation arrangements that can pose conflicts also have other, positive market
effects, such as helping to extend investment advice and encouragement to save to lower-income
market segments. But some comments affirm the prevalent use of a wide variety of
compensation arrangements that have the potential to introduce bias into investment advice
regarding IRAs.315
Economic theory predicts that adviser conflicts such as those enumerated above can bias
advice and harm advice recipients.
For example, Stoughton, Wu, and Zechner (2011) model a market where financial
advisers act as intermediaries between individual investors and portfolio managers, and find that
non-conflicted financial advisers improve the welfare of investors. However, when conflicts of
interest are introduced – the authors model a “fee rebate” or “kickback” from the portfolio
manager to the financial adviser – individual investors are harmed. The investors are now not
only worse off than they were without the conflict of interest, they are worse off than they would
have been if the investment adviser did not exist at all. The authors find that, “kickbacks are
always associated with higher portfolio management fees and negatively impact fund
performance” [italics added]. Some in the industry have made the claim that although fees are
hidden and advice is conflicted, consumers are still better off in these advice arrangements than
getting no advice at all. Results like those from Stoughton, Wu, and Zechner (2011) cast doubt
on that assertion.

315

See for example, 2015 NPRM Comments from Franklin Templeton, (the Proposed Rule will “impact the ability of a financial advisor to
obtain certain types of compensation, including commissions and trails fees, from a fund that the advisor recommends to its clients.”);
Insured Retirement Institute (“the levelized distribution compensation structures that appear to be compelled by the Proposed BIC
Exemption are incompatible with well-functioning individual annuity product distribution models.”); Litan/Singer Report (“commissionbased compensation creates incentives for brokers to offer beneficial advice to investors,”); Transamerica (“differential compensation exists
in the advice models that serve small accounts and small businesses.....and is necessary to preserve access to advice across all account
sizes.”); AALU (the “new definition” of “insurance commission” is problematic as it “does not include revenue sharing payments,
administrative fees or marketing payments, or payments from parties other than the insurance company or its affiliates… it is important for
the Department to provide a broad definition that will encompass common and appropriate compensation practices.”); and Prudential
(“Moreover, the BIC Exemption does not explicitly address existing advice programs that are offered to IRAs and plans, such as wrap fee
programs.”) Comment letters are available at: http://www.dol.gov/ebsa/regs/cmt-1210-AB32-2.html.

133

In an October 2013 report, FINRA stated that “conflicts of interest can arise in any
relationship where a duty of care or trust exists between two or more parties, and, as a result, are
widespread across the financial services industry.” The report goes on to review many types of
conflicts that can bias retail investment advice. Broker compensation structures typically favor
some products over others. Many include production thresholds that trigger large rewards that
can encourage mis-selling or churning. FINRA reviews various strategies to mitigate conflicts,
including the adoption of less variable compensation structures, and monitoring advisers’ sales
for evidence of bias, particularly near compensation thresholds and at major investor lifecycle
events, such as rollovers at retirement. The FINRA report also notes that brokers often are
conflicted with respect to investors’ choice between commission- or fee-based relationships.
Finally, it summarizes regulation of broker conflicts in the U.S. and abroad, noting strong bars
against conflicts that have been implemented or proposed in some jurisdictions.316 FINRA also
has expressed concerns about broker conflicts that can arise from recruitment compensation
practices that can encourage mis-selling or churning.317

3.2.3.2

Costly Pursuit of Customers

IRA advisers (and their employers and affiliates) pursuing IRA advice customers incur
costs to produce marketing materials, place advertisements, hold seminars, or make “cold” phone
calls or knock on doors to speak with potential customers. Unfortunately, these costs are
unlikely to yield commensurate benefits for IRA customers.
Some BD representatives (and insurance agents and brokers) are compensated entirely or
primarily by commissions resulting from product sales. This creates an incentive to aggressively
maximize sales, which is likely to result in costly and potentially economically inefficient318
efforts to attract new customers. The average BD representative working for a BD firm receives
60 percent of his/her revenue through commissions.319 Cerulli Associates determined that RIAs
and BD representatives spent 18 percent of their time acquiring new clients in 2013,320 and that
this time share had increased from 15 percent in 2008.321
In efficient, competitive markets, advertising should be used as a means to reduce
information costs and promote transparency (Sirri and Tufano 1998). However, in the U.S.,
mutual fund advertisements rarely highlight one of the best predictors of performance-fees
(Gallaher, Kaniel, and Starks 2006; Barber, Odean, and Zheng 2005). Both theory and ample
empirical evidence show that fees are strong predictors of future fund performance, while past
performance is not. Active investing often is, in large part, a zero-sum game, wherein for each
investor who wins, a counterparty investor usually must lose. In securities markets that are very

316

317

318

319
320
321

FINRA, “Report on Conflicts of Interest” (Oct. 2013); last accessed at:
https://www.finra.org/web/groups/industry/@ip/@reg/@guide/documents/industry/p359971.pdf.
FINRA Regulatory Notice 13-02, “Recruitment Compensation Practices” (Jan. 2013); available at:
http://www.finra.org/web/groups/industry/@ip/@reg/@notice/documents/notices/p197599.pdf.
There may be instances where costly pursuit of customers improves social welfare, for example by overcoming a consumer’s myopia that
would otherwise lead them to save too little. Such instances are likely to be outweighed, however, by instances where the marginal cost to
pursue customers exceeds associated social value, as in the case of some sales efforts to attract IRA rollovers that merely move, rather than
increase savings. Some advisers may expend more effort on the latter than the former because the latter can yield far larger immediate
rewards for advisers.
FSI, “2013 Broker-Dealer Financial Performance Study,” 32.
Cerulli Associates, “Advisor Metrics 2013: Understanding and Addressing a More Sophisticated Population” (2013), Exhibit 5.15.
Cerulli Associates, “Cerulli Quantitative Update: Retail Investor Provider Relationships 2011” (2011), Exhibit 6.05.

134

efficient, securities prices very quickly reflect essentially all information, and there is very little
mispricing to be found and exploited. Historically, the excess cost of active management –
trying to identify and buy (sell) underpriced (overpriced) securities – has been higher on average
than any gain in performance over a lower-cost, passive management approach. Moreover, it
appears that past superior performance by an active manager more often reflects luck than skill
(Sharpe 1966; 1991; and 2013; Fama and French 2010; French 2008).
Instead, advertisements often focus on performance, or even suggest that advice is “free”
(when it is not) or that 401(k) accounts are “old” relative to the retail mutual funds available in
an IRA. That advertisements focus on poor predictors of future results, rather than on fees (a
strong predictor), is indicative of a costly pursuit of customers that does not promote welfare
gains – but the advertisements do seem to achieve their aim of promoting particular products.322
Inderst and Ottaviani (2009) develop a theoretical model of a sales transaction where an
agent of the seller must pursue customers and provide advice to those customers. This agency
setup is representative of much of the financial industry where insurance agents sell insurance
products and BD representatives sell securities and mutual funds, etc. The researchers find that
as agents require more effort to pursue customers, harm to the customer increases. These costs
can only be offset by firms lowering their advice standards. They explain the implications of
their result:
“[T]his suggests that one should expect the standard of advice to be lower when the roles
of consumer acquisition and advice provision are performed by the same agent, and when
performance cannot be easily measured and rewarded in isolation by separating the two
tasks. We should expect the need for policy intervention to increase when incentives for
customer acquisition become more important to firms. Intuitively, the more agents are
expected to actively prospect for new customers, the more scope there is for [mis-selling]
to occur at the advice stage, even when consumers are wary and product providers
directly bear costs following unsuitable advice” (Inderst and Ottaviani 2012, 509; Inderst
and Ottaviani 2009, 893 – 895).

3.2.3.3

Obstacles to Distinguishing Good and Bad Advice

It is sometimes argued that, under certain conditions, reputational concerns might compel
conflicted advisers to act in their customers’ interest. This theoretical result, however, rests on
the assumption that customers can distinguish impartial advice from biased advice. The
importance of this assumption to the theory of reputational effects is detailed in Section 3.2.3.3.1
below.
There is compelling evidence that most IRA investors are ill-equipped to assess the
quality of advice they receive, or even the investment performance they achieve. Most do not
understand what they pay for advice and for investments, how their advisers are compensated
and regulated, the conflicts their advisers might face, nor how those conflicts might affect their
advice (see Section 3.2.3.3.3 below). Investors have a difficult time understanding whether their
adviser is acting as a broker-dealer or as an RIA, and generally do not know which regulatory

322

Evidence indicates that past performance has little or no signaling power in predicting future performance – though it does have power to
influence fund flows (Jain and Wu 2000).

135

regime applies or how the regulatory standards differ between regimes. As a result, advisers
have both an opportunity and an incentive to preferentially recommend products that increase
their profits, and/or those of the vendors whose products they recommend, at IRA investors’
expense, without fear that their reputation or market share will suffer much if at all.
There is also compelling evidence that additional or different disclosure practices are
unlikely to fill in these gaps in IRA investors’ skills and knowledge. Many investors ignore
disclosures. Many simply lack the financial sophistication and/or the time and attention
necessary to master the complex information such disclosures would have to communicate.
Moreover, there is no clear basis on which even sophisticated, attentive IRA investors could
translate a thorough understanding of recommended and other available investments and their
advisers’ compensation and conflicts into optimal decisions about advice and investing. In
particular, it is unclear how an IRA investor could determine whether or how a conflict has
influenced her adviser’s recommendation. And there is reason for concern that disclosure of
conflicts can even have negative, unintended consequences. Section 7.4 summarizes the bases
for these conclusions.

3.2.3.3.1

Obstacles to Assessing Advice Quality

Detecting lapses in the quality of investment advice is not easy.323 IRA investors
typically have access only to information on their own experience – the advice they received, the
investments they chose, and perhaps the results they achieved. In all likelihood they can neither
directly observe the quality of the advice, nor infer it from their investment results. Moreover,
IRA investors often do not know what they pay for advice. Without a good understanding of the
quality and price of advice, they cannot make optimal decisions about purchasing it, and are
vulnerable to paying too much for bad advice and to incurring financial losses by following it.
Almost certainly, the great majority of IRA investors cannot directly assess the quality of
the investment advice they receive. It is the nature of an advisory relationship that the adviser
has an informational advantage over the advisee. Bluethgen, Meyer, and Hackethal (2008) note
that, “as financial advice is an expert service just as the ones provided by lawyers or doctors, the
ordinary investor will hardly be able to determine the quality of the advice given even ex-post
because the investor simply lacks the knowledge or the information to assess the quality of the
advice.” Lusardi, Mitchell and Curto (2009, 15) found that older Americans “lack even a
rudimentary understanding of stock and bond prices, risk diversification, portfolio choice, and
investment fees.”
While gaps in IRA investors’ financial sophistication alone provide sufficient basis to
conclude that most cannot directly assess the quality of advice, available empirical evidence
lends additional support. In one study, auditors were trained to mimic actual clients and to
record their advice interactions. The auditors were not trained to evaluate advice quality,
however, and it appears that they overwhelmingly failed to recognize problems with the advice.
Advisers failed to mention fees to one-half of the auditors, failed to recommend index funds to
92 percent, and tended to recommend that auditors chase returns and/or choose actively managed

323

One of the obstacles of assessing advice quality is the time and cost of investigating the advice. Individuals often purchase advice so that
they don’t have to worry about their investments. Those individuals, whose time-cost of investing is such that they choose to purchase
advice, likely also have a prohibitively high time-cost of investigating that advice.

136

funds. Yet 70 percent of the auditors said they would go back to the adviser with their own
money (Mullainathan, Noeth and Schoar 2012). In a study of actual retirement investment
advice interactions in Australia, investors “were rarely able to tell whether or not the advice they
received had a reasonable basis.” In most cases where the Australian authority found “major
shortcomings in the advice,” the investors “thought the advice was satisfactory and said they
intended to follow it.324
Agnew et al. (2014), in an experimental setting, found that clients’ opinions of adviser
quality are easily manipulated. If an adviser first provides good advice on a financial decision
that is easy to understand, the client will subsequently trust bad advice on a more difficult or
complicated topic. Clients rely too much on advisers’ stated credentials. The authors offer
policy recommendations: credentialing should be improved, advisers’ interests should be aligned
with their clients’, and all advisers should be subject to a uniform fiduciary standard of conduct.
Inferring the quality of advice based on investment results is also problematic, for several
reasons. First, the investment results themselves often are not transparent to the IRA investor.
FINRA’s suitability rules325 do not require BDs to disclose their customer’s personal rates of
return. Many account statements show only transaction details and beginning and ending asset
values for specified periods. Translating these into rates of return requires sophisticated
calculations, well beyond the capability of all but the most sophisticated IRA investors. For
example, Lusardi and Mitchell report that only one-half of individuals aged 50 and older in the
United States can correctly answer two simple financial questions that involve calculations.
Many respondents failed to correctly conclude that $100 would grow to more than $102 after
five years if interest accrues at 2 percent per year, while others were unable to determine that an
account earning interest at 1 percent while inflation was 2 percent would lose buying power
(Lusardi and Mitchell 2011).
Second, even if the IRA investor knows her rate of return, she will be hard pressed to
determine whether it is favorable. Selecting an appropriate benchmark for comparison requires
financial sophistication about asset classes, among other things. Yet only about one-half of
individuals age 50 or older correctly state that a single stock is usually riskier than a stock mutual
fund.326 In addition the investor may have followed only some of the adviser’s
recommendations, in which case the results of followed recommendations would be blended
with other results, and the results of recommendations not followed (and possibly not
remembered) would be invisible to most investors. Finally, if the investor simply follows a
recommendation to buy and hold a mutual fund, the fund’s disclosure will report its returns net
of fees and provide benchmark for comparison. But even in this simple case, the investor might
need to adjust for loads paid, and if she buys or sells shares during the reporting period, her
personal, asset-weighted return will differ from the time-weighted return reported by the fund,
sometimes substantially.
Third, even if the IRA investor can determine whether her rate of return was favorable,
this is not tantamount to determining whether her adviser gives good advice. Investment returns
are noisy, and even several years of experience cannot reveal with high confidence whether the

324
325
326

Australian Securities and Investments Commission, “Shadow Shopping Survey on Superannuation Advice,” ASIC Report 69, 2006.
FINRA Rule 2111.
These findings are affirmed by research funded by the FINRA Investor Education Foundation, 2009.

137

performance difference between an adviser’s recommendations and a benchmark are due to
chance or skill, unless the difference is substantial and persistent.
For these reasons, IRA investors are unlikely to successfully assess the quality of their
advisers’ recommendations based on past investment results.
In addition, investors often do not know what they pay for advice. Hung et al. (2008, 9597) reports that many investors exhibit confusion about fees. For example, in one survey, among
investors who receive advisory services from an advisory firm that is not also a brokerage firm,
23 percent report paying for the services by commission, while 19 percent report paying a fee
specified as a percentage of assets. This appears to conflict with information provided by the
firms themselves. Among SEC-registered advisory firms that are not also brokerage firms, 97
report that they are compensated with asset-based fees, and only 10 percent report that they
receive commissions. Substantial numbers of investors receiving advisory services from either
advisory or brokerage firms either fail to report how much they pay for the services or report that
they pay nothing for the services. Why do investors fail to understand what or even whether they
pay for advice? Although fees and prices are not inherently complex financial concepts that
require sophistication to understand, in practice, as elaborated earlier in this analysis (see Section
3.2.3.1 above), payments for investment advice are often highly complex, indirect, and not
readily transparent. IRA investors who do not know what they are paying for advice cannot
make sound decisions about which or how much advice to purchase.
Edelen, Evans, and Kadlec (2012) provide direct evidence that consumers have difficulty
observing fees and accounting for them in their financial decisions. The authors observe that
hidden fees have a more negative impact on returns than transparent fees. But hidden fees are
less likely than transparent ones to chase investors away. The evidence shows that investment
managers and brokers benefit from hiding fees – for example through commission bundling – at
the expense of the consumer.
IRA investors are likely to be even more hard pressed to assess the quality of advice
related to insurance products, mainly fixed, fixed-indexed and variable annuities. These
products are often complex. Their features vary widely across both products and insurers,
making comparisons difficult for consumers. Their fees likewise are complex and difficult to
interpret. Most IRA investors therefore have the ability to judge neither the suitability nor the
price of any recommended product.
The National Association of Insurance Commissioners (NAIC)’s “Buyer’s Guide for
Deferred Annuities” sets out to summarize the types of annuities available to retirement investors
and the factors consumers should consider before buying one.327 According to the guide,
different types of deferred annuities generally share certain features including specified
accumulation and payout periods, surrender or withdrawal charges, tax deferral, and riders that
add features at additional cost. The guide cautions that different annuities differ with respect to
fees, charges and adjustments that can reduce their value, and with respect to premium or interest
bonuses (which may be lost upon early surrender). Fees can include contract fees, loads
deducted as a percentage of each premium payment, premium taxes, transaction fees, mortality
and expense risk charges, and (in the case of variable annuities) underlying fund expenses, the

327

National Association of Insurance Commissioners, “Buyer’s Guide for Deferred Annuities” (2013).

138

guide explains. The guide goes on to distinguish fixed, fixed-indexed, and variable annuities.
Fixed annuities provide a fixed interest rate for a specified period, after which the rate may
increase or decrease. The rate credited to fixed-indexed annuities generally is determined with
reference to a market index, often subject to participation rates, caps, and/or spread rates. The
value of variable annuities can go up or down with the value of funds in associated subaccounts.
The guide explains that annuities’ value may be annuitized or fully or partly withdrawn.
Annuitized payments may continue for the consumer’s (or his or her spouse’s) lifetime, for a
specified period, or the longer of both. Variable annuities often offer additional guaranteed
benefits for additional fees. These include a guaranteed minimum accumulation benefit, a
guaranteed minimum income benefit, and a guaranteed lifetime withdrawal benefit. The guide
cautions consumers about risks, including the risk of not getting all of the investor’s money back
and not being able to withdraw money without incurring fees, adding that the risks vary among
different annuities. The guide directs consumers to seek information available on specific
products in 3 documents: namely, each product’s associated “contract,” “disclosure,” and
“illustration.” It recommends 12 detailed questions for consumers to ask before buying an
annuity. Referencing multiple potentially technical documents and discussing a dozen
potentially technical questions almost certainly leaves most consumers vulnerable to confusion
and entirely reliant on the advice they receive.
FINRA takes on annuities’ complexity in several “Investor Alerts” aimed at retail
investors. One Alert, targeting variable annuities,328 in three dense pages explains that variable
annuities resemble mutual funds, but with additional features including tax-deferred earnings, a
death benefit, and annuity payout options that can provide guaranteed income for life. It
distinguishes the accumulation phase, during which premiums are allocated across subaccounts,
from the distribution phase; and deferred annuities from immediate annuities. It explains
associated sales and surrender charges, and ongoing fees and expenses including mortality and
expense risk charges, administrative fees, underlying funds’ expenses, and charges for special
features such as stepped-up death benefits, guaranteed minimum income benefits, long-term
health insurance, and principal protection. Noting that ongoing fees can exceed two percent of
the annuities’ value annually, the Alert recommends that “if you don’t need or want these
features, you should consider whether this is an appropriate investment for you.” It explains
some tax considerations. It observes that “in an attempt to attract investors, many variable
annuities offer bonus credits,” such as a one percent to five percent addition to each premium
payment – but cautions that these are offset by other charges. It warns that promised guarantees
“are only as good as the insurance company that gives them.” Finally, it provides special
considerations for IRA investors (for whom investing in a variable annuity “may not be a good
idea”), including that “a variable annuity will provide no additional tax savings” but will
increase costs and profit the adviser, and that mandatory IRS withdrawals beginning at age 70 ½
might trigger surrender charges.

328

FINRA Investor Alert, “Variable Annuities: Beyond the Hard Sell,” 2012; available at:
http://www.finra.org/web/groups/investors/@inv/documents/investors/p125846.pdf; and SEC Investor Bulletin, “Variable Annuities - An
Introduction” (Nov. 24, 2015); available at: https://www.sec.gov/investor/alerts/ib_var_annuities.pdf.

139

A second FINRA Alert targets fixed-indexed annuities.329 According to this Alert, fixedindexed annuities (in this case, more specifically, equity-indexed annuities), “are anything but
easy to understand.” It echoes many of the points made in NAIC’s Guide, but fills in more
detail. Fixed-indexed annuities’ guaranteed minimum return is “typically at least 87.5 percent of
the premium paid at 1 to 3 percent interest.” However, early surrender can result in surrender
charges and tax penalties that will “reduce or eliminate any return.” The Alert explains that
fixed-indexed annuities’ index-linked interest rate generally is computed by applying a
participation rate, a spread/margin/asset fee, and interest rate caps. It points to advantages and
disadvantages of different indexing methods including annual reset (ratchet), high water mark,
and point-to-point. Calculations also vary with respect to index averaging and use of simple v.
compound interest crediting, and dividends are generally excluded, the Alert says.
Notwithstanding their complexity, annuities can play a very important and beneficial role
in retirement planning, particularly in the management of individual longevity risk. More
generally, as with other forms of insurance, the transfer or pooling of various risks provided by
annuity products and issuers can enhance individual and social welfare. However, annuities’
complexity heightens the risk posed by adviser conflicts and makes it doubly important that
advice be both expert and impartial. The Federal Insurance Office within the US Treasury
Department has stated that “As unprecedented numbers of seniors reach retirement age with
increased longevity, and as life insurers continue to introduce more complex products tailored to
consumer demand, the absence of national annuity suitability standards is increasingly
problematic.”330
It is doubtful whether IRA investors can determine what value if any they should place
on the insurance benefits associated with any particular variable annuity product. Consumers’
degree of aversion to various possible losses is subject to a number of behavioral biases
(Schwarcz 2010) and vulnerable to manipulation by advisers. In addition, whether a consumer’s
insurance coverage for any particular risk is adequate is often not apparent to the consumer until
after a (potentially) insured loss occurs. It is possible that only a fraction of investors will ever
elect, or perhaps even qualify for, any particular benefit. For those that do, the ex post value of
the benefit will vary widely (depending, for example, on age at death, or financial market
conditions). For these reasons it will be difficult for an IRA investor to assess the quality of past
recommendations, even after benefits are claimed (Schwarcz 2009).

3.2.3.3.2

Lack of Reputation Effects

In economic theory, efficiency often requires perfect and costless information. The retail
market for financial products and services, however, is beset by high information costs – i.e.,
investors are ill equipped to evaluate the quality of advice. Given the combination of high
information costs and adviser conflicts, the potential for social welfare losses is high. IRA
investors are likely to make inefficient decisions about their purchases of advice and/or,
following suboptimal advice, about their investments. Suboptimal investment decisions erode
risk-adjusted net returns for investors and allocate capital inefficiently in the national economy.

329

330

FINRA Investor Alert, “Equity-Indexed Annuities: A Complex Choice,” 2012, available at: https://www.finra.org/file/alert-equity-indexedannuities-complex-choice; and SEC Investor Bulletin, “Indexed Annuities” (April 1, 2011); available at:
https://www.sec.gov/investor/alerts/secindexedannuities.pdf.
U.S. Department of the Treasury, Federal Insurance Office, ”Annual Report on the Insurance Industry” (Sept. 2015).

140

Theory also predicts transfers, as advisers and producers of the products they recommend
capture surplus from investors in IRAs characterized by conflicts of interest. Both of these
effects can be expected to erode IRA investors’ retirement security. In addition, reputational
concerns may be less likely to impose discipline on adviser behavior to the extent the adviser is
not dependent on repeated interactions with the customer, but rather can accrue large earnings
based on a one-time sale, as is often the case with rollover advice.331 However, as the remainder
of this section demonstrates, even repeated interactions do not necessarily ensure that an advisor
will act in the best interest of the customer because of the difficulties that inexpert customers
have in assessing the quality and value of the advice they receive.
High information costs limit advisees’ ability to act as a check on adviser misbehavior.
The inability to act as a check on adviser misbehavior can manifest itself in different ways,
relating to an advisee’s lack of important information or the advisee’s inability to interpret
important information.
Bolton, Freixas, and Shapiro (2007) model a relationship between advisers and advisees
where reputational concerns prevent advisers from acting on their conflicts of interest and ensure
that advice is in the best interest of the client. However, their model reveals an important
characteristic that can distinguish advisory markets with harmful conflicts from advisory
markets with harmless conflicts. The authors explain:
“To model the reputational concern we assume that an [adviser] suffers a reputation
loss… when a lie told to a customer leads to a purchase by that customer. This loss arises
because the financial product is an experience good; the customer realizes a return from her
investment and can compare that with the initial expected return promised her by the [adviser].”
In other words, the model assumes that soon after making an investment decision, the
customer can determine whether the advice that was given was in her best interest. If the
customer could not determine the quality of the advice in a timely manner, the adviser would not
be bound by reputational concerns to act in the client’s best interest. Thus, one key element in
an advice market with harmful conflicts is the inability of the advisee to assess the quality of the
advice soon after the advice is given. As previously noted, the data show that consumers are not
able to make this type of an assessment in today’s advice market.
Other models that also generate the conclusion that firms produce high-quality goods due
to reputational concerns rely on similar assumptions. In MacLeod 2007’s model, the buyer
observes the seller’s level of performance after the good is received (MacLeod 2007). Klein and
Leffler (1981, 618-619) assume that, “if a particular firm supplies less-than-contracted-for
quality to one consumer, the next period all consumers are assumed to know.”
Krausz and Paroush (2002, 57-58) don’t assume that customers directly observe the
quality of advice, but they do require that all customers are able to perform detailed financial
calculations on their own:
“At the end of the period when the actual return is observed, the investor will assess
whether her initial decision… was based on sound information. If the return is below [the
reported expected return on the risky asset] then she has received a lower income than expected

331

See Chapter 4 for further discussion of the rollover market.

141

and if it is above then she has invested less than she would have liked to. … She computes a new
[proportion of wealth invested in the risky asset] by using the new information she has to update
the expected [return], by giving a lower weight to reported expected return on the risky asset…
the greater the deviation of the actual realization from [the reported expected return on the risky
asset], relative to the riskiness of the asset as announced by the advisor.”
Fischel and Kendall’s 2011 Comment Letter to DOL echo many findings on reputational
and competitive effects from the academic literature in their post-hearing comment on the
Department’s 2010 Definition of Fiduciary Investment Advice Proposal. The authors do not put
forth their own model of an advice relationship. Their conclusions necessarily presume,
however, that IRA investors can police the quality of advice and make efficient decisions as to
what advice to buy, how much to pay for it, and what investments to make pursuant to it. The
Department rejects this presumption, based on the evidence to the contrary presented herein.
Rogerson (1983, 508-509) recognizes that the previous literature on reputational
concerns had not accurately depicted markets where the customer has difficultly assessing the
quality of service:
“Consumers are, however, often capable of performing only very partial and vague
evaluations of the quality of professional services they receive from doctors, lawyers, banks,
mechanics, opticians, etc. Furthermore, the quality of a service from a given professional may
vary from time to time. This combination of observer error and actual quality variance makes it
difficult for consumers to evaluate correctly the quality of service that a firm produces.”
Rogerson’s model is relevant to the IRA market because it allows customers to make
mistakes in assessing the quality of a good or service, such as advice. While the author’s
conclusions are supportive of reputation effects in general, the model demonstrates that
reputation effects fail in markets where customers have more difficulty assessing the quality of
the service. The result is intuitive. If a customer mistakes poor service for quality service,
they’ll likely return as a repeat customer and may even recommend the firm to others.
Consumers’ understanding of the nature of conflicts of interest was directly tested in the
setting of disclosures by mortgage brokers.332 A recent study found that consumers failed to
fully comprehend the nature of conflicts of interest in broker compensation despite repeated
attempts to address the issue through revisions of disclosures. In particular, consumers did not
understand how lender payments to brokers created an incentive for brokers to recommend loans
with higher interest rates. Even those who understood the broker’s incentive to obtain higher
interest rates tended to assume that brokers would work in their best interest. One of the main
impediments to consumer appreciation of the significance of the conflict of interest was a lack of
understanding of how the interest rate on their loan was determined. Consumers assumed that
the interest rates were set by the lender based on their creditworthiness alone and did not realize
that the broker could have latitude in deciding which loans and what interest to offer. Although
the study examined the mortgage brokerage industry, not the financial advice industry, the
findings of the study are relevant to the IRA market because of the light it sheds on the dangers
posed by conflicts of interest and opaque fee structures, which are often also characteristics of
investment products.

332

MACRO International Inc., July 10, 2008, “Summary of Findings. Consumer Testing of Mortgage Broker Disclosures.”

142

Inderst and Ottaviani (2012) present a second model which shows that harm to
consumers depends on how “wary” they are of conflicts present in the market.333 Wary
consumers are unharmed because they recognize that advisers are more likely to recommend
products for which they receive commissions and they discount those recommendations.
However, the model requires that wary consumers “form rational expectations about the level of
these payments and the resulting quality of advice.”334 On the other hand, naïve customers –
those who do not understand how a conflict of interest might bias the adviser’s recommendations
– can be taken advantage of. This means that for a consumer to be considered wary, both of the
following must be true: 1) commissions or other conflicting payments must be disclosed and
must be salient at the time a decision is made; and 2) given this knowledge, the consumer must
correctly adjust for the probability that the adviser will act on his or her conflicts at the
consumers’ expense.335
A key question then becomes whether IRA investors are “wary.” As elaborated
elsewhere in this analysis, most IRA investors lack attention to and understanding of their
advisers’ compensation and attendant conflicts. Most are unable to assess effectively the quality
of their advice and of consequent investment results. Moreover, research suggests that
disclosure of advisers’ conflicts can backfire, leading both advisers and consumers to act
contrary to consumers’ interests.336 For example, researchers conducted a randomized
controlled experiment to examine the effects of the mortgage broker compensation disclosure.337
In this study, consumers were randomly assigned to one of five groups – three were disclosure
groups and two were control groups. Consumers in the disclosure groups received disclosure
concerning the compensation of the mortgage brokers, whereas the control groups did not
receive such disclosures. Consumers in the disclosure groups chose more expensive mortgage
loans whereas consumers in the control groups correctly identified less expensive loans. This
study exemplifies how disclosing the conflicts of interest can make consumers worse off.
Therefore, it is highly likely that few IRA investors would qualify as “wary” consumers in this
model – rather, most would be naïve and therefore vulnerable to abuse.
Based on the foregoing, one defining characteristic of harmful advice markets appears to
be the advisee’s inability to act as a check on adviser misbehavior.338 The IRA advice market
exhibits this characteristic, as elaborated immediately below.

3.2.3.3.3

Obstacles to Understanding Conflicts

Similar to advice quality, IRA investors are equally hard pressed to understand the
potential for bias associated with adviser conflicts. Even an IRA investor who knows exactly
how and how richly his or her adviser is compensated is unlikely to understand the conflicts of

333

334
335

336
337

338

Inderst and Ottaviani (2012), supra, at 494. Unlike the model by these authors discussed above, this model is not specific to transactions
where the product is sold through an agent.
Ibid., 499.
Inderst and Ottaviani (2012, 500) also allow for the possibility that a wary consumer could form rational expectations that are correct in
equilibrium even when commissions are not disclosed. The Department agrees that the scenario is possible in theory, but recognizes that it
is highly unrealistic.
See Sections 3.2.1.2 and 7.4.
James Lacko and Janis Pappalardo, February 2004. “The Effect of Mortgage Broker Compensation Disclosures on Consumers and
Competition: A Controlled Experiment.” Federal Trade Commission Bureau of Economics Staff Report.
Also see Egan, Matvos, and Seru (2016) who suggest that some financial advisory firms “‘specialize’ in misconduct and cater to
unsophisticated consumers.”

143

interest that are associated with the adviser’s compensation arrangements or how such conflicts
could affect the quality of the adviser’s service.
Adviser compensation often is not fully transparent, even to an attentive investor. In the
earlier diagram depicting some common conflicts in advice (see Figure 3-15), different adviser
compensation streams and relationships are shown in different colors. Those shown in dark blue
generally are not disclosed and are invisible to IRA investors. Those shown in light blue are
disclosed in a mutual fund’s prospectus and therefore visible to IRA investors who read,
understand and remember that document. Those shown in yellow are more directly visible to
IRA investors. Not shown in the diagram are certain, more qualitative, disclosures. BDs are
required under certain circumstances, such as when making a recommendation, to disclose
material conflicts of interest to their customers, in some cases at the time of the completion of the
transaction. A RIA that has a material conflict of interest must either eliminate that conflict or
fully disclose to its clients all material facts relating to the conflict. But such disclosures tend to
include only general descriptions of arrangements that do not illuminate the amount of adviser
compensation that might be motivating a particular recommendation.
The potential conflicts affecting insurance intermediaries are likewise varied, complex,
and difficult for consumers to discern. As Beh and Willis (2009) observe, “Determining what
intermediaries do and for whom they work has not leant itself to easy answers; definitive
characterizations have been elusive. The intermediary’s relationship with the insurer and the
insured must often be determined on a case-by-case basis.” The authors describe how these
relationships vary along several dimensions, each with implications for potential conflicts.
These include their degree of independence v. exclusivity, the extent of their role in the
distribution of various products (relative to alternative distribution channels), and their authority
as an agent of either the insurer or the insured. Because of these variations, any characterization
of insurance intermediaries’ loyalties and duties is “imperfect at best, because whether the
insured or the insurer serves as the principal can depend on the actual tasks performed… the
intermediary, the insured, and the insurer cannot be certain for whom the intermediary is
working.”
Because most IRA investors cannot determine the quality of the advice they receive and
often do not understand or beneficially react to their advisers’ potential conflicts, it seems
unlikely that they could act as an effective check on adviser misbehavior. Therefore reputational
concerns alone are unlikely to sufficiently mitigate adviser conflicts. Additional or different
disclosure alone is unlikely to help much if at all.

3.2.3.3.4

Summary and Conclusions

Based on the foregoing, it appears that the IRA advice market exhibits the characteristics
that economic theory associates with harm from adviser conflicts. Serious, material conflicts are
widespread. The supply chain devotes substantial resources to pursuing customers in ways that
are not consistent with efficient price competition. IRA investors are ill equipped to police the
quality of advice so reputational concerns cannot be expected to ensure adviser impartiality.
In light of these facts, it is safe to predict that conflicted investment advisers to IRA
investors will act on their conflicts, and when they do, IRA investors will suffer as a result. The
conflicts therefore likely offer advisers ample opportunities to secure large profits at IRA
investors’ expense (while also causing further losses due to inefficient asset allocation).

144

3.2.3.4

Advisers Act on Their Conflicts

The economic models discussed above share one assumption: advisers will act on their
conflicts of interest when it is in their self-interest to do so. In reality, people do not always
behave according to pure financial self-interest. For example, an adviser may provide advice
that is in the best interest of a client because she genuinely cares about the client’s retirement
security or feels a moral obligation to do so. However, empirical evidence indicates that
financial advisers do act on conflicts in ways that harm IRA investors.
One strand of research literature looks directly at the recommendations made by advisers
by asking advisees to record certain aspects of their interaction with the adviser. This allows for
a direct examination of whether an adviser’s recommendation reflects his or her client’s best
interest. A second examines how inflows to mutual funds are affected by the amount of
commissions or revenue that they pass on to the advisers that recommend their funds. Other
things equal, inflows that increase as commissions or revenue sharing increase would indicate
that advisers are choosing to recommend the funds that provide more financial benefit to
themselves, rather than to their clients.

3.2.3.4.1

Questionable Recommendations

There is evidence that advisers often recommend investments that they should know are
not the best alternative for their customer. Numerous academic studies have found that, as a
group, passively managed mutual funds (i.e. index funds) consistently outperform actively
managed funds, largely due to their low fees, (Gruber 1996; French 2008; Fama and French
2010). Therefore it is likely that IRA advisers who honor their customers’ best interests would
widely recommend index funds with low fees.
Yet there is evidence that advisers do not widely recommend diversified low-fee
portfolios. One study’s authors sent trained auditors339 to financial advisers in the Boston area
and observed whether the advisers acted in their own interest or in the interest of the client
(Mullainathan, Noeth, and Schoar 2012). Auditors were each assigned one of four different
styles of portfolios. One portfolio style in particular was designed such that the adviser could
only profit by recommending an action that was clearly not in the best interest of the advisee.
Auditors came into the session with a diversified portfolio of low-fee index funds. According to
the authors, “Moving the low-fee portfolio to an actively managed portfolio with the same
risk/return profile but average management fees would result in additional costs of about one
percentage point per year, i.e., between U.S. $500 and U.S. $1,000 in our scenario.”340
However, the adviser would typically stand to profit only if the investor purchased an activelymanaged fund that returned some commissions or revenue to the adviser’s firm.

339

340

The auditors were professionals who were trained to impersonate regular customers seeking advice on how to invest their retirement savings
outside of their 401(k) plan. To implement the actual logistics of the visits, a financial audit firm was hired that specializes in identifying
and training auditors. To ensure that auditors were able to understand the advice that was given to them, they had to know at least some
basics of financial products and received some guidelines on how to ask for specific advice. Auditors were trained first about basic
financial literacy through an online manuscript. Then, they participated in a training session via video conference. Finally, audit candidates
had to take a short online test to qualify for the study (about 10 percent of the pre-selected auditors failed and were excluded from this
study.
Ibid., 7.

145

Presented with a client invested in index funds, the advisers overwhelmingly put their
own interests ahead of their clients. Less than 3 percent of advisers were supportive of the
auditor’s existing portfolio, while 85 percent were against the strategy. Across all scenarios, less
than 8 percent of advisers recommended index funds, while almost 50 percent of advisers
recommended actively-managed funds. Put differently, in this study, for every adviser who
provided advice that is likely to be in their client’s best interest, there were seven who gave
advice that likely is not in their client’s best interest, but in their own best interest.
While the auditors did not present themselves as IRA investors, the study closely
mimicked advice interactions that are typical of IRA investors. Auditors met face-to-face with
actual advisers for about one hour, usually in the adviser’s office, to seek advice on investing
between $45,000 and $105,000. The advisers did not know the auditors were impersonating
actual investors.
An additional audit study’s results corroborate the findings of this initial study.
Antoinette Schoar, a professor of finance at MIT Sloan School of Management, previewed these
additional findings during the Department’s August 11, 2015, hearing regarding the regulatory
impact analysis for the 2015 Proposal. Professor Schoar testified that she conducted an audit
study where mystery shoppers made 250 client visits to RIAs and BDs in the greater Boston and
Cambridge, Massachusetts area and replicated the study with more than 450 visits in the New
York City area. She stated that in half of the visits mystery shoppers presented mistaken beliefs
about financial markets, indicated that they wanted to chase past returns or exhibited other welldocumented biases. The results were very concerning. The advice received from BDs failed to
correct such biases, and actually encouraged return chasing while vigorously discouraging
investments in low-cost index funds. Professor Schoar’s research also found that BDs favored
high-fee funds, such as actively managed funds over lower-cost index funds. They encouraged
the misconceptions of clients if it made it easier for them to sell more expensive, higher fee
products. In contrast, RIAs, who have a fiduciary duty to act in their clients’ best interest, were
less likely to engage in such activity.341
Research from Australia provides additional evidence to the same effect. The Australian
Securities and Investments Commission (ASIC) recruited participants in Australia’s retirement
system who intended to seek out investment advice, and had the participants answer survey
questions and provide written materials from the adviser following meetings.342 Based on the
information collected, researchers were able to determine 1) if the adviser had a conflict of
interest, such as receipt of trailing commission from the sale of a fund, and 2) if the advice given
had a reasonable basis (as required by law). (It seems safe to assume that if the advice given did
not have a reasonable basis, then it was also not in the client’s best interest.) An adviser who
had a conflict of interest was three to six times more likely to give advice that did not have a
reasonable basis. Many advisers had a conflict of interest stemming from fees that the investor
pays flowing back to the adviser. Of these 123 advisers, 35 percent gave advice that did not
have a reasonable basis, whereas just 6 percent of the 139 advisers that did not have this conflict

341

342

Testimony of Antoinette Schoar from August 11, 2015, Department of Labor. Pages 376-385 of hearing transcript that may be accessed at:
http://www.dol.gov/ebsa/pdf/1210-AB32-2-HearingTranscript2.pdf.
Australian Securities and Investments Commission, “Shadow Shopping Survey on Superannuation Advice,” ASIC Report 69, 2006.
Available at: https://dv8nx270cl59a.cloudfront.net/media/1347026/shadow_shop_report_2006.pdf.

146

gave such advice. Another (potentially overlapping) set of advisers had a conflict of interest
insofar as they recommended products that were associated with their employer. Out of these 96
advisers, 32 percent were judged to have given advice that did not have a reasonable basis,
whereas out of the 161 advisers that did not have this conflict, only 11 percent gave advice that
lacked a reasonable basis. Many clients of conflicted advisers were advised to switch funds,
predominantly to funds with higher fees, or falsely told that further contributions could not be
made to a current fund.
Additional, overseas audit-style studies reached similar conclusions with respect to
insurance intermediaries. Intermediaries in Germany provided low quality information.
Intermediaries in India provided little useful information and steered customers toward products
that advanced their own interests’ at their customers’ expense (Schwarcz and Siegelman 2015
forthcoming). In Chile, when consumers received advice from an insurance agent whose
commissions depend on the sales of annuity products, only 20% of those consumers chose the
most appropriate annuity offered. In contrast, 60% of consumers who received advice from
independent advisers who are not compensated by commissions chose the best annuity offered
(Stanko & Paklina, 2014).
Two other audit-style examinations provide further evidence that conflicts of interest
negatively influence adviser recommendations. The SEC investigated a series of “free lunch
seminars” that they concluded “were intended to result in the attendees’ opening new accounts
with the sponsoring firm and, ultimately, in the sales of investment products.” In 23 percent of
their targeted examinations, the SEC observed that recommendations from BDs and RIAs
appeared “unsuitable” for the individual consumer.343 These advisers were clearly providing
advice that was not in the best interest of their customers, likely a direct result of their inherent
conflict of interest as an employee of the firm sponsoring the seminar. An audit study of
advisers in the United Kingdom found that 1 in 5 failed to recommend the optimal product for
the customer, often instead recommending a product that returns higher commissions to the
adviser (Charles River Associates 2002).
With respect to advice on insurance products specifically, there is evidence that insurance
professionals themselves believe agents sometimes act on their conflicts at their customers’
expense. According to surveys conducted among life insurance professionals in 1990, 1995 and
2003, insurance professionals identified several issues as major ethics problems. Insurance
professionals identified the top four major problems as follows (Cooper and Frank 2005):

343

•

False or misleading representation of products or services in marketing, advertising or
sales efforts (rank 4/32 in 2003);

•

Conflicts between opportunities for personal financial gain (or other personal
benefits) and proper performance of one’s responsibilities (rank 2/32 in 2003);

•

Lack of knowledge or skills to competently perform one’s duties (rank 3/32 in 2003);

•

Failure to identify the customer’s needs and recommend products and services that
meet those needs (rank 1/32 in 2003);

SEC, “Protecting Senior Investors: Report of Examinations of Securities Firms Providing “Free Lunch” Sales Seminar,” 2007, p. 5.

147

Each of these four were scored on average as 3 or higher on a 1 to 5 scale where 1
indicated that this ethical issue was not a problem and 5 that it was a major problem.
The ethics environment did not change much over 13 years as these four issues
persistently ranked as top major issues. In addition, insurance professionals reported the
following conflict-related issues, although they did not rank consistently high:


Misrepresenting or concealing limitations in one’s abilities to provide services
(ranks 5/32 in 2003);



Conflicts of interest involving business or financial relationships with
customers, suppliers or competitors that influence, or appear to influence,
one’s ability to carry out his or her responsibilities (rank 9/32 in 2003);



Conflicts of interest involving the marketing of products and services
competing with those of one’s own company (rank 10/32 in 2003).

Another survey results suggest that these conflicts are likely systematic problems, not
necessarily caused by the faults in personal characters of professionals. When the professionals
were asked what programmatic changes would result in response to a new ethical standard, they
identified the following as the top four:
•

Influencing the senior managers of the principal life insurance companies that the
professionals represent to more strongly encourage and support ethical market
conduct (rank 1/9 in 2003);

•

Influencing the senior managers of life insurance companies in general to more
strongly encourage and support ethical market conduct (rank 2/9 in 2003);

•

Improving the ethical environment/culture at the principal life insurance companies
that the professionals represent in the sale of individual life insurance and annuity
products (rank 3/9 in 2003);

•

Improving the ethical environment/culture at the life insurance companies in general
(rank 4/9 in 2003).

The findings above suggest structural and cultural issues deeply embedded in the
insurance business model. Therefore, it might be extremely difficult for insurance professionals
to voluntarily eliminate or reduce conflicts in their practice and align their interests with
customers, in the absence of regulatory changes.
Similar to the audit study about the quality of financial advice in the U.S. by
Mullainathan, Noeth and Schoar (2012), there is another audit study examining the quality of
advice about life insurance products in India (Anagol, Cole and Sarkar 2013). Although this
study examines life insurance products sold in India, this study is relevant in examining the
annuity market in the US because it examines the same core issue – conflicts of interest – and
how these conflicts of interest influence the advice that insurance agents give to their prospective
customers. As in the U.S., insurance agents in India also receive compensation from
commissions. The amount of commissions varies by the type of products that insurance agents
sell. The products that are clearly worse for consumers – higher premiums and lower pay-outs –
often pay higher commissions to agents. In this field experiment audit study, the researchers
find that insurance agents recommended more expensive products for consumers 60 percent to
90 percent of the time, even when these products were clearly not suitable for consumers based
on consumer’s stated needs. A more troubling result was that the quality of advice varied by the
sophistication of the customers. When insurance agents realized that the prospective customers
148

were less sophisticated, agents were more likely recommend unsuitable products. In this
experiment, insurance agents could easily tell the financial sophistication of the customer based
on the statements that the customer made during their conversation. Another interesting finding
of this study is that disclosure of the information about commissions was not sufficient to
address conflicts of interest. In 2010, India’s insurance regulator required insurance agents to
disclose the commissions they earned from one particular product. Comparing the
recommendations by agents before this rule to recommendations made after this rule, Anagol,
Cole and Sarkar found that agents were less likely to recommend the product requiring the
disclosure and more likely to recommend the product paying higher commissions.
These findings support other literature cited in this Regulatory Impact Analysis
examining the effects of conflicts of interest in financial advice and documenting its harmful
effects on consumers in general.

3.2.3.4.2

Questionable Investments

The audit study literature provides convincing evidence that conflicts of interest
negatively influence adviser recommendations. Other studies using broader, nationwide data
produce corroborating results by finding that investor dollars tend to flow toward mutual funds
that send a large portion of their revenue back to the investor’s adviser.
Christoffersen, Evans, and Musto (2013) (CEM) find that payments to brokers influence
the advice they provide to clients. Examining U.S. mutual funds from 1993 through 2009 the
authors find that mutual funds that make larger load-sharing payments to brokers attract more
investor dollars. Unaffiliated brokers, in particular, appear to be strongly influenced by these
payments. “For each $1 increment in the load payment to the broker there is a $14.20 increase
in flows.”
Other researchers arrive at a similar conclusion. Using data on U.S. equity, bond, and
hybrid mutual funds from 1992 through 2001, Zhao (2008) finds that front-end-loads and backend loads paid to mutual funds are positively associated with flows into those funds. He
interprets this finding to suggest that “brokers and financial advisers apparently serve their own
interests by guiding investors into funds with higher loads.” Hackethal, Haliassos, and Jappelli
(2012) find that advisers are influenced by conflicts of interest in Germany as well. In their
dataset, customers who relied on advice traded more frequently and were more likely to
purchase a product that helped the adviser reach a sales target. These results all indicate that the
influence of conflicts of interest on brokers’ advice is widespread. Taken together, the two
strands of literature presented above provide ample evidence that conflicts of interest influence
the advice provided to IRA investors. The audit study literature offers explicit examples of
advisers who act in their own interest rather than the interest of their clients. The econometric
literature shows that these are not isolated incidents and that conflicts of interest are sufficiently
widespread to meaningfully alter flows into mutual funds on a national scale.

3.2.3.4.3

Eroded IRA Returns

There is substantial evidence that conflicts in advice lead to eroded IRA investment
returns. A series of academic papers finds lower returns for mutual fund share classes and
distribution channels that are more prone to conflicts of interest. Australia’s ASIC study
discussed above projected inferior investment returns attributable to conflicted advisers’
recommendations that lacked reasonable bases.
Comments on the 2015 NPRM Regulatory Impact Analysis suggest that DOL
inappropriately interpreted results presented in some of the academic papers referenced in this
149

section and other sections of the Regulatory Impact Analysis. Of course, data can be interpreted
in a multitude of ways, and reasonable minds can disagree. However, DOL continues to
strongly believe that readings contained in the 2015 NPRM Regulatory Impact Analysis and
carried over into the current Regulatory Impact Analysis are the most appropriate interpretations
of these studies given the available data. All indications are that the authors of the cited studies
generally agree with DOL’s interpretations. For example, Jonathan Reuter wrote that, “These
papers have been used by the Council of Economic Advisers and the Department of Labor to
argue that conflicted advice is both common and costly. This is an accurate description of my
main findings.”344 Antoinette Schoar remarked that, “While surely [the 2015 NPRM] alone does
not solve all the problems that might arise in retail financial services, my research suggests that
it will actually help to improve the quality of the advice that people receive.”345 Finally, Susan
Christoffersen and Richard Evans did not publicly disagree with anything written in the 2015
NPRM Regulatory Impact Analysis which extensively cited their 2013 Journal of Finance
paper.346 In contrast, these same authors wrote the following in response to an ICI comment: “In
the ICI’s recent letter (dated July 21, 2015) to the Office of Regulations and Interpretations at
the US Department of Labor, the ICI makes several incorrect claims about the results and
interpretation of our paper… the claims made by ICI are incorrect. Regarding the first point, our
methodology accounts and adjusts for the variation in funds’ assets. Regarding the second, the
statistics the ICI chose for its letter are misleading, as is apparent in statistics from the ICI’s own
website that show that investments subject to loads have grown significantly. Lastly, with
regards to the latter two points, both are wrong.”347
ASIC found substantial harm to investors from conflicted advice.348 The authors identify
40 cases where advisers recommended switching funds and the advice did not have a reasonable
basis.349 In 23 of these cases, all of which involved a conflict of interest, the advisers provided
sufficient information to calculate the cost of the fund. Projections suggest that the high fees
charged by the recommended funds will reduce future retirement benefits for 20 of the 23
participants. If the projections bear out, the participants who received conflicted advice will
have their future retirement benefits reduced by as much as 38 percent relative to the benefit that
would be projected if they did not switch funds. The average projected benefit reduction is
approximately $37,000 Australian dollars, or 16 percent of the participant’s future benefit.
There are strong commonalities between the choices facing U.S. IRA investors and those facing
Australians when they save for retirement, suggesting that conflicts of interest are likely to be
similarly harmful in each arena.
Bergstresser, Chalmers, and Tufano (2009) find inferior mutual fund performance in
more conflicted distribution channels. Individuals can purchase mutual fund shares directly
from a mutual fund company, (“direct channel”) or through an intermediary or broker (“broker

344

345

346
347

348
349

Written testimony submitted by Jonathan Reuter to Department of Labor’s Conflict of Interest Public Hearing (Aug. 11); available at:
http://www.dol.gov/ebsa/pdf/1210-AB32-2-WrittenTestimony10.pdf.
August 11 Transcript of Department of Labor’s Conflict of Interest Public Hearing, Antoinette Schoar verbal testimony, p. 378; available at:
http://www.dol.gov/ebsa/pdf/1210-AB32-2-HearingTranscript2.pdf.
See Christoffersen, Evans and Musto (2013).
See Susan Christoffersen and Richard Evans’ comment letter, (September 10); available at: http://www.dol.gov/ebsa/pdf/1210-AB32-202766.pdf.
Australian Securities and Investments Commission, “Shadow Shopping Survey on Superannuation Advice,” ASIC Report 69, 2006.
The 40 cases where a recommendation to switch funds did not have a reasonable basis represent 32% of the 124 cases where an adviser
made a recommendation to switch funds and 14% of the 284 instances where any advice was given.

150

channel”). The distinction is useful for assessing the impacts of advice because both conflicts
and individualized investment advice are prevalent in the broker channel, but rare in the direct
channel. The authors examine mutual fund returns between 1996 and 2004 without factoring in
distribution costs (loads or 12(b)-1 fees). They find that funds distributed through the more
conflicted broker channel perform worse. Domestic equity funds sold through the direct channel
outperform brokered equity funds by between 0.33 percent and 0.88 percent on a risk adjusted
basis. Likewise, bond funds and money-market funds sold through the direct channel
outperform their full-service counterparts by 0.56 percent to 0.90 percent and 0.040 percent to
0.043 percent, respectively. In all three cases, it appears that the conflicted advice that is given
by brokers has a harmful effect on the individual’s financial situation, including, in many cases,
the individual’s retirement benefit. Unlike the other fund categories, foreign equity mutual
funds sold through the broker channel outperform direct foreign equity funds by 1.53 percent to
2.05 percent, but this result may not be generalizable because it is attributable to favorable
performance within just one large mutual fund family.
Overall, the authors calculate that the cost of using the broker channel, in terms of
reduced returns alone, was $4.6 billion in 2004. This cost is in addition to the estimated $9.8
billion per year that the same customers paid in 12b-1 fees, and neither of these numbers
includes the loads paid by customers who purchase funds through brokers.
Del Guercio and Reuter (2014) find that broker-sold funds underperform direct-sold
funds by an average of 1.15 percentage points per year after accounting for risk and other
factors. The authors identify misaligned incentives in the broker-sold market as the cause of the
underperformance. In the direct-sold market, asset managers are incentivized to generate alpha
(superior performance above and beyond that of the market). As a result, the authors find that
within the direct-sold market, actively managed funds perform similarly to index funds.
However, in the broker-sold market, asset managers are not sufficiently incentivized to produce
alpha. Instead, mutual funds can sell more of their product by making higher payments to
brokers through load-sharing and revenue-sharing.350 In this broker-sold market, actively
managed funds underperform index funds by 1.12 – 1.32 percentage points per year.
Del Guercio, Reuter, and Tkac (2010) present similar evidence on mutual fund
performance across distribution channels. In the sample of domestic equity funds between 1996
and 2002, direct channel funds outperform brokered funds by 0.08 percentage points to 0.12
percentage points per month, or about 1.0 percentage points to 1.5 percentage points per year.351
The authors hypothesize that the returns difference is due to the lack of incentive for mutual
funds in the broker channel to find and pay for top-quality portfolio management in order to
maximize risk-adjusted investor returns. This hypothesis is supported by the finding that, while
actively-managed funds perform poorly across the entire sample, within the direct channel, the
performance of actively-managed funds is equal to that of index funds. In the direct channel,
investors are typically more sophisticated and more attentive to performance, giving activelymanaged mutual funds in this channel a strong incentive to invest in portfolio management.
Conversely, in the broker-sold channel, investors rely on brokers to evaluate the quality of
actively-managed funds. Brokers may be attentive to performance, but they often have a

350
351

See Christoffersen, Evans and Musto (2013).
The Department’s calculation assumes a 6.00 percent annual return for direct channel funds.

151

competing interest – the load-sharing and revenue-sharing that they receive as a result of the sale
of particular funds. Thus, in the broker-sold channel, a mutual fund may find it more profitable
to induce the sale of their fund by increasing broker payments rather than by investing in
performance. The evidence bears out this hypothesis.
Harm from adviser conflicts is also evident in a comparison of returns across mutual fund
share classes. Class A, B, and C shares all include one or more type of load and often a
distribution fee. As described earlier, many of the dollars from these loads and fees end up
being returned to the broker recommended the purchase of the fund. Where individualized
investment advice is given, these loads and fees can create a conflict of interest for the broker.
There is evidence that load fund investors fare worse than no-load fund investors, which strongly
suggests that conflicts harm IRA investors.
Morey compares the performance of load and no-load domestic equity mutual funds
between 1993 and 1997 (Morey 2003). Without taking the load into account, no-load funds
outperformed load funds 0.03 percentage points or 0.06 percentage points per month, or 0.43
percentage points to 0.82 percentage points per year on a risk adjusted basis. This result alone
suggests that the conflicted advice received from brokers is harmful to individual investors,
including IRA investors. However, adjusting for the actual loads that investors pay reveals that
the magnitude of the problem is much larger. Factoring in the loads paid, load funds
underperform no-load funds by 1.6 to 2.0 percentage points per year on a risk-adjusted basis.352
The load-adjusted returns differences are a more complete estimate of the potential cost to
consumers of harmful conflicted advice.
Friesen and Sapp (2007) investigate how actual investor performance (asset-weighted) in
load and no-load funds combined differs from the performance reported in the funds’
prospectuses (time-weighted). Additional estimates from what appear to be the same data are
presented in a second paper with co-author Bullard (Bullard, Friesen, and Sapp 2008). In the
sample of domestic equity fund returns between 1991 and 2004, actual investor performance
generally lags the performance reported in the prospectuses because investors have poor timing
– they tend to have more money invested in funds when returns are low and less money invested
when returns are high.353 For the purpose of this impact analysis, differences in performance
between load and no-load funds are more of a focus than differences between actual investor
performance and reported performance. However, the latter may play a part in the former if
investor timing in load funds is better or poorer than investor timing in no-load funds. Bullard,
Friesen, and Sapp (2008) find that the difference in performance between load and no-load funds
has two components: first, the difference in prospectus returns across share classes; and second,
an additional difference in investor returns resulting from differences in investor timing.

352
353

The Department’s calculations are based on Morey (2003), Table 3, p. 1261.
This phenomenon is sometimes characterized as a “disposition effect” whereby investors sell winning investments too soon and hold losing
investments too long (Shefrin and Statman 1985; Odean 1998). Such investor tendencies have been well documented (Weber and Camerer
1998). The disposition effect is often explained by prospect theory and/or cognitive dissonance. Prospect theory suggests that investors
value gains and losses relative to the initial purchase prices and investors become risk averse with respect to protecting gains but riskseeking with respect to recouping losses (Della Seta and Gryglewicz 2015). Consequently investors sell the winners too soon and hold the
losers too long. Cognitive dissonance suggests that investors are reluctant to realize their losses because they cannot admit that they made
poor investment decisions. Thus they keep losers too long. This effect may be absent with respect to actively managed mutual funds,
because investors may blame the fund manager rather than themselves for the poor result. It might be more likely to be manifest with
respect to passive funds or single-issue stocks (Chang, Solomon and Westerfield 2016).

152

Consistent with the other studies presented in this section, the researchers find that investors who
use brokers have poorer investment results. Looking only at prospectus returns, no-load funds
outperform Class A load funds by 0.03 percentage points to 0.06 percentage points per month,
Class B load funds by 0.11 percentage points to 0.13 percentage points per month, and Class C
load funds by 0.04 percentage points to 0.06 percentage points per month on a risk-adjusted
basis. In addition to this underperformance, the researchers find that the gap between prospectus
returns and actual (poorer) investor returns is larger for load funds (0.14 percentage points, 0.19
percentage points, and 0.11 percentage points per month for Class A, B, and C shares,
respectively) than for no-load funds (0.07 percentage points per month).354
This result sheds light on one of the paths through which conflicted advice can be
harmful to IRA investors. Friesen and Sapp (2007) find that “timing underperformance is
consistent with investor return-chasing behavior.” Conflicts in advice appear to exacerbate the
tendency for IRA investors to chase returns and trade excessively, and the results presented here
suggest that the consequences can be large. When prospectus returns and investor timing are
both considered, the data reveal that investors in load funds underperform investors in no-load
funds by 1.9 percentage points to 2.2 percentage points per year.
Christoffersen, Evans, and Musto (2013) (CEM) estimate the impact of load-sharing –
payments from the mutual fund to the broker – on mutual fund returns. In contrast to the studies
reviewed above that compare returns across distribution channels or across fund share classes,
these authors compare returns within a particular share class – Class A, with front-end-loads.
The data reveal that as the size of the load-share increases, mutual fund returns decrease. This
suggests that the greater the magnitude of the adviser’s conflict of interest, the worse off the IRA
investor can expect to be. For “the average 2.3 [percentage points] payment to unaffiliated
brokers” an IRA investor or other customer can expect “a 1.13 [percentage point] reduction in
annual performance” of the mutual fund. If the payment to the broker is higher than 2.3
percentage points, as is often the case, the IRA investor will likely suffer even more.355
The evidence discussed above on balance strongly supports the conclusion that
individuals who seek advice from conflicted brokers have substantially worse outcomes than
those who invest directly in mutual funds. There is also evidence that consumer harm from
adviser conflicts extends to advisers other than BD representatives and to markets beyond the
US.
Findings from Chen, Yao, and Yu (2007) suggest that brokers who are affiliated with
insurers (and therefore are likely to be insurance agents as well) also act on conflicts at IRA
investors’ expense. The authors investigate the performance of mutual funds managed by
insurance companies or their affiliates. They note that “insurance funds are often cross-sold
through the extensive broker/agent network of their parent firms.” This relationship between the
broker, who in many cases provides individualized investment advice, and the mutual fund, can
create a conflict of interest, particularly when differential compensation is paid by the insurance
company to the broker to promote the sale of one or more funds. In a sample of actively-

354

355

The performance differences presented in Friesen and Sapp (2007) and Bullard et al. (2008) do not account for the actual loads paid by
investors in load and no-load funds.
Similar to Friesen and Sapp (2007) and Bullard et al. (2008), the performance reduction presented in CEM does not include loads paid by
investors in front-end-load funds.

153

managed domestic equity funds’ returns between 1990 and 2002, the authors identify funds
owned by insurance companies and compare their returns to returns for the remainder of the
funds in the sample. Note that this is not a clean comparison of funds that do and do not involve
conflicts of interest in their distribution. Many of the non-insurance funds in the sample will
also be distributed by brokers who face conflicts of interest. As such, any observed
underperformance of insurance funds could be viewed as an underestimate of the harm to
insurance fund investors from conflicted advice. The data show that insurance funds
underperform non-insurance funds by 0.85 percentage points to 1.4 percentage points per year
on a risk-adjusted basis. The authors are able to confidently rule out the possibility that lower
insurance fund returns are a result of insurance companies reducing systematic risk or that they
reflect “rational learning about managerial ability,” and argue that they are due to “lack of
investor oversight on poorly performing insurance funds.” This lack of oversight allows
advisers to act on their conflicts of interest without negative market consequences, as discussed
earlier. The authors conclude that “underperformance due to lack of investor monitoring is quite
likely a universal problem in the fund business,” and advise that similar conflicts of interest
“may affect mutual funds sponsored by other types of financial institutions, such as commercial
banks and investment banks.”
Chalmers and Reuter (2014) study investment performance in the Oregon University
System’s defined contribution retirement plan and find that participants who receive advice from
brokers underperform relative to self-directed portfolios (by 1.54 percentage points) and also
relative to the default target-date fund. The underperformance relative to self-directed portfolios
costs each advice recipient an average of $530 per year. The authors also find that the brokeradvised portfolios are riskier than self-directed portfolios, despite the underperformance.
Hackethal, Haliassos, and Jappelli (2012) utilize datasets from a large German brokerage
firm and from a large German commercial bank to investigate whether conflicted advice harms
customers. The datasets, on the level of the individual customer, include portfolio performance
between 2003 and 2005, demographic characteristics of the customer, and an indicator of
whether the customer received investment advice. The data show that brokerage clients who
receive investment advice have inferior portfolio returns relative to those who do not receive
advice, in the amount of 5.0 percent per year after fees have been factored in. The demographic
characteristics in the dataset allow the researchers to examine whether the underperformance
could be caused by inherent differences between customers who seek advice and those who do
not. However, after controlling for personal and regional characteristics, the estimated
underperformance of advised accounts remained virtually unchanged. The authors also find
evidence of churning among advised accounts; the average turnover rate is more than double that
of self-managed accounts. Because advisers get commissions based on the volume of purchases,
this churning can be viewed as additional evidence that the harm – the underperformance of
advised accounts – is a result of conflicted advice. Finally, the authors find that the results from
the commercial bank dataset are consistent with those from the brokerage firm dataset, pointing
“to systematic negative effects of financial advisors rather than to statistical flukes or sample
peculiarities.”

154

Some comments on the 2015 NPRM Regulatory Impact Analysis suggest that the
Department should not be regulating in the insurance market because there is less evidence that
conflicts are harmful in the sale of annuity products (compared to evidence that conflicts are
harmful in the sale of securities).
When the expert knows more about goods and services than consumers, the goods and
services are often called “credence goods.”356 Examples of credence goods are services
provided by lawyers, doctors and mechanics. In the market for credence goods, an expert has an
incentive to exploit the asymmetric information in the market, which sometimes results in fraud,
overcharges, undertreatment and overtreatment. The remedies for the informational problems in
the market for credence goods vary by specific goods and conditions. For example, in medical
treatments and with individual physicians, the Hippocratic Oath can be viewed as a partial
remedy to address the issues with respect to credence good.
Biased recommendations regarding variable annuities can be especially costly for IRA
investors. The SEC’s online “Investor Information” resources provide a consumer primer on
variable annuities.357 It punctuates the issues with five “Caution!” boxes that warn:


Variable annuities may be disadvantageous as IRA investments,



Various benefits add to costs, might not be needed, and might be available separately
elsewhere at better prices,



Exchanging one variable annuity for another may be disadvantageous,



Bonus credits may cost more than they are worth, and



Exchanging products to gain bonus credits is likely to be disadvantageous.

Schwarcz and Siegelman (2015 forthcoming) argue that insurance “agents can
inefficiently withhold information and distort consumer choices by providing misleading
information or operating in their own self-interests.” They conclude “that neither market forces
nor legal or regulatory rules substantially constrain insurance agents’ capacity to advance their
own interests by providing biased advice, though direct empirical evidence about the frequency
of such misbehavior is limited.”

3.2.3.4.4

Alternative Explanations for
Underperformance

Above, the Department presents evidence of the underperformance of retail assets held as
a result of conflicted investment advice. In many cases, the underperforming assets are brokersold mutual funds. Before reaching any strong conclusions about harms caused by conflicts of
interest, the Department considered other possible explanations for the underperformance of
these assets. In the same paper reviewed above, Bergstresser et al. discuss and ultimately
dismiss several possible alternative explanations for the returns discrepancies (Bergstresser,
Chalmers, and Tufano 2009).

356

357

Uwe Dulleck and Rudolf Kerschbamer, 2006, “On Doctors, Mechanics, and Computer Specialists: The Economics of Credence Goods”
Journal of Economic Literature Vol . XLIV pp.5-42.
SEC, “Variable Annuities: What You Should Know,” available at: http://www.sec.gov/investor/pubs/varannty.htm.

155

First, brokers do not appear to provide for superior asset allocation advice across asset
classes. All of the results presented above have examined the performance of mutual funds
within broad asset classes, such as domestic equity, foreign equity, and bond funds. But brokers
also provide advice on how to allocate assets across these asset classes over time. If broker
channel assets are more often in equity funds when equity markets do well and more often in
bond funds when equity markets do poorly, then customers, including IRA investors, will
benefit. Moreover, the benefit to IRA investors will not show up in within-asset-class returns
discrepancies. To the extent that brokers provide high quality asset allocation advice, the benefit
to customers may offset or even outweigh the inferior returns generated within the asset classes.
To test whether brokers provide superior asset allocation advice, (Bergstresser, Chalmers, and
Tufano 2009) simulate the growth of direct channel and broker channel assets between 1981 and
2002 using the actual aggregate asset mix from each channel over the time period. Statistical
tests on their data find no evidence that broker channel funds have superior asset allocation.
Also, recall that for a sample of domestic equity funds, (Bullard, Friesen, and Sapp 2008) find
that load fund investors have significantly poorer investment timing than no-load fund investors.
Second, brokers do not appear to recommend less expensive funds to their clients.
Distribution fees, expense ratios, and loads are all generally higher for broker channel funds than
for direct channel funds (Bergstresser, Chalmers, and Tufano 2009, 4148, Table 5).
Third, while brokers may serve a different set of customers, the differences appear to be
limited and in any event seem unlikely to explain the observed results. As noted by
Bergstresser, Chalmers, and Tufano (2009), a significant fraction of customers purchase mutual
funds through both the direct channel and the broker channel.358 The customers who choose
only one channel or the other appear to not be very different across observable characteristics.
In general, broker clients have only slightly lower average incomes, they are only a bit more risk
averse, and have similar investing goals. Bergstresser, Chalmers, and Tufano (2009) suggest
that investors in the two distribution channels are more similar than different, stating that, “by
any standard, mutual fund investors in both channels are disproportionately drawn from upper
ranks of national wealth, income, and educational attainment.” Customers across the two
channels may differ in other, non-observable ways, but the authors find that it is “problematic to
explain how these traits lead investors to continue to accept poorer pre-distribution-fee
investment performance.” Chalmers and Reuter (2014) find that measured underperformance of
broker advised portfolios decreases by only 7 percent to 11 percent when controlling for
observable, individual-level characteristics.359 This result suggests that the difference in
performance across distribution channels is not driven by differences in the individuals choosing
each channel. Hackethal, Haliassos and Jappelli (2011) similarly find that the measured
underperformance of advice recipients does not change after controlling for observable,
individual characteristics.
Fourth, it appears that brokers fail to help investors overcome important “behavioral
biases” that impair their financial decisions. Mullainathan, Noeth, and Schoar (2012) provide

358

359

ICI Research Report, “Profile of Mutual Fund Shareholders, 2014” (Feb. 2015), p. 19, available at:
http://www.ici.org/pdf/rpt_15_profiles.pdf.
Across OLS (Fama-MacBeth) regressions, measured broker underperformance is 2.62 percent (2.52 percent) without controlling for
investor characteristics and 2.44 percent (2.24 percent) with investor characteristics included – a difference of 0.18 (0.28) percentage points.

156

additional evidence, at the level of the individual adviser, that brokers intensify this same
returns-chasing bias.
After ruling out the above explanations for the returns discrepancies presented in this
section, there remains the possibility that broker customers receive some other benefit or benefits
that are not observed by the researchers. These may include both non-financial benefits, such as
peace of mind and time savings, and benefits with a financial component not directly related to
the performance of a mutual fund, such as help understanding various investment options, estate
planning, and help establishing savings goals. A 2006 ICI survey finds that all of these benefits
are important to at least some customers of financial professionals.360 (Bergstresser, Chalmers,
and Tufano 2009) call these unobserved benefits “intangible benefits” and suggest that
intangible benefits and conflicted advice are two alternative hypotheses that can explain the
underperformance of broker channel funds. What evidence is there on each of these hypotheses?
There is a great deal of persuasive evidence to suggest that conflicts of interest are
harmful to IRA investors. Much of that evidence is presented in the preceding sections.
Conflicts of interest are prevalent in the market. The majority of investors are not sophisticated
and do not have the necessary skill and information to act as a check on adviser misbehavior.
Finally, there is substantial evidence that conflicts of interest do in fact influence adviser
recommendations, as both the recommendations themselves and the investments made pursuant
to them appear to be compromised in ways that harm IRA investors.
In contrast, evidence favoring the “intangible benefits” hypothesis is limited. As
mentioned above, there is some industry-generated survey evidence to support the notion that
broker customers receive additional benefits from dealings with brokers beyond mutual fund
performance. The evidence indicates that survey respondents received some peace of mind, time
savings, help understanding various investment options, estate planning, or help establishing
savings goals. In addition, Foerster et al. (2014) find that advised Canadian investors
underperform primarily because they pay higher fees. In light of the underperformance, they
conclude that investment advice services alone cannot justify the fees. However, they also find
robust evidence that advice affects savings styles and levels, and suggest that the higher fees
paid by advised investors might reflect payment for broader financial advice. Both the industrygenerated investor survey results and Foerster et al.’s finding of savings impacts suggest that at
least some of advised investors’ excess fees (and associated underperformance) can be
interpreted as fair payment for financial services that yield consumer benefits other than
improved investment performance. Neither of these, however, challenges the more extensive
and robust evidence, presented above, that investors often cannot understand the cost of their
advisers’ services and generally cannot determine whether the value of those services justify
their cost. As such, both of these findings are consistent with the proposition that advised
investors’ higher fees and underperformance are excessive relative to the services their advisers
provide.
Taken as a whole, the evidence strongly favors conflicts of interest as the primary cause
of returns differences across distribution channels and share classes. That is not to say that

360

ICI Research Fundamentals, “Why Do Mutual Fund Investors Use Professional Advisers?” (April 2007), available at:
http://www.ici.org/pdf/fm-v16n1.pdf.

157

benefits such as savings goals, estate planning, and time savings can be completely ruled out as
factors contributing to returns differences. But to suggest that these unobserved benefits explain
the entirety or large part of the returns differences and that conflicts of interest play little or no
role would be to turn a blind eye to a robust body of evidence on conflicts of interest.
Furthermore, as discussed earlier, CEM find that even within a share class, larger
conflicts of interest imply greater harm to the individual investor. For the results to be explained
by benefits such as savings goals, estate planning, and time savings, these unobserved benefits
would have to increase as load-sharing increases. However, it is hard to justify why that
relationship should occur. Customers of brokers are very unlikely to know the amount of loadsharing that is involved in their mutual fund purchases, and therefore to have any mechanism to
demand greater services from the broker. Higher load-sharing does not create any incentive for
the brokers to provide a higher level of service. Or conversely, brokers have no incentive to
reduce the load-share that they receive just because a customer requires a lower level of service.
In sum, the weight of the evidence supports the finding that biased advice, rather than
unobserved benefits, is the primary cause of the inferior returns suffered by IRA investors in
conflicted load/distribution channels.

3.2.4 Magnitude of Harm
A wide body of economic evidence supports a finding that the impact of these conflicts
of interest on investment outcomes is large and negative. The supporting evidence includes,
among other things, statistical analyses of conflicted investment channels, experimental studies,
government reports documenting abuse, and economic theory on the dangers posed by conflicts
of interest and by the asymmetries of information and expertise that characterize interactions
between ordinary retirement investors and conflicted advisers. A careful review of this data,
which consistently point to a substantial failure of the market for retirement advice, suggests that
IRA holders receiving conflicted investment advice can expect their investments to
underperform by an average of 50 to 100 basis points per year over the next 20 years. The
underperformance associated with conflicts of interest – in the mutual funds segment alone –
could cost IRA investors between $95 billion and $189 billion over the next 10 years and
between $202 billion and $404 billion over the next 20 years. While these expected losses are
large, they represent only a portion of what retirement investors stand to lose as a result of
adviser conflicts. Data limitations impede quantification of all of these losses, but there is ample
qualitative, anecdotal, and in some cases empirical evidence that they occur and are large both in
instance and on aggregate.
Strong evidence ties adviser conflicts to investments in higher-load, more poorly
performing mutual funds. Other evidence strongly suggests that adviser conflicts inflict
additional losses, possibly of a similar magnitude, by prompting IRA investors to trade more
frequently, which will increase transaction costs and multiply opportunities for chasing returns
and committing timing errors. Adviser conflicts likely are also associated with excessive price
spreads in principal trades between IRA investors and BDs. Other types of investments such as
single-issue securities, banking or insurance products also are likely to be subject to
underperformance due to conflicts (Evans and Fahlenbrach 2012).
Academic literature available at the time of publication of the 2015 NPRM Regulatory
Impact Analysis showed that investments distributed through conflicted channels underperform
peer comparison investments. The results from each of these studies are concerning. Each study
identifies investment underperformance that could cost IRA investors tens or hundreds of
158

billions of dollars over a ten-year period. Figure 3-17 summarizes the most relevant studies
identified by the Department (also see Burke et al. 2015, 13).361

Figure 3-17 Literature Investigating the Performance of Assets Held as a Result of Conflicted
Investment Advice
Paper

Sample

Methodology

Annual Impact

Bergstresser,
Chalmers, and
Tufano (2009)

Domestic equity,
foreign equity, bond,
and money market
mutual funds; 19962004

Compares annual
performance – prior to
distribution fees – of
broker-sold funds with
direct-sold funds

Broker-sold domestic equity funds
underperform by 0.27-0.88
percentage points on an assetweighted basis and by 0.93-2.50
percentage points on an equalweighted basis.
Broker-sold foreign equity overperform by 1.45-3.26 percentage
points on an asset-weighted basis,
but underperform by 1.13-2.08
percentage points on an equalweighted basis.362
Broker-sold bond funds
underperform by 0.14-0.90
percentage points on an assetweighted basis and by -0.10-0.45
percentage points on an equalweighted basis.

361
362

Bullard, Friesen,
and Sapp (2008)

Domestic equity
mutual funds; 19912004

Investigates how load and
no-load fund investor
returns compare to a buyand-hold strategy

Load funds underperform a buyand-hold strategy by 1.82
percentage points, more than
double the underperformance for
no-load investors.

Chalmers and
Reuter (2014)

Oregon University
System’s defined
contribution
retirement plan
accounts; 1996-2007

Estimate the causal impact
of brokers on their clients’
portfolio

Broker clients underperform selfdirected investors by 1.54
percentage points.

Christoffersen,
Evans, and Musto
(2013)

Mutual funds with
front-end-loads;
1993-2009

Investigates the effect of
load sharing and revenue
sharing on performance

Investment returns decrease by
about 50 basis points for every 100
basis points of load-share for
mutual funds that pay load-shares
exclusively to unaffiliated brokers.

Del Guercio,
Reuter, and Tkac
(2010)

Domestic equity
mutual funds, 19962002

Compares returns for
broker-sold funds with
comparable direct-sold
funds

Broker-sold actively-managed funds
underperform direct-sold funds by
approximately 1 percentage point.

Figure 3-17 was adapted from (Burke et al. 2015).
The authors report that the asset-weighted broker-sold over-performance is “attributable to a small number of very large international funds
sold through one specific broker-channel fund family.”

159

Del Guercio and
Reuter (2014)

Domestic equity
mutual funds, 19922004

Compares returns for index
funds with actively
managed funds in the direct
channel and comparable
funds in the broker channel

Direct-sold actively managed funds
do not underperform index funds,
but broker sold actively managed
funds underperform index funds by
approximately 1 percentage point.

Friesen and Sapp
(2007)

Domestic equity
mutual funds, 19912004

Investigates how load and
no-load fund investor
returns compare to a buyand-hold strategy

Load funds underperform a buyand-hold strategy by about 2
percentage points, approximately
double the underperformance for
no-load investors.

Hackethal,
Haliassos and
Jappelli (2011)

Customers of a large
German brokerage
firm and customers of
a large German
commercial bank;
2003-2005

Compares net returns of
advised and self-managed
accounts

Annual returns for advised accounts
are lower by approximately more
than 4 percentage points.

Morey (2003)

Domestic equity
mutual funds; 19931997

Compares performance of
load and no-load funds
before and after adjusting
returns for loads paid.

Load funds underperform no-load
funds by 0.4-0.8 percentage points,
prior to adjusting for loads. After
adjusting returns for loads paid,
load funds underperform no-load
funds by 1.3-2.0 percentage
points.363

The Department reviewed two papers that consider the performance difference across
mutual fund distributions channels – the direct brokerage channel and the full-service brokerage
channel – and three papers that consider performance differences between load funds and noload funds. Beyond the groups that are compared, there are additional differences in the
methodology of the academic studies. One paper (Morey 2003) accounts for the actual loads
paid by load fund investors, while the others do not. One set of papers (Bullard, Friesen, and
Sapp 2008; Friesen and Sapp 2007) investigates the timing gap – the difference between the
performance of actual investors and the performance of the funds that is captured in the
prospectus that is due to the general poor timing of investors – while the others do not. One
paper (Bergstresser, Chalmers, and Tufano 2009) measures returns differences on an assetweighted basis thereby giving large funds more impact on the result relative to small funds,
while the others do not. None of the papers adjust for the selection of investors into distribution
channels or load/no-load funds, but one paper (Chalmers and Reuter 2014) has information on
individual investors, and is able to control for observable investor characteristics. Finally, two
papers (Bergstresser, Chalmers, and Tufano 2009; Christoffersen, Evans, and Musto 2013)
include bond mutual funds and international equity mutual funds in their analysis, while the
other papers estimate underperformance in domestic equity mutual funds only.
The evidence regarding broker-sold mutual fund underperformance presented in Figure
3-17 and in the 2015 NPRM Regulatory Impact Analysis focuses heavily on domestic equity
mutual funds. Domestic equity is the largest segment of the IRA mutual fund market. Conflicts
of interest that cause underperformance in the domestic equity segment alone would cause

363

DOL calculations. Morey (2003) presents returns on a monthly, rather than annual, basis.

160

substantial harm to IRA investors. However, there may be reason to believe that
underperformance in the domestic equity segment could be an indicator of harm due to
conflicted advice regarding other mutual funds as well.364 Across the spectrum of mutual funds
– domestic equity, foreign equity, and bond funds included – brokers face similar conflicts of
interest when advising clients where to invest their savings.
None of these papers listed in Figure 3-17 attempts to detect some major possible sources
of underperformance of IRA assets attributable to conflicts of interest. None accounts for other
potential sources of loss from conflicts such as mark ups in principal transactions, transaction
costs associated with the purchase of securities other than mutual funds, investment timing losses
associated with such other securities, and excessive premiums and/or unfavorable mortality
tables associated with insurance company products. If conflicted recommendations lead
investors to hold underperforming assets and incur timing losses, it seems likely they would also
lead investors to trade securities excessively and thereby generate more mark ups or
commissions. Investors’ tendency to chase returns and pay insufficient attention to expenses,
and advisers’ incentives to exploit these tendencies for personal gain, likely are not limited to
mutual fund sales and recommendations.
On the other hand, as discussed earlier, some of the performance gap identified in many
of these papers may reflect the fair value of unobservable or intangible benefits of advice. Also
as discussed earlier, however, available evidence suggests that only a fraction of the
performance gap can be attributed to fair compensation for services. The majority likely reflects
harm from adviser conflicts, comprising transfers from IRA investors to conflicted advisers and
others in the supply chain and social welfare losses from capital misallocation.
Comment letters on the 2015 NPRM Regulatory Impact Analysis and testimony
delivered at the DOL hearing on conflicts of interest in investment advice in August 2015
demonstrate that the harm from conflicts is ongoing and can be expected to continue into the
future in the absence of regulatory action. At the hearing, Antoinette Schoar described new
research that finds that conflicted investment advisers continue to act on their conflicts of
interest.365 A comment letter from academics Susan Christoffersen and Richard Evans clarifies
that the mutual fund market has not undergone a fundamental change in recent years.366 Mercer
Bullard’s testimony at the hearing provides recent examples of conflicted compensation schemes
that incentivize brokers to provide bad advice to clients.367

364

365

366

367

Domestic equity is unique in that there is almost full information available to all market participants, especially for large cap domestic
stocks. This information makes the market highly efficient and limits the ability of asset managers to produce alpha – returns above and
beyond that of the market. When conflicts of interest harm investors in domestic equity mutual funds, that harm is largely unobscured by
other factors that affect performance, such as the ability of asset managers. By contrast, in other markets such as foreign equity, the ability
of skilled asset managers to produce relatively strong performance may hide the harmful effects of conflicts of interest on performance,
especially when performance is measured on an asset-weighted basis. Findings from Bergstresser et al. (2009) support this hypothesis. As
presented in Figure 3-17 and in the 2015 NPRM Regulatory Impact Analysis, Bergstresser et al. (2009) found that broker-sold foreign
equity mutual funds overperformed direct-sold foreign equity funds when performance was averaged on an asset-weighted basis, but
underperformed direct-sold foreign equity funds when performance was averaged on an equal-weighted basis. The authors stated that the
asset-weighted outperformance was “attributable to a small number of very large international funds sold through one specific brokerchannel fund family.”
Testimony of Antoinette Schoar from August 11, 2015, Department of Labor. Pages 376-385 of hearing transcript; available at:
http://www.dol.gov/ebsa/pdf/1210-AB32-2-HearingTranscript2.
See Susan Christoffersen and Richard Evans’ comment letter, (September 10); available at: http://www.dol.gov/ebsa/pdf/1210-AB32-202766.pdf.
Testimony of Mercer Bullard from August 10, 2015, Department of Labor. Pages 167-176 of hearing transcript; available at:
http://www.dol.gov/ebsa/pdf/1210-AB32-2-HearingTranscript1.pdf.

161

Furthermore, the Department fully recognizes the trends that have been occurring in the
mutual funds market. A decline in the average size of front-end-loads and a decline in the
frequency with which retail mutual fund assets are subject to front-end-loads are trends that were
incorporated into the gains-to-investors calculations in the 2015 NPRM Regulatory Impact
Analysis and continue to be incorporated into the gains-to-investors calculations in this
document.
Available empirical evidence, while broadly consistent in finding that adviser conflicts
harm investors, varies widely with respect to both the type of harm considered and the
magnitude of such harm. Therefore, the Department developed a number of different estimates
of the overall performance gap associated with conflicted advice, reflecting different
assumptions and methods, all of which are grounded in one or more academic empirical
studies.368 The estimates are described briefly below. A fuller explanation for these
underperformance estimates is provided in Appendix B.
Various studies (see Figure 3-17 above) consistently show that broker-sold mutual funds
underperform direct-sold mutual funds, and some of these studies suggest a modest
underperformance of approximately 50 basis points per year. Applying a 50 basis point
performance gap to the current IRA marketplace implies an earnings deficit of $9 billion per
year, $95 billion over 10 years, and $202 billion over 20 years.369 Many studies estimate a larger
amount of underperformance of approximately 100 basis points per year. Applying this
performance gap to the current IRA marketplace implies an earnings deficit of $17 billion per
year, $189 billion over 10 years, and $404 billion over 20 years. Some studies suggest that the
underperformance of broker-sold mutual funds may be even higher than 100 basis points,
possibly due to loads that are taken off the top and/or poor timing of broker sold investments.
Put differently, if underperformance is 50 or 100 basis points per year, an ERISA plan investor
who rolls her retirement savings into an IRA could expect to lose 6 to 12 percent of the value of
her savings over 30 years of retirement by accepting advice from a conflicted financial adviser.
Unfortunately, some investors will fare worse than average. An ERISA plan investor who
receives particularly poor advice from a conflicted adviser and moves her savings into assets that
underperform by 200 basis points per year could expect to lose 23 percent of the value of her
savings over 30 years of retirement.370
This final Regulatory Impact Analysis differs from the 2015 NPRM Regulatory Impact
Analysis by providing a range (50-100 basis points) for the underperformance of broker-sold
mutual funds rather than a point estimate (100 basis points). The lower end of the range is added
for two reasons. First, as discussed later in this section, data provided subsequent to the
publication of the 2015 NPRM Regulatory Impact Analysis suggest that broker-sold mutual
funds may, at times, underperform on average by less than 100 basis points. Second, many of

368

369
370

Reviewing the literature available around the same time as the 2015 NPRM Regulatory Impact Analysis, the Council of Economic Advisers
(CEA) generated estimates of similar magnitude to those in the 2015 NPRM Regulatory Impact Analysis for the underperformance
associated with conflicted advice. See “The Effects of Conflicted Investment Advice on Retirement Savings,” Feb. 2015, available at:
https://www.whitehouse.gov/sites/default/files/docs/cea_coi_report_final.pdf.
See Appendix B for details on these calculations.
For example, an ERISA plan investor who rolls $200,000 into an IRA earns a 6 percent nominal rate of return with 2.3 percent inflation,
and aims to spend down her savings in 30 years, would be able to consume $11,034 per year for the 30 year period. A similar investor
whose assets underperform by 0.5, 1, or 2 percentage points per year would only be able to consume $10,359, $9,705, or $8,466 per year,
respectively, in each of the 30 years.

162

the studies presented in Figure 3-17 do not back out 12b-1 fees, a practice that often reduces
measured underperformance by about 25 basis points.371 12b-1 fees are often paid from the
mutual fund to intermediaries for the distribution of the funds and can sometimes represent
compensation to a BD for advice services provided to a client. To the extent that 12b-1 fees are
fair compensation to a BD in exchange for advice, it may be appropriate to back out these fees
when estimating underperformance. However, where 12b-1 fees are charged on retail investor
assets for any other purpose, this charge is appropriately reflected in the performance of the
assets. Therefore, any adjustment made to this end should account for some, but not all of the
12b-1 fees paid by retail investors. In the Department’s view, the 50 to 100 basis points range
likely contains the true underperformance of broker-sold mutual funds after adjusting for fair
compensation paid in the form of 12b-1 fees.
While consistent with the broader literature, one particular study distinguishes the effect
of conflicts of interest from other factors that impact performance. CEM looked only at brokersold mutual funds with loads and found that brokers directed clients to funds with unusually high
payments to brokers. The authors also found that these mutual funds which paid high loadshares to brokers subsequently underperformed other broker-sold funds with more moderate
levels of load sharing.372 Every 100 basis points in load sharing paid to an adviser reduced future
returns by about 33 to 35 basis points. In another specification of the model, among funds
distributed exclusively by unaffiliated brokers, every 100 basis points in load sharing paid to an
adviser reduced future returns by about 50 basis points, while among funds distributed
exclusively through captive brokers, the reduction in future returns was 15 basis points.
Projecting these results onto the current IRA marketplace suggests that load fund holders could
lose about $9 billion per year in loads and underperformance as a result of these conflicts of
interest. The accumulation of the loads and underperformance over time could cost load fund
holders $99 billion over 10 years and $212 billion over 20 years. However, while the mutual
fund industry has been trending away from front-end-load mutual funds, it has not been trending
away from conflicts of interest more generally. More and more of the conflicted revenue streams
received by brokers come through channels that are not observed by IRA investors, regulators, or
researchers. If the industry has simply shifted conflicted revenue streams, rather than reducing
conflicts, these conflicts of interest could cost IRA mutual fund holders approximately $25
billion per year, $273 billion over 10 years, and $583 billion over 20 years.
In addition to the academic literature available at the time of the 2015 NPRM Regulatory
Impact Analysis, the Department has considered more recent data provided in a comment letter
submitted by ICI on the 2015 NPRM Regulatory Impact Analysis.373 Similar to many of the
studies presented in Figure 3-17 above, ICI examined the performance of front-end-load mutual
funds relative to the performance of direct-sold no-load funds. Using returns data on a broad
range of mutual funds between 2008 and 2014, ICI found that, on average, broker-sold frontend-load mutual funds underperformed their no-load counterparts by 43 basis points per year.

371

372

373

See ICI comment letter, (July 21, 2016), p. 20: available at: http://www.dol.gov/ebsa/pdf/1210-AB32-2-00749.pdf.]. Adjusting for 12b-1
fees reduces underperformance by 22 basis points (43 basis points – 21 basis points = 22 basis points). See also Reuter (2015), p. 8.
Adding back 12b-1 fees shrinks underperformance by 28 basis points (65 basis points – 37 basis points = 28 basis points and 70 basis points
– 42 basis points = 28 basis points).
ICI affirms this finding using newer data (2010-2013). See ICI’s letter to the Department dated December 1, 2015 at:
http://www.dol.gov/ebsa/pdf/ ici-letter-to-supplement-comment-12-01-2015.pdf .
ICI comment letter, (July 21, 2016); available at: http://www.dol.gov/ebsa/pdf/1210-AB32-2-00749.pdf.

163

ICI’s underperformance measure was a weighted average of the relative performance
experienced by front-end-load mutual funds in the domestic equity, foreign equity, and bond
markets. Based on the previous literature, one might expect that the front-end-load domestic
equity mutual funds in ICI’s sample underperformed by substantially more than 43 basis points
while the front-end-load foreign equity funds in ICI’s sample overperformed (using assetweighted averaging) their no-load counterparts.
The Department appreciates the data provided by the ICI, but rejects ICI’s contention that
the data presented in its comment letter contradict the claims made in the 2015 NPRM
Regulatory Impact Analysis. The Department bases this rejection on the following findings.
First, ICI’s analysis is not a direct measure of the impact of conflicts of interest on
investment performance. Like other evidence discussed in this chapter, ICI compares the
performance of funds sold through brokers to the performance of direct-sold funds. This is an
imperfect measure of the impact of conflicts of interest; other factors, aside from conflicts of
interest, affect the relative performance of mutual funds sold through the two distribution
channels. In contrast to ICI’s analysis, CEM employs a methodology that the Department
interprets as directly identifying the impact of conflicts of interest on performance by examining
the size of loads within the universe of front-end-load funds. CEM finds that as the severity of
the conflict of interest (as measured by the size of the load-share paid by the mutual fund to the
broker) increases, the performance of the mutual fund decreases.374 Even if a cross-distributionchannel analysis were to find that overall front-end-load funds outperform direct-sold funds, the
finding would not necessarily contradict the results from CEM demonstrating that more severe
conflicts cause more harm to investors.
Second, ICI’s treatment of all broker-sold mutual funds in a single analysis likely
obscures meaningful differences in underperformance across different types of broker-sold
funds. The previous literature consistently demonstrates that broker-sold domestic equity mutual
funds underperform their direct-sold counterparts (See Figure 3-17). On the other hand, the
literature presents more mixed findings in the case of foreign equity mutual funds. In fact, as
mentioned in Figure 3-17 both here and in the 2015 NPRM Regulatory Impact Analysis,
Bergstresser et al. (2009) find that when aggregated on an asset-weighted basis, broker-sold
foreign equity mutual funds overperform direct-sold foreign equity mutual funds by a substantial
margin. Like Bergstresser et al. (2009), ICI measures performance on an asset-weighted basis.
Thus, based on the previous literature, it should be expected that the foreign equity broker-sold
funds in ICI’s sample would overperform based on ICI’s chosen metric while the domestic
equity broker-sold funds in the sample would underperform. It is consistent with some previous
literature, that the aggregate underperformance of broker-sold mutual funds within ICI’s sample
is smaller than previously published estimates of domestic equity broker-sold mutual fund
underperformance.
Third, data included in ICI’s comment letter and in a report from NERA suggest that
ICI’s time-horizon for estimating the underperformance of broker-sold mutual funds likely
renders their estimates far less reliable than if they had chosen a longer time-horizon. ICI claims

374

This result is replicated by ICI, as presented in a letter from ICI to the Department dated December 1, 2015 at: http://www.dol.gov/ebsa/pdf/
ici-letter-to-supplement-comment-12-01-2015.pdf. For more discussion of the CEM result and its interpretation by the Department, see
Padmanabhan, Panis and Tardiff (2016).

164

that the academic studies relied on in the 2015 NPRM Regulatory Impact Analysis are “out of
date” and implies that these studies should not be relied upon by the Department. Just as mutual
fund returns are highly volatile, ICI and NERA data demonstrate that the underperformance of
broker-sold mutual funds is highly volatile as well. There are likely other variables that explain
some of the variability in mutual fund performance. In the absence of a single, long-term
measure of broker-sold mutual fund underperformance, the Department must consider evidence
from multiple studies, which, in aggregate, span a long time horizon. This necessitates that the
Department continue to rely on the data from the academic studies presented in Figure 3-17 in
combination with the new data provided by ICI.
ICI contends that the market for mutual funds has recently undergone a “fundamental”
transformation and chose to examine broker-sold mutual fund underperformance from the
beginning of the financial crisis to 2014. ICI cites a change in the percentage of mutual funds
with a front-end-load-share class that also have a no-load-share class as evidence of this
“fundamental” transformation. Other commenters disagree. The Consumer Federation of
America explains: “by focusing on the number of share classes instead of the amount of money
in the various share classes, ICI creates a misleading impression that the market is less
segmented than it actually is. According to ICI’s own data, the amount of money in front-endload-share classes has grown from approximately $1.7 trillion in 2005 to approximately $2.1
trillion in 2014.”375 Furthermore, dynamics of the market for financial advice do not seem to
have changed in recent years.
Numerous commenters and witnesses at DOL’s public hearing confirmed that conflicts
of interests continue to harm retirement investors. For example, Antoinette Schoar discussed
new research showing that conflicted broker-dealers “reinforce erroneous beliefs about the
market” and “guide people towards high-fee funds.”376 Mercer Bullard provided examples of
compensation models that incentivize “financial advisers to give investors bad advice.”377
Jonathan Reuter finds that “conflicted advice is readily observed in real-world data and, in the
settings that we study, associated with significantly lower after-fee, risk-adjusted returns.”378
Broker-sold mutual funds compete by incentivizing brokers to sell their products. It would be
difficult to explain a fundamental change in the nature of competition in the mutual fund market
without a corresponding fundamental change in the dynamics of the market for financial advice.
Finally, the ICI comment letter makes several comparisons that provide no relevant
information. ICI makes several statements in its comment letter and elsewhere379 in which they
compare a performance metric using a simple-weighted (also known as equal-weighted) average
to a metric that uses an asset-weighted or sales-weighted average. ICI finds that the average
performance of broker-sold front-end-load funds is greater when calculated using an asset- or

375

376

377
378
379

See Consumer Federation of America comment letter, (Sept. 24, 2015); available at: http://www.dol.gov/ebsa/pdf/1210-AB32-2-03031.pdf.
Also see Susan Christoffersen and Richard Evans’ comment letter, (Sept. 10, 2015); available at:
http://www.dol.gov/ebsa/pdf/1210-AB32-2-02766.pdf.
Testimony of Antoinette Schoar from August 11, 2015, Department of Labor. Pages 376-385 of hearing transcript; available at:
http://www.dol.gov/ebsa/pdf/1210-AB32-2-HearingTranscript2.
See Mercer Bullard’s written testimony, (August 13); available at: http://www.dol.gov/ebsa/pdf/1210-AB32-2-WrittenTestimony28.pdf
See Jonathan Reuter’s written testimony (August 11); available at: http://www.dol.gov/ebsa/pdf/1210-AB32-2-WrittenTestimony10.pdf.
See ICI comment letter, (July 21, 2015); available at: http://www.dol.gov/ebsa/pdf/1210-AB32-2-00749.pdf; and ICI’s June 17, 2015
testimony provided to the House Committee on Education and the Workforce, Subcommittee on Health, Employment, Labor and Pensions;
available at: http://edworkforce.house.gov/uploadedfiles/testimony_reid.pdf .

165

sales-weighted average than when using a simple-weighted average. The Department agrees that
this result suggests that front-end-load fund investors tend to concentrate their assets in funds
that perform. However, ICI then goes on to conclude that the result constitutes evidence that
front-end-load funds do not “underperform.” The 2015 NPRM Regulatory Impact Analysis
presents evidence that broker-sold mutual funds underperform direct-sold mutual funds, on
average. ICI’s findings regarding the relative values of simple-weighted and asset- or salesweighted averages do not contradict the 2015 NPRM Regulatory Impact Analysis.
Following the initial comment period for the 2015 NPRM, the Department sought
assistance from an outside consultant, Advanced Analytical Consulting Group (AACG), to help
review the ICI comment letter and other comments related to the Regulatory Impact Analysis for
the 2015 NPRM.380 AACG’s analysis includes the ICI’s initial comment letter regarding the
2015 NPRM Regulatory Impact Analysis, submitted in July 2015, and two follow-up letters,
submitted in September 2015 and December 2015. The Department agrees with AACG’s
conclusions that “(1) ICI’s criticisms of the academic literature and front-end-load performance
results do not undermine DOL’s estimates of the benefits from reducing conflicted advice and
(2) ICI’s estimates of the costs to investors of having to pay more for and/or losing financial
advice are based on unsupported assumptions that are contradicted by information provided by
other commenters.”
As a supplement to the Department’s and AACG’s evaluations of the ICI letters
described above, the Department conducted its own analysis of broker-sold mutual fund
performance using data obtained from Morningstar. The analysis confirms the Department’s
finding that, contrary to ICI’s claim, the nature of competition in the mutual fund market has not
fundamentally changed in recent years. The analysis also confirms the Department’s
characterization of the academic literature presented in the 2015 NPRM Regulatory Impact
Analysis. Specifically, the Department finds that broker-sold domestic equity mutual funds
underperformed direct-sold domestic equity mutual funds, on average, by 59 to 85 basis points
per year over the entire sample period 1980-2015, and underperform by 81 to 101 basis points
over the recent period, 2008-2015. Details of the Department’s analysis of broker-sold mutual
fund performance using data obtained from Morningstar can be found in Appendix A of this
document.
Also updating information in the record is a working paper by Jonathan Reuter which
was released in November 2015.381 Using data from 2003 through 2012, Reuter finds that
actively-managed broker-sold domestic-equity funds underperform index funds by 64 basis
points per year. This result is smaller in magnitude, but consistent with the previous literature
showing underperformance in broker-sold domestic equity mutual funds.

3.2.5 Conclusion
This section has considered whether there is evidence of a need for regulatory action to
reduce or mitigate conflicts of interest in advice on the investment of IRA assets. The foregoing
analysis has established that IRAs have special importance and that IRA investors are vulnerable

380
381

See AACG’s response to the industry comment letters (Padmanabhan, Panis and Tardiff 2016).
See Reuter (2015).

166

to abuse. Changes in the market have overtaken the 1975 regulation, rendering obsolete its
omission of much advice from coverage under ERISA and/or IRC fiduciary standards. The IRA
advice market displays the characteristics economic theory indicates predict harm from adviser
conflicts: serious conflicts are widespread, pursuing customers is costly, and IRA investors are
poorly equipped to police advice. There is evidence that advice is biased, and that this bias hurts
IRA investors. Losses to IRA investors from conflicted advice are expected to amount to tens or
hundreds of billions of dollars over the next 10 years. As discussed further below, the final
regulatory package published today with this Regulatory Impact Analysis addresses conflicts
and combats the bias in order to protect IRA investors.

3.3

Gains to Investors

The Department expects the final rule and exemptions to deliver large gains for IRA
investors by reducing, over time, the losses identified above. Because of data limitations, as
with the losses themselves, only a portion of the expected gains are quantified in this analysis.
In particular, the Department has quantified estimated gains associated with only one subset of
the market - front-end-load mutual funds - and only with respect to one category of conflicts
associated with those funds. Even so, the Department estimates that the gains to IRA front-endload mutual fund investors from improved returns alone have the potential to be worth between
$33 billion and $36 billion over 10 years and between $66 and $76 billion over 20 years. These
quantified gains do not include additional large, expected gains to IRA investors resulting from
reducing or eliminating other effects of conflicts in IRA advice.
Under the new rule, advisers to IRA investors generally must either avoid compensation
arrangements that involve conflicts, or contractually bind themselves to act in their customers’
best interests. The new rule and exemptions are likely to affect investors’ investment choices,
savings decisions, and use of advisory services. By limiting or mitigating IRA advisers’
conflicts, the new rule and exemptions are intended to ensure that their advice is impartial and
thereby reduce the IRA performance gap otherwise attributable to conflicted advice. The
Department expects the financial gain to IRA investors to amount to tens of billions of dollars or
more over the next 10 years, even if one only considers the rule’s impact on front-end-load
mutual funds.
By holding advisers to a fiduciary standard of care, the final rule and exemptions may
also deliver additional investor gains beyond those associated with mitigation of adviser
conflicts. Independent research demonstrates that the regulatory regime under which an
investment adviser acts and the standards of conduct to which an adviser must adhere affect the
advice given. Kozora (2013) isolates the impact of a 2007 SEC rule, and observes that fiduciary
status increases the sale of investment grade municipal bonds. While the study does not directly
observe whether this increase benefits investors, it is reasonable to presume that
recommendations made pursuant to a higher standard of care on average will be more favorable
to investors. For example, in Kozora (2013), if the clients would otherwise have purchased noninvestment grade bonds with risks higher than would be optimal for them, or with yields that do
not make up for their added risk, an adviser’s fiduciary status would improve investment
recommendations and client decisions.

167

The misaligned incentive system – conflicts of interest - in the annuity market can result
in unsuitable sales of annuity products to investors. The conflicts of interests in the annuity
market can be more pronounced than the mutual fund market because commissions in the
annuity market, ranging from 6 to 9 percent of premiums, are generally higher than commissions
earned in connection with the recommendation of mutual funds.382 Previous research suggested
that this incentivizes insurance agents to steer consumers toward insurance products with higher
commissions. In the annuity market, this may have led consumers to purchase annuities that
were not in their best interest. This rule will ensure the best interest standard for the purchasers
of annuity products and reduce inappropriate sales. Due to data limitations, the gains to
purchasers of annuities cannot be easily quantified. However, this rule is expected to create
benefits in the annuity market by enhancing efficiencies through better matches between
consumers and the annuity product.
The Department’s focus in the retirement space, and a core part of the mission of the
Employee Benefits Security Administration (EBSA), is to assure the security of retirement
benefits, including those derived from account-based plans: defined contribution (DC) plans and
IRAs. Excessive fees and substandard investment performance in DC plans or IRAs, which can
result when advisers’ conflicts bias their advice, erode benefit security. The aim of the rule and
exemptions is to ensure that advice is impartial, thereby rooting out excessive fees and
substandard performance otherwise attributable to advisers’ conflicts, producing gains for
retirement investors. Delivering these gains will entail some compliance cost – namely, the cost
incurred by new fiduciary advisers to avoid prohibited transactions and/or satisfy relevant PTE
conditions. The Department expects investor gains to be very large relative to compliance costs,
and therefore believes this proposal is economically justified and sound.
In the language of social welfare economics and reflected in Circular A-4, the investor
gains which are the aim of this proposal generally can be said to comprise two parts: pure social
welfare “benefits” attributable to improvements in economic efficiency, and “transfers” of
welfare to retirement investors from the financial industry. A full accounting of a rule’s social
welfare effects would encompass all of the rule’s direct and indirect effects as would be manifest
in general market equilibrium. Likewise, that full accounting would consider pure social welfare
costs – that is, reductions in economic efficiency – which may not be the same as simple
compliance costs. The Department considered this proposal’s potential indirect effects and
associated social welfare implications, particularly its potential supply and demand side impacts
on the market for advisory services, and for financial products and services more generally.
The quantitative focus of this analysis, however, is on the proposal’s most direct, and
directly targeted, effects: gains to retirement investors, and compliance costs to advisers and
others. The available data do not allow the Department to fully quantify the expected gains to
investors nor break down those gains into component social welfare “benefits” and “transfers.”
Therefore, the Department has quantified a subset of the potential gains to IRA investors, which

382

According to “American Council of Life Insurers: Life Insurers Fact Book 2014,” US life insurers’ aggregate commission payments
amounted to 9 percent of total premiums in 2013. Wink’s Sales and Market Report 4th Quarter, 2014 reported that commissions of fixedindexed annuities were on average 6.1 percent of total premium paid in the 4th quarter in 2014.

168

include both some pure economic efficiency gains (benefits)383 and some transfers from the
financial industry.384

3.3.1 Quantified Gains to Investors
Several comments on the 2015 NPRM Regulatory Impact Analysis criticized the
Regulatory Impact Analysis for failing to provide an analysis of the performance of commissionbased accounts relative to that of fee-based accounts using account-level data.
In the 2015 NPRM Regulatory Impact Analysis, the Department recognized the lack of
comprehensive data on the potential harmful effects of conflicts of interest in the ERISA plan
and IRA marketplaces. Following a December 2011 request from the Department, industry
sources indicated that the data necessary to fully address this question would be prohibitively
expensive to compile or obtain.385 Within the 2015 NPRM Regulatory Impact Analysis, the
Department again requested that interested parties provide the data or analysis that might shed
additional light on the harmful effects of conflicts of interest in the ERISA plan and IRA
marketplaces.
The Department received no comments on the 2015 NPRM Regulatory Impact Analysis
providing raw account-level data that could be used to assess the harmful effects of conflicts of
interest in the ERISA plan and IRA marketplaces and only one comment including such an
analysis based on account level data.386 The Department found this commenter’s analysis to be

383

384

385

386

Impartial advice leads to better investment decisions. Better decisions in turn free some resources from sub-optimal uses in the financial
sector (which may include excessive trading, and duplicative or sub-standard research or asset management) to other, more productive uses.
They also lead to more optimal deployment of capital in financial markets (toward more productive enterprises), increasing overall returns
to capital investment.
Transfers to retirement investors will largely consist of reduced levels and/or volumes of fees paid for financial products or services. Other
transfers to retirement investors will derive from improved trading decisions relative to counterparties from whom the transfers will come.
The Department believes that transfers to IRA investors in conflicted advice arrangements are likely to come mostly from professional asset
managers and others in the financial industry rather than from other retail investors (non-IRA retail investors and IRA investors that are not
in conflicted advice arrangements). One possible impact of closer-to-optimal investment of the assets of conflicted-advice-receiving IRA
investors is to lessen the opportunity for counterparties to take advantage of abnormally positive investment opportunities. However, to the
extent that those counterparties are retail investors or mutual funds held by retail investors, those retail investors typically do not profit from
such investment opportunities. Instead, the academic literature (e.g., Del Guercio and Reuter 2014) finds the profit is captured by the retail
investors’ advisers or their funds’ asset managers. Retail investors only break even relative to index investment. Therefore, retail investors
likely would not experience much of a loss if such abnormally positive investment opportunities dried up. Instead, this transfer would come
mostly from reduced opportunities for skilled asset managers. In the 2015 NPRM Regulatory Impact Analysis, the Department invited
commenters to provide evidence that would confirm or refute this conclusion and more generally allow for characterization of rule-induced
transfers. The Department received no comments on this issue.
On December 15, 2011, the Department sent a letter to six trade groups requesting that they voluntarily provide data that could help the
Department evaluate the impact, if any, of conflicts of interest faced by brokers or others who advise IRA investors. The Department met
with the groups to clarify the data request, and sent a follow-up letter on February 10, 2013, requesting the trade groups to notify the
Department regarding whether they would be able to provide any of the requested data elements, and if so, when the Department could
expect to receive such data. If the groups were not able to provide any of these requested data, the Department asked them to inform the
Department of any other available data sources that could help the Department to evaluate the impacts of conflicts of interest.
None of the groups provided any relevant data in response to these requests. In general, their responses indicated that these data were not
available and would be prohibitively expensive to collect or compile. Moreover, the groups asserted that even if such data were made
available, these data would not allow the Department to determine whether conflicts bias advice and harm investors. Generally, they also
did not provide the Department with any other data sources that could be used to assess the impact, if any, of conflicts of interests on IRA
investors. The Department remains interested in receiving these data and hereby requests the industry or any other groups that have such
data to provide it.
The Financial Service Institute sent its Broker-Dealer Financial Performance Study to the Department for several years. The Department
used data from the study where relevant and appreciates receiving it. The study focuses on the operation and management of independent
broker-dealer firms. However, the reports do not contain individual account data for IRAs or other products that could be used to access the
impact, if any, of conflicts of interest on IRA investors.
See NERA comment letter, (Sept. 24. 2015); available at: http://www.dol.gov/ebsa/pdf/1210-AB32-2-03079.pdf.

169

unpersuasive. Some of the Department’s main concerns are that the account data spanned too
short of a time horizon, the investment returns were not risk-adjusted (the commission-based
accounts appeared to be owned by older individuals and therefore likely took on less risk), the
data do not appear to be representative of IRA’s or retail accounts more generally, and the
analysis lacked a strategy to distinguish the impact of conflicts and compensation on
performance from other reasons the accounts would differ.387
In the absence of comprehensive data, research, and analysis that would allow for an
estimate of the harmful effects on investors of conflicts of interest in the ERISA plan and IRA
marketplaces, the Department has assessed the harm to investors by specifically quantifying
harms in one area of the IRA market, namely front-end-load mutual funds, where the conflicts
are well measured. This segment of IRA assets currently amounts to approximately 13 percent of
overall IRA assets. Narrowly focusing only on how load-shares paid to brokers affect the size of
loads IRA investors holding load funds pay and the returns they achieve, the underperformance
attributable to these practices alone amounts to $67 billion over 10 years and $142 billion over
20 years.388 These estimates of potential harm from conflicts of interest in the absence of
regulatory intervention are largely consistent with the estimates presented in Section 3.2.4
reflecting a large body of academic literature, comments on the 2015 NPRM Regulatory Impact
Analysis, and testimony at the DOL hearing in August 2015.
The final rule and exemptions have the potential to go a long way to reducing these and
other losses by requiring fiduciary IRA advisers to forgo conflicted fee structures when
providing fiduciary advice to IRA investors or to provide advice that is in their clients’ best
interest and impartial as a condition of relying on certain PTEs that provide flexibility to
continue a wide range of compensation practices subject to protective conditions. Even taking
into account the gradual movement of IRA assets into more optimal investments, and backing
out improvements in cost-effectiveness that might be expected without the new rule and
exemptions, the Department estimates that the new rule and exemptions have the potential to
restore to IRA investors approximately $33 billion to $36 billion over 10 years and $66 billion to
$76 billion over 20 years even in spite of existing regulations protecting investors. These
quantitative estimates are calculated with an assumption that the rule will eliminate (rather than
just reduce) underperformance associated with the practice of incentivizing broker
recommendations through front-end-load sharing; if the rule’s effectiveness in this area is
substantially below 100 percent, these results may correspondingly overstate rule-induced gains
to investors from mitigation of conflicts in the front-end-load mutual fund segment of the IRA
market.389 The rule’s effectiveness, especially in the short-term, may be below 100 percent
because several provisions of the Best Interest Contract Exemption will not go into effect until

387
388
389

For a more detailed analysis of the NERA comment, see Padmanabhan, Panis and Tardiff (2016).
See Appendix B, Section B.4, for details on these calculations.
The gains estimates reflect the Department’s assumption that, as a result of the final rule and exemptions, advisers’ recommendations will
not be influenced by variation in the share of mutual fund front-end-loads paid to them. If, to the contrary, the influence of variation in
load sharing on advisers’ recommendations were to be reduced by one-half rather than eliminated, then perhaps only one half of the
estimated gain would be realized. These estimates do not take into account any additional gains attributable raising standards for advisers,
independent of their impact on conflicts of interest, however. Nor do they consider the impact of the rule and exemptions in other segments
of the market or on conflicts of interest other than those directly tied to variations in load-shares.

170

approximately nine months after the initial applicability date.390 If the rule and exemptions are
only 50 percent effective in the first year following the initial applicability date (which includes
the approximately nine-month transition period when some Best Interest Contract Exemption
provisions are not yet in effect), the quantified subset of gains – specific to the front-load mutual
fund segment of the IRA market – would amount to between $30 billion and $33 billion over 10
years. These gain estimates exclude additional potential gains to investors resulting from
reducing or eliminating the effects of conflicts in financial products other than front-end-load
mutual funds. These potential additional gains include improvements in the performance of IRA
investments other than front-end-load mutual funds and potential reductions in excessive
trading391 and associated transaction costs, and timing errors (such as might be associated with
return chasing), as well as additional potential gains attributable to the application of fiduciary
standard of care.392
The quantified gains for investors in IRAs currently characterized by conflicts of interest
are composed of benefits to society as a whole and transfers of value between members of
society. Fund underperformance, the reduction of which is key to the quantification about to be
described, can result from excessive transaction costs (associated with labor and other resources
being used for excessive trading within a fund), or from active asset management costs in excess
of resulting excess returns,393 and the freeing of these resources for other uses would be a benefit
of this proposed rule. Fund underperformance may also represent sub-optimal allocation of
financial capital in the national economy; improvement in this area would represent another
investor gain (and social welfare benefit) of the rule. Further social welfare benefits will accrue
as the rule is expected to deter firms from engaging in the costly and inefficient pursuit of
customers.394 On the other hand, fund underperformance can also result from assets being
purchased at a relatively high price and sold at a relatively low price; the effect of the rule for the
fund’s counterparties in these transactions would be a reduction in the abnormally good returns
they currently experience, an effect that would be categorized, from the perspective of society as
a whole, as a transfer. This chapter’s approach to quantification produces estimates of investor
gains that do not distinguish between social welfare benefits and transfers to investors from
financial firms, and thus will be referred to with the term “gains to investors.”
The Department began its effort to quantify gains to investors by conducting a thorough
review of the economic literature that investigates the relationship between retail investment

390

391

392

393

394

Gains-to-investors begin accruing once new advice is given following the initial applicability date - scheduled for April 2017, or one year
following the finalization of the rule. For a discussion of the rule and exemptions’ effective date, initial applicability date, transition period,
and full applicability date, see Sections 2.9.1 and 2.9.2.
Excessive trading generally means trading at a level above that which generates optimal expected returns for the investor. Excessive trading
sometimes takes the form of “churning,” or repeated recommendations to trade that are intended to generate more commissions for advisers
rather than benefit investors. Such churning generally violates securities law. However, excessive trading can take other, generally smaller,
forms that may not run afoul of securities law. For example, an adviser might recommend that an IRA investor construct a diversified
portfolio by buying several mutual funds (and periodically trading to rebalance the portfolio) in circumstances where the same
diversification and expected return could be achieved with less transaction cost by buying a single, internally-diversified fund that offers
ongoing, internal rebalancing.
The requirements that attach with a broker-dealer’s suitability obligation vary based on the facts and circumstances. Activities such as
excessive trading and churning have been found to violate the suitability obligation in some circumstances, but not in others. See SEC Staff
Dodd-Frank Study 2011, 65.
As noted in Chapter 2, reform in the United Kingdom to eliminate adviser conflicts accelerated sales of lower cost funds, including funds
that track market indices. Some or all of the associated cost savings is likely to represent increases in social welfare. The final rule and
exemptions are expected to produce some similar social welfare gains.
See Section 3.2.3.2 for a discussion of the costly pursuit of customers.

171

advice and retail investment performance.395 Across a broad array of data samples and
methodologies, these papers consistently show that assets in retail accounts that are invested
directly perform better than assets invested based on advice from a broker who gets paid
(directly or indirectly) by the mutual fund. One study in particular, discussed immediately below
isolates the effect of front-end-load-shares paid to brokers recommending mutual funds on fund
performance, distinguishing the effect of conflicts of interest from other factors that impact
performance. This research provides a basis for estimating the potential benefit to investors of
mitigating this particular type of conflict in this particular subset of the IRA market.396
Christoffersen, Evans, and Musto (2013) (CEM) find a significant relationship
(statistically significant and economically meaningful) between the amount of sales load that the
mutual fund shares with the broker (“load-share”) and the inflows into the mutual fund.397 In
other words, the more the mutual fund pays the broker, the more likely the broker is to
recommend that mutual fund to clients. This result is an immediate concern regarding conflicted
advice – that the investment recommendations are influenced by conflicted payments, rather than
strictly being constructed in the best interest of the IRA investor. The load-share itself is harmful
because the incentive provided to the broker tilts investment recommendations toward higherload funds and helps to keep loads higher than what they otherwise would be. However,
according to the Department’s estimates, higher-than-otherwise loads are likely to constitute only
a small fraction of the total losses to IRA investors due to conflicted investment advice.
Investment underperformance also appears to be a result of conflicted investment advice
as the authors find a significant, negative relationship between load-shares and mutual fund
investment performance. Importantly, CEM find that conflicts of interest skew brokers’
recommendations and that, as a result, such investments underperform, even in spite of existing
regulations intended to protect investors against conflicts of interest.
Several comments on the 2015 NPRM Regulatory Impact Analysis disagree with the
Department’s interpretation of the CEM result linking conflicts of interest to lowered investment
performance. These comments contend that the underperformance observed by CEM applies
only to funds with above average front-end-loads.398 The Department’s original interpretation of
this CEM result – suggesting that, at any level of front-end-loads, an increase in load-shares paid
to the broker correlates with a decrease in future performance – aligns with the interpretation of

395
396

397

398

The Department’s efforts included an internal review by staff and an external review by the RAND Corporation. See Burke et al. (2015).
As discussed above, due to limitations in data availability, the Department has been able to quantify only a portion of the gains that are
expected to accrue to IRA investors under this proposal. Specifically, the Department quantified only those gains expected to accrue to IRA
investments in front-end-load mutual funds due to the remediation of only one type of advisers’ bias, namely, bias heretofore attributable to
variations in load sharing (a load is a fee paid to the mutual fund company by the investor in order to invest in (certain share classes of) the
mutual fund)). Substantial additional gains are expected to accrue to IRA investors with respect to investments other than front-end-load
mutual funds and with respect to the remediation of adviser bias heretofore attributable to conflict-laden compensation other than loadshares. However, data limitations prevented the Department from confidently quantifying these gains so they are omitted from the
estimates presented here. In addition, requiring advisers to adhere to a higher standard of conduct can be expected to result in improved
investment outcomes independent of the impact of conflicts of interest, but those potential impacts are not quantified.
The authors find a similar result with regard to 12(b)-1 fees; larger 12(b)-1 fees are associated with higher levels of flows into a fund. The
result is statistically significant when load-sharing is not included in the regression. When load-sharing is added into the regression, the
12(b)-1 fee result is no longer statistically significant, but the magnitude of the coefficient is still economically meaningful.
E.g. see ICI comment letter (July 21, 2015); available at: http://www.dol.gov/ebsa/pdf/1210-AB32-2-00749.pdf; and NERA comment letter
(July 20, 2015); available at: http://www.sifma.org/issues/item.aspx?id=8589955443.

172

the CEM authors. The CEM authors submitted a comment letter responding to an ICI comment
addressing the interpretation of their results: 399
“…we are economically trying to evaluate how a change in the load paid to the broker
would relate to changes in the future performance of the fund. We chose 2.3 percent to
evaluate this effect economically since this represents the change of moving from no
payment to the broker to average payment to the broker (2.3 percent in our sample). If we
multiply this change in broker fees (2.3 percent) by the coefficient (-0.49), we get the
expected effect on changes in the future performance of the fund (-1.13 percent). [ICI’s]
proposed use of scaling by zero is economically meaningless and implies the obvious: if
brokers are not given any additional compensation to sell a fund, then there are no
additional incentive problems as to which fund the broker sells an investor. The negative
and significant coefficient relating excess loads to future performance implies that any
positive changes in load payments to the broker (additional compensation for the broker)
associate with decreased expected future performance realized by the investor over the
following year.”
The Department’s application of the CEM result mirrors the result itself. CEM estimates
the relationship between front-end-load-shares and performance. The fact that CEM estimates
“excess” load-shares in a first-stage regression does not limit the applicability of the result. The
word “excess” may be misleading; these excess load-shares can be positive or negative. (One
might alternatively think of the excess load-share as a “relative” load-share.) Excess load-shares
are an estimate of the difference between the actual load-share of a particular mutual fund and
the mutual fund’s expected load-share based on the sample of front-end-load mutual funds and
the characteristics of the particular fund. CEM’s two-stage regression methodology retains the
full variation in load-shares, save for the variation explained by covariates in either the first or
second stage.400 Therefore, CEM’s second stage coefficient on the excess load-share variable is
a measure of the relationship between front-end-load-shares and performance. This estimated
relationship applies to all load-shares, not just those in excess of the average.
One possible explanation for the link between load-sharing and underperformance is that
a mutual fund company faces a tradeoff between incentivizing its brokers to recommend a
product (e.g., by offering higher compensation for selling that product) and investing sufficient
resources in fund management to produce performance that covers fund expenses. Del Guercio
and Reuter (2014) provide an empirical investigation of this relationship and demonstrate that
funds providing broker compensation appear to invest less in fund management.
The subset of expected gains to investors that the Department has quantified was
estimated by comparing projections of front-end-load mutual fund IRA assets under different
scenarios: (1) a baseline scenario where the 1975 regulation remains in place, and (2) a series of
alternative reform scenarios where the new rule and exemptions are finalized and affect the

399

400

See Christoffersen and Evans comment letter (Sept. 10, 2015); available at: http://www.dol.gov/ebsa/pdf/1210-AB32-2-02766.pdf. Also
see AACG’s response to the industry comment letters (Padmanabhan, Panis and Tardiff 2016).
The coefficient on excess load-shares produced by CEM’s two-stage approach would, under certain conditions, be very similar to the
coefficient on total load-shares that would be produced by a one-stage approach consisting of a single regression including all of the first
and second stage variables, and potentially some interactions between them, as controls (adjusted for collinearity and related practical
considerations in the specification). The conditions that would determine the coefficients’ similarity include the relationships between the
control variables used in the first and second stages and sample differences between the first and second stages.

173

market in particular ways. The scenarios do not differ with respect to contribution rates or
rollover rates into IRAs or withdrawal rates from IRAs. (The question of whether or how
contributions, rollovers, and withdrawals might be affected by the rule is discussed in Section 8.3
below.) The scenarios also do not differ with respect to the proportion of IRA assets invested in
mutual funds. Finally, the scenarios do not differ by the percentage of mutual fund IRA assets
that incur front-end-loads.
The scenarios differ only in the net rate of return experienced by IRA investors in frontend-load mutual funds. The net rate of return is higher in scenarios where the rule is
promulgated for two reasons. First, under some of the reform scenarios, loads decrease at a faster
rate than under the baseline because the incentive for a broker to place IRA clients in mutual
funds with a higher load-share is effectively mitigated. The second and larger reason is
increased investment performance under the reform scenarios.
Comments on the 2015 NPRM Regulatory Impact Analysis point to decreasing fees and
decreasing assets subject to front-end-loads as indicators that the market for financial advice is
changing. The Department appreciates these trends. The partial gains-to-investor estimates
presented in the 2015 NPRM Regulatory Impact Analysis incorporated both the decreasing sizeof-loads and decreasing assets-subject-to-loads trends. The current analysis also incorporates
these trends.
The baseline scenario includes a projection of loads that decreases over time, at a rate
similar to that of recent years. Under the first reform scenario, loads decrease over time at the
same rate as the baseline scenario. Under the second reform scenario, loads decrease over time
at twice the rate of the alternative scenario. This is where removing the brokers’ adverse
incentive comes into play.
Historical data show that loads have fallen over time, likely due to growing demand for
less-expensive investment options and supply-side factors such as technological improvements.
However, a counterforce has kept loads from falling faster. CEM show that investment assets
tend to flow away from funds with high loads, but toward funds with high load-shares. These
pressures can act independently only when the load-share is a small fraction of the load.
However, the data indicate that this is not the case: the average load-share is 81 percent of the
average load. Because load-shares (that is, the portion of the load paid to the distributing BD), in
general, are a very high fraction of the load, the load-share cannot significantly increase without
increasing the load. Likewise, the load cannot decrease without decreasing the amount of the
load that is shared. The brokers’ incentive to collect a high load-share tends to push loads up
while demand for less expensive investment options pushes loads down. Recent history has
shown that the downward force has outweighed the upward pressure. But removing the upward
pressure, by mitigating the brokers’ incentive, should cause loads to fall faster than they
otherwise would.
Under the third reform scenario, loads paid by investors immediately fall to zero.
Because the Department does not believe loads will disappear, this scenario represents an upper
limit to the direct gains-to-investors that could accrue from the rule’s impact on front-endmutual-fund-loads paid. It is also illustrative of the amount that IRA investors can expect to pay
in loads over the next 10 and 20 years if no action is taken by the Department.
While this decrease in load size is important, the second and larger reason that the net
rate of return is higher under the reform scenarios is increased investment performance for
broker-sold mutual funds. By requiring advisers to act in the best interest of their IRA clients
and including safeguards to that effect, the new rule and exemptions will ensure that adviser
recommendations are not biased by load-sharing or other variable compensation. The estimates
174

discussed below assume that the proposal’s exemption conditions will be fully effective in
mitigating adviser conflicts of interest. Rather than reflecting the adviser’s interests, future
recommendations will be based on factors that appropriately reflect investors’ interests with
respect to risk and expected future returns of investment alternatives. This, in turn, will direct
flows toward mutual funds that invest in performance, rather than relying on payments to
advisers, to sell their product. The estimated magnitude of this effect on investment performance
is the same across all three alternative scenarios.
The new rule and exemptions’ expected positive impact on broker-sold mutual fund
performance constitutes a large majority of its partial, quantified gains to investors. Under the
first reform scenario, there is no direct effect of the rule on loads, so the effect on investment
performance is the entirety of the quantified investor gain. Under the second reform scenario,
the effect on investment performance constitutes approximately 90 percent of the estimated gain.
The Department anticipates that the rule will immediately improve broker-provided
investment advice and will consequently improve investment performance for IRA investors
who seek new advice; however, the 10- and 20-year estimates of the portion of IRA investors’
expected gains that the Department has quantified reflect the likelihood that the gains to
investors will be back-loaded. When the rule becomes applicable, brokers will be required to
adhere to new standards for any advice given from that point forward. As a result, some IRA
investors will remain in underperforming, broker-sold funds after the initial applicability date.
The farther removed from the initial applicability date of the rule, the greater the number of IRA
investors who will have received new advice, and the larger the benefits that will accrue from
improved investment performance.
Figure 3-18 presents the estimated, quantified
Figure 3-18 Front-load-mutual-fund-gainsportion of expected gains to investors by time
horizon (10- and 20-year) and by alternative scenario to-investors Estimates by Time Horizon
and Alternative Scenario ($ billion)
(1, 2, and 3). Alternative scenario 1 generates 10year estimated partial gains of $33 billion and 20Alternative scenario
year estimated partial gains of $66 billion.401 The
Time horizon
1
2
3
Department considers this to be in many respects a
10-year
32.5 35.9
53.8
conservative estimate of the quantifiable portion of
IRA investor gains from the proposal. Alternative
20-year
66.4 75.5
102.9
scenario 2, which includes an acceleration of the
decline in loads, generates a 10-year gain of $36 billion and a 20-year gain of $76 billion.402 The
Department considers this to be a reasonable high estimate of the quantifiable portion.
Alternative scenario 3 projects a future where brokers no longer advise IRA investors to invest in
front-end-load funds. In this case, the quantified subset of IRA investors’ expected gains is
estimated to be $54 billion over 10 years and $103 billion over 20 years. The Department offers

401

402

The Alternative Scenario 1 estimates can be thought of as the sum of counterparties’ reduced performance (transfer), asset allocation that is
closer to optimal for society (benefit), and—because CEM measure underperformance net of most expenses—net real cost savings (i.e.,
resources previously used for management, or for excessive within-fund trading, being freed for other uses, minus resources newly used for
seeking out higher returns).
Front-end-loads have been declining over time, both in magnitude and in the frequency with which mutual fund assets are subject to a load.
All of the Department’s alternative scenarios are presented relative to a baseline scenario where front-end-loads continue to decline in both
magnitude and frequency. Alternative scenario 2 presents the case where the provision of fiduciary advice causes loads to decline faster
than in the baseline scenario. See Appendix B for more details on these calculations.

175

this as an illustration of possible gains if loads decline even faster than in alternative scenario 2,
but does not expect the proposal to immediately result in the elimination of all load funds from
IRA recommendations.
This quantified subset of IRA investors’ expected gains, as presented in Figure 3-18, is
expected to be achieved by requiring broker-dealers who engage in self-dealing to adhere to
conditions designed to mitigate the impact of those conflicts of interest. While brokers will still
be allowed to receive 12b-1 payments and other forms of revenue sharing, the exemption
conditions will ensure that brokers are not incentivized to recommend products that provide the
broker more revenue at the expense of the customer.403 Currently, mutual fund companies are
able to sell their products by enticing brokers to give recommendations that benefit the broker
but are not in the best interests of the clients. The change in broker incentives will, in turn, force
mutual fund companies to invest more in performance in order to sell their products.404 The
increased performance has the potential to result in gains for front-end-load mutual fund
investors of $33 billion over 10 years and $66 billion over 20 years (under scenario 1).
The quantified subset of IRA investors’ expected gains presented in Figure 3-18 assumes
that the proposal’s exemption conditions will be fully effective in mitigating adviser conflicts of
interest. The rule’s effectiveness, especially in the short-term, may be below 100 percent,
however, because some provisions of Best Interest Contract Exemption will not go into effect
until approximately nine months after the initial applicability date.405 If the rule and exemptions
are only 50% effective in the first year following the initial applicability date (which includes the
approximately nine-month transition period when some Best Interest Contract Exemption
provisions are not yet in effect), the quantified subset of gains – specific to the front-load mutual
fund segment of the IRA market – would amount to between $30 billion and $33 billion over 10
years. In addition, the Department understands that: (a) the industry may not achieve full
compliance with the rule and exemption conditions, and (b) the combined rule and exemption
conditions may not be fully effective at ensuring advisers’ impartiality in the manner anticipated
by the Department. If advisers identify ways to circumvent the protections in the rule, they
would continue to impose costs on their customers and — because of their ability to continue
subordinating their clients’ interests to their own — the anticipated gains to investors would be
reduced. For example, if only 75 percent of anticipated gains were realized, under scenario 1,
the quantified portion of such gains would amount to $24 billion over 10 and $50 billion over 20
years. This question is addressed further in Chapter 8. Similar to other quantified estimates
presented herein, however, these estimates – specific to the front-end-load mutual fund market
segment – omit large, unquantified expected investor gains.
CEM “find load sharing, but not revenue sharing, to predict poor performance.” While
CEM’s evidence against load sharing is stronger than the evidence against ongoing conflicted
payment streams such as revenue sharing, the CEM results, along with larger body of literature,
suggests that ongoing conflicted payment streams are also harmful to IRA investors.
The CEM results should not be interpreted to absolve ongoing conflicted payment
streams for several reasons. First, the signal from revenue-sharing may have been drowned out

403
404
405

See the preambles to the proposal’s exemptions for more details on the anticipated effects of the exemptions conditions.
See Appendix B, Section B.3.1 for a more detailed discussion of broker and mutual fund incentives.
According to the Department’s estimates, gains-to-investors begin accruing once new advice is given following the effective date of the rule

176

by noise. When CEM attempt to simultaneously evaluate the effects of load sharing and revenue
sharing on investment results, they confidently identify a negative effect from load sharing.
With respect to revenue sharing, their point estimate, or best guess, is that every 100 basis points
in revenue sharing is associated with more than 100 basis points poorer performance. This
revenue sharing finding, however, is not statistically significant – that is, its margin for error is
too wide to rule out a zero effect.
This lack of statistical significance might mean that a relationship between revenue
sharing and results is small and weak, or even absent. But it might instead mean that the signal
from revenue sharing was simply drowned out by noise.
Second, CEM’s measure of revenue sharing is incomplete. CEM did not directly test the
effects of revenue sharing at all, but rather the effect of 12(b)-1 fees on performance. BD firms
frequently receive other types of payments from mutual funds, such as sub-accounting fees, and
frequently receive revenue sharing payments from investment advisers to the mutual funds.
These other, non-12(b)-1, payments are not recorded on the SEC forms N-SAR that CEM use to
estimate the load sharing and revenue sharing paid by mutual funds.406
Third, the sample used to generate the non-finding on revenue sharing was not
representative of all funds with either 12(b)-1 fees or revenue sharing. That sample was
restricted to funds with both positive load sharing and positive revenue sharing (as measured by
12(b)-1 fees) and amounted to less than 6 percent of CEM’s original sample.407 It is possible that
funds with front-end loads systematically charge less in other fees and pay less to brokers in
revenue sharing than broker-sold funds that don’t charge front-end loads. For example, Class B
and Class C funds often charge 12(b)-1 fees of 100 basis points, while Class A funds rarely
charge 12(b)-1 fees of that magnitude (often charging 25 basis points instead). In that case, more
meaningful revenue sharing activity would fall mostly outside the scope of the statistical test
used by CEM to evaluate the relationship between revenue sharing and performance, so it would
not be surprising that CEM did not find any revenue sharing effects. A study using more robust
data on conflicted payments to brokers might find both a statistically significant relationship
between those payments and underperformance and a stronger relationship than is suggested by
the point estimates from the CEM study. To the extent that these other fees are correlated with
load-sharing, they could introduce omitted variable bias into the regressions showing a
relationship between load-sharing and underperformance. In either case (omitted variable bias or
not omitted variable bias), the implication would be that higher levels of conflicted payments are
associated with poorer performance.
The Department also notes that compensation paid one way today might be paid another
tomorrow. If new rules or market developments were to reduce load sharing, revenue sharing
might increase to take its place, and increased revenue sharing might have larger (and more
easily detected) effects on investment results.
The quantified subset of IRA investors’ expected gains are estimated relative to a
baseline scenario in which (in the absence of Department action) the IRA front-end-load mutual

406

407

The Form N-SAR is a semi-annual filing by mutual funds to the SEC containing information on mutual fund operations, including 12(b)-1
payments.
By contrast, the sample used to reduced performance was approximately 54 percent of CEM’s original sample.

177

fund segment is expected to shrink over time and harmful conflicts of interest within the segment
are expected to become less severe. Were the Department to project a baseline scenario where
the IRA front-end load mutual fund segment retains its current market share and harmful
conflicts of interest within the segment maintain their current severity, the estimated gains would
be significantly larger. Appendix B below provides a more detailed explanation of these
estimates.
As discussed above, the quantified subset of IRA investors’ expected -gains represent
only part of the gains to IRA investors that the new rule and exemptions are estimated to deliver.
First, the quantified gains pertain only to the 13 percent of all IRA assets that are invested in
front-end-load mutual funds, and only to the subset of conflicts associated with front-end loads.
Second, as noted in Section 3.2.4 above, other evidence strongly suggests that adviser conflicts
will inflict additional losses of a similar magnitude by prompting IRA investors to trade more
frequently, which will increase transaction costs and multiply opportunities for losses from
chasing returns. Third, adviser conflicts also are likely to be associated with excessive price
spreads in principal trades between IRA investors and BDs. Finally, other types of investments
such as single-issue securities, banking products, or insurance products could also be subject to
underperformance due to conflicts. The new rule and exemptions, by limiting or mitigating all
adviser conflicts, will help IRA investors if it substantially reduces the $95 to $189 billion loss
they might otherwise suffer over the next 10 years – and $202 to $404 billion over 20 years.408
Benefits of the new rule and exemptions are expected to extend well beyond
improvements in IRA investment results. The market for fiduciary advice and other services
may become more efficient as a result of more transparent pricing and greater certainty about the
fiduciary status of advisers and about the impartiality of their advice. There may be benefits
from the increased flexibility that the new PTEs will provide with respect to fiduciary investment
advice currently falling within the ambit of the 1975 regulation. The new rule and exemptions
would extend guidance on the boundaries between fiduciary advice and education, heretofore
provided only with respect to plan participant investments, to plan distributions and thereby will
not impair and may improve access to IRA investor educational services. Innovation in new
advice business models, including technology-driven models, may be beneficially accelerated,
and nudged away from conflicts and toward transparency, thereby promoting healthy
competition in the fiduciary advice market. The Department believes that these benefits,
together with gains to IRA investors (both those quantified here and others) will justify the
proposal’s compliance cost.

3.3.2 Qualitative Discussion
The new rule and exemptions are designed to effectively limit or mitigate adviser
conflicts without diminishing IRA investors’ access to beneficial advice or other effective
support for sound saving and investing decisions. It balances revisions to the 1975 regulation
that extend fiduciary status to essentially all personal advice on IRA investments (but not pure
sales or educational activity) with PTEs that allow fiduciary advisers to receive compensation

408

See Appendix B, Section B.4 for details on these calculations.

178

that can introduce conflicts subject to protective conditions designed to mitigate those conflicts
so advice is impartial.
Under the new rule, a person would be an investment advice fiduciary if he provides a
recommendation to an IRA investor regarding the advisability of acquiring, holding, disposing,
or exchanging of securities or other property pursuant to a written or verbal agreement,
arrangement or understanding that the advice is specifically directed to the advice recipient for
consideration in making investment decisions with respect to securities or other property. As
fiduciaries, pursuant to the IRC prohibited transactions provisions, advisers to IRA investors
generally would have to refrain from transactions that introduce conflicts of interest unless
covered by a PTE. The PTEs and amendments to applicable existing PTEs carry strong
protective conditions – in particular a requirement that the adviser be loyal to IRA investors’
interests. Consequently, IRA underperformance otherwise attributable to adviser conflicts would
be greatly reduced.
Many comments on the 2010 and 2015 NPRMs express concern that as fiduciaries,
advisers’ cost to provide advice would be higher. Commenters argued that lower-income IRA
investors, and or those with smaller IRA balances, would be unable to afford or unwilling to pay
enough to cover that cost, so their access to advice would diminish, and their investment results
would suffer. Moreover, they asserted that because advisers help IRA investors not only with
investment decisions but also with setting and achieving savings goals – even with opening an
IRA and beginning to save – retirement savings itself might suffer. Such negative consequences
could more than offset the benefits of eliminating bias from advice, the comments said.
As discussed in Chapter 8, the Department believes the potential for such severely
negative consequences is limited. The Department believes the market already shows the
potential to serve small accounts with quality, impartial, affordable advice or other effective
support for sound saving and investing decisions.409 Under rules that largely ban adviser
conflicts, advisers would compete to offer consumers better service at lower prices, and product
vendors would compete to offer quality, low-fee products that advisers would recommend. This
contrasts with today’s market, where product vendors compete for advisers’ shelf space at
consumers’ expense.
Nevertheless, the Department takes seriously the risk that banning adviser conflicts could
reduce access to advice for some IRA investors, and that not only their investing but also their
saving might suffer as a result. Even if the potential for this result is limited, its severity is
sufficiently great that the Department agrees caution is required.
In addition, the Department is committed to harmonizing its rules with intersecting rules
under securities law. Comments on the 2010 Proposal argued that the Advisers Act may present
an obstacle to BDs’ provision of advice to IRA investors if adviser conflicts are banned.
According to these comments, taking such direct, arguably “special” compensation for advice
would force BDs to register under the Advisers Act as RIAs, and as RIAs they would be

409 Concerns also were expressed regarding an “advice gap” within the UK when the RDR became effective. However six months after the
effective date the FCA commissioned research showing that investment advisers continue to serve clients with savings and investments
between £20,000 and £75,000 and that a third of advisers continue to serve clients with less than £20,000. For a further discussion, see
Section 2.10.1.7 above.

179

required to provide a higher level of service to small accounts than those accounts now receive
or can afford.
In light of these considerations the Department has included certain PTEs that allow
fiduciary advisers to receive certain otherwise prohibited compensation subject to certain
protective conditions. The new rule and exemptions will deliver its potential benefits if these
PTEs provide sufficient flexibility to advisers to avoid substantial erosion of IRA investors’
access to advice, and their protective conditions sufficiently mitigate any conflicts associated
with compensation covered under the PTEs so that advice is impartial. As elaborated below, the
Department believes that a large proportion of the performance gap attributable to adviser
conflicts would be eliminated by the new rule and exemptions due to the calibrated scope of the
exemptions and the strength of the attached conditions.
Comments on the 2010 Proposal requested relief for the receipt by investment advice
fiduciaries of various fees and compensation resulting from transactions involving plans and
IRAs. The final rule and exemptions include a “Best Interest Contract Exemption” covering the
receipt of fees and compensation by an individual investment advice fiduciary, a financial
institution with respect to which the individual is an employee or representative, and any
affiliates and related entities, resulting from investment transactions engaged in pursuant to the
fiduciary adviser’s advice. This PTE includes strong protective conditions requiring that advice
be provided in accordance with certain impartial conduct standards that are fundamental
obligations of fair dealing and fiduciary conduct, and include obligations to act in the plan’s or
IRA’s best interest, avoiding misleading statements, and receive no more than reasonable
compensation. Additionally, the firm must provide written acknowledgment that it and its
individual advisers are fiduciaries; adopt policies and procedures designed to ensure compliance
with the impartial conduct standards; and disclose material conflicts of interest and fees. Firms
must generally enter into a written contract documenting these obligations when they transact
with IRAs and other investors in plans not covered by Title I of ERISA.
Commenters responding to the 2010 Proposal also requested exemptive relief for
principal transactions between a plan and a BD that is an investment advice fiduciary. Many
BDs view the ability to execute principal transactions as integral to the economically efficient
execution of a variety of transactions including fixed income securities. The Department
adopted a final exemption covering principal transactions in certain investments between a plan
or IRA and a fiduciary adviser if the principal transaction is a result of the provision of fiduciary
investment advice.
The final rule and exemptions include amendments to existing PTEs, attaching the same
impartial conduct standards as required in the Best Interest Contract Exemption and narrowing
the scope of some of the existing exemptions. Together these new and amended PTEs (see
Section 2.2) offer advisers targeted flexibility to structure compensation arrangements, including
many of the more common forms of compensation available to brokers today, subject to
conditions including acting in the client’s best interest. As long as applicable conditions are met,
advisers can offer customers the choice to compensate them via commissions, product-related
fees, mark-ups on bond prices, or direct fees. Advisers also would have the option of avoiding
prohibited transactions and with them any need to satisfy PTE conditions. IRA investors would
continue to have access to advice and a more than adequate range of investment options.
At the same time, the calibrated scope of the new and amended as PTEs and the
protective conditions attached to them should largely prevent covered compensation
arrangements from biasing advice. Because of the targeted scope and strong protections
provided in the new and amended PTEs, the Department believes that the final regulation and
180

exemptions will deliver the quantified gains to investors and other positive effects qualitatively
discussed in the preceding section. The Department intends to monitor compliance and market
developments under the new rule to assess whether it is achieving its goals and inform possible
future changes to the regulation and/or the PTEs’ scope or conditions.
Some of the IRA investor gains from this new rule and exemptions will be realized
quickly, while others will be realized more gradually. Some of the underperformance
attributable to conflicts results from excessive trading and is manifested as excess loads,
commissions and some timing errors associated with return chasing. This activity should
immediately be reduced, delivering large gains to investors in the near term. The rest of the
underperformance is associated with holding of “dominated” investments (insofar as available
alternatives would promise higher returns without greater risk, or lower risk with equal or
superior returns) such as those that are inadequately diversified or carry excessive fees. Some
investors might be slow to swap these investments for more efficient ones. They may be slow to
seek, be offered, or follow advice to do so. Gains to investors associated with such swapping
will be realized more gradually.
Under the final rule, IRA investment education that does not include specific personal
recommendations would not be treated as fiduciary advice. Currently IB 96-1 clarifies this
principle only for plan investment education, not for IRA investment education.
IRA investors also will benefit from the clarity and certainty that will be associated with
the application of uniform standards to all professional advice on retirement investing, regardless
of its source.
As noted later in connection with the benefits for plans, plan participants will benefit
from wider availability of education on plan distribution options. The new rule clarifies that
such education is not fiduciary investment advice. IB 96-1 provided clarity only with respect to
investment and savings education, not education regarding distribution options. Because the
new rule and exemptions are expected to improve decisions about the disposition of plan
distributions, its benefits will extend via roll overs to IRA investors as well.
Some additional benefits may accrue to existing fiduciary advisers – those who are
fiduciaries under the 1975 regulation – and to their plan and IRA clients, because of new
flexibility available pursuant to the proposed PTEs.
Still another benefit is expected to be healthier development of business models that rely
heavily on technology to generate and deliver advice and/or that build advice into financial
products themselves, as is the case with target date funds. So-called “robo advisers” and
products (such as target date funds) that minimize the need for complex advice are already
rapidly gaining market share. They promise to make advice far more affordable for small
investors, especially young investors who generally are more accustomed to technology-based
tools. More traditional advisory firms are scrambling to develop, partner with, or acquire such
innovative tools, and to combine these with more traditional services to deliver tailored services
to more market segments at far lower cost than that historically associated with traditional

181

approaches alone.410 The new rule and exemptions will help ensure that these new approaches
evolve toward less conflicted and more innately impartial business models, rather than
succumbing to the competitive pressures that have led more conflicted models to dominate
today’s highly imperfect marketplace.
The securities industry has itself identified certain important benefits of reform. These
includes the benefits of ensuring that advice honors investor’s interests, reducing investor
confusion over what to expect from advisers, and promoting trust in advisers.411 The
Department expects that the new rule and exemptions would deliver all of these benefits for plan
and IRA investors.
Consumers’ strong support for a clear fiduciary standard for all advisers has been
recognized by both consumer and some industry advocates. 412

410

411

412

Joyce Hanson, Investment News, “Robo startup will work with advisers exclusively” (July 15, 2014) and Steve Sandusky, Investment News,
“The reasons why human and robo-advisers will soon converge” (July 14, 2014). Because models are not susceptible to unhelpful human
behavioral biases, model-based advice may be superior to that formulated by an adviser without benefit of a model (Gray 2014).
SIFMA October 11, 2013 comment to SEC, “Recommendations of the Investor as Purchaser Subcommittee on Broker Dealer Fiduciary
Duty,” available at: https://www.sec.gov/comments/265-28/26528-39.pdf.
September 15, 2010 joint letter from AARP, Consumer Federation of America, Certified Financial Planner Board of Standards, Financial
Planning Association, National Association of Professional Financial Advisers, Investment Adviser Association, and North American
Securities Administrators Association to the SEC, transmitting survey results demonstration that a “vast majority” of U.S. Investors support
a clear fiduciary standard, available at: https://www.sec.gov/comments/4-606/4606-2748.pdf.

182

183

4. ERISA-Covered Plans
As noted in the introduction, ERISA-covered plans are critical to the retirement security
of most U.S. workers. In March 2015, about one-half – 49 percent – of private-sector employees
were participating in a job-based retirement plan. Fifteen percent participated in defined benefit
(DB) plans, and 43 percent participated in defined contribution (DC) plans (nine percent
participated in both).413 These numbers are just a snapshot, so they understate the reach of the
plans across employees’ careers. Employees who are young, part-time or low-paid are less
likely to participate. Over a full career many of these employees will at some point hold fulltime, higher-paying jobs that come with retirement benefits. In March 2015, 59 percent of fulltime private sector employees participated in some form of retirement plan.414
By the third quarter of 2015, plan assets totaled $8.0 trillion, including $2.8 trillion in DB
plans and $5.2 trillion in DC plans.415 This also understates the plans’ role in U.S. workers’
retirement security, because a large fraction of DC and some DB assets are transferred at some
point, usually upon leaving a job, to IRAs, providing a large majority of the flows into such
accounts. IRAs held $7.3 trillion by the third quarter of 2015.416
Both plan officials and plan participants rely heavily on professional advisers to assist
them with the investment of plan assets. DB plans, which promise a specific benefit to each
participant based on a specified formula and manage assets centrally, typically hire external or
internal asset managers to exercise their own discretion in making investment decisions. But
plan officials often rely on advisers to help them select these asset managers and assess their
performance. The asset managers, who themselves are plan fiduciaries, may also consult outside
advisers for help with various investment decisions. DC plans, in which employers and/or
employees contribute to separate employee accounts, often divide responsibility for investing
plan assets between plan officials and participants. Officials typically select a menu of
investment choices, often consisting of mutual funds or other diversified investment vehicles,
and designate one of these options as the default. They may rely on advisers to help them
construct the menu and select the default. Participants usually are responsible for allocating the
assets in their accounts among the available options. They may seek help from plan-provided
advisers, where available, or outside advisers, in making this allocation. They may also seek
advice when they are eligible to withdraw assets from their accounts, such as when changing
jobs or retiring, as to whether or not they should withdraw the assets, whether to transfer the
assets to an IRA, and how to invest the assets after such a transfer. Both DB and DC plan
officials sometimes rely on appraisers – essentially advisers who specialize in determining the
value of assets for which no market price can be observed – to determine the price that should be
paid or demanded for a hard-to-value asset that the plan will buy or sell.

413

414
415

416

U.S. Bureau of Labor Statistics, “National Compensation Survey: Employee Benefits in the United States, Table 2 and 3: Retirement
benefits, March 2015” (Sept. 2015); available at: http://www.bls.gov/ncs/ebs/benefits/2015/benefits_retirement.htm.
Ibid.
Board of Governors of the Federal Reserve System, “Financial Accounts of the United States, Third Quarter 2015,” Federal Reserve
Statistical Release Z.1 (Dec. 2015). DB assets do not include claims of pension fund on sponsor.
ICI, “The U.S. Retirement Market, Third Quarter 2015,” 2015.

184

4.1

Affected Entities

The final rule and exemptions will affect plan service providers who under its provisions
would be fiduciary advisers. It will also affect the plans and plan participants that they serve.

4.1.1 Service Providers
The Department used data from Schedule C of the 2013 Form 5500 to estimate the
universe of plan service providers that would
be affected by the final rule and exemptions.
Figure 4-1 Service Providers to ERISA Plans
Generally, plans with 100 or more participants
are required to report persons who rendered
Type of Service Provider
Number
services to or who had transactions with the
plan during the reporting year if the person
1,977
Recordkeepers
received, directly or indirectly, $5,000 or more
1,280
in reportable compensation in connection with
Consulting (general)
services rendered or their position with the
1
Consulting (pension)
plan. The types of services provided by each
service provider also must be reported unless
379
Insurance agents and brokers
the service provider only received certain types
1,311
Investment Advisory
of indirect compensation. Based on these
(Participants)
Schedule C service codes, the Department
estimates that 6,040 unique service providers
2,581
Investment Advisory (Plans)
most likely provide investment and valuation15
Real Estate Brokerage
related services covered under the proposed
415
rule that could cause them to be fiduciaries. In
Securities Brokerage
order to provide a reasonable estimate, service
236
Valuation (appraisals)
providers reporting service codes
662
corresponding to recordkeeping, consulting
Participant Communication
(general and pension), insurance agents and
6,040
All Types
brokers, investment advisory services (both
plans and participants), brokerage (real estate
Source: 2013 Form 5500 Schedule C
and securities), valuation services and those
providing participant communication were assumed to provide services that potentially could be
covered by the final rule and exemptions.
Although some small plans file Schedule C, small plans generally are not required to
complete Schedule C. Therefore, the Department’s estimate could underestimate the number of
covered services providers to small plans if any of these service providers only perform services
for small plans. The Department, however, believes that its estimated number of covered service
providers is reasonable, because most small plans use the same service providers as large plans.

185

4.1.2 Plans and Participants
The final rule and exemptions might not affect all plans because it is possible that not all
plan investors receive fiduciary investment advice. However, for purposes of this analysis, the
Department assumes that all plans and plan participants will be affected. This includes 681,000
plans, of which 44,000 are DB plans and 637,000 are DC plans. Gross participation in these
plans totals 132 million, 39 million, and 93 million, respectively.417
Some individuals participate in two or more plans, so the number of individual persons
affected is smaller than gross participation.418
Many participants will be affected indirectly in connection with fiduciary advice
rendered to plan officials. In particular, participants in the nearly 600,000 smaller plans (less
than 100 participants) are likely to be affected indirectly. Smaller plans may be more exposed to
conflicts of interest on the part of service providers, because they are less likely than larger plans
to receive investment assistance from a service provider that is acting as a fiduciary. Smaller
plans also often receive investment assistance from insurance brokers or broker-dealers (BDs),
who may be subject to conflicts of interest.419 The conflicted advice received by smaller plans
from service providers is particularly troubling in the context of the number of smaller plans,
since smaller plans make up the vast majority of the retirement plan universe.
The 69.9 million participants in the nearly 518,000 DC plans that allow at least some
participant direction may be more directly affected in connection with fiduciary advice rendered
directly to them.420
Although the vast majority of defined contribution plans allow participants to direct the
investment of at least a portion of assets, only about 30 percent of plans offer investment advice
to plan participants, and only 22 percent of plan participants with access to investment advice
through their plans avail themselves of this advice.421
Participants with access to optional lump sum distributions also may be directly affected
when they become eligible for such distributions. This includes participants in essentially all
DC plans and approximately one-third of DB plans.422 When workers change jobs and receive
distributions from their retirement plans, the average distribution is over $20,000, while almost
75 percent of distributions are less than $37,500.423 Almost nine in ten retiring workers who
receive a lump sum distribution take their entire account balances as a distribution. For those
workers, the median account balance is over $90,000.424 For retiring workers who take only a

417
418

419

420

421
422

423

424

2013 Private Pension Plan Bulletin; available at: http://www.dol.gov/ebsa/publications/form5500dataresearch.html#planbulletins.
See U.S. Department of Labor, Employee Benefits Security Administration, “Private Pension Plan Bulletin Historical Tables and Graphs”
(Sept. 2015); available at: www.dol.gov/ebsa/pdf/historicaltables.pdf.
U.S. Government Accountability Office, GAO-11-119, “401(K) Plans: Improved Regulation Could Better Protect Participants from
Conflicts of Interest” (2011); available at: http://www.gao.gov/products/GAO-11-119.
See U.S. Department of Labor, Employee Benefits Security Administration, “Private Pension Plan Bulletin, “Abstract of 2013 Form 5500
Annual Reports” (Sept. 2015); available at: www.dol.gov/ebsa/pdf/2013pensionplanbulletin.pdf.
PSCA, “58th Annual Report Reflecting 2014 Plan Experience,” Tables 104 and 107.
See Table 28. Traditional defined benefit plans: Availability of lump sum benefits at retirement, private industry workers, National
Compensation Survey, 2014; available at: http://www.bls.gov/ncs/ebs/detailedprovisions/2014/ownership/private/table28a.txt.
EBRI Notes, “Lump-Sum Distributions at Job Change,” Vol. 34, No. 11 (Nov. 2013); available at:
https://www.ebri.org/pdf/notespdf/EBRI_Notes_11_Nov-13_LSDs-WBS.pdf.
ICI Research Series, “Defined Contribution Plan Distribution Choices at Retirement: A Survey of Employees Retiring between 2002 and
2007” (Fall 2008), 33; available at: http://www.ici.org/pdf/rpt_08_dcdd.pdf.

186

partial lump-sum distribution, the median distribution received is almost $40,000.425
Participants who take lump sum distributions also may be affected by this rule’s application to
IRAs. Almost 90 percent of retiring workers who receive a lump sum distribution choose to
rollover at least some of the proceeds into an IRA, and 65 percent choose to rollover the entire
distribution into an IRA.426
Participants who do not have target-date funds or who choose not to use target-date funds
may be directly affected when they seek advice on alternative allocations for their retirement
assets. Among workers of all ages, target-date funds comprise 9.5 percent of all 401(k) assets.
The other 90.5 percent of assets may require investment advice. Further, when 401(k) assets are
broken down by accountholder age, retirement plan participants in their 20s invest 23.5 percent
of their assets in target-date funds. All other age groups invest 13.5 percent or less in target-date
funds.427

4.2

Need for Regulatory Action

As noted above, in 1975, the Department and the IRS issued parallel regulations428 that
defined the scope and meaning of the term “investment advice” under ERISA. The 1975
regulation substantially narrowed the broad statutory language conferring fiduciary status on all
persons rendering investment advice to a plan or an IRA for a fee.429
In the decades since its issuance, the Department has observed that as a result of its
narrow scope, the current regulation has effectively functioned as an “escape hatch” from
fiduciary status for advisers to plan investors in instances where, for example, the investment
advice was rendered to the plan on a single occasion, or in cases where the adviser has
disavowed any understanding that the advice would serve as a “primary basis” for the plan’s
investment decision. As a result, the 1975 regulation failed to provide adequate protection to
plan participants and beneficiaries from the effects of conflicts of interest and self-dealing on the
part of persons providing investment advisory services. The Department’s efforts to obtain
satisfactory remedies on behalf of plan participants and beneficiaries whose retirement security
had been jeopardized because of the misconduct of such advisers had been repeatedly thwarted.
The 1975 regulation has been overtaken by subsequent dramatic changes in the design,
operation, and marketing of employer–sponsored retirement plans. The variety and complexity
of financial products have increased, widening the information gap between advisers and their
clients and increasing the need for expert advice. Consolidation in the financial industry and
innovations in products and compensation practices have multiplied opportunities for selfdealing and made fee arrangements less transparent to clients and regulators. At the same time,
much of the responsibility for investing retirement savings has shifted from large private pension
fund managers to individual DC plan participants, many with low levels of financial literacy.

425
426
427

428
429

Ibid.
Ibid., 43.
EBRI Issue Brief, “401(K) Plan Asset Allocation, Account Balances, and Loan Activity in 2010,” Number 366 (Dec. 2011); available at:
http://www.ebri.org/pdf/briefspdf/EBRI_IB_12-2011_No366_401(k)-Update.pdf.
29 C.F.R. 2510.3-21(c); and 29 C.F.R. 4975-9(c).
The scope of the 1975 regulation was further limited by the Department in AO 76-65, in which it concluded that, under the facts described
therein, a valuation of closely held employer securities that would be relied on in the purchase of the securities by an employee stock
ownership plan (ESOP) would not constitute investment advice under the regulation.

187

These trends were not foreseen when the existing regulation was issued in 1975. 401(k) plans
did not yet exist when the 1975 regulation was promulgated. Between 1975 and 2013, the share
of total plan participation attributable to DC plans grew from 29 percent to 83 percent. In 2013,
84 percent of DC plan participation was attributable to 401(k) plans, and 97 percent of 401(k)
plan participants had responsibility for directing some or all of their account investments. 430
Participants in 401(k) plans have more control over the investment of their retirement
assets, but also bear the risk of loss from poor investment decisions.

4.2.1 Plan Level Advice
Plan sponsors have a fiduciary duty to ensure that plan assets are managed prudently and
in the exclusive interest of plan participants. Many rely on professional advisers to help them
discharge this duty. For example, DC plan sponsors may seek professional advice regarding the
selection of investment alternatives that will be available to plan participants and beneficiaries.
Plan trustees often rely on appraisers and valuation experts to attach fair values to so-called
hard-to-value assets. DB plan sponsors often rely on consultants to help them oversee plan
investments. Recently, concerns have been raised about the impartiality of the advice provided
by service providers to plan officials due to conflicts of interest and confusion regarding the
fiduciary status of the service providers. These issues are further discussed below.

4.2.1.1

Pension Consultants

Due to the increased complexity of investment opportunities available to DB plans, plan
sponsors often seek investment advice from pension consultants regarding matters such as: (1)
identifying investment objectives and restrictions; (2) allocating plan assets among various
objectives; (3) selecting money managers to manage plan assets in ways designed to achieve
objectives; (4) monitoring performance of money managers and mutual funds and making
recommendations for changes; and (5) selecting other service providers, such as custodians,
administrators and BDs. There also is a greater potential for conflicts of interest to exist in the
DB pension plan service provider market than when the current regulation was promulgated.
Many pension consulting firms provide services both to pension plan investors who are their
advisory clients and to money managers, and many have added brokerage and/or money
management affiliates, increasing the opportunities for self-dealing. As further discussed below,
the Department’s Consultant/Adviser Project (CAP) focused on the receipt of improper or
undisclosed compensation by plan consultants and other investment advisers. Through the CAP
program, the Department uncovered numerous cases where pension consultants and investment
advisers abused their relationship of trust with plan investors by recommending investment
managers or strategies in exchange for undisclosed compensation from third parties. The 1975
regulation has impeded the Department’s ability to redress what could be service provider
abuses.

430

See U.S. Department of Labor, Employee Benefits Security Administration, “Private Pension Plan Bulletin Historical Tables and Graphs”
(Sept. 2015), p. 9, p. 25, and p. 32; available at: www.dol.gov/ebsa/pdf/historicaltables.pdf. Please note that the number of active
participants in 1975 and 2013 are not directly comparable because of adjustments in the definition of a participant. This adjustment is
explained in detail in the historical tables and graphs.

188

An SEC staff study431 found that 13 of the 24 pension consultants examined, or their
affiliates, provided products and services to pension plan advisory clients, money managers, and
mutual funds on an ongoing basis without adequately disclosing these conflicts. The SEC staff
also found that the majority of examined DB plan consultants had business relationships with
BDs that raised a number of concerns about potential harm to plans. The report concludes that
consultants with conflicts of interest may steer plan investors to hire certain money managers or
other vendors based on a consultant’s (or an affiliates’) other business relationships and receipt
of fees from these firms rather than because the money manager is best suited to the plan’s
needs. Using data from the SEC staff study and other DB pension data, a GAO study concluded
that conflicts of interest that were not disclosed by pension consultants were associated with 130
basis points of underperformance.432

4.2.1.2

Confusion about Service Providers’ Fiduciary
Status

The service providers that DB and DC plan officials engage to perform many types of
plan services are subject to different regulatory regimes. For example, some plan investors work
with RIAs that are subject to SEC jurisdiction under the Advisers Act. RIAs must seek to avoid
conflicts of interest or, at a minimum, make full disclosure of material conflicts of interest.433
Other service providers may not be subject to ERISA fiduciary duty requirements or SEC
regulation, and therefore, may not be required to act in their clients’ best interest or to disclose
all conflicts of interest. GAO has reported that “there is a considerable amount of confusion
among plan sponsors about whether or not they are receiving investment advice subject to
ERISA fiduciary standards.”434 GAO found that plan sponsors are often not aware when a
service provider is not an ERISA fiduciary and often assume that the advice they receive from
the service provider is subject to ERISA standards and safe from harmful conflicts.
“Consequently, plan sponsors may not be aware that service providers can have a financial
incentive to recommend certain funds that would be prohibited if they were ERISA
fiduciaries.”435 The problem is particularly acute for smaller plan investors, because, as GAO
reported, “Smaller plans may be more exposed to conflicts of interest on the part of service
providers because they are less likely than larger plans to receive investment assistance from a
service provider that is acting as a fiduciary.”436
Another anomaly associated with the current regulation is that even persons who
represent themselves to plan sponsors as fiduciaries in rendering investment advice may not be
subject to ERISA’s fiduciary or prohibited transaction provisions if they fail to satisfy one or
more elements of the five-part test. For example, a consultant could hold itself out as a plan
fiduciary in a written contract with the plan, render investment advice for a fee, and still argue

431

432
433

434
435
436

“SEC Staff Report Concerning Examination of Select Pension Consultants” (May 16, 2005), available at:
http://www.sec.gov/news/studies/pensionexamstudy.pdf. The report’s findings were based on a 2002 to 2003 examination of 24 pension
consultants.
GAO Publication No. GAO-09-503T.
15 U.S.C. § 80b-1 et seq.; SEC v. Capital Gains Research Bureau, Inc., 375 U.S. 180, 194 (1963); General Instruction 3 to Part 2 of Form
ADV.
GAO Publication No. GAO-11-119, 28.
Ibid., 27.
Ibid., 28.

189

that its advice was insufficiently “regular,” was not mutually understood to serve as a “primary
basis” for the investment decision, or otherwise failed to meet some element of the five-part test.
The current test also makes it easy for consultants to structure their actions to avoid
fiduciary status. The SEC found evidence of this practice in its pension consultant examinations
and made the following statement regarding this issue in its report: “[m]any pension consultants
believe they have taken appropriate actions to insulate themselves from being considered a
‘Fiduciary’ under ERISA. As a result, it appears that many consultants believe they do not have
any fiduciary relationships with their advisory clients….”437 GAO also found that many service
providers structure their business arrangements “with a 401(k) plan to avoid meeting one or
more parts of the current five-part test….”438 For example, GAO states that some service
providers providing investment advice “include a provision in their contract that states that the
investment recommendations provided are not intended to be the primary basis for decision
making.”439 A report by the Department of Labor’s Office of Inspector General found that some
service providers that have significant undisclosed conflicts of interest attempted to avoid
ERISA fiduciary status under the current five-part test simply by stating in their investment
adviser contract that they were not fiduciaries.440 This problem confronts sponsors of both large
and small pension plans.

4.2.1.3

Platform Providers

Plan sponsors often retain financial advisers to assist in the provision of investment
alternatives to participants in 401(k) plans. For example, many plan sponsors seek advice from
“platform providers,” who are service providers, such as recordkeepers and third-party
administrators, that make available a menu of investments from which a plan sponsor typically
selects a more limited menu that will be available as designated investment alternatives under
participant-directed DC plans. The provider may simply offer a “platform” of investments from
which the plan sponsor selects those appropriate for the plan, or alternatively may select or assist
the plan fiduciary in selecting the plan’s designated investment alternatives.441
These platform providers vary significantly in their compensation arrangements. Some
work on a fee-only basis and receive compensation only directly from plans or sponsors for the
direct services they provide to the plan investors; others may receive indirect third-party revenue
sharing payments442 from other service providers, such as an investment fund provider, rather
than (or in addition to) direct payments from the plan sponsor for plan services. These fee

437
438
439
440

441

442

See SEC Staff Report on Select Pension Consultants, 2005, 6.
GAO Publication No. GAO-11-119, 24.
Ibid.
DOL Office of Inspector General, “EBSA Needs to Do More to Protect Retirement Plan Assets from Conflicts of Interest,” Report No. 0910-001-12-121(Sept. 30, 2010); available at: http://www.oig.dol.gov/public/reports/oa/2010/09-10-001-12-121.pdf.
The final rule provides a limitation for platform providers that makes clear that persons would not act as investment advice fiduciaries
simply by marketing or making available a platform of investments without regard to the individualized needs of the plan or its participants
and beneficiaries, as long as they disclose in writing that they are not undertaking to provide impartial investment advice or to give advice
in a fiduciary capacity. Similarly, a separate provision recognizes certain common activities that platform providers may carry out to assist
plan fiduciaries in selecting and monitoring the investment alternatives that they make available to plan participants. Therefore, merely
identifying offered investment alternatives meeting objective criteria specified by the plan fiduciary or providing objective financial data
regarding available alternatives to the plan fiduciary would not cause a platform provider to be a fiduciary investment adviser.
Payments can take several forms, for example 12b-1 fees, sub-transfer agency fees that reimburse the plan’s record keeper for services that
otherwise would be provided by a mutual fund, or payment of the mutual fund investment adviser’s compensation to the financial adviser,
its firm or an affiliated firm for promotion, marketing, or distribution.

190

arrangements among providers can introduce conflicts of interest. Some platform providers
include on their platform proprietary products, and/or products that are proprietary to an affiliate.
The inclusion of proprietary or affiliated products in platforms can also introduce conflicts.
Platform providers may have a financial incentive to recommend that their proprietary
funds be included as designated investment options on DC plan menus. Researchers have found
evidence that platform providers act on this conflict of interest, and that plan participants suffer
as a result. In a study examining the menu of mutual fund options offered in a large sample of
DC plans, underperforming non-propriety funds are more likely to be removed from the menu
than propriety funds. Similarly, the study found that platform providers are substantially more
likely to add their own funds to the menu, and the probability of adding a proprietary fund is less
sensitive to performance than the probability of adding a non-proprietary fund. The study also
concluded that proprietary funds do not perform better in later periods, which indicates that they
are left on the menu for the benefit of the service provider and not due to additional information
the service provider would have about their own funds (Pool, Sialm and Stefanescu
forthcoming).443
GAO has found that revenue sharing is a widespread practice among 401(k) plan service
providers. GAO stated that consequently, service providers that assist plan investors with
selecting funds to be included in a 401(k) menu “may suggest funds that have poorer
performance or higher costs for participants compared with other available funds.”444 The
financial impact of conflicts of interest can be substantial; fees for plans that have been managed
by service providers with conflicts of interest could be reduced by 30 percent or more according
to a representative of a service provider GAO spoke with.445 These arrangements can be harmful
to plan sponsors and plan participants, because the plan may pay excessive fees for the provided
services, which could lower returns. Participants in participant-directed 401(k) plans are
especially vulnerable in these situations, because they must rely on the assets in their individual
accounts to meet their retirement income needs.

4.2.2 Plan Participant Advice
As discussed above, with the growth of participant-directed DC plans, a substantial
proportion of plan participants now direct the investment of their pension plan assets and assume
more responsibility for ensuring the adequacy of their retirement income. At the same time,
there has been an increasing interest on the part of the Department, employers, and others to
ensure that participants have sufficient ability or support to make their own sound investment
decisions. This is especially true in the areas of asset allocation and the disposition of assets that
are rolled over or distributed from a plan.

443

444
445

Other researchers have found that, controlling for risk and other factors, evidence indicates that 401(k) plan funds outperformed what a
random selection across all funds would generate by more than 50 basis points annually. However, the authors found that those selections
would underperform analogous index funds by 31 basis points (Elton, Gruber, and Blake 2013). Other studies have found evidence that
menu selection as a whole is sometimes less than optimal, with sponsors offering an inadequate range of funds and index funds that are
more expensive than investors select in other settings (Elton, Gruber, and Blake 2006). The authors study the characteristics of plans that
are associated with adequate investment choices, and find that for 62 percent of plans, the selection offerings are inadequate. Additionally,
when examining one category of investment choices, S&P 500 index funds, they found that the index funds chosen by 401(k)-plan
administrators are on average inferior to the S&P 500 index funds selected by the aggregate of all investors. See also Tang et al. (2010).
GAO Publication No. GAO-11-119, 16.
Ibid., 30.

191

Many participants and beneficiaries receive information and assistance regarding asset
allocation and rollovers from plan service providers. In some cases, service providers may steer
participants toward purchasing products that benefit the service provider but are not in the
participants’ best interest. These issues are further discussed below.

4.2.2.1

Advice Regarding Asset Allocation

Plan participants often receive assistance from plan service providers regarding how to
allocate their 401(k) plan assets among their plans’ designated investment alternatives. This
assistance is provided through a variety of sources, such as brochures and other print materials,
call centers or help desks, group seminars, one-on-one sessions, and computer models. If service
providers deliver investment advice to participants for a direct or indirect fee, they are
fiduciaries under the 1975 regulation. However, if service providers provide only investment
education, they are not fiduciaries. As previously stated, in IB 96-1446 the Department identified
four specific categories of information and materials – plan information, general financial and
investment information, asset allocation models, and interactive investment materials – that
would be considered investment education and not result in rendering fiduciary investment
advice within the meaning of the 1975 regulation if they are furnished to plan participants or
beneficiaries alone or in combination.
IB 96-1 allows a suggested asset allocation using specific investment alternatives
available under the plan to be treated as investment education as long as the model or asset
allocation is accompanied by a statement indicating that other investment alternatives having
similar risk and return characteristics may be available under the plan and identifying where
information on those alternatives may be obtained.447 When the Department issued IB 96-1, it
addressed concerns that such use of investment alternatives available in the plan could allow
service providers to effectively steer participants to specific investment alternatives by
identifying only one particular fund in connection with an asset allocation example. To address
this concern, the bulletin encourages plan sponsors to identify other investment alternatives
within an asset class as part of an example when possible.
GAO has raised concerns about potential steering abuses in cases where specific plan
investment alternatives are used in asset allocation examples pursuant to IB 96-1. GAO found in
its 2011 report that, under this practice, “funds in which the service provider has a financial
interest can be highlighted and participants may perceive this information as investment advice.”
GAO concluded that “[p]articipants who confuse investment education for impartial advice may
choose investments that do not meet their needs, pay higher fees than with other investment
options, and have lower savings available for retirement.”448 In a subsequent report, GAO stated
that “[e]ven with disclosure statements as required in [IB 96-1], participants may interpret
information about their plans’ providers’ retail investment products contained in their plans’
educational materials as suggestions or recommendations to choose those products.”449

446
447
448
449

29 C.F.R. 2509.96-1(d).
29 C.F.R. 2509.96-1(d)(3)(iii) and (4)(iv).
GAO Publication No. GAO-11-119.
GAO Publication No. GAO 13-30.

192

Both investment education and impartial, expert fiduciary investment advice can help
participants make sound investment decisions. However, if a service provider, as part of an
education program, singles out specific investment alternatives in which it has a financial
interest, there is a risk that participants will suffer if there are no protective conditions in place.

4.2.2.2

Advice Regarding Plan Distributions

Plan participants also seek advice from plan service providers and other advisers
regarding whether to take a distribution from their plan account and roll over the distributed
amounts into an IRA. Under the 1975 regulation, participants often do not receive adequate
protection from conflicted advice about distributions and rollovers.
In 2005, the Department issued AO 2005-23A, which addressed whether a
recommendation that a participant take a distribution from his or her DC plan and roll over the
funds to an IRA was subject to ERISA’s fiduciary standards and associated prohibited
transactions provisions of ERISA and the IRC. Specifically, the AO addressed whether a
recommendation that a participant roll over an account balance to an IRA to take advantage of
investment options not available under the plan would constitute “investment advice” with
respect to the plan or the participant. AO 2005-23A concluded that advising a plan participant to
take an otherwise permissible plan distribution, even when that advice is combined with a
recommendation as to how the distribution should be invested, does not by itself constitute
“investment advice” within the meaning of the 1975 regulation. The Department stated that the
1975 regulation defines when a person is a fiduciary by virtue of providing investment advice
with respect to assets of an employee benefit plan. The Department expressed the view that a
recommendation to take a distribution is not advice or a recommendation concerning a particular
investment (i.e., purchasing or selling securities or other property) as contemplated by the 1975
regulation, and that any investment recommendation regarding the proceeds of such a
distribution would be advice with respect to funds that are no longer plan assets. However, in
instances where a plan officer or someone who is already a plan fiduciary responds to participant
questions concerning the advisability of taking a distribution or the investment of amounts
withdrawn from the plan, the Department opined in AO 2005-23A that the fiduciary is
exercising discretionary authority respecting management of the plan and must act prudently and
solely in the interest of the participant.
As a result of the Department’s position in AO 2005-23A, many plan participants are
vulnerable to being harmed by conflicted advice when they receive recommendations regarding
whether to take a plan distribution and rollover the assets into an IRA. This problem is
especially acute, because most IRA assets are attributable to rollover distributions,450 and the
amount of assets rolled over to IRAs is large and projected to increase substantially. In 2015,
new IRA rollover contributions amounted to about $376.5 billion, and by 2020, new IRA
rollover contributions are projected to total over $517 billion.451 Given the structural advantages
of retirement plans – larger investible asset balances may provide access to better asset
management and lower costs – plan participants often can expect lower net returns after rolling
their account into an IRA. Performance is especially likely to suffer when the rollover choice

450
451

ICI, “The U.S. Retirement Market, Third Quarter, 2015,” 2015.
Cerulli Associates, “Retirement Markets 2015,” Exhibit 9.08.

193

and asset selection results from conflicted advice. The sheer magnitude of current and future
rollovers renders any loss in performance economically impactful. If plan participants who
receive conflicted rollover advice suffer on average by a modest 50 basis points on future
returns, a single year’s worth of rollovers would cause participants to lose out on about $1
billion in returns over that year. Rollovers accumulating over 10 years could lose out on $48
billion over that time period, while rollovers accumulating over 20 years could lose out on about
$152 billion over those 20 years. If conflicted rollover advice instead causes performance to
suffer by 100 basis points on average, the losses would double to $2 billion, $96 billion, and
$305 billion over one, 10, and 20 years, respectively.452
Moreover, many plan sponsors and participants are not aware that participants lose
important protections after rolling over funds into an IRA. As IRA investors, they no longer
have the benefit of a plan fiduciary, such as the plan sponsor, representing their interests in
selecting a menu of investment options or structuring advice arrangements. They also are not
able to sue fiduciary advisers under ERISA for losses arising from fiduciary breaches, nor can
the Department sue on their behalf.
GAO confirmed the perils faced by plan participants in this area in a 2013 report where it
found that plan participants may not be adequately protected from plan service providers who
provide distribution recommendations that subordinate participants’ interests to the advisers’
own interests.453 For example, non-fiduciary plan service providers can recommend that
participants take distributions from their 401(k) plan and roll over their funds into the service
providers’ products outside the plan, thereby increasing the service provider’s compensation
without violating ERISA. According to GAO, “much of the information and assistance
participants receive is through the marketing efforts of service providers touting the benefits of
IRA rollovers and is not always objective.”454 In many cases, rolling over funds into these
products might not be in the participant’s best interest, because the products are not appropriate
for the participant’s needs or have higher fees than products that are available within the 401(k)
plan. In the 2013 report, GAO also discussed the practice of steering participant rollovers into
IRAs from plan service providers’ call centers. The report states that “service providers may
offer their call center representatives financial or other incentives for asset retention when
separating plan participants leave their assets in the plan or roll over to one of the providers’ IRA
products, which could lead to representatives promoting the providers’ products over other
options.”455
GAO describes how 401(k) service providers sell non-plan products and services, such as
IRA rollovers to participants outside their 401(k) in a practice known as “cross-selling,”
sometimes steering workers towards higher cost funds.456 The amount of additional fees that are
attributable to these rollovers can be substantial. For example, in its 2011 report, GAO stated
that according to an industry professional, “cross-selling” IRA rollovers to participants provides
an important source of income for service providers, and “a service provider could earn $6,000

452
453
454
455
456

See Section B.4 in Appendix B for details on the calculation of these estimates.
GAO Publication No. GAO-13-30.
Ibid., 22.
Ibid., 25-26.
GAO Publication No. GAO-11-119, 36.

194

to $9,000 in fees from a participant’s purchase of an IRA, compared with $50 to $100 in fees if
the same participant were to invest in a fund within a plan.”457
FINRA also has opined that recommendations regarding whether to take a distribution
and roll over plan assets into an IRA present an inherent conflict of interest. In a regulatory
notice issued in December 2013, FINRA stated that “[f]irms and their registered representatives
that recommend an investor roll over plan assets to an IRA may earn commissions or other fees
as a result. In contrast, a recommendation that an investor leave his plan assets with his old
employer or roll the assets to a plan sponsored by a new employer likely results in little or no
compensation for a firm or a registered representative…. Thus, a financial adviser has an
economic incentive to encourage an investor to roll plan assets into an IRA that he will represent
as either a broker-dealer or an investment adviser representative.”458
In the Notice, FINRA urges broker-dealers to review their retirement service activities to
assess conflicts of interest, and requires them to supervise these activities to reasonably ensure
that conflicts of interest do not impair the judgment of a registered representative or another
associated person about what is in the customer’s best interest.
In a January 2014 letter announcing its 2014 regulatory and examination priorities,
FINRA stated that it would evaluate securities recommendations made in rollover scenarios to
determine whether they comply with the suitability standards under FINRA Rule 2111.459
In a related development, the SEC Office of Compliance Inspections and Examinations
(OCIE) also announced that its 2014 examination priorities included (1) examining BDs and
RIAs for conflicts when recommending the movement of assets from a retirement plan to a
rollover IRA account in connection with a client’s change of employment, and (2) examining
broker-dealers and investment advisers for possible improper or misleading marketing and
advertising conflicts, suitability, churning, and the use of potentially misleading professional
designations when recommending the movement of assets from a retirement plan to a rollover
IRA account in connection with a client’s change of employment.460 More recently, in its 2016
enforcement priorities, the SEC stated that protecting investors and retirement savers remain a
priority and will likely remain a focus for the foreseeable future. The 2016 examination
priorities stated that OCIE would continue its multi-year Retirement-Targeted Industry Reviews
and Examinations (ReTIRE) Initiative established in 2015, which include examining the
reasonable basis for recommendations made to investors, conflicts of interest, supervision and
compliance controls, and marketing and disclosure practices.461
Research shows that many individuals making contributions to an IRA spend very little
time scrutinizing disclosure statements.462 Thus, plan participants may not understand the

457
458

459

460

461

462

Ibid.
FINRA Regulatory Notice 13-45, “Rollover to Individual Retirement Accounts” (Dec. 2013); available at:
https://www.finra.org/web/groups/industry/@ip/@reg/@notice/documents/notices/p418695.pdf.
FINRA “2014 Regulatory and Examinations Priorities Letter” (Jan. 2. 2014); available at:
http://www.finra.org/web/groups/industry/@ip/@reg/@guide/documents/industry/p419710.pdf.
SEC National Exam Program, “Examination Priorities for 2014,” Office of Compliance Inspections and Examinations (Jan. 9, 2014);
available at: https://www.sec.gov/about/offices/ocie/national-examination-program-priorities-2014.pdf.
SEC National Exam Program, “Examination Priorities for 2016,” Office of Compliance Inspections and Examinations (Jan. 11, 2016);
available at: https://www.sec.gov/about/offices/ocie/national-examination-program-priorities-2016.pdf.
Press Release on LIMRA Study, “Many Americans Don’t Fully Read Retirement Plan Disclosures; Few Know What Fees they Pay” (Dec.
18, 2014); available at: http://www.limra.com/newscenter/newsarchive/archivedetails.aspx?prid=259.

195

differences between fees they would incur if they left their money in the plan compared to the
fees they would incur if they rolled over the funds into an IRA. Moreover, even when presented
with information on the difference in fees among 401(k)-type plans and IRAs, participants may
have difficulty understanding the information or the implications for their retirement income
security.
A substantial body of academic research suggests that consumers pay inadequate
attention to mutual fund fees; that their advisers and the RIAs managing the mutual funds their
advisers recommend deliberately exploit this tendency (Cici and Boldin 2010);463 and that
consumers’ returns suffer as a result.
In fact, many service providers do not make fee information accessible and
understandable for participants and beneficiaries that are considering IRA rollovers. In a 2013
report, GAO reviewed websites of numerous large IRA providers to locate fee information and
concluded that IRA fee information was “generally scattered across the providers’ websites in
multiple documents, making it difficult to identify all applicable fees.”464 GAO noted that in one
rollover application, the schedule of fees was located in the last section of a 49-page document,
and the fee information was covered over four-and-one-half pages in eight-point typeface.465
GAO concluded that “misleading statements … make it difficult to understand IRA fees” and
presented an example where GAO investigators made calls to 401(k) service providers, most of
whom offer IRA products, and found that “7 of 30 call center representatives (representing firms
administering at least 34 percent of IRA assets at the end of the 1st quarter in 2011) said that
their IRAs were ‘free’ or had no fees with a minimum balance, without clearly explaining that
investment, transaction, and other fees could still apply, depending on investment decisions.”466
GAO also reviewed ten IRA websites, and found that “5 providers … made similar claims, often
with certain conditions such as a $50,000 minimum balance or consent to receive electronic
statements explained separately in footnotes. For example, an IRA provider’s website [GAO]
reviewed stated that the provider would waive annual custodial fees if the balance exceeded an
unspecified amount and only referred vaguely to other fees that might still apply, which were
disclosed in multiple separate documents available upon request. Accurate information on when
IRA providers will waive fees and what fees they will waive can be difficult for participants both
to locate and understand.”467

463

464
465
466
467

The authors found that a measurable number of investors select index funds with excessive fees and uncompetitive returns. They identify a
naïve group of investors who seem to be unduly influenced by brokers and financial advisers. See also Choi, Laibson, and Madrian (2010).
In their experiments, subjects selecting among similar index funds overwhelmingly fail to minimize fees, even when fee information is
costless. They reject the hypothesis that subjects buy high-fee index funds because of bundled non-portfolio services. See also Palmiter
and Taha (2008). They find that mutual fund investors are unaware of the basics of their funds, pay insufficient attention to fund costs, and
chase past performance despite little evidence that high past fund returns predict future returns. See Houge and Wellman (2007). The
authors find that as the industry becomes more adept at segmenting customers by level of investment sophistication, load mutual fund
companies take advantage of this ability and charge higher expenses to their target customer: the less knowledgeable investor. No-load
fund companies, which tend to attract the more sophisticated investor, offer lower expenses. See Gil-Bazo and Ruiz-Verdu (2009). They
present evidence consistent with this strategic fee setting argument. Mutual funds with worse before-fee performance charge higher fees.
The authors posit that funds expected to perform poorly (or that have performed poorly in the past) raise fees and target less performance
sensitive (less sophisticated) investors, often through increased marketing efforts (which increase distribution costs).
GAO Publication No. GAO-13-30, 34.
Ibid., 34-35.
Ibid., 36.
Ibid., 37.

196

FINRA shares GAO’s concern that BDs’ marketing campaigns on television and radio,
print, websites, and social media may not be fair and balanced and could be misleading, because
they frequently emphasize that fees are not charged in connection with their IRAs, but only
disclose in a footnote that certain fees apply. In July 2013, FINRA stated that “referring to an
IRA account as a ‘free IRA’ or ‘no-fee IRA’ where costs exist would fail to comply with
[FINRA] Rule 2210’s prohibition of false, exaggerated, unwarranted, promissory or misleading
statements or claims.”468 FINRA concluded that a “headline statement to the effect that a firm
does not charge annual maintenance fees should include an explanation in close proximity to the
headline of the conditions associated with the offer and the other fees that would apply.” In a
January 2014 letter announcing its 2014 regulatory and examination priorities, FINRA stated
that reviewing firms’ rollover practices was an examination priority, and that it will examine
firm’s marketing materials and supervision in this area.469 The SEC Office of Compliance
Inspections and Examinations also announced that its 2014 priorities included examining the
sales practices of investment advisers targeting retirement-age workers to rollover their
employer-sponsored 401(k) plan into higher cost investments, including whether advisers are
misrepresenting their credentials or the benefits and features of IRAs and other alternatives.470
More recently, in a January 2015 letter announcing its regulatory and examination priorities,
FINRA stated that it will continue to focus on IRA rollovers. FINRA also provided guidance
regarding specific steps firms should take if they do not intend to provide security
recommendations as part of rollover transactions or only intend to provide educational materials
with respect to such transactions. In this regard, FINRA stated that “[i]f a broker-dealer does not
intend for its registered representatives to recommend securities transactions as part of the IRA
rollovers of their customers, then the broker-dealer should have policies, procedures, controls
and training reasonably designed to ensure that no recommendation occurs. Similarly, if
registered representatives are authorized to provide educational information only, a firm’s
written supervisory procedures should be reasonably designed to ensure that recommendations
are not made.”471

4.2.3 Department Enforcement Challenges
The 1975 regulation’s narrow approach to fiduciary status sharply limited the
Department’s ability to protect plan investors from conflicts of interest that may arise from the
diverse and complex fee practices existing in today’s retirement plan services market and to
devise effective remedies for misconduct when it occurs. In recent years, the Department has
observed in its investigations that certain non-fiduciary service providers – such as consultants,
appraisers, and other advisers – have abused their relationships with plan investors by
recommending investments in exchange for undisclosed kickbacks from investment providers,
engaging in bid-rigging, misleading plan fiduciaries about the nature and risks associated with
plan investments, and by giving biased, incompetent, and unreliable valuation opinions. Yet, no
matter how egregious the abuse, plan consultants and advisers had no fiduciary liability under
ERISA, unless they met every element of the five-part test.

468

469
470
471

FINRA Regulatory Notice 13-23, 2013. FINRA Rule 2210 requires broker-dealer communications to be fair and balanced and not omit
material information that would cause them to be misleading.
FINRA, “2014 Regulatory and Examinations Priorities Letter,” 2014.
SEC, “Examination Priorities for 2014.”
FINRA Regulatory and Examination Priorities Letter, 2015.

197

In instances where a plan had relied upon abusive investment advice from a self-dealing
consultant concerning an investment product on a single occasion, the Department generally
could not bring an action for fiduciary breach against the consultant, because the “regular basis”
element of the 1975 regulation’s five-part test was not satisfied. For example, a plan’s purchase
of annuity contracts is a major transaction, but it may occur only in connection with the plan’s
termination. Accordingly, one-time advice on the expenditure of virtually all of a plan’s assets
on the purchase of an annuity to cover all of the plan’s obligations was typically not treated as
fiduciary advice despite its clear importance to the plan participants who depended upon the
annuity for their retirement benefits. As a result of the five-part test, rather than focus on the
impartiality or prudence of advisers’ recommendations, investigators had to first gather evidence
on a series of factors not set out in the text of the statute and with little or nothing to do with the
legitimate interest of plan investors in being able to rely on the recommendations of persons who
held themselves out as trustworthy advice professionals.

4.2.3.1

EBSA Consultant/Adviser Project

EBSA seeks to protect plans and their participants and beneficiaries by concentrating
significant enforcement resources on carefully selected areas with significant potential for abuse.
The evaluation and determination of fiduciary status was particularly important to one of
EBSA’s major enforcement projects: the Consultant/Adviser Project (CAP). CAP focused on
the receipt of improper or undisclosed compensation by employee benefit plan consultants and
other investment advisers. EBSA’s investigations sought to determine whether the receipt of
such compensation, even when disclosed, violates ERISA because the adviser/consultant may
have leveraged its position with a benefit plan to generate additional fees for itself or its
affiliates.
One of the most critical elements in bringing enforcement actions under the CAP
initiative was establishing whether a service provider is a fiduciary. In order to make a fiduciary
determination, investigators had to gather evidence to support a finding for each element of the
five-part test. In all cases, the analysis necessary to determine fiduciary status was very factintensive and require extensive review of plan documents and contracts, client files, emails,
investment documentation, accounting records, and interview statements to be obtained from
service providers and their affiliates. Consequently, EBSA investigators routinely devoted
considerable time and resources to establishing all elements of the five-part test, rather than
focusing on the actual misconduct at issue in particular cases.
In recent years, the Department has uncovered cases of pension consultants and
investment advisers abusing their relationship with plan investors by recommending investments
in exchange for undisclosed compensation, misleading plan fiduciaries about the true risks
associated with plan investments, and giving biased investment advice. These cases typically
involved the production of extensive documents and the conducting of many interviews, because
the services had to be evaluated for each plan client. Since fiduciary status had to be established
on a transaction-by-transaction basis, individual client files had to be reviewed against all five
parts of the fiduciary test to evaluate whether advice had been given based on the particular
needs of the plan. Production and review of the individual client files was a time-consuming
process but one that was essential to provide evidentiary support for the five-part test.
The requirement under the 1975 regulation that the plan consult the adviser “on a regular
basis” presented one of the greatest obstacles to holding investment advisers to fiduciary
standards. This was because plan investors often hired investment managers, advisers, or
consultants to render advice for specific investment decisions. Despite the size or nature of the
198

transaction, if the adviser did not provide advice on a “regular basis,” the adviser would not be
deemed a fiduciary no matter what percentage of plan assets were involved in the transaction.
The 1975 regulation also creates significant barriers to establishing fiduciary status by
requiring that advice be rendered pursuant to a “mutual agreement, arrangement, or
understanding that the advice would serve as a primary basis for investment decisions.” Under
the test, the Department must devote considerable resources in order to meet the burden of
showing that the parties had a mutual understanding that the advice would be a “primary” basis
for investment decisions. Absent such proof, the adviser could not be deemed a fiduciary,
regardless of the plan’s or participants’ actual reliance on the advice. In cases where prudent
fiduciaries consult multiple advisers, or the advisers included boilerplate disclaimers of any
“mutual understanding” as to the primacy of the advice, the test serves only to defeat legitimate
plan expectations - and imposes one more set of investigative hurdles for holding advisers
accountable for biased or imprudent advice.

4.3

The Final Rule and Exemptions’ Impact on Plan Participants

Plan participants gain value when a plan’s investment advisers, in competition to provide
the best value to the plan, deliver high quality advice in plan participants’ exclusive interest at
competitive prices. Harm can result, however, if advice is tainted by unmitigated conflicts of
interest, which may occur when a plan’s advisers strike deals with other service providers for
additional consideration at the plan’s expense or subordinate the plan participants’ interest to
someone else’s.
Participants in participant-directed DC plans also benefit from various kinds of support
for their own decisions about the investment of the assets in their own accounts. As detailed
above in the discussion of IRAs, most consumers are not financially sophisticated and are prone
to costly investment errors. Consequently, they can benefit from personalized, competitively
priced fiduciary investment advice. They also can benefit from investment education that does
not provide specific, individualized investment recommendations.
The final rule and exemptions include a number of measures designed to ensure that
investment advice is aligned with plan participants’ interests, without impairing plan sponsors’
ability to make available investment education. These measures are found in the revisions to the
1975 regulation and in conditions attached to the accompanying PTEs.
The final rule also provides several important categories of services and materials that
are not treated as recommendations. This is based on the Department’s belief that Congress did
not intend to include them as fiduciary “investment advice,” because they do not present
opportunities for service providers to self-deal and are not characterized by a relationship of trust
where clients reasonably expect service providers to act solely in their best interest.

4.3.1 Promoting Good Advice and Education
Under the final rule, more of the investment advice that plan officials and participants
rely on will be treated as fiduciary advice under ERISA and the IRC. As a result, much more
advice will have to be prudent, loyal to participants’ interests, and free from bias. Such good
advice will promote sound investment decisions and better retirement security results. The final
rule also includes measures to promote financial education, including education that supports
plan participants’ decisions about plan distributions. These provisions are discussed below.

199

4.3.1.1

Advice to Plan Sponsors and Other Plan Officials

As discussed above, investment advice rendered to plan sponsors and other plan officials
is sometimes conflicted, and these conflicts sometimes bias the advice. Sponsors and plan
officials following biased advice may make poor investment decisions, which can compromise
participants’ retirement security. The final rule and exemptions include a number of measures
calibrated to ensure that advice to plan sponsors and officials is prudent, loyal to participants’
interests and unbiased.
The final rule substantially relaxes certain parts of the 1975 regulation’s five-part test.
For example, advice is fiduciary in nature if it consists of a single recommendation given once
(relaxing the 1975 regulation’s requirement that the advice be given on a regular basis). Advice
also is subject to a fiduciary standard if it is (i) rendered pursuant to a written or verbal
agreement, arrangement, or understanding that the advice is based on the particular investment
needs of the plan, or (ii) specifically directed to the plan sponsor regarding the advisability of a
particular investment or management decision with respect to a plan’s securities or other
investment property, or if the adviser represents or acknowledges that he or she was acting as a
fiduciary with respect to the advice (relaxing the 1975 regulation’s requirement that the advice
be individualized and mutually agreed to serve as a primary basis for investment decisions).
Plan sponsors and officials will gain value from these provisions, because they will
reduce the potential for harm from biased advice being provided to them. Under the 1975
regulation, plan sponsors and officials sometimes relied on recommendations that were
presented, implicitly or explicitly, as trustworthy advice, but where the advice was not provided
regularly, or where the adviser maintained that there was no mutual agreement, arrangement, or
understanding that the advice was individualized or intended to serve as a primary basis for
investment decisions, there was no fiduciary advice. This is especially true in the current
investment marketplace where plans invest not only in stocks and bonds but also in more
sophisticated investment products such as partnerships, private equity funds, real estate, and
hedge funds.
As noted above, because under the 1975 regulation such recommendations were not
fiduciary advice under ERISA or the IRC, the adviser providing the recommendations may have
been conflicted, and the advice may have been biased. The adviser owed no fiduciary duty of
prudence and loyalty to the plan’s or investors’ interests, and faced no liability under ERISA for
harm that may result from the plan or investor following biased investment advice. Under the
final rule, such recommendations will now constitute fiduciary advice under ERISA and the
IRC. Accordingly, the adviser will have to avoid conflicts, or mitigate them by satisfying the
protective conditions of an applicable PTE, and will owe plan sponsors, officials, participants,
and beneficiaries’ duties of prudence and loyalty. If the adviser breaches these duties, plan
sponsors, officials, participants, beneficiaries, or the Department can hold the adviser
accountable for any resultant losses.
The provisions will also ensure that advice that is sold as fiduciary advice can be trusted
to be so. The 1975 regulation applied the five-part test even to persons that have held
themselves out to plan sponsors as fiduciary advisers. Thus, an adviser could have held itself out
as a plan fiduciary in its written contract with the plan sponsor, rendered advice about
investments for a fee, and still argued that its advice was insufficiently “regular” or “primary,”
or otherwise failed to meet each and every one of the five elements of the test. The final rule
provides important protections by ensuring that an adviser is a fiduciary if it represents or
acknowledges that it is acting as a plan fiduciary with respect to providing investment advice
covered by the rule. This will produce value for plan sponsors and plan participants by ensuring
200

they can rely on the expert guidance provided by consultants and other advisers who represent
that they are fiduciaries providing impartial investment advice, without being concerned that the
adviser will disavow fiduciary status if something goes wrong.
The final rule will also treat as fiduciary advice certain recommendations of investment
managers or investment advisers to manage securities or other property. As noted previously,
the SEC has documented numerous instances in which pension consultants recommended
investment managers with whom they did business to plan sponsors without disclosing these
conflicts, and GAO found that these conflicts were associated with 130 basis points of
underperformance. While the Department has made special efforts to target such abusive
situations for enforcement action, these efforts have been impaired by the difficulty of
establishing the consultants’ fiduciary status under the 1975 regulation. This provision of the
final rule will produce value for plan sponsors and participants by ensuring that consultant
recommendations of investment managers are unbiased and by holding consultants accountable
for ensuring that such recommendations are prudent and loyal to plan sponsors’ interests.
The final rule extends fiduciary status to certain service providers that provide investment
advice to plan officials, and by subjecting them to the full extent of remedies under ERISA. This
will create more beneficial arrangements in the pension plan service provider market by ensuring
that advice is prudent, loyal to participants’ interests and unbiased. The final rule will produce
value for plan sponsors and participants by enabling more optimal decisions regarding plan
investments as the risk of receiving and then acting on conflicted advice will be lessened. In
instances where advisers commit abuses, the new rule will additionally produce value for plan
sponsors and other plan fiduciaries by making it possible to recover losses from the advisers
rather than solely from the plan fiduciaries that in good faith relied on the advice.

4.3.1.2

Advice to Plan Participants

Many plan participants currently are at risk of receiving and following biased investment
advice. They are especially vulnerable to being harmed by conflicted advice when they receive
recommendations on whether to take a plan distribution and roll over plan assets into an IRA,
because the Department interpreted the 1975 regulation to provide that such recommendations
generally do not constitute fiduciary investment advice in AO 2005-23A.
To ensure that plan participants are protected from conflicted advice with respect to one
of the most important financial decisions regarding their retirement assets, the final rule
specifically includes recommendations to take a distribution of benefits or regarding the
investment of securities or other property to be rolled over or otherwise distributed from a plan,
as fiduciary investment advice. Participants will gain value from this provision, because it will
limit their exposure to harm caused by advisers’ conflicts of interest by clearly placing
recommendations to take distributions (and thereby withdraw assets from existing plan
investments) or to entrust the investment of securities or other property that is rolled over or
otherwise distributed from the plan to particular money managers or advisers within the scope of
covered fiduciary investment advice.
Additionally, the final rule draws a critical distinction between fiduciary investment
advice and non-fiduciary investment information and educational materials by preserving the
provisions of IB 96-1 that facilitate general investment education. Therefore, furnishing plan
information, general investment and financial information, asset allocation models, and
interactive investment materials to a plan, plan fiduciary, participant or beneficiary will not
constitute the rendering of investment advice, irrespective of who provides the information (e.g.,
plan sponsor, fiduciary or service provider), the frequency with which the information is shared,
201

or the form in which the information and materials are provided (e.g., on an individual or group
basis, in writing or orally, or by way of video or computer software). Such educational support
will benefit plan participants by helping them make better decisions about plan distributions and
by leading to better investment outcomes.
In the 2015 Proposal, the Department concluded that certain provisions in IB 96-1 which
were included in the 2010 Proposal needed to be modified to protect participants from conflicts
of interest. The provisions involved the presentation and use of asset allocation models and
interactive materials that identified specific investment products available under the plan. IB 961 allowed asset allocation models and interactive materials to include specific investment
products available under an employee benefit plan as long as the model or materials were
accompanied by a statement indicating that other similar alternatives are available under the plan
and identifying where plan participants and beneficiaries could obtain information on those
alternatives. The Department observed that such asset allocation models and interactive
materials can be constructed or presented in ways that reflect the service providers’ own
interests. The risk those conflicts present to retirement investors was magnified by the fact that
the 2015 Proposal expressly extended the investment education provisions to include IRAs. In
order to protect plan participants from these conflicts, the 2015 Proposal expressly excluded
from the investment education provision asset allocation models and interactive materials that
identified specific investment alternatives. The Department’s intention was to ensure that
neither plan service providers nor outside advisers would be able to claim that asset allocation
models or interactive materials were “education” but use them to steer plan and IRA investor
toward investment products that benefit the advisers at the plan’s or investor’s expense. This
also was a concern identified by GAO in two reports discussed above.472
The Department received many comments regarding this departure from IB 96-1. Some
commenters agreed that participants are particularly vulnerable to subtle, yet powerful,
influences by advisers when they receive asset allocation information and use interactive
investment tools. They believe that plan participants and beneficiaries and IRA owners may
view these models and interactive tools as providing investment advice even if the investment
provider includes the general statement described in IB 96-1 about the availability of other
products. The commenters asserted that these models and interactive materials seem sufficiently
“individualized,” that they will influence retail investors to be favorable to certain products, and
that they should be treated as fiduciary investment advice.
On the other hand, many other commenters stated that it is a mistake to prohibit the use
of specific investment options in asset allocation models and interactive investment tools used
for educational purposes. They believe it is critical to “connect the dots” for retail investors in
understanding how to apply educational tools to the specific options available in their plan, and
that the rule as proposed would greatly undermine the effectiveness of these models and
materials as educational tools. They asserted that otherwise, participants will be led to “guess”
or try to figure out an investment strategy on their own (or do nothing based on inertia).
The Department agrees with commenters that adjustments could be made to the final rule
in a way that could facilitate delivery of educational information to participants and beneficiaries

472

GAO Publication No. GAO-11-119, and U.S. Government Accountability Office, GAO-13-30, “401(K) Plans: Labor and IRS Could
Improve the Rollover Process for Participants” (2013); available at: http://www.gao.gov/products/GAO-13-30.

202

in ERISA plans, but also addresses the concern about improper influence and steering of
investor choices. Therefore, the final rule allows asset allocation models and interactive
materials to include reference to specific investment products or alternatives available under an
ERISA-covered plan, if certain conditions are met. Specifically, in addition to meeting the
general conditions for educational asset allocation models and interactive materials, such models
or materials may include or identify a specific investment product or alternative if the product or
alternative is a designated investment alternative under the employee benefit plan subject to
fiduciary oversight by a plan fiduciary who is independent of the person who developed or
markets the investment product or alternative and if the model or materials: (i) identify all other
designated investment alternatives available under the plan with similar risk and return
characteristics, if any and (ii) are accompanied by a statement indicating and identifying where
information on those investment alternatives may be obtained, including information described
in paragraph (d) of 29 C.F.R. 2550.404a-5.
In this regard, it is important to emphasize that a responsible plan fiduciary would also
have, as part of the ERISA obligation to monitor plan service providers, an obligation to evaluate
and periodically monitor the asset allocation model and interactive materials being made
available to the plan participants and beneficiaries.473 That evaluation should include an
evaluation of whether the models and materials are in fact unbiased and not designed to
influence investment decisions towards particular investments that result in higher fees or
compensation being paid to parties that provide investments or investment-related services to the
plan. In this context and subject to the conditions above, the Department believes such a
presentation of a specific designated investment alternative (DIA) in a hypothetical example or
interactive tool can be appropriately treated as non-fiduciary “education” rather than fiduciary
investment recommendations, and the interests of plan participants are protected by fiduciary
oversight and monitoring of the DIAs as required under Title I of ERISA and the Department’s
regulations.
The Department does not agree that the same conclusion applies in the case of
presentations in asset allocation models or interactive materials of specific investments to IRA
owners because of the likelihood that such asset allocation models and interactive investment
tools would often amount to specific investment recommendations in the IRA context. The IRA
context lacks the Title I protections of review and prudent selection of the presented options by
an independent plan fiduciary. The Department also was not able to reach the conclusion that it
should create a broad safe harbor from fiduciary status for circumstances in which the IRA
provider narrows the entire universe of investment alternatives available to IRA owners to just a
few coupled with asset allocation models or interactive materials. Nor could the Department
readily import the conditions applicable to plan communications regarding designated
investment alternatives to IRA communications. For example, the Department was not able to
conclude that information analogous to the disclosures regarding DIAs under 29 C.F.R.
2550.404a-5 could be made available about other comparable investment alternatives available
under an IRA.

473

See 29 C.F.R. 2550.404a–5(f) and 2550.404c–1(d)(2)(iv).

203

Similarly, because the provision is limited to DIAs available under employee benefit
plans, the use of asset allocation models and interactive materials with specific investment
alternatives available through a self-directed brokerage account would not be covered by the
“education” provision in the final rule. Such communications lack the safeguards associated
with DIAs, and pose many of the same problems and dangers as identified with respect to IRAs.
The 2015 Proposal also provided a carve out such that advice provided by an employee
of a plan sponsor or employee organization sponsoring the plan to a plan fiduciary was not
investment advice covered by the proposal if the person providing the advice does not receive
any compensation beyond the normal compensation received for work performed for the
employer or employee organization. Some commenters asserted that the carve-out should not be
limited to employees’ statements to a plan fiduciary. They stated that the provision should also
apply to advice provided to a plan sponsor’s (or an affiliate’s) employees or participants in the
sponsor’s (or an affiliate’s) plans. DOL could condition this broader exclusion on the employee
not receiving any additional (beyond salary) or variable compensation with respect to such
advice.
The Department considered these comments and has included a provision in the final rule
providing that a communication is excluded from fiduciary status if an employee of the plan
sponsor of an employee benefit plan (or an affiliate) provides advice to a plan fiduciary, an
employee (other than in his or her capacity as a participant or beneficiary) or independent
contractor of such plan sponsor (or affiliate) as long as the employee receives no compensation,
direct or indirect, in connection with the advice beyond the employee’s normal compensation for
work performed for the employer or an affiliate. The final rule also provides an exclusion for
employee communications to other employees as participants or beneficiaries of the plan but
includes conditions to prevent the provision from being used as a loophole for programs clearly
designed and intended to provide investment advice. Specifically, the employee’s job
responsibilities cannot involve the provision of investment advice, the employee cannot be
registered under federal or state securities or insurance laws, the advice the employee provides
does not require the person to be registered or licensed under federal or state securities or
insurance laws, and the employee cannot receive any fee or other compensation beyond their
normal compensation. The Department believes that employees will derive value from this
revision, because certain employees of the employer who may be receiving salaries for
evaluating investments for the employer or for the employer’s human resource department, will
not be deemed to provide fiduciary advice if they provide recommendations to fellow employees
regarding investments in casual conversations.

4.3.2 More Effective Enforcement
By amending the 1975 regulation to broaden the scope of plan services that would be
considered fiduciary investment advice, the final rule enhances the Department’s ability to
redress service provider abuses that currently exist in the market, such as undisclosed fees and
misrepresentation of indirect compensation arrangements that harm participants and
beneficiaries. The Department will also be able to more effectively and efficiently allocate its
enforcement resources. More effective Department enforcement activity will benefit plan
fiduciaries, participants and beneficiaries, both by deterring abuse and improving loss recoveries
when abuse does occur. The final rule and exemptions also will empower other plan fiduciaries
and plan participants to exercise their own legal rights to redress more adviser abuses, thereby
further deterring abuse and improving loss recoveries.

204

205

5. Cost
Summary of section: The Department estimates that the compliance cost associated with
the final rule and exemptions will total between $10.0 billion and $31.5 billion over the first 10
years with a primary estimate of $16.1 billion. These costs are primarily comprised of the costs
incurred by new fiduciary advisers to satisfy relevant PTE conditions. Costs are substantially
higher in the first year as firms first implement changes, then fall in later years. These cost
estimates rely primarily on unverifiable cost estimates submitted by financial services industry
trade groups due to the lack of data from other sources that would present a more neutral
perspective.

5.1

Background

As discussed above, the 2010 Proposal prompted over 200 comments. Several of these
comments asserted that the Department’s economic analysis did not provide a robust estimate of
the likely costs that would be imposed on the financial services industry, particularly brokerdealers (BDs), if the proposal were adopted. On several occasions, the Department requested
data from the regulated community that would allow it to quantify these costs. The
Department’s objective in making such requests was to obtain detailed and reliable data to
support the economic analysis. Generally, the Department did not receive such data.474
In developing the economic analysis for the 2015 Proposal, the Department conducted
thorough research to ascertain whether any relevant data were available that would inform the
Department’s cost estimates. As part of this process, the Department reviewed comments the
SEC received in response to its March 2013 request for data and other information (RFI) relating
to the benefits and costs that could result from various alternative approaches regarding the
standards of conduct and other obligations of broker-dealers and investment advisers.475 The
Department found two comments to be particularly relevant.476 The first comment from SIFMA
provided estimated costs that would be incurred by broker-dealers if the SEC promulgated a
regulation establishing a uniform fiduciary standard pursuant to Section 913 of the Dodd-Frank
Act. These estimates were based on a survey of 18 SIFMA members. The Department used
these data to analyze broker-dealer costs associated with the 2015 Proposal, because of the
similarities between the cost components necessary to satisfy Section 913 of the Dodd-Frank Act
and those necessary to satisfy the requirements of the 2015 Proposal. A second comment from
the Investment Adviser Association (IAA) reported actual costs incurred by its RIA members in
different size categories to comply with the Advisers Act based on a recent survey of investment
advisers. The Department used these data as a basis to adjust the SIFMA-estimated brokerdealer costs for what appeared to be an over-estimate of the cost of compliance. This led to an
estimate of a range for the likely aggregate costs of the 2015 Proposal.
Similar to the 2010 Proposal, the Department received a large volume of comments on
the 2015 Proposal. The Department reviewed and closely considered all of the comments and

474

475
476

FSI sent its Broker-Dealer Financial Performance Study to the Department for several years. The Department used data from the reports
where relevant and appreciates receiving the reports.
SEC Release No. 69013, IA-3558, “Duties of Brokers, Dealers and Investment Advisers,” 2013.
SIFMA Oct. 5, 2011 comment and Investment Adviser Association July 3, 2013 comment on SEC RFI for fiduciary standard; available at:
http://www.sec.gov/comments/4-606/4-606.shtml.

206

hearing testimony that addressed the Department’s methodology for generating the 2015
Proposal’s cost estimates. The Department found two comment letters particularly relevant
because they provided alternative cost estimates for parts of the BD industry to comply with the
2015 Proposal.477 One comment was from SIFMA (SIFMA Report) which engaged Deloitte &
Touche LLP to facilitate a study of the operational impacts of the proposed rule on the financial
services community in conjunction with a working group composed of over 140 senior
operations, technology, and legal professionals from SIFMA member firms. As part of this
project, SIFMA conducted a cost survey of 18 firms that were a part of the working group that
asked them to estimate the compliance costs associated with the 2015 Proposal. Using the same
small, medium, and large firm size breakdown the Department used in the 2015 Proposal, the
SIFMA firms participating in the survey were grouped into nine large firms, four medium firms,
and five small firms. In completing the survey, the firms were asked to consider “Key Cost
Components” such as (1) information technology and suppliers cost, (2) outside legal counsel
and compliance consultant costs, (3) communications, marketing and training costs, (4) costs
associated with reviewing and updating existing client contracts and disclosures, reviewing and
updating sales policies and procedures and surveillance tools, and finalizing editing and
publishing revised marketing materials.
In its comment, SIFMA did not identify the firms that participated in the survey but
indicated that they “represent a diverse business mix; including both clearing and non-clearing
firms and range from full-service broker-dealers to smaller retail oriented broker-dealers.”
Using the SIFMA survey, the surveyed firms indicated that there would be significant costs for
implementation and ongoing maintenance for operations to comply with the 2015 Proposal.478
The report stated that in basing the costs associated with the 2015 Proposal on SIFMA’s
response to the SEC’s RFI, the Department relied on a narrow dataset that never was intended to
measure costs for complying with the 2015 Proposal. According to the report, the SIFMA
working group viewed the results of the new cost survey as a better guide to understanding the
costs associated with the 2015 Proposal, because the survey focused on understanding the
broker-dealer industry’s cost estimates to comply with the 2015 Proposal. Although submitted
as a comment upon which the Department could rely, the report included what appeared to be
Deloitte’s boilerplate disclaimer of reliance, stating that the report “[should not] be used as a
basis for any decision or action that may affect your business.”
In an August 26, 2015, letter to Kenneth E. Bentsen, Jr., SIFMA President and CEO, the
Department requested supplemental information from SIFMA regarding the survey. Specifically,
the Department asked for the following information: (1) the identities of the 18 SIFMA
members that participated in the survey, or a characterization of how closely these firms
generally represent or differ from the industry at large; (2) a description of the methodology used
to select the firms for participation in the survey and any other information that might assist in
assessing the generalizability of the survey findings; (3) any scripts, questionnaires, and/or
survey instruments used to illicit response from the survey participants; (4) raw survey data in

477

478

Deloitte report on behalf of SIFMA, “Report on the Anticipated Operational Impacts to Broker-Dealers of the Department of Labor’s
Proposed Conflicts of Interest Rule Package,” July 17, 2015; available at: http://www.sifma.org/issues/item.aspx?id=8589955444; Oxford
Economics on behalf of FSI, “Economic Consequences of the US Department of Labor’s Proposed New Fiduciary Standard.” August 2015,
available at: http://www.financialservices.org/uploadedFiles/FSI_Content/Advocacy_Action_Center/DOL/FSI-OE-Economic-ImpactStudy.PDF.
Ibid.

207

csv format, or in the absence of such data, descriptive statistics, including means and ranges by
firm size category for each question posed to respondents; (5) whether the survey elicited
separate estimates for costs attributable to the 2015 Proposal’s specific provisions or the “Key
Cost Components” listed in Figure 2.9 of the comment.
Mr. Bentsen responded to the Department in a letter dated September 24, 2015. With
respect to the first request, Mr. Bentsen stated that the survey participants could not be
identified, because they required anonymity as a precondition to participating in the survey. He
did not characterize the firms other than to restate the report’s observations that the firms
represent a diverse business mix, and that the large- and medium-sized firms that completed the
survey are representative of the firms in their respective categories.
With respect to the second request, Mr. Bentsen stated that SIFMA requested the
working group members to complete the survey, and that 18 of the 40 SIFMA member firms
responded. In response to the Department’s request to receive the survey instrument, SIFMA
provided a two page document that set forth the purpose and assumptions of the survey, target
profiles for the survey, and key cost components. The document stated that the purpose of the
survey was “to obtain a high-level firm cost estimate of (1) the total dollar amount it would cost
your firm to build and implement the current DOL Proposal, and (2) the total dollar amount it
would cost for your firm to maintain the current DOL Proposal on a yearly basis.” The
document also said that “[r]esults are based on respondents’ best estimates based on current
understanding of the proposed rule’s requirements,” and cautioned that the “results should not be
relied upon or used for purposes of budgeting and planning.”
Mr. Bentsen stated that SIFMA could not provide raw survey data beyond what was
provided in the report due to SIFMA’s agreement with its members that completed the survey.
He also said the survey enabled, but did not require, participants to submit separate cost
estimates in categories, and that some, but not all, participants provided more granular cost
estimates. He said SIFMA could not provide granular detail on survey responses other than
what is outlined in the report without risking revealing the identity of survey respondents.
Another comment was received from the Financial Service Institute (FSI) which engaged
Oxford Economics to prepare a report (FSI Report) on the impacts of the 2015 Proposal and
exemptions on independent broker-dealers and clearing firms. After reviewing the Department’s
cost estimate for the 2015 Proposal, the comment states that the Department vastly
underestimated the compliance cost for all firms, but especially smaller firms, and questioned
the Department’s assertion that the costs for RIAs will be trivial. Oxford Economics prepared
cost estimates based on interviews with nearly three dozen executives from 12 firms (typically
identified as independent broker-dealer firms) and detailed cost estimates obtained from a survey
completed by seven of those firms who were willing to attempt to break startup costs into
detailed categories. The report did not include ongoing costs.
The FSI Report identified the following cost categories that the 2015 Proposal would
impose on broker-dealer firms: (1) data collection; (2) modeling future costs and returns; (3)
disclosure requirements; (4) recordkeeping; (5) implementing Best Interest Contract Exemption
contracts; (6) training and licensing; (7) supervisory, compliance and legal oversight; and (8)
litigation. It then provided estimates of the likely compliance costs in each category.
In an August 26, 2015, letter to David T. Bellaire, Esq., FSI Executive Vice President
and General Counsel, the Department requested the following supplemental information with
respect to data presented in the FSI report: (1) the type of information that was collected via the
interviews and survey and how each support the report’s cost estimates; (2) the identities of the
financial institutions whose executives were interviewed; (3) a description of the sample
208

captured in the online survey and other information that might assist in assessing the
generalizability of the survey findings; (4) any scripts, questionnaires, and/or survey instruments
used to elicit response from the survey participants; (5) raw survey data in csv format, or in the
absence of such data, descriptive statistics, including means and ranges by firm size category for
each question posed to respondents; (6) whether the survey elicited estimates for the 2015
Proposal’s various specific provisions within the different cost categories.
Mr. Bellaire responded to the Department in a letter dated September 24, 2015. With
respect to the first question, he stated that the cost estimates were derived entirely from the email
survey, and that the interviews informed the email survey because the categories were hashed
out through back-and-forth discussions. Mr. Bellaire did not identify the survey participants. He
characterized the participants as representing a cross-section of FSI’s member independent
financial services firms. Mr. Bellaire provided the initial email survey and the basic script that
organized the interviews with financial institutions to the Department. He did not provide raw
survey data but stated that the report included the overall mean and median for each cost
category and the average total costs by firm size and that the email survey split out costs by
business process required to comply with the proposal, not by provision of the rule.
The Department also received comments that criticized the SIFMA and FSI Reports.
One commenter, responding to the SIFMA report offered the following criticisms:
•

As stated in the report “Deloitte has aggregated and summarized these views, but was
not asked to and did not independently verify, validate or audit the information
presented by the SIFMA Working Group.”

•

While at least some SIFMA members appeared to have submitted detailed cost
estimates, only aggregated total cost estimates were submitted in the report. They
also did not present their underlying assumptions. This lack of transparency raises
questions about the validity of the information.

“[T]he report has no credibility, particularly in light of the gross misrepresentation of the
rule included in SIFMA’s comment letters.” The commenter further argued that misstatements
about the regulatory package’s requirements and associated burdens resulted in a significant
overstatement of costs. The same commenter, responding to the FSI report, offered the
following criticisms:
•

Oxford Economics collected, but did not independently analyze the data to make sure
it was accurate and reliable.

•

The report was not clear on why the particularly firms were chosen to be interviewed
and how representative they were of the industry as a whole

•

There was no explanation of why data were obtained only from seven of the twelve
firms interviewed.

•

Interviewed firms were not identified.

•

The results are highly speculative as respondents to the survey had expressed an
uncertainty about the requirements of the new rule and what the costs would be.

The Department largely shares the commenters’ concerns about both the SIFMA and FSI
reports. These data are problematic due to small sample sizes, selection issues, lack of peer
review, lack of independent verification, and the reports’ omission of details about sample
composition, survey design, and data collection. All of these factors cast doubt on the accuracy
and reliability of these data. (The Department notes that the data from the SIFMA comment
209

letter to the SEC that were used in estimating the costs associated with the 2015 Proposal suffer
from the same issues.) Moreover, the reports do not appear to account for the prospect of market
adjustments that will favor business models and firms that achieve compliance at lower cost.
The Department has particular concerns about the cost estimates provided for small
firms. The SIFMA Working Group chose to exclude small firms from their report of total costs
due to concerns about how representative the small firms in their sample were. While FSI did
not express similar concerns, the same issues seem likely to be present. Small firms are
numerous and very diverse, and it seems particularly unlikely that the three firms that responded
to FSI’s survey that were categorized as small could be representative of the small firm
population as a whole. Additionally, it appears to the Department that the small firms surveyed
by FSI were likely on the large end of the size group, given that the start-up cost estimate
provided - $1,118,000 per small firm - would appear to be exceptionally high for many small
firms to implement the requirements of the 2015 Proposal.
Many small firms have business relations with larger partners that provide them with
compliance services and training. For example, many small firms purchase or lease software
from vendors. The vendors compete for clients in part by ensuring that their products are up-todate and in compliance with the regulatory environment. The software changes necessary to
accommodate new regulations may, in some or many cases, already be reflected in pricing and
considered as part of the service purchased.479
Despite the estimates’ limitations, the Department nevertheless found them useful and
used the cost estimates provided by SIFMA and FSI in developing its methodology to estimate
the costs associated with the final rule and exemptions. For example, the FSI report broke down
the costs into categories, which the Department used to identify parts of the rule that were
particularly burdensome. The Department also used the categories to evaluate certain regulatory
alternatives. However, while the Department has decided to use the two reports’ estimates as
benchmarks for its own cost estimates, it believes caution is warranted. It is likely that the
reports err on the side of overestimating compliance costs.
As discussed in the regulatory impact analysis for the proposal, the Department
originally used the costs submitted to the SEC in response to its RFI to develop its cost
methodology to estimate the costs associated with the proposal. However, for purposes of that
analysis, the submissions to the SEC had many of the same flaws as the SIFMA and FSI reports
submitted to the Department in this rulemaking, as well as the additional flaw of being prepared
in connection with a different rulemaking by a different agency under a different statutory
scheme. Accordingly, the Department has relied on the reports from SIFMA, FSI, and other
effected entities that provided comments and cost estimates specific to the 2015 Proposal. The
Department’s ability to interpret the SIFMA and FSI reports also benefited from other
stakeholders’ comments on the reports. By using these cost estimates to inform the
methodology to estimate the costs associated with the final rule and exemptions, the Department
took a particularly conservative approach. Had it relied on the same SEC submissions that it
relied upon in the proposal, it would have arrived at smaller cost estimates.

479

FSI’s report suggested that many small firms shared this view. However, the report also stated that at least some large technology firms that
support small firms were not as clear that this would be the case.

210

The discussion immediately below addresses the Department’s estimate of the entities
affected by the rule. It then discusses the cost methodology, describes how SIFMA and FSI data
were used to develop the analysis, and concludes with a presentation of the Department’s cost
estimate for the final rule and exemptions.

5.2

Affected Entities

As a first step in estimating the cost of the final rule and exemptions, the Department
estimated the number of firms that will be affected. To improve the accuracy of its estimates,
the Department grouped firms by market segment as in the analysis for the 2015 Proposal. This
section discusses how the Department estimated the number of BDs, RIAs, insurers, and ERISA
plan service providers that are affected by the rule. In an effort to be responsive to comments
and utilize the most current and accurate data available, the Department has used more current
data than was used in the analysis of the 2015 Proposal where possible.
It should also be noted that in its analysis of the proposed rule, the Department did not
quantify the costs that insurers (as opposed to advisers selling insurance products) could incur to
comply with the final and exemptions. During the comment period, however, the Department
received a number of comments suggesting that costs to insurers could be significant and that the
Department had underestimated total costs by not explicitly including in its cost estimates
quantified insurer costs. In response to these comments, the Department has attempted to
quantify the costs that could be incurred by insurers to comply with the final rule and
exemptions, such as the Best Interest Contract Exemption, if they want to continue current
business practices. The exact approach taken by the Department is described in Section 5.2.4
and 5.3.3 below.

5.2.1 Number of BDs
In the analysis of the proposed rule, the Department relied on estimates of the number of
brokerage firms registered with the SEC as of year-end 2013 (4,410 firms) provided to the
Department by the SEC. Several comment letters and studies used the estimates reported in the
2015 Proposal when providing responses without raising serious concerns. One commenter said
it did not generally disagree with those estimates, but thought it would be preferable to have
better data. Since the release of the analysis of the proposal, additional data that are in line with
the previously used estimates, but that are more current, have become available.
As of February 2016, FINRA reported that 3,957 brokerage firms had registered with
The Department is not aware of more accurate data than these numbers, which are
them.
taken from required filings with FINRA, and has utilized them in this final analysis. This
represents a decrease in the number of firms relative to the estimate used in the analysis of the
2015 Proposal.
480

5.2.2 Number of RIAs
The analysis of the proposed rule used data from the SEC Staff Dodd-Frank Study (see
Section 2.6.4) which reported that more than 11,000 RIAs were registered with the SEC and

480

FINRA Newsroom; available at: http://www.finra.org/newsroom/statistics. Last accessed February 29, 2016.

211

more than 15,000 RIAs were registered with the states. Because of a lack of more precise
estimates, the Department then rounded up the total number of RIAs assumed to exist in the
marketplace to 30,000. A more recent report estimated that there were 17,000 RIAs under state
oversight and 10,500 under Federal oversight.481 While the report contains the most recent data
on RIAs under state oversight, the SEC’s 2014 Investment Adviser Information Report contains
more recent data on RIAs under Federal oversight and reports that 10,609 RIAs are registered
with the SEC.482 Thus, the total number of RIAs estimated to be affected by the rule is 27,609,
which represents a decrease in the number of RIAs assumed to exist in the marketplace relative
to the total used in the analysis of the proposal.

5.2.3 Number of ERISA Plan Service Providers
Other service providers,483 primarily for ERISA plans, also will be affected by the final
rule and accompanying exemptions. The analysis of the proposed rule used data from the 2012
Form 5500 Schedule C that showed there were 5,760 such service providers that could be
affected. More recent data from the same source have since become available and have been
incorporated into this analysis. Using data from the 2013 Form 5500 Schedule C, the
Department estimates that there are approximately 6,040 service providers to ERISA-covered
plans that could be affected, representing a slight increase over the number used in the analysis
of the proposal.484

5.2.4 Number of Insurers
Insurers are primarily regulated by the states; therefore, a national level count of insurers
is not collected by regulators. Although state regulators track the insurers operating in their
states, the sum of all insurers cannot simply be calculated by aggregating individual state totals
because insurers often operate in multiple states. A simple sum of state insurer counts would
likely result in counting insurers that operate in multiple states multiple times, leading to an
overestimate of the number of insurers.
However, SNL Financial provides nationwide data on the amount of premiums written in
specified business lines, including individual and group annuities. The data are compiled from
insurance companies’ financial filings with state insurance regulators. Based on review of the
SNL Financial data, it appears that 398 life insurance companies reported receiving either
individual or group annuity considerations in 2014. Accordingly, the Department estimates that
398 or fewer annuity writers could be affected by the rulemaking. The actual number could be
lower than 398 because the SNL Financial data include firms that no longer sell annuities but
continue to receive premiums under closed blocks of business, as well as companies that
reported very small amounts of annuity considerations (approximately 75 companies reported
total annuity considerations lower than $100,000). Also, the SNL data count affiliates

481

482

483

484

North American Securities Administrators Association 2012/2013 Report, “The Pillars of Protection;” available at:
http://www.nasaa.org/wp-content/uploads/2011/08/NASAA-2012-2013-Report.pdf.
The Investment Adviser Information Reports, June 2, 2014, http://www.sec.gov/foia/docs/invafoia.htm#.U6negzYpC73. Note that RIAs
that reported zero assets under management or that did not report a figure were not counted in this total.
In order to provide a reasonable estimate, service providers with reported service codes corresponding to recordkeeping, consulting
(general and pension), insurance agents and brokers, brokerage (real estate), brokerage (stocks, bonds, commodities), valuation appraisals,
participant communications, investment advisory (participants and plans) were used. See section 4.1.1 for more details.
Form 5500 data last accessed Dec. 10, 2015, at: http://www.dol.gov/ebsa/foia/foia-5500.html.

212

separately. Some affiliates operate like one firm and would share compliance costs, while others
operate as two separate firms and would each have separate compliance costs. To be
conservative in the estimates each affiliate is counted as incurring the cost.

5.2.5 Banks
In developing its cost estimates for the final rule, the Department considered the rule’s
impact on banks providing fiduciary investment advice services to ERISA plans, participants,
beneficiaries and IRA investors. There are of 6,182 insured depository institutions in the U.S.485
Of those banks supervised by the OCC, approximately 3 percent are large, 3 percent are midsize,
and 93 percent are community banks.486
For investments in bank products such as Federal Deposit Insurance Corporation (FDIC)
insured deposits, the cost impact of the rule on banks is mitigated by existing ERISA statutory
exemptions.487
The discussion below focuses on banks’ activity in effectuating transactions regarding
retail non-deposit investment products (RNDIPs) including, among other products, equities,
fixed income securities, exchange-traded funds, and variable and fixed rate annuities.
Bank activity related to RNDIPs is subject to various conditions.488 In particular, the
Gramm-Leach-Bliley Act of 1999 (GLBA),489 replaced a blanket exception for banks from
broker registration under the Securities Exchange Act of 1934 with 11 specific exceptions from
broker registration for banks.490 Under these exceptions, certain bank securities activities are
permitted to be conducted within a bank and under bank regulator supervision, whereas others
must be “pushed out” under a “networking arrangement” to a person (e.g., a securities broker or
dealer) that is subject to SEC supervision.491

485

486

487

488

489
490

491

The FDIC reports there are 5,338 and 844 Savings Institutions (thrifts) for a total of 6,182 FDIC- Insured Commercial Banks as of
December 31, 2015. In addition to insured depository institutions, there are an additional 60 uninsured national trust banks regulated by the
OCC that do not show up in the insured commercial bank count. Finally, there are also state-chartered, uninsured trust companies that are
not included in the FDIC’s count as well. Each of these institutions potentially offers retail non-deposit products; available at:
https://www.fdic.gov/bank/statistical/stats/2015dec/industry.pdf.
The OCC supervises national banks and federal savings associations and federal branches and agencies of foreign banks in the United
States. These institutions comprise nearly two-thirds of the assets of the commercial banking system. Comptroller of the Currency 2015
Annual Report; available at: http://www.occ.gov/publications/publications-by-type/annual-reports/2015-annual-report.pdf
See, e.g., ERISA § 408(b)(4); Code § 4975(d)(4). This exemption covers deposits bearing a reasonable rate of interest in a bank or similar
financial institution supervised by the United States or a State. As discussed in Appendix C, the Department understands that according to
available data, small savers appear to hold their IRAs at banks more often than with brokers, and get certain financial advice from banks
more often than brokers. This is based on data from a household survey asking respondent which sources of information they used to make
decisions about savings and investments. The Department believes that households reporting to obtain advice from banks may, in fact,
receive advice regarding a wide range of topics, such as college or emergency savings that do not involve retirement savings. If they
receive advice about retirement savings, such advice may relate to products covered by the statutory exemption.
See Interagency Statement on Retail Sales of Nondeposit Investment Products (Feb. 1994); 15 U.S.C. § 78c(a)(4)(B) (exception from the
term “broker” for certain bank activities); 12 C.F.R. parts 14, 208, 343 and 536 (Consumer Protection in Sales of Insurance).
Pub. Law. No. 106-102, 113 Stat .1338.
15 U.S.C. § 78c(a)(4)(B)(i)-(xi). The 11 exceptions are (i) third-party brokerage arrangements (commonly referred to as “networking
arrangements”); (ii) trust and fiduciary activities; (iii) permissible securities transactions (e.g., U.S. Treasury and U.S. Agencies
obligations); (iv) certain stock purchase plans; (v) sweep accounts; (vi) affiliate transactions; (vii) private securities offerings; (viii)
safekeeping and custody activities; (ix) identified banking products; (x) municipal securities; (xi) a de minimis number of other securities
transactions.
See OCC Comptroller’s Handbook, Retail Nondeposit Investment Products (January 2015), p.17; available at:
http://www.occ.treas.gov/publications/publications-by-type/comptrollers-handbook/o-rnip.pdf.

213

The Department understands that “networking arrangements” are the most widely used
program structure for banks to sell RNDIPs to their clients.492 Under such arrangements, the
bank enters into a contractual or other written arrangement with a registered BD, for example,
under which the broker or dealer offers brokerage services to the bank’s interested customers. In
such arrangements, the BD must be clearly identified as the person performing the brokerage
services in an area that is clearly marked and, to the extent practicable, physically separate from
the routine deposit-taking activities of the bank. Any materials used by the bank to advertise or
promote generally the availability of brokerage services under the arrangement must clearly
indicate that the brokerage services are being provided by the broker or dealer and not by the
bank and be in compliance with the Federal securities laws before distribution.
Under such arrangements, bank employees may not receive incentive compensation for
any brokerage transaction unless such employees are associated persons of a broker or dealer
and are qualified pursuant to the rules of a self-regulatory organization, except that the bank
employees may receive compensation for the referral of any customer that is generally restricted
to a nominal one-time cash fee of a fixed dollar amount, if the payment of the fee is not
contingent on whether the referral results in a transaction.493 The bank employees must perform
only clerical or ministerial functions in connection with brokerage transactions including
scheduling appointments with the associated persons of a broker or dealer, except that bank
employees may forward customer funds or securities and may describe in general terms the
types of investment vehicles available from the bank and BD under the arrangement.494 Bank
employees referring a customer to a broker-dealer under the exception may not provide
investment advice concerning securities or make specific securities recommendations to the
customer under OCC guidance.495 Similar compensation restrictions exist with respect to bank
employees’ referrals regarding insurance products496 and investment advisers.497
Public Call Report Data suggests that about 1,495 banks utilizing networking
arrangements.498 Because of the limitations on the activities of bank employees in making
referrals, the Department believes in most cases such referrals will not constitute fiduciary
investment advice because they will not constitute a “recommendation” within the meaning of

492

493
494
495

496
497

498

See OCC Comptroller’s Handbook, Retail Nondeposit Investment Products (January 2015) (“Most bank RNDIP sales programs rely on the
GLBA exception from broker registration for third-party arrangements with affiliated or unaffiliated securities broker-dealers.”); Federal
Deposit Insurance Corporation. “Uninsured Investment Products: A Pocket Guide for Financial Institutions;” available at:
https://www.fdic.gov/regulations/resources/financial/.
15 U.S.C. § 78c(a)(4)(B)(i)(VI).
15 U.S.C. § 78c(a)(4)(B)(i)(I)-(V).
See Federal Reserve Board and Securities Exchange Commission Release, Definitions of Terms and Exemptions Relating to the ‘‘Broker’’
Exceptions for Banks, 72 FR 56514 (Oct. 3, 2007); see also OCC Comptroller’s Handbook, Retail Nondeposit Investment Products (Jan.
2015).
See 12 C.F.R. parts 14, 208, 343 and 536(Consumer Protection in Sales of Insurance).
See OCC Comptroller’s Handbook, Retail Nondeposit Investment Products (“While the provision of financial planning services and
investment advice to bank customers is not a sale of an RNDIP, the OCC treats these services as if they were the sale of RNDIPs if provided
to bank customers outside of a bank’s trust department. Therefore, if a bank chooses to provide financial planning or investment advice
through an RIA or other provider, in order to provide a high level of customer protection, the bank should meet all of the risk management
standards contained in the Interagency Statement [on Retail Sales of Nondeposit Investment Products] and third-party relationship guidance
contained in OCC Bulletin 2013-29, ‘Third-Party Relationships: Risk Management Guidance.’”)(citing OCC Interpretive Letter #850,
January 27, 1999).
Public bank Call Report data available on the Industry Analysis section of the FDIC’s website https://fdic.gov/bank/ contains the number of
banks that report retail brokerage income (used as a proxy for networking arrangements) as of December 31, 2015. Institutions reporting
retail brokerage income is as follows: 288 national banks, 77 federal thrifts, 1,130 state banks. This data does not include information about
state-chartered, uninsured trust companies that do not file Call Reports.

214

the Final Rule or because they will not involve a covered recommendation to hire a nonaffiliated third party. Nevertheless, to the extent banks and their employees provide fiduciary
investment advice as part of the referral process, the resulting compensation can raise prohibited
transactions issues. To deal with these situations, the Best Interest Contract Exemption provides
an exemption for bank referral compensation if the bank and bank employee(s) satisfy the
impartial conduct standards contained in the exemption when making referrals.499 These
standards would not impose costs on banks beyond an incremental increase in the training and
oversight already required of bank management under the networking arrangement exception.500
The Department is aware that some banks may provide investment advice services using
their own employees to perform activities covered by the other 10 GLBA exceptions, and
therefore the banks could incur costs to comply with the final rule and exemptions. The
Department does not have sufficient data to estimate such costs, because it does not know how
frequently banks use the other 10 exceptions and the extent to which they involve the provision
of investment advice that would be covered by the 2016 Final Rule. It is reasonable to believe
that these banks would incur costs similar to the compliance costs of other firms that must
comply with the 2016 Final Rule and exemptions. However, due to the prevalence of banks
using networking arrangements to effectuate transaction related to RNDIPs, the Department
believes that only a small number of these banks will be affected.

5.2.6 Dividing Firms into Small, Medium, and Large Size
Categories
The Department expects that firms will incur different costs to comply with the
requirements of the final rule and related exemptions based on their size. For example, larger
firms have more employees, more assets, multiple business locations, and more complex
arrangements and systems, all of which increase total costs of compliance even given that such
firms might also experience some economies of scale in compliance. In its analysis of the
proposal, to improve the accuracy of its cost estimates, the Department grouped firms by size
and market segment. The Department has retained this approach in its final analysis.
In the analysis of the proposal, the Department divided BD firms into large, medium, and
small size categories based on a firms’ capital using data reported by the SEC in its 2011
FOCUS Report.501 While there was more than one way to group the firms based on the FOCUS
Report, the Department believes the approach it took was reasonable. First, the Department
categorized firms with capital greater than $1 billion as large, firms with capital between $50

499
500

501

See Best Interest Contract Exemption, Section II(i).
For example, national banks are currently expected to implement an effective initial due diligence process when selecting a third party for
the bank’s RNDIP sales program, as well as adopt an effective ongoing due diligence process to monitor the third-party’s activities, which
may include requiring the third party to provide various reports and provide access to the third-party’s RNDIP sales program records. See
OCC Comptroller’s Handbook, Retail Nondeposit Investment Products; OCC Bulletin 2013-29. In addition, a bank’s management is
responsible for overseeing its vendors regardless of whether they are operating on or off-site. Typical oversight would include reviewing:
(1) the types and volume of products being sold; (2) the number of opened and closed accounts; (3) new products being offered; (4)
discontinued products; and (5) customer complaints and their resolution. See Federal Deposit Insurance Corporation. “Uninsured
Investment Products: A Pocket Guide for Financial Institutions;” available at: https://www.fdic.gov/regulations/resources/financial/.
Under SEC Rule 17a-5, broker-dealers are required to file with FINRA reports concerning their financial and operational status using SEC
Form X-17A-5, also known as a Financial and Operational Combined Uniform Report or “FOCUS” Report. The Department used FOCUS
Report data contained in the SEC’s Financial Responsibility Rules for Broker-Dealers (78 Fed. Reg. 51824, 51869 (Aug. 21, 2013)).

215

million and $1 billion as medium-sized, and firms with less than $50 million in capital as small.
The Department then counted how many firms fell into each of these three size categories in the
2011 FOCUS Report data in order to determine an initial size distribution that could be applied
to the SEC’s more recent (year-end 2013) estimate of the total number of BDs. The FOCUS
data reported a total of 4,709 BDs in the marketplace, while the SEC data reported a lower total
of 4,410. The source of the reduction in the number of firms was not known: some firms
(especially small firms) may have gone out of business, while others may have merged forming
larger firms. Given this uncertainty, the Department took a conservative approach in producing
its estimates of the number of BDs in each size category. Instead of simply applying distribution
percentages derived from the FOCUS Report to the SEC’s estimate of total BDs which would
result in the number of BDs in each size category falling relative to the counts seen in the
FOCUS data, the Department kept the estimate of large and medium firms the same as in the
FOCUS data --42 and 233 respectively. The only size category whose total was adjusted
downward was the small category – the category in which firms’ are assumed to incur the lowest
compliance costs – which fell from 4,434 in the FOCUS data to 4,135.
The Department did not receive any comments opposing this approach, and the
Department continues to use this measure of firm size in this analysis of the final rule and
exemptions, primarily to allow it to use the data that were received in comments, which also
grouped firms using these size categories. As discussed above, however, more recent data on the
number of BDs have become available showing a further decline in the total number of BDs
from 4,410 to 3,957. Consistent with the conservative approach taken in the analysis of the
proposal, the Department again adjusted downward only the number of firms in the small size
category, leaving the total number of firms in the large and medium size categories – those with
the highest anticipated compliance costs – at 42 and 233 respectively. As reported in the first
column of Figure 5-1, this results in an estimate of 42 large firms, 233 medium firms, and 3,682
small firms.
Figure 5-1 Affected Firms

Large BD
Medium BD
Small BD

42
233
3,682

Firms
Serving
ERISA
Plans or
IRAs
42
147
2,320

Large RIAs
Medium RIAs
Small RIAs

113
3,021
24,475

113
1,903
15,419

105
1,877
15,001

Plan SP Large Impact
Plan SP Medium Impact
Plan SP Small Impact

302
906
4,832

302
906
4,832

169
506
2,700

22
175
201

22
175
201

22
175
201

Firms
Divided By
Size

Large Insurers
Medium Insurers
Small Insurers

216

Double
Counting
Removed
42
147
2,320

In its analysis of the proposal, the Department also divided RIAs into large, medium, and
small categories. The Department divided RIAs registered with the SEC into size categories
based on a firm’s assets under management by using a percentage distribution of firms across
three size categories derived from data reported by the SEC in its 2014 Investment Adviser
Information Report.502 While there was more than one way to group the SEC-registered firms
based on this report, the Department believes it took a reasonable approach by categorizing firms
with more than $100 billion in assets under management as large, firms with between $1 billion
and $100 billion in assets under management as medium, and firms with less than $1 billion in
assets under management as small. The Department also assumed that all RIAs registered with
the states are small. The Department did not receive any comments opposing this approach. As
reported in the first column of Figure 5-1, applying this distribution to the updated estimate of
total RIAs in the marketplace (27,609) results in an estimate of 113 large firms, 3,021 medium
firms, and 24,475 small firms.
Dividing ERISA plan service providers into size categories posed challenges. While the
Form 5500 Schedule C contains fee information, it does not include information on all revenue
sources, meaning the filings cannot be used to accurately measure firms’ relative sizes.
Therefore, the Department assumed a size distribution for service provider firms of 5 percent
large, 15 percent medium, and 80 percent small in its analysis of the proposed rule. Using this
distribution represented an attempt by the Department to take a conservative approach to
measuring potential costs among plan service providers, as this distribution is more skewed
towards large firms – which have the highest estimated compliance costs – than the distribution
used for BDs and RIAs. No comments opposing this assumption were received in the comments
on the proposal and the Department continues to use this assumed distribution in its analysis of
the final rule and exemptions. As reported in the first column of Figure 5-1, applying this
distribution to the updated estimate of total plan service providers in the marketplace (6,040)
results in estimates of 302 large firms, 906 medium firms, and 4,832 small firms.
For insurers, their total assets are provided in the SNL financial data. Therefore, the
Department applies size categories to insurers based on assets. Firms with more than $100
billion in assets are considered large, firms with between $1 billion and $100 billion in assets are
medium sized firms, and firms with less than $1 billion in assets are small firms. As reported in
the first column of Figure 5-1, applying this distribution to the estimate of total insurers that may
be affected by the rule (398) results in estimates of 22 large firms, 175 medium firms, and 201
small firms.

5.2.7 Determining the Share of Firms Servicing Plan or IRA
Investors
Not all BDs and RIAs serve ERISA plan or IRA investors. The Investment Adviser
Association found that 41 percent of RIAs responding to a private survey advise ERISA plans or
are pension consultants, while 22 percent advise retail individuals.503 The Department has not

502
503

SEC Investment Adviser Information Reports (June 2, 2014); available at: http://www.sec.gov/foia/docs/invafoia.htm#.U6negzYpC73.
Investment Adviser Association, “2015 Investment Management Compliance Testing Survey,” 2015.
https://www.investmentadviser.org/eweb/docs/Publications_News/Reports_and_Brochures/Investment_Management_Compliance_Testing
_Surveys/2015IMCTppt.pdf.

217

been able to ascertain how much overlap there is between these two groups. To provide a
conservative estimate, the Department assumes that there is no overlap and therefore that 63
percent of RIAs advise either an ERISA plan or IRA investors and will be affected by the rule.504
However, given that large RIAs are more likely to service multiple markets, when
calculating the final number of firms servicing ERISA plans and/or IRA investors, the
Department further conservatively assumes that all large RIAs will service at least one of these
two markets. In other words, the Department assumes that 100 percent of large RIAs will be
affected by the rule, but that 63 percent of medium and small RIAs will be affected. Assuming a
uniform distribution of 63 percent could lead to an underestimate of the number of medium
firms servicing both markets if medium firms are more likely to service both markets relative to
small firms. As reported in the second column of Figure 5-1, the application of these
percentages to the total number of RIAs in each size category results in an estimate of 113 large
firms, 1,903 medium firms, and 15,419 small firms.
Similar data were not available for BDs, so the Department assumes that an identical
percentage of BDs are servicing ERISA plans and/or IRA investors as RIAs. This approach was
used in the analysis of the proposal and the Department did not receive any comments opposing
it. As reported in the second column of Figure 5-1, applying these percentages to the total
number of BDs in each size category results in an estimate of 42 large firms, 147 medium firms,
and 2,320 small firms.
Similar data on the share of insurers that sell annuities to ERISA plans or IRA investors
were also not available. While the Department lacks certainty, it is probable that most of these
insurers sell annuities tot ERISA plans and IRAs. In light of this uncertainty it is assumed that
100 percent of the insurers will be affected by the final rule and exemptions. As reported in the
second column of Figure 5-1, this results in an estimate of 22 large firms, 175 medium firms,
and 201 small firms being affected.
All 6,040 plan service providers are assumed to be affected by the rule given that all are
servicing ERISA plans.

5.2.8 Accounting for Overlap Between Firm Types
Eighteen percent of BDs are dually registered as RIAs. For a dually-registered firm, the
costs for each part of its business are not mutually exclusive; therefore, they can be shared as
both parts of the firm will have to comply with the same regulations and exemptions. This
means that for the purposes of this analysis, dually-registered affected firms should only be
counted once. However, to ensure that no potential costs are omitted, if a firm is dually
registered as both a BD and a RIA, the Department counts it as a BD and assigns it the higher
per firm compliance costs associated with BDs. While both BDs and RIAs will have to make
changes to comply with the final rule and exemptions, the Department expects that BDs will
have higher compliance costs, because RIAs are already subject to a fiduciary standard under
federal securities law and have less conflicted compensation arrangements.

504

This is slightly higher than the percent of RIAs servicing these markets assumed in the analysis of the proposal (59 percent) because a more
recent version of the same survey is being utilized in this analysis.

218

To execute this adjustment, the Department first calculated how many BDs in each size
category can be expected to also be registered as RIAs by multiplying the number of BDs in
each category by the overlap percentage (18 percent). The resultant numbers were then
subtracted from the number of RIAs in the same size categories. As reported in the third column
of Figure 5-1, this results in an estimate for RIAs of 105 large firms, 1,877 medium firms, and
15,001 small firms.
There is also significant overlap between service providers for ERISA plans as reported
on Schedule C of the Form 5500 and BDs, RIAs, and insurers. To account for this overlap,
service providers reporting codes that relate to these categories in the Form 5500 data are
removed from the service provider counts.505 This effectively results in service providers that
are dually registered either as BDs, RIAs, or insurers being counted as BDs, RIAs, or insurers –
groups with higher anticipated compliance costs – in this analysis. As reported in the third
column of Figure 5-1, this removal of double counting results in an estimate for service
providers of 169 large firms, 506 medium firms, and 2,700 small firms.
Based upon the data available at the time of this analysis, the Department was not able to
determine what percentage of insurers selling annuity products might also be registered as BDs
or RIAs. Consequently, while the Department does know that there is overlap between these
groups, for the purposes of this analysis, the Department has nevertheless conservatively left the
number of insurers in each size category unchanged.

5.3

Methodology for Cost Estimates

As discussed above, the Department is primarily using cost data provided in the SIFMA
and FSI comments on the 2015 Proposal to develop cost estimates for the final rule and
exemptions. The Department notes from the outset that relying on the cost estimates provided
by SIFMA and FSI has both strengths and weaknesses. With respect to strengths, these groups
are potentially in the best position to provide data regarding the compliance costs associated with
the final rule and exemptions, because their members are directly engaged in the affected lines of
business. Potential weaknesses include the following: (i) the Department does not have access
to detailed information about the design, content, execution and implementation of the surveys,
interviews, and discussions upon which these data are based; (ii) the SIFMA survey had 18
participants and FSI had seven, and the Department does not have sufficient information to
determine whether these firms are representative of the affected industry as a whole; (iii) the
costs estimates are reported at an aggregate level, which makes it difficult to estimate the cost
impact of changes the Department made in the final rule and exemptions; (iv) SIFMA reported
small firm costs, but did not include them in their aggregate cost estimate, because they stated
that costs for small firms may not be representative; and (v) the disclaimer of reliability by
Deloitte. To compensate for these weaknesses, it was necessary to exercise judgment informed
by the totality of the public record (including comments on the SIFMA and FSI survey reports),
and to make certain adjustments. The adjustments, detailed below, were necessary to reflect the

505

One of the codes used was if investment advisory services were reported. Due to the construction of the codes, firms other than RIAs or
BDs could be offering investment advisory services. To avoid removing other types of firms 25 Schedule C filings were reviewed. It was
found that 20 percent of service providers with the related code were not BDs or RIAs. Therefore, when removing overlap between BDs,
RIAs, and other service providers, the number of unique firms reporting investment advisory was adjusted by a factor of 0.80.

219

scope of the universe affected by the final rule and exemptions and the changes to the 2015
Proposal that are embodied in the final rule and exemptions.
Based on the results of a survey of 18 of its members, SIFMA estimated per firm average
start-up costs of $38.1 million, $23.1 million, and $3.4 million for large, medium, and small
firms respectively, and annual average ongoing costs of $9.5 million, $5.0 million, and $2.6
million for large, medium, and small firms respectively using the same firm sized categories the
Department used in its analysis of the 2015 Proposal.506
Based on the results of a survey of seven of its members, FSI estimated per firm average
start-up costs of $16.3 million, $3.4 million, and $1.1 million for large, medium, and small firms
respectively using the same firm size categories the Department used in its analysis of the 2015
Proposal. In contrast to the SIFMA cost analysis, the FSI analysis reported no estimates for
ongoing costs, but did offer estimates of costs in several detailed categories.
There are significant differences in the SIFMA and FSI cost estimates. First, the SIFMA
estimates are two, seven, and three times greater for large, medium, and small firms respectively
than FSI’s. There are many reasons why the estimates could be so varied. The estimates were
based on small sample sizes (18 SIFMA members and seven FSI members) and the members of
each organization are different: SIFMA comprises broker-dealers, banks, and asset managers
while FSI primarily comprises independent broker-dealers. SIFMA and FSI members may also
have had different understandings of the 2015 Proposal’s requirements, and the requirements
may have different impacts on them. Within the firm size categories, the firm size on average
could be significantly larger for SIFMA than FSI. The Department attempted to obtain
information that may have shed light on these issues by sending supplemental letters to SIFMA
and FSI, but no additional data were provided to the Department that would have allowed it to
better understand the cause of the significant differences in their estimates.507
Another important difference between the estimates can be found in their estimation of
compliance costs for small firms. While FSI offered cost estimates for small firms and
incorporated them into their calculation of total costs, the SIFMA report stated, “The Working
Group chose to exclude small firms from this exercise because of the broad and diverse make-up
of the industry’s small firm population, which the SIFMA survey respondents may not have been
representative of.” Consequently, in SIFMA’s own calculations of the total compliance costs
associated with the proposed rule, SIFMA declined to incorporate their small-firm cost estimates
into their total cost estimate. Given these facts, the Department did not believe it could
justifiably utilize SIFMA’s estimates of small-firm costs either and has also left these costs out
of all of the cost estimates calculated using SIFMA figures. Because of the methodology utilized
by the Department and described in detail below, this means that small-firm compliance costs

506

507

SIFMA members include securities firms, banks, and asset managers. While the report says it represents the views of the BD industry, at
least some of SIFMA’s membership are or have affiliates that are RIAs or insurers, so it is plausible the comments also could speak to these
industries as well.
Fidelity Investments and TIAA-CREF reported start-up costs for the Best Interest Contract Exemption to be $46 million and $24.7 million,
with annual ongoing costs to be $18 million and $37.8 million respectively. (http://www.dol.gov/ebsa/pdf/1210-AB32-2-00540.pdf and
http://www.dol.gov/ebsa/pdf/1210-AB32-2-03089.pdf). Northwestern Mutual listed its total cost to implement the proposed rulemaking to
be $13-15 million for start-up costs and $3-4 million for ongoing costs (http://www.dol.gov/ebsa/pdf/1210-AB32-2-0076.pdf). These costs
were not used in the cost estimation. Fidelity’s estimate was focused solely on the annual disclosure. The TIAA-CREF estimate was
different from all others as it had annual ongoing cost higher than start-up costs, but provided no explanation as to why that was the case.
All three firms are members of SIFMA and could be part of the SIFMA costs estimates already.

220

have been excluded from the SIFMA-based estimates of exemption compliance costs for BDs,
the subset of RIAs that will comply with exemptions, and the subset of insurers that will comply
with exemptions. However, all FSI-based compliance cost estimates for BDs, RIAs, and insurers
do incorporate small-firm costs, as do all cost estimates that are calculated not using SIFMA- or
FSI-submitted data, such as the costs incurred by BDs as a result of rising insurance premiums.
A final key difference between SIFMA and FSI cost estimates already mentioned above
is that SIFMA provided estimates of start-up and ongoing costs, while FSI only provided start-up
cost estimates. In order to estimate ongoing costs for the FSI data, the Department applied the
ratio of start-up to ongoing costs in the SIFMA data by firm size category to the FSI data based
on the assumption that the ratios would be the same for both groups. For example, to produce an
estimate of large-firm ongoing costs based on FSI data, the Department first calculated the ratio
of large-firm ongoing costs to large-firm start-up costs in the SIFMA data ($9,500,000 /
$38,100,000 = 0.249) and then multiplied this ratio by FSI’s estimate of large-firm start-up costs
($16,266,000) to arrive at a large-firm ongoing cost estimate of $4,056,000. However, because
the Department is not utilizing SIFMA’s estimate of small firm costs, to produce an FSI-based
estimate of small-firm ongoing costs the Department simply used the ratio of FSI small firm
start-up costs to FSI medium firm start-up costs to adjust the newly calculated estimate of
medium firm ongoing costs. The results of these calculations can be seen in column d of Figure
5-1.
Based on the foregoing, the start-up and ongoing costs for SIFMA and FSI are
summarized in Figure 5-2 and Figure 5-3 below. As noted above, the estimated number of BD
firms has decreased from the estimate used in the 2015 Proposal. The estimates of total costs
reported by SIFMA and FSI as shown in the figures are adjusted to account for the updated
number of firms.
Figure 5-2 Start-up Costs Using Updated Number of Firms
# of Firms

Firm Size
Category

SIFMA
Per Firm Average Costs

Total Costs

FSI
Per Firm Average Costs

Total Costs

(a)

(b)

(c)

(d)

(e)

Large BD

42

Medium BD

147

$38,100,000
$23,100,000
NA

$1,600,200,000
$3,395,700,000
NA

$16,266,000
$3,350,000
$1,118,000

$683,172,000
$492,450,000
$2,593,760,000

2,320

Small BD
Total

2,509

$4,995,900,000

$3,769,382,000

Figure 5-3 Ongoing Costs Using Updated Number of Firms
Firm Size
Category
Large BD
Medium BD
Small BD
Total

# of Firms
(a)

SIFMA

FSI (DOL Estimated)
(d)

(b)

(c)

Per Firm Average Costs

Total Costs

Per Firm Average Costs

Total Costs

42

$9,500,000

$399,000,000

$4,055,827

$170,344,724

147
2,320

$5,000,000
NA

$735,000,000
NA

$725,108
$241,991

$106,590,909
$561,419,913

2,509

$1,134,000,000

221

(e)

$838,355,547

The FSI data also presented firm mean and median costs grouped into specific cost
categories. The medians and means presented were for the entire surveyed population. Figure
5-4 below provides a summary of the cost categories and the mean and median value for each of
the categories.
Figure 5-4 FSI Categories of Costs Related to the Proposed Rules Requirements
Cost Category

Mean

Median

(a)

(b)

$570,000

$250,000

Not estimated

Not estimated

Disclosure Requirements

$870,000

$870,000

Record Keeping (Data Retention)

$200,000

$150,000

$4,500,000

$400,000

Training and Licensing

$800,000

$580,000

Supervisory, compliance, and legal oversight

$210,000

$138,000

Not estimated

Not estimated

Data Collection
Modeling Future Returns and Costs

Implementing BICE Contracts

Litigation
Total

$7,150,000

Source: Oxford Economics on behalf of Financial Services Institute, “Economic
Consequences of the US Department of Labor’s Proposed New Fiduciary Standard.”
August 2015.

There is a discrepancy between the estimate of total cost in and the FSI reported total
firm cost reported in Figure 5-2. As show in Figure 5-4, adding up the individual averages of the
cost categories results in an estimate of $7.15 million in total costs per firm. However,
calculating the weighted average per firm cost using the average per firm total costs by firm size
presented in Figure 5-2 yields a $7.9 million per firm average cost ((3/7)*$16.3 million +
(1/7)*$3.4 million + (3/7)*$1.1 million).508 The difference between these two estimates
suggests that either some costs that were included in the total per-firm average costs by size
categories were not included in the cost categories or that not all firms reported costs in every
category. The Department does not have sufficient data to determine why the discrepancy
occurred.
The FSI report did not present costs for the category of modeling future returns and costs.
The final rule and exemptions do not include this requirement, so not having a cost estimate for
this category does not pose a problem for estimating costs associated with the final rule and
exemptions. The costs of litigation are discussed separately in Section 5.4.

508

The FSI estimates were based off of surveys of three large firms, one medium firm, and three small firms, which is why the cost totals are
weighted as such.

222

5.3.1 Estimating BD Firm Costs for Complying With Final Rule and
Exemptions
The cost categories provided in the FSI estimate provide a useful framework for the
Department to estimate the total costs associated with the final rule and exemptions. However,
in order to use this framework to determine total costs, the Department had to develop a method
to break the reported mean and median costs for each cost category for the seven surveyed firms
into per firm average costs for the small, medium, and large firm size categories. Therefore, the
Department developed a methodology to use the sample universe mean and median cost
estimates to derive estimates of the mean firm costs for small, medium, and large firms based on
the following assumptions: (1) all firm costs should be correlated with firm size and the reported
median cost should be the per firm cost for the one medium-sized firm in the sample; (2) the
average cost for small firms cannot be larger than the median cost; (3) the ratio of large firm
average total start-up costs to small firm average total start-up costs must be preserved; and (4)
the average of the estimated per firm costs for each category of cost must equal the reported
average costs. Figure 5-5 presents the results that meet those four conditions.509 It should be
noted that when the individual cost category estimates for large, medium, and small firms
produced using this method are added up, the resultant total per firm costs do not match the
average total costs by firm size reported by FSI due to the reporting discrepancies discussed
above.
Figure 5-5 Calculations of Average Costs by Firm Size
Large BD
(a)
3
42
$1,106,265
$1,544,038
$369,741
$9,966,667
$1,484,880

Medium BD
(b)
1
147
$250,000
$870,000
$150,000
$400,000
$580,000

Small BD
(c)
3
2,320
$140,402
$195,962
$46,926
$400,000
$188,453

$393,996

$138,000

$50,004

Per Firm Total Cost Using this Method

$14,865,587

$2,388,000

$1,021,746

Industry Total Costs Using this Method

$624,354,653

$351,036,000

$2,370,451,548

Firm Size Category
Number of Firms in Sample
Number of Firms in Universe
Data Collection
Disclosure Requirements
Record Keeping (Data Retention)
Implementing BICE Contracts
Training and Licensing
Supervisory, Compliance, and Legal
Oversight

509

Total
(d)
7
2,509

$3,345,842,201

Estimates are developed using the following method and in the following order 1) medium firms, 2) small firms, 3) large firms. Define: a =
3 = the number of large firms in the FSI sample; b = 1 = the number of medium size firms in the FSI sample; c = 3 = the number of small
firms in the FSI sample; d = 7 = the total number of firms in the FSI sample; p = $16,266,000 = large firm average total costs in the FSI
sample; q = $1,118,000 = small firm average total costs in the FSI sample; m = median costs at the component level in the FSI sample; and
n = average cost at the component level in the FSI sample. The values of interest are x, y, and z, where x = large firm average costs at the
component level, y = medium firm average costs at the component level, and z = small firm average costs at the component level. Assume y
= m. Conceptually, for each component, there are two equations, (a*x + b*y + c*z = d*n and x/z = p/q), and two unknowns, (x and z).
Solving yields x = (d*n-b*y-c*z)/a and z = (d*n-b*y)/(c + a*p/q). This solution, however, results in an inappropriate estimate for z (namely
z>y) in one component (implementing Best Interest Contract Exemption contracts). Thus, for all components, while the above formula is
used to calculate x, the following is used for z: z = min[r*(d*n-b*y)/(c + a*p/q),m]. In this formula, r is defined to be the scalar, 1.7512, that
yields values for z such that ∑x/∑z = p/q. Consider data collection costs, for which m = $250,000 and n = $570,000. Then y = $250,000, z
= 1.7512*(7*$570,000 -1*$250,000)/(3+3*$16,266,000/$1,118,000) = $140,404, and x = (7*$570,000-1*$250,000-3*$140,404)/3 =
$1,106,265.

223

The Department cannot, however, directly use these estimates derived from the FSI
report or the cost estimates provided in the SIFMA report to estimate the costs for BDs
associated with the final rule and exemptions. This is due to the fact that these cost estimates
assessed the cost of the 2015 Proposal, and, in response to comments, the Department has made
substantial changes to the proposal in the final rule and exemptions that reduce costs, facilitate
compliance, and enhance workability. To reflect uncertainty in how much cost reduction has
occurred due to the changes made in the final rule and exemptions, the Department has
developed a high, medium, and low estimate of the cost reductions of the final rule and
exemptions relative to the proposed rule and exemptions.
Figure 5-6 provides a summary of the estimates of the reduction in costs by category, and
the discussion below describes the Department’s rationale for estimating these percentage cost
reductions. It should be noted that these adjustments are intended solely to account for changes
to the final rule and exemptions. These changes do not adjust for any overestimate or upward
bias in the original cost estimate. Any upward bias that existed in the original cost estimates
remains in the final estimates, leading to what is believed to be an overestimate of per firm costs,
perhaps by a significant amount.
Figure 5-6 Percent Reduction of Costs of Final Rule Relative to the 2015 Proposal Estimates by FSI

60%
70%
0%
50%
0%

Medium
Assumptions of
Percent Reduction in
Costs
(b)
Not Estimated
70%
80%
0%
60%
10%

0%

10%

Low Assumptions of
Percent Reduction in
Costs
(a)
Modeling Future Returns and Costs
Disclosure Requirements
Data Collection
Record Keeping (Data Retention)
Implementing BICE Contracts
Training and Licensing
Supervisory, Compliance, and Legal
Oversight

High Assumptions of
Percent Reduction in
Costs
(c)
80%
90%
0%
70%
20%
20%

Not Estimated

Litigation

Disclosure Requirements: The Department has substantially revised the disclosure
requirements that were in the 2015 proposed Best Interest Contract Exemption and the Proposed
Principal Transactions Exemption to make compliance less burdensome and costly. The changes
made to each disclosure are discussed below.
Best Interest Contract Exemption Pre-transaction Disclosure (modified): The
proposed pre-transaction disclosure would have required a retirement investor to receive a chart
setting forth the “total cost” of the recommended investment for 1-, 5- and 10- year periods,
expressed as a dollar amount, assuming an investment of the dollar amount recommended by the
Adviser and reasonable assumptions about investment performance. A number of commenters
raised significant objections to the proposed transaction disclosure. They generally indicated the
disclosure would be costly to implement and an extensive transition period would be required for
financial institutions to comply. In this vein, several commenters stated that financial institutions
224

do not currently assemble or maintain all of the required information and that current systems
could not deliver the disclosures. Thus, costly system modifications would be necessary for
financial institutions to obtain the data necessary to comply with the requirement. In addition,
the required disclosure required an individually tailored mathematical calculation for each
recommendation, including projections and accounting for the present value of different costs
paid at different times.
In response to the commenters, the Department has significantly revised the transaction
disclosure requirement in the final exemption to reduce burden, focus on the most salient
information about the contractual relationship and material conflicts of interest, and eliminate the
highly tailored calculation above. More detailed disclosures are required only upon request to
retirement investors who are interested in receiving such detail. The final exemption requires the
transaction disclosure to:
•

State the Best Interest standard of care owed by the adviser and financial institution to
the retirement investor; and disclose any material conflicts of interest with respect to
the recommended transaction;

•

Inform the retirement investor that the investor has the right to obtain copies of the
financial institution’s written description of its policies and procedures, as well as
specific disclosure of costs, fees and other compensation associated with the
purchase, sale, exchange and holding of the investment product, including third-party
payments regarding direct and indirect compensation payable to the adviser and
financial institution in connection with the recommended transaction.510 The
information required under this section must be provided to the retirement investor
before the transaction, if requested prior to the transaction, and if the request occurs
after the transaction, the information must be provided within 30 business days after
the request; and

•

Advise the retirement investor of the financial institution’s website required by the
exemption and its address, and inform the retirement investor that: (i) model contract
disclosures updated as necessary on a quarterly basis are maintained on the website,
and (ii) the financial institution’s written description of its policies and procedures are
available free of charge on the website.

These disclosures do not require individually tailored calculations, permit the preparation
of standardized materials that can be uniformly presented to numerous investors, and do not have
to be repeated for subsequent recommendations by the adviser and financial institution of the
same investment product within one year, unless there are material changes in the subject of the
disclosure.
As revised, the pre-transaction disclosure adopted in the final exemption involves
significant reductions in burden and costs, compared with the proposed pre-transaction
disclosure. The proposal would have required a customized disclosure for each recommended
investment and the adviser and financial institution would have been required to calculate cost

510

The costs, fees, and other compensation may be described in dollar amounts, percentages, formulas, or other means reasonably designed to
present materially accurate disclosure of their scope, magnitude, and nature in sufficient detail to permit the retirement investor to make an
informed judgment about the costs of the transaction and about the significance and severity of the material conflicts of interest.

225

projections based on the retirement investor’s dollar amount, and convert the costs into dollar
figures over the three holding periods. In comparison, the final pre-transaction disclosure is
more general and requires more specific information only to be provided upon request. Even if
more specific information is requested, the final exemption does not require calculation of a
specific amount expressed in dollars, but rather allows the information to be disclosed in dollar
amounts, percentages, formulas, or other means reasonably designed to present materially
accurate disclosure of their scope, magnitude, and nature in sufficient detail to permit the
retirement investor to make an informed judgment about the costs of the transaction and about
the significance and severity of the material conflicts of interest.
Best Interest Contract Exemption Annual Disclosure (eliminated): In addition to the
pre-transaction disclosure, the proposed exemption would have required the adviser or Financial
Institution to provide the following written information to the retirement investor, annually,
within 45 days of the end of the applicable year, in a succinct single disclosure:
(1) A list identifying each asset purchased or sold during the applicable period and the
price at which the asset was purchased or sold;
(2) A statement of the total dollar amount of all fees and expenses paid by the plan,
participant or beneficiary account, or IRA (directly and indirectly) with respect to
each asset purchased, held or sold during the applicable period; and
(3) A statement of the total dollar amount of all compensation received by the adviser
and financial institution, directly or indirectly, from any party, as a result of each
asset sold, purchased or held by the plan, participant or beneficiary account, or IRA
during the applicable period.
The Department received numerous comments expressing concerns about the burden,
cost, and utility of the annual disclosure requirement. In response to such comments, the
Department did not adopt the annual disclosure requirement in the final exemption. The
Department is confident that the elimination of the annual disclosure results in a substantial cost
reduction from the proposed annual disclosure. The potential magnitude of the reduction was
illustrated in a comment from one of the world’s largest financial services providers. The
commenter estimated that it would incur total cost to implement the annual disclosure
requirement of more than $46 million the first year and more than $18 million dollars annually
thereafter based on a detailed cost assessment process for each affected area of its business.511
The Department notes that these cost estimates exceed the anticipated costs for large firms to
comply with the rule and exemptions that were contained in the SIFMA report. The commenter
noted that some of the magnitude of the expense is related to the firm’s large size, but many of
the same expenses would be incurred by smaller firms as well.
Best Interest Contract Exemption Website Disclosure (modified): The proposed
exemption would have required a Financial Institution to maintain a Web page, freely accessible
to the public, which shows the following information:

511

The comment states that many employees participated in the cost assessment process including, among others, personnel in finance,
technology, risk and compliance, product management, analytics, digital communications, distribution services, and platform support.

226

(A) The direct and indirect material compensation payable to the adviser, financial
institution and any affiliate for services provided in connection with each asset (or,
if uniform across a class of assets, the class of assets) that a plan, participant or
beneficiary account, or an IRA is able to purchase, hold, or sell through the adviser
or financial institution, and that a plan, participant or beneficiary account, or an IRA
has purchased, held, or sold within the last 365 days. The compensation may be
expressed as a monetary amount, formula or percentage of the assets involved in the
purchase, sale or holding; and
(B) The source of the compensation, and how the compensation varies within and
among assets.
The financial institution's website would have been required to provide access to the
information described in (A) and (B) above in a machine readable format.
The Department’s intent in proposing the web disclosure was to provide broad
transparency about the pricing and compensation structures adopted by financial institutions and
advisers. The Department contemplated that the data could be used by financial information
companies to analyze and provide information comparing the practices of different advisers and
financial institutions. This information would allow retirement investors to evaluate and
compare the practices of particular advisers and financial institutions.
A number of commenters viewed the proposed web disclosure as too costly, burdensome,
and unlikely to be used by IRA investors, or expressed confidentiality and privacy concerns. In
particular, commenters opposed disclosure of adviser-level compensation. A few commenters
misinterpreted the proposal to require disclosure of the precise total compensation amounts
earned by each individual adviser, and strongly opposed such disclosure. Other commenters
took the position that the requirements of the proposed web disclosure would violate other legal
or regulatory requirements applicable to advertising, and antitrust law.
The Department has reworked the final web disclosure requirement to be based on a
more principles-based approach to address commenters’ concerns. The Department accepted the
suggestion of a commenter that the web disclosure should contain: a schedule of typical account
or contract fees and service charges, and a list of product manufacturers with whom the financial
institution maintains arrangements that provide payments to the adviser and financial institution,
and a description of the arrangements and their impact on adviser compensation. Another
commenter suggested that the Department require disclosure of the financial institution’s
business model and the material conflicts of interest associated with the model. The commenter
further suggested the Department should require disclosure of the financial institution’s
compensation practices with respect to advisers, including payout grids and non-cash
compensation and rewards. The Department has adopted these suggestions as well.
With respect to the level of detail required, the Department has modified the web
disclosure requirement by providing financial institutions with considerable flexibility regarding
how best to present the information subject to the following principle: the website must “fairly
disclose the scope, magnitude, and nature of the compensation arrangements and material
conflicts of interest in sufficient detail to permit visitors to the website to make an informed
judgment about the significance of the compensation practices and material conflicts of interest
with respect to transactions recommended by the financial institution and its advisers.”
In response to comments, the final web disclosure requirement also reduces cost and
burden by permitting financial institutions to rely on other public disclosures, including those
227

required by the SEC and/or the Department to provide information required by the exemption by
posting them to its website.512
The Department is confident that the revision to the web disclosure requirement in the
final exemption will result in significant cost savings. The proposed web disclosure required the
financial institution to calculate and disclose compensation payable to itself, its advisers and its
affiliates with respect to each asset recommended or a class of assets. In the final exemption, the
Department reduced this burden by minimizing the specificity of the information provided. The
financial institution must disclose its third-party compensation arrangements with investment
product providers, and its compensation and incentive arrangements with advisers. The final
exemption allows such disclosures to be grouped together based on reasonably-defined
categories of investment products or classes, product manufacturers, advisers, and arrangements,
and financial institutions may disclose reasonable ranges of values, rather than specific values,
as appropriate. The final exemption also makes clear that individual adviser compensation is not
required to be disclosed. The final exemption also did not adopt the requirement that the
information in the web disclosure be made available in machine readable format.
Principal Transactions Exemption Disclosures – The proposed Principal Transactions
Exemption would have required financial institutions to provide a pre-transaction disclosure that
included pricing information about the security to be purchased or sold in the principal
transaction. The pricing information included two price quotes with respect to the transaction,
obtained from ready and willing non-affiliated counterparties, as well as the mark-up or markdown to be charged with respect to the transaction. The mark-up or mark-down also would have
been required to be disclosed in the confirmation statement and proposed annual disclosure.
Commenters stated that the disclosure of the two price quotes and mark-up or mark-down on the
transaction would be difficult to implement operationally. In the final exemption, the
Department retained the pre-transaction disclosure, confirmation statement and annual
disclosure, but did not require the disclosures to include these specific items of pricing
information.
Based on the foregoing – particularly including the elimination of the requirements in the
Best Interest Contract Exemption for highly-tailored dollar cost calculations at the point of sale,
elimination of the annual disclosure requirement, and the less burdensome approach to web
disclosure, as well as the revision to the disclosure in the Principal Transactions Exemption - the
Department’s high, medium, and low estimates of the percentage cost reductions related to the
changes in the disclosure requirements are 80 percent, 70 percent, and 60 percent, respectively.
Data Collection – The proposed Best Interest Contract Exemption would have required
financial institutions to collect and maintain detailed data relating to inflows, outflows, holdings,
and returns for retirement investments for six years from the date of the applicable transactions
and to provide that data to the Department upon request within six months. The Department
reserved the right to publicly disclose the information provided on an aggregated basis, although

512

These commenters argued that the information required to be disclosed as part of the exemption may already be part of other existing
disclosures, such as those provided pursuant to ERISA Sections 404(a)(5) and 408(b)(2) and the SEC’s required mutual fund summary
prospectuses and Form ADV. The Department has accepted these comments insofar as the information required disclosed pursuant to other
requirements also satisfies the conditions of the exemption, and so long as the Financial Institution provides an explanation that the
information can be found in the disclosures and a link to where it can be found.

228

it made clear it would not disclose any individually identifiable financial information regarding
retirement investor accounts.
Commenters expressed concerned about the burden and costs of obtaining and
maintaining the necessary data and responding to the Department within the timeframe set forth
in the proposal. In response to the comments, the Department eliminated the data request in its
entirety, and leaving in place only the more streamlined general recordkeeping requirement that
requires financial institutions to maintain for six years records necessary for the Department and
certain other entities – including plan fiduciaries, participants, beneficiaries, and IRA owners –
to determine whether the conditions of the exemption have been satisfied. In addition, the final
exemption eliminated the proposed annual disclosure requirement, and the pre-transaction and
web disclosures were revised in a way that eliminates much of the need for data collection by the
firm.
The net effect of all these changes is a large reduction in the data firms have to collect to
comply with data retention and disclosure requirements. Based on the foregoing, the
Department’s high, medium, and low estimates of the percentage cost reductions related to the
changes in the data collection requirements are 90 percent, 80 percent, and 70 percent
respectively.
Recordkeeping (Data Retention) – This provisions of the Best Interest Contract
Exemption and the Principal Transactions Exemption requires the financial institution to
maintain for six years records necessary for the Department and certain other entities, including
plan fiduciaries, participants, beneficiaries and IRA owners, to determine whether the conditions
of the exemption have been satisfied. These records include, for example, records concerning
the financial institution’s incentive and compensation practices for its advisers, the financial
institution’s policies and procedures, any documentation governing the application of the
policies and procedures, the documents required to be prepared for proprietary products and
third-party payments, contracts entered into with retirement investors that are IRAs and nonERISA plans, and disclosure documentation. The Department believes that it will not be costly
for firms to comply with this requirement, because they already maintain records similar to those
required under the final exemption as part of their usual and customary business practices.
The requirement to maintain the records necessary to determine compliance with the
exemption both encourages thoughtful compliance and provides an important means for the
Department and retirement investors to assess whether financial institutions and their advisers
are, in fact, complying with the exemption’s conditions and fiduciary standards. Although the
requirement does not lend itself to the same sorts of statistical and quantitative analyses that
would have been promoted by the data collection requirement, it too assists the Department and
retirement investors in evaluating compliance with the exemption, but at substantially less cost.
The Department’s high, medium, and low estimates of the percentage cost reductions related to
the changes in the recordkeeping requirements are all zero.
Implementing Contracts – The Department has made several burden reducing changes
to the contract requirement in the proposed Best Interest Contract Exemption and proposed
Principal Transactions Exemption. The contract need not be executed by each person at a firm
that renders advice to the customer; does not have to be executed before each instance of advice;
and can be incorporated in existing contracts at the time of the transaction or entered into through
a negative consent process for existing customers. Additionally, the contract requirement has
been eliminated for advice to ERISA plans and plan participants.
For example, for “new contracts,” the final exemptions provide flexibility for the contract
terms to occur in a standalone document or in an investment advisory agreement, investment
229

program agreement, account opening agreement, insurance or annuity contract or application, or
similar document, or amendment thereto. For retirement investors with “existing contracts,” the
final exemptions permit assent to be evidenced either by affirmative consent, as described above,
or by a negative consent procedure in which the financial institution delivers a proposed contract
amendment to the retirement investor prior to January 1, 2018, and deems the retirement
investor’s failure to terminate the amended contract within 30 days as assent. An existing
contract is defined in the exemptions as “an investment advisory agreement, investment program
agreement, account opening agreement, insurance contract, annuity contract, or similar
agreement or contract that was executed prior to January 1, 2018, and remains in effect.” If the
financial institution elects to use the negative consent procedure, it may deliver the proposed
amendment by mail or electronically, but it may not impose any new contractual obligations,
restrictions, or liabilities on the retirement investor by negative consent. The final Best Interest
Contract Exemption additionally provides a method of complying with the exemption in the
event that the retirement investor does not open an account with the adviser but nevertheless acts
on the advice through other channels.
Moreover, the Department has clarified the required timing of the contract. As proposed,
the exemptions generally required that, “[p]rior to recommending that the Plan, participant or
beneficiary account, or IRA purchase, sell or hold the asset, the adviser and financial institution
enter into a written contract with the retirement investor that incorporates the terms required by
[the exemption]….” A large number of commenters responded to various aspects of this
proposed requirement. Many commenters objected to the timing of the contract requirement.
They said the timing “prior to” any recommendations would be contrary to existing industry
practices. The commenters indicated that preliminary discussions may evolve into
recommendations before a retirement investor has decided to work with a particular adviser and
financial institution. Requiring a contract upfront would chill such preliminary discussions,
marketing, and an investor’s ability to shop around, commenters said. Many commenters on this
subject suggested that the proper timing of the contract would be before execution of the
investment transaction at issue. Several commenters that strongly supported the contract
requirement nevertheless agreed that the timing could be adjusted without loss of protection to
the retirement investors.
In the Department’s view, the critical aspect of the requirement for IRAs and non-ERISA
plans is that all instances of advice be covered by an enforceable contract. Therefore, the
Department adjusted the exemptions by deleting the specific requirement that the contract be
entered into before the advice recommendation. Instead, the exemptions provide that the advice
must be subject to an enforceable written contract entered into before or at the same time as the
execution of the recommended transaction. The final exemptions also allow the contract to be
incorporated into other documents to the extent desired by the financial institution.
Additionally, as requested by commenters, the Department confirmed in the final exemptions
that the contract requirement may be satisfied through a master contract covering multiple
recommendations and does not require execution prior to each additional recommendation.
The Department eliminated the proposed contract requirement with respect to ERISA
plans in the final exemptions in response to public comments on this issue asserting that the
contract requirement in the context of ERISA-covered plans introduced unnecessary cost and
complexity to the exemption. A number of commenters indicated that the contract requirement
was unnecessary for ERISA plans due to the statutory framework that already provides
enforcement rights to such Plans, their participants and beneficiaries, and the Secretary of Labor.
Some commenters questioned the extent to which the contract provided additional rights or
remedies, and whether it would be preempted under ERISA’s preemption provisions. The final
230

exemption retains the contract requirement with respect to IRAs and non-ERISA plans. Based
on the foregoing, the Department’s high, medium, and low estimate of the percentage cost
reduction related to these changes in implementing the contract requirements are 70 percent, 60
percent, and 50 percent respectively.
Training and Licensing – Similar to the 2015 Proposal, the final rule and exemptions
will, as a practical matter, require firms to implement training programs for employees to learn
how the final rule and exemptions will affect their advisory work. The Department believes that
these costs will be less than they were under the 2015 Proposal due to the burden reducing
changes that have been incorporated into the final rule and exemptions that provide additional
flexibility and reduce complexity in complying. In particular, the final exemptions permit the
contract to be entered into prior to or at the same time as the execution of the recommended
transaction, rather than prior to the recommendation; the final exemptions include disclosures
that are much less complex than the disclosures in the proposals, and the final Best Interest
Contract Exemption eliminated the limited “Asset” list. The Department’s high, medium, and
low estimate of the percentage cost reduction related changes in the final rule and exemptions are
20 percent, 10 percent, and 0 percent respectively.513
Supervisory, Compliance, and Legal Oversight – Similar to the 2015 Proposal, the
final rule and exemptions will require firms to implement significant supervisory, compliance
and legal oversight to ensure compliance with the rules and exemptions. The Department
believes that these costs will be less than they were under the 2015 Proposal due to the burden
reducing changes that have been incorporated into the final rule and exemptions. The final Best
Interest Contract Exemption provides a method of complying with the exemption if the
retirement investor does not open an account but nevertheless acts on the advice through other
channels. The final exemptions will not be violated if certain disclosure requirements are not
satisfied, provided the financial institution acted in good faith and with reasonable diligence and
corrects the error within the required time periods. Finally, the final exemptions did not adopt
the proposed compliance with laws warranty. The Department’s high, medium, and low
estimates of the percentage cost reduction with respect to supervision, compliance, and legal
oversight related to changes in the final rule and exemptions are 20%, 10%, and 0% respectively.
Figure 5-17 at the end of Chapter five contains in tabular form a comparison of the 2015
Proposal and the final rule and exemptions, highlighting cost reducing changes.
The Department’s new estimates for BD start-up costs that account for the cost-reducing
changes made between the 2015 Proposal and the final rule and exemptions are obtained by first
multiplying the cost estimate for each category of cost for each firm size presented in Figure 5-5
by the applicable percentage cost reduction presented in Figure 5.6. The new cost category
estimates are then summed up to produce new estimates of total costs per firm by firm size. The
results of these calculations for the medium cost reduction scenario are presented in Figure 5-7.

513

The Department notes that neither the rule nor the exemptions require any particular licensing of Advisers or Financial Institutions.
Therefore, the Department does not agree that any cost is properly assigned to licensing.

231

Figure 5-7 Start-up Costs Before Adjusting for Data Discrepancy: Medium Reduction Scenario
Firm Size Category

Large BD

Medium BD

Small BD

Disclosure Requirements

$463,212

$261,000

$58,788

Data Collection

$221,253

$50,000

$28,080

Record Keeping (Data Retention)

$369,741

$150,000

$46,926

Implementing BICE Contracts

$3,986,667

$160,000

$160,000

Training and Licensing

$1,336,392

$522,000

$169,608

$354,596

$124,200

$45,004

Per Firm Total Cost Using this Method

$6,731,860

$1,267,200

$508,406

Industry Total Costs Using this Method

$282,738,139

$186,278,400

$1,179,502,399

Supervisory, Compliance, and Legal Oversight

Total

$1,648,518,938

However, as described above, there is a discrepancy between the weighted average of the
per firm start-up cost estimates provided by FSI (see Figure 5-2, column d) and the sum of the
means in Figure 5-4. They should be equal, but they are not. The weighted average is
(3/7)*$16,266,000 + (1/7)*$3,350,000 + (3/7) *$1,118,000=$7,929,000 while the sum of the
means is $7,150,000. If costs are estimated using the cost categories, cost would be understated.
An adjustment is made to account for this difference in costs. To make the adjustment, the
Department multiplies the new sum of the cost categories for each firm size by the ratio of FSI’s
original total per firm cost estimate for that firm size to the pre-cost-reduction sum of the cost
category estimates for that firm size. The resultant final per firm cost estimates are presented in
Figure 5-8.
Figure 5-8 Start-up Costs After Adjusting for Data Discrepancy: Medium Reduction Scenario
Firm Size Category

Large BD

Medium BD

Small BD

Disclosure Requirements

$506,848

$366,143

$64,327

Data Collection

$242,096

$70,142

$30,726

Record Keeping (Data Retention)

$404,572

$210,427

$51,346

Implementing BICE Contracts

$4,362,231

$224,456

$175,073

Training and Licensing

$1,462,287

$732,286

$185,586

$388,001

$174,234

$49,243

Per Firm Total Cost Using this Method

$7,366,036

$1,777,688

$556,301

Industry Total Costs Using this Method

$309,373,493

$261,320,201

$1,290,617,454

Supervisory, Compliance, and Legal Oversight

Total

$1,861,311,148

For example, to determine how much a large firm will pay in start-up costs in the
medium cost reduction scenario, first the estimated large firm cost for data collection under the
proposed rule ($1,106,265) is reduced by the percentage cost reduction seen in the final rule and
exemptions in the medium cost reduction scenario (80 percent) to produce a new per firm data

232

collection cost of $221,253. This same exercise is conducted for each of the cost categories for
large firms, producing the new per firm cost estimates seen in the first column of Figure 5-7. All
of these costs are then summed together to produce a total large firm cost of $6,731,860. This
sum is then multiplied by 1.0942 – the ratio of FSI’s original total large firm start-up cost
estimate ($16,266,000) to the pre-cost-reduction sum of the cost category estimates for large
firms ($14,865,587) – to produce a final per firm cost estimate of $7,366,036 for large firms.514
This exercise is done for each firm size category and each cost reduction scenario, and the per
firm costs are then multiplied by the number of firms in each size category to produce the total
BD start-up cost estimates presented Figure 5-9.
Figure 5-9 Total Costs for BDs (In Millions of Dollars)
SIFMA

FSI

Start-up

On-going

Start-up

On-going

High Reduction Scenario

$2,052

$463

$1,501

$333

Medium Reduction Scenario

$2,527

$571

$1,861

$413

Low Reduction Scenario

$3,001

$679

$2,222

$493

To produce new total start-up cost estimates based on the data supplied by SIFMA, the
Department first calculates by what percentage total start-up costs are reduced for each firm size
category in the calculations using the FSI data. The Department then uses these percentage
reductions to lower the SIFMA-based estimates of total costs by firm size category presented in
Figure 5-2. The resultant total cost estimates by firm size category are then added together to
produce the new SIFMA total BD start-up cost estimates also presented in Figure 5-9. Using this
approach means that the cost reductions seen in each firm size category are identical in
percentage terms in the SIFMA and FSI estimates.
To produce new estimates of ongoing costs, the ongoing costs for each firm size category
presented in Figure 5-3 are reduced by the same percentage that the start-up costs were reduced
by. The new total ongoing costs for each firm size category are then added up and are presented
in Figure 5-9.

5.3.2 RIA Firm Costs For Complying With Rule and Exemptions
As discussed above, the Department does not expect most RIAs to incur significant costs
to comply with the new rule and exemptions. This is because RIAs already operate under a
fiduciary standard that requires them to act in the best interest of their clients and most are
compensated through compensation arrangements that would require only compliance with the
more streamlined conditions in the Best Interest Contract Exemption for “level-fee” fiduciaries.
In its analysis of the proposed rule, the Department estimated a cost for RIAs to receive
initial compliance reviews and ongoing training and assigned the full cost of compliance for BDs

514

The ratio used for large firms – as described above – is $16,266,000/$14,865,587 = 1.0942. The ratio used for medium firms is
$3,350,000/$2,388,000 = 1.4028. The ratio used for small firms is $1,118,000/$1,021,746 = 1.0942. The totals used to calculate these ratios
can be found in Figures 5-2 and 5-5.

233

to all RIAs that were dually registered as BDs, but it did not attempt to calculate separate
exemption compliance costs for the subset of non-dually-registered RIA firms. After receiving
comments on the analysis of the proposed rule suggesting that the Department might have
underestimated costs for RIAs that choose to comply with an exemption such as the Best Interest
Contract Exemption, the Department quantified these costs in the final analysis as described
below. The Department also continues to assign full BD compliance costs to dually-registered
RIAs and estimate costs of compliance reviews and training and some costs to comply with
more streamlined conditions in the Best Interest Contract Exemption for the vast majority of
RIAs that will either not use an exemption or only use the stream-lined process for rollovers.

5.3.2.1

RIAs Needing to Comply with Full Exemption

While comments on the proposal suggested that the cost for at least some RIAs to
comply with the rule could be significant, no cost estimates were provided specifically for RIAs,
and the Department has had to develop its own approach to calculating these costs. As stated
above, because RIAs already operate under a fiduciary standard they should have lower costs to
comply with the final rule and exemptions than do BDs. What costs they do incur will also
depend on the nature of their current revenue streams and whether they are receiving level or
variable forms of compensation for their advice.
RIAs that receive variable compensation for investment advice provided once the money
is out of an ERISA plan would need to comply with an exemption such as the Best Interest
Contract Exemption in order to continue to receive the variable compensation. These firms
could incur costs that are similar to BDs. To estimate the costs for RIAs that would need to
comply with the Best Interest Contract Exemption or another exemption on an ongoing basis, the
Department assigned them the full value of the cost categories in the FSI report for data
collection, disclosure requirements, record keeping, and implementing Best Interest Contract
Exemption contracts. Because RIAs already operate under a fiduciary standard, they should
have to make significantly fewer supervisory, compliance, or legal oversight changes and
consequently the costs assigned to RIAs for this cost category have been reduced by half relative
to the costs assumed for BDs. RIA representatives are also not likely to need additional
licensing given that most already have a Series 65 license, and because they are already
complying with a fiduciary standard they will likely need less training.515 Given these facts, the
costs assigned to RIAs for training and licensing have also been reduced by half relative to the
costs assumed for BDs.
In order to determine what share of RIAs are likely to require the ongoing use of an
exemption and that these costs should be applied to, the Department turned to a Rand study for
the SEC which reported that 13 percent of RIAs surveyed reported receiving commissions while

515

The Series 65 license is a securities license required by most U.S. states for individuals who act as investment advisers. Successful
completion of the Series 65 exam permits an investment professional to function as an Investment Adviser Representative in certain states.
The Series 65 exam, called the Uniform Investment Adviser Law Examination, covers laws, regulations, ethics and topics such as
retirement planning, portfolio management strategies and fiduciary responsibilities.

234

8 percent reported specifically receiving performance-based fees.516 Taking a conservative
approach that, all else equal, creates a tendency towards overestimation of costs, the Department
assumes no overlap between these two groups and estimated that 21 percent of affected RIAs
will have to incur the costs described above. This leads to 22 large, 394 medium, and 3,150
small RIAs incurring these costs to comply.
Total costs for these firms are then obtained using the same procedure as for BDs with
the category modifications discussed above.517 Figure 5-10 summarizes these costs.
Figure 5-10 Total Costs for RIAs using an Exemption (In Millions of Dollars)
SIFMA

FSI

Start-up

Ongoing

Start-up

Ongoing

High Reduction Scenario

$3,138

$688

$1,615

$353

Medium Reduction Scenario

$3,930

$862

$2,046

$448

Low Reduction Scenario

$4,723

$1,036

$2,477

$542

5.3.2.2

RIAs Complying with Stream-lined Exemption

The exemption contains conditions applicable to level fee fiduciaries, which are advisers
and financial institutions who (with their affiliates) will receive a level fee that is disclosed in
advance to the retirement investor, for the provision of advisory or investment management
services to the plan or IRA. The following streamlined conditions apply to such level fee
Fiduciaries: (1) prior to or at the same time as the execution of the recommended transaction,
the financial institution must provide the retirement investor with a written statement of the
financial institution’s and its advisers’ fiduciary status; (2) the financial institution and adviser
comply with the impartial conduct standards of Section II(c) of the exemption; and (3) the
financial institution must document the specific reason or reasons why a rollover from an ERISA
plan to an IRA, a rollover from another IRA or a switch from a commission-based account to a
fee-based account, was considered to be in the best interest of the retirement investor, prior to
the transaction. The costs for such level fee fiduciaries to comply with the streamlined
conditions are included in cost estimates discussed in Section 5.6.
While the requirements of the final rule and exemptions should impose no significant
costs on RIAs already operating under a fiduciary standard that do not wish to utilize one of the
rule’s exemptions, to be cautious the Department has continued to estimate a cost for these firms
to receive a compliance review from outside legal counsel. Also, the Department has continued
to estimate the cost associated with providing a half-day of additional training to the firms’

516

517

Hung, et.al. 2008. The 13 percent of RIAs was based on RIAs that have individuals as clients. Applying the 13 percent to RIAs effected by
the final rule and exemptions assumes that percent holds for those who are affected by the final rule and exemptions. This assumption
appears valid as firms without individual clients would have been excluded from the estimates. The eight percent for firms with
performance based fees was obtained by multiplying the reported 20 percent of RIAs reporting having performance based fees times the 40
percent that are not likely to be hedge funds.
This includes the conservative approach the Department has taken to account for the discrepancy seen between FSI’s reported total per firm
costs by firm size and the sums of the category costs by firm size described in Section 5.3.1.

235

representatives to ensure that they are complying fully with the requirements of the final rule and
exemptions and not taking actions that would require the use of an exemption. A consultation to
evaluate and ensure a firm’s compliance with the rule was assumed to be conducted by a senior
partner of an outside legal firm with an hourly fee of $480. More recent data have since become
available, and the hourly fee used in this final analysis is $490.518 Small firms are more likely to
have less complex arrangements, so eight hours of consultation was assumed for them, with 16
hours for medium firms, and 40 hours for large firms. These are the same number of hours
assumed for this task in the analysis of the proposed rule, and the Department received no
comments opposing these estimates. This leads to estimated costs for the consultation for small
firms of $3,920, for mid-sized firms of $7,840, and large firms of $19,600.
Following the consultation, firms might choose to provide training to their
representatives to ensure that they understand when they are giving advice and how to comply
with firm policies and procedures regarding investment advice. There are many alternatives for
how a firm could conduct such training. Many firms already use outside vendors to train
employees on many issues, including compliance issues. The training could be on-site, through
webinars, or online according to the firm’s preferences. Costs could vary based on vendor and
mode of delivery. A common price for such vendor training was found to be $1,500 and this is
the price small firms are assumed to pay for their initial training in this analysis. Also, this is a
price similar to four hours of a mid-level attorney’s time, which costs roughly $1,340 (4 x $335)
which could be an alternative way to provide training. Later years would only require training to
help insure compliance. A cost of $500 is assumed to be incurred for small firms for these
ongoing annual trainings. This is a price similar to one and a half hours of a mid-level attorney’s
time, which costs roughly $503 (1.5 x $335).
While small firms could provide training in a single session for all their employees, midsized and large firms would probably need to hold multiple sessions to keep class size low.
Training costs for mid-sized firms are estimated to be $4,000 for the first year and $1,500 in
each subsequent year. Training costs for large firms are estimated to be $30,000 for the first
year and $10,000 in later years. The costs assumed for initial and ongoing training for small,
mid-sized, and large firms are identical to those assumed in the analysis of the proposed rule.
The Department received no comments opposing these assumptions other than more general
comments that expressed a feeling that overall costs for RIAs had been underestimated which
the Department has addressed above by incorporating into its analysis costs of compliance with
the rule’s exemptions for the subset of RIAs that may have to comply with the full exemption.
In total, the estimated costs in the first year for large, medium, and small RIAs to
evaluate their compliance with the rule and provide training to their representatives are $49,600,
$11,840, and $5,420, respectively. For subsequent years, the estimated costs per RIA are
$10,000 for large firms, $1,500 for medium firms, and $500 for small firms per year. Subsequent
year costs could be even lower as firms already conduct training of their staff, and training to
ensure compliance with the final rule and exemptions could be integrated into this training,
resulting in minimal increases in costs.

518

The cost of outside legal work is taken from the Laffey Matrix. The senior partner rate is the average of the 11-19 and 20+ years of
experience categories. The mid-level attorney’s rate is the average of the 4-7 year and 8-10 years of experience categories.

236

Applying these costs to the 79 percent of RIAs assumed not to need to incur the higher
costs described in Section 5.3.2.1 leads to total costs for legal consultations and training in the
first year of $85.9 million and for each subsequent year of $9.0 million. As discussed above
costs for complying with the disclosure requirements are included in Section 5.6 costs estimates.

5.3.3 Insurer Costs For Complying With Rule and Exemptions
Insurers will also incur costs to comply with the final rule and exemptions. In comment
letters, TIAA-CREF reported start-up costs for the Best Interest Contract Exemption to be $24.7
million with annual ongoing costs of $37.8 million, and Northwestern Mutual listed its total
costs to implement the proposed rulemaking at $13-15 million for start-up costs and $3-4 million
for ongoing costs. As with others’ submitted cost estimates, no additional information was
provided to evaluate the accuracy of the reported numbers, how they would relate to other
insurers, or how they would change given changes made in the final rule and exemptions. It
should be noted that SIFMA reports both firms as members and it could be that both of these
firm’s cost estimates were also included in the SIFMA report’s cost estimates.
While the financial products they sell might vary, BDs and insurers should incur costs
that are similar to comply with the final rule and exemptions as they will have to do many of the
same things to comply with the final rule and exemptions. It is even possible that insurers would
have significantly lower costs than BD firms, particularly to the extent insurers rely on BD firms
that are selling their products for implementation of the Best Interest Contract Exemption. In
this scenario BDs would already be incurring the costs of the Best Interest Contract Exemption
for other clients, and thus only have small added costs to do the same for those individual
purchasing insurance products, while insurers could experience lower costs. Due to this
similarity in burden and lacking comparable data for insurers, similar cost estimates are used to
estimate costs for the insurers as were used for BDs. The Department assigned to insurers the
full value of the cost categories in the FSI report for data collection, disclosure requirements,
record keeping, implementing Best Interest Contract Exemption contracts, and supervisory,
compliance, and legal oversight. However, the cost category reported in the FSI report for
training and licensing is problematic. While additional licensing may not be needed for insurers,
additional training similar to that required by BDs could be, but the report does not break down
the costs between training and licensing. Accordingly, the Department allocated 50 percent to
each category and uses only the portion for training in its estimates.
There are an estimated 398 insurers that could be affected by the rulemaking: 22 large,
175 medium, and 201 small insurers. However, as previously discussed in Section 5.2.8, an
additional issue is that a number of insurers may also either be dual-registered as BDs or have
affiliates that are BDs. These insurers will already have had costs of complying with the final
rule and its exemptions attributed to them in Section 5.3.1, but due to a lack of reliable
information that would allow the Department to determine exactly how much overlap there is
between BDs and insurers, they are assigned costs in this section as well. This means there is
likely double counting of some firms which will likely result in an over-estimation of costs for
insurers. Also as discussed in section 5.2.4 the actual number could be lower than 398 because
the SNL Financial data include firms that no longer sell annuities but continue to receive
premiums under closed blocks of business, as well as companies that reported very small
amounts of annuity considerations (approximately 75 companies reported total annuity
considerations lower than $100,000). Also, the SNL data count affiliates separately. Some
affiliates operate like one firm and would share compliance costs, while others operate as two
separate firms and would each have separate compliance costs. There are at least 43 groups
237

involving about 100 firms that could be affiliated. To be conservative in the estimates each
affiliate is counted as incurring the cost to comply with the final rule and exemptions.
Independent insurance agents could also be affected. Many of these agents are also
registered as BDs and are consequently captured in the estimates for BDs. Reliable data are not
available to estimate how many are not registered as BDs. Insurance agents are licensed with the
states, but a single agent can be licensed in multiple states, therefore it is difficult to obtain a
national number.519
Total costs for these firms are then obtained using the same procedure as for BDs with
the category modifications discussed above.520 Figure 5-11 summarizes these costs.
Figure 5-11 Total Costs for Insurers using an Exemption (In Millions of Dollars)
SIFMA

FSI

Start-up

Ongoing

Start-up

Ongoing

High Reduction Scenario

$1,638

$363

$386

$87

Medium Reduction Scenario

$2,045

$454

$486

$110

Low Reduction Scenario

$2,453

$545

$586

$133

5.3.4 Plan Service Provider Costs For Complying With Rule and
Exemptions
The Department believes that many of the other plan service providers will incur
minimal costs to comply with these regulations. Many of them may largely fall outside the
scope of the fiduciary investment rule, already meet the terms of one of the exceptions from the
rule, or do not rely for profits on the sort of conflicted fee arrangements that would require relief
under a PTE.
One of the largest costs that these firms face could be training their employees to
recognize when they are offering advice, so that they do not cross the line between education
and advice and become fiduciaries unintentionally.
Costs for these firms are estimated in the same way as the costs for RIA firms not
needing to comply with an exemption. The estimate includes costs for a consultation with an
outside attorney to evaluate the firms’ practices and procedures and then training for the
employees of the firms that could be giving investment advice. The per-firm costs are identical
to the per-firm costs discussed above for RIAs using the streamlined exemption. This is the same
approach to estimating service provider costs used in the analysis of the proposed rule which the
Department received no comments objecting to the approach.

519

520

Due to this data limitation, costs for these independent insurance agents may not be accounted for in the total costs. However, to the extent
insurers provide support some costs could be accounted for in the total costs for insurers.
This includes the conservative approach the Department has taken to account for the discrepancy seen between FSI’s reported total per firm
costs by firm size and the sums of the category costs by firm size described in Section 5.3.1.

238

Applying these costs to 2,700 small service providers, 506 medium service providers,
and 169 large service providers, results in aggregate start-up costs of $29.0 million and ongoing
costs of $3.8 million annually.

5.4

Additional Costs of Assuming Fiduciary Status

Based on industry-provided data, the Department also estimated additional expenses
some firms may incur such as the cost of increased insurance premiums. These cost estimates are
discussed below.

5.4.1 Increased Insurance Premiums/Litigation
Some advisers purchase insurance, such as errors and omissions insurance, to help
protect themselves from financial exposure for liability and litigation costs based on claims
made by clients alleging negligence, errors, or fiduciary beaches resulting from rendered
services. As further discussed below, the Department expects some insurance premiums to
increase for certain advisers under the broader fiduciary investment advice definition provided in
the final rule and exemptions.
Commenters on the 2015 Proposal said that some insurance policies cannot be used to
pay for penalties, and some do not cover litigation costs if the covered individual loses their
case. The commenters cite these situations to suggest that the cost from claims is not fully
captured by the increased cost of the insurance premiums. For those fiduciaries that are accused
of wrong doing and successfully defend against the claim, the insurance coverage pays for the
litigation. These costs should be captured in this analysis and are discussed below. The costs
that are not quantified in the following analysis are penalties paid by the advisers who lose their
cases.
In the IRA space, some of the additional costs of higher premiums comprise transfers
from insured service providers to IRA investors through the payment of recoveries. In both the
ERISA plan space and IRA space there are also transfers from those insured service providers
without claims to service providers covered by the insurance that receive coverage of litigation
costs.521 The Department estimates that 50 percent of the cost reflects the expenses and profits
of insurance carriers and agents who sell the policies, while the remainder is not a cost but a
transfer.522 Particularly in the IRA space, this transfer could be considered as contributing to a
just outcome because those harmed are now compensated. The change in profits could be a
transfer as well (to insurance carriers and agents from advisers and also from investors if
advisers can pass the additional premium payments onto their clients). However, due to
limitations of the literature and available evidence, the Department is not able to estimate the
fraction of the profits that could be a transfer.
The final rule broadly defines fiduciary investment advice subject to certain carve-outs
that exclude special circumstances that the Department believes should not be treated as

521
522

The difference in the treatment of insurance in the ERISA plan market verses the IRA market is due to Section 410 of ERISA.
Based on conversations with industry consultants the Department uses $3,000 as the cost for an insurance policy, 10 percent as the increase
in premiums due to the rulemaking, and a 50-50 split of the premium increase between insurer profits and a transfer.

239

fiduciary in nature.523 Under the broad definition, a person renders investment advice by (1)
providing investment or investment management recommendations to an employee benefit plan,
a plan fiduciary, participant or beneficiary, or an IRA investor or fiduciary and (2) broker-dealer
representatives that provide advice to plans, plan fiduciaries, participants, beneficiaries, and IRA
investors regarding the advisability of a particular investment or management decisions with
respect to securities or other property.
The Department believes that, under the revised investment advice definition, the
following categories of advisers may experience higher premiums: (1) consultants providing
investment or investment management recommendations to employer-sponsored plans, plan
fiduciaries, participants, and beneficiaries; and (2) broker-dealer representatives that provide
advice to plans, plan fiduciaries, participants, beneficiaries, and IRA investors regarding the
advisability of a particular investment or management decisions with respect to securities or
other property.
The Department understands that (1) premiums for these affected advisers could be
expected to increase by approximately 10 percent due to their new fiduciary status,524 (2)
insurance is priced on a per-representative basis; and (3) the average insurance premium is
approximately $3,000 per representative.525 Based on the foregoing, the estimated 10 percent
premium increase would be approximately $300 per insured representative.

5.4.1.1

Broker-dealers

According to industry data, BDs employed 643,322 representatives as of 2015.526 Not all
of these representatives, however, service ERISA plans or retail individuals. To arrive at the
number that do, the Department assumed that the share of BD representatives that service
ERISA plans and/or retail individuals is the same as the share of BDs the Department estimates
advise ERISA plans or retail individuals, or 63 percent (see Section 5.2.7). Multiplying 643,322
by 63 percent produces an estimate of 405,293 BD representatives that provide brokerage
services to at least one of these two groups.

523

524

525

526

For example, the regulation would not treat recommendations made by a party on the opposite side of a large plan in an arm's length
transaction as fiduciary investment advice provided that the carve-outs’ specific conditions are met. Similarly, the proposal does not treat
specified communications to employees of the plan sponsor and platform providers as fiduciary advice. Additionally, the rule draws a
distinction between covered fiduciary investment advice and non-fiduciary investment or retirement education. As the regulation makes
clear, a person does not render investment advice merely by providing educational materials or information on generally recognized
investment principles or by furnishing objective financial reports or information on investment alternatives. All of the rule’s carve-outs are
subject to conditions designed to draw an appropriate line between fiduciary and non-fiduciary communications, consistent with the text and
purpose of the statutory provisions.
The Department notes that parties providing investment advice to plans, participants, and beneficiaries for a fee must act prudently, solely in
the interest of participants and beneficiaries, for the exclusive purpose of paying benefits and defraying reasonable expenses, and must
diversify assets. If they breach their fiduciary duty, such service providers can be held personally liable by the Department, other
fiduciaries, participants, and beneficiaries for any losses that arise from the breach and also are subject to a prohibited transaction excise tax.
Service providers that provide investment advice to IRA investors and deal with IRA assets for their own interest or their own account or
are paid by a third-party in connection with a transaction involving IRA assets only are subject to the prohibited transaction excise tax. For
purposes of this analysis, the Department has assumed that the service providers to plans, participants, beneficiaries, and IRA investors will
be the same even though their liability exposure is different.
This is consistent with insurance premium amounts provided by the FSI in its 2009 Annual Report provided to the Department. FSI is a
trade group for independent providers of advice (those that are not unaffiliated with any funds). Their member statistics indicate that
average errors and omissions premiums per individual were approximately $2,300 in 2009. The Department used $3,000 because premiums
would be expected to increase since 2009. Additionally, $3,000 was on the high end of costs for representative insurance according to a
professional liability insurance distributor with whom the Department consulted. Thus, the $3,000 premium is a reasonable and
conservative estimate for estimating the increase in premiums.
FINRA Newsroom; available at: http://www.finra.org/newsroom/statistics. Last accessed February 29, 2016.

240

However, some of these BD representatives will also be registered as RIA
representatives and will consequently already have insurance that has largely priced in the
liability associated with operating under a fiduciary standard. These individuals are not expected
to experience a significant increase in premiums, although the Department does acknowledge
that depending on how the fiduciary liability insurance is priced, some premiums for duallyregistered representatives could increase to reflect additional exposure associated with the
broker-dealer portion of their business. To account for this overlap, the Department assumes
that the share of BD representatives that are dually registered as RIA representatives is the same
as the share of BDs that are dually registered as RIAs, or 18 percent (see Section 5.2.8). Once
this overlap is accounted for, it is estimated that 332,340 broker-dealers could experience higher
premiums.527
The Department estimates that the total per-year premium increase would be
approximately $99.7 million ($300 x 332,340) if each of these representatives sees a premium
increase of $300.528 However, as discussed above, about 50 percent, or $150, is paid to the
insuring firms and the other 50 percent is paid out as claims, which is counted as a transfer.
Therefore, there would be a total cost increase of $49.9 million per year for BD representatives
and a transfer of $49.9 million per year from BD firms to plans or retail investors and the
lawyers that represent them.

5.4.1.2

Plan Service Providers Excluding Broker-Dealers
and RIAs

There are less data available to estimate the number of employees at other firms that
service ERISA plans that will experience higher premiums. A review of 2013 Form 5500
Schedule C filings showed 3,375 unique service providers, not including BDs and RIAs, as
providing services that could be impacted by the new rule and exemptions. For the purposes of
this analysis, the Department conservatively assumes that all of these employee benefit plan
service providers have aspects of their business that would involve providing fiduciary
investment advice under the new rule and exemptions but not under the 1975 regulation.
If firms providing other plan services, for example, consulting services, are staffed
similarly to BD firms, then the average number of representatives at a BD firm can be used as a
proxy. As the data in the SEC Staff Dodd-Frank Study were presented in ranges, estimates of
the weighted average number of representatives per firm were obtained using the high ends
(51.3) and mid-points of the ranges (38.4). Out of concern that using the high-end of the ranges
would produce an over-estimate and using the mid-points an underestimate, the average of the
two numbers was used (45). Therefore an estimated 151,875 employees (3,375 x 45) were
estimated to possibly experience higher premiums totaling $45.6 million ($300 x 151,875) if
each of these employees sees a premium increase of $300. Splitting this into costs and transfers
as was done above results in a total cost increase of $22.8 million per year for service provider

527

528

For purposes of this analysis, the Department assumes that insurance premiums will increase by 10 percent for both plan and IRA service
providers even though fiduciary breaches are treated differently in the two markets.
Whether (and to what degree) the costs are paid directly by the broker-dealer firm or by the representative varies by firm. See FSI Annual
Report 2009. The FSI finds that the median percentage of liability costs covered by broker-dealers is 28 percent with the average being 48
percent. One would expect if the service provider pays a higher share, then other compensation for the representative would be lower.

241

employees and a transfer of $22.8 million per year from service provider firms to plans or retail
investors and the lawyers that represent them.

5.4.2 BD Conversion to RIA Status
In the analysis of the 2015 Proposal, estimates were provided for the costs for some BD
representatives to become licensed as RIA representatives. The FSI report included a cost
category for “training and licensing” and included a short discussion about the proposal’s cost
estimates. As the transition costs are included in the estimates above, they are not included here.
A short discussion of possible conversions from BD to RIA representative status is discussed
below.
In response to the 2010 and 2015 Proposals, representatives of the financial services
industry expressed concerns that the proposal would impair access to commission-based
advisory relationships for IRA investors and that such investors would be left with higher-cost,
fee-based advice as their only alternative. Their comments argued that investors (especially
small investors), therefore, would be harmed; they would receive less help in establishing IRAs,
less investment advice, and be subject to higher minimum account balances, because providing
such help to small accounts is expensive for BDs and they could not afford to do so without
receiving revenue sharing payments. The financial services industry also argued that many BDs
would no longer be able to service the IRA market, because they would have to become certified
as RIAs, and the cost to obtain the Series 65 license required for such certification would be
prohibitive.
In response to these concerns, the Department has taken steps to ensure that BDs can
receive a variety of transaction-based fees and compensation as a result of investments by plan
and IRA investors subject to conditions that provide appropriate protections. Fiduciary advisers
that comply with the conditions set forth in the exemptions and other guidance issued by the
Department may continue using many of their current business models without incurring the
costs associated with converting from a commission-based business model to a RIA asset-based
fee model.
Moreover, fiduciary investment advisers may take advantage of existing relief to receive
additional fees without violating the prohibited transaction rules. Of particular note is the
statutory exemption under Section 408(b)(14) of ERISA (and Section 4975(d)(17) of the IRC),
which applies to the provision of investment advice under an “eligible investment advice
arrangement,” as defined in paragraph (2) of ERISA Section 408(g) to participants and
beneficiaries of participant-directed individual account plans and IRA investors; and AO 200109A in which the Department concluded that the provision of fiduciary investment advice would
not result in prohibited transactions where the advice provided by the fiduciary results from the
application of methodologies developed, maintained and overseen by a party independent of the
fiduciary. Certain relevant relief may also be available under PTEs 86-128 and 84-24.
Some BD registered representatives that convert to RIA status might incur additional
frictional cost, not accounted for here, to set up a new RIA firm. Others, however, might already
be affiliated with or might join an existing RIA firm.
The cost of complying with PTEs might motivate some dual-registered BDs/RIAs to
convert customer accounts from brokerage, commission-based accounts to advisory, fee-based
accounts. However, as the selection of invest account (e.g. brokerage versus advisory) is
fiduciary advice and the Department has designed the PTEs to provide advisers with significant
flexibility in choosing the business model that allows them to best serve their clients, including
the choice of whether or not to receive direct and indirect variable payments, and thus expects
242

the degree of conversion to fee-based advisory accounts to be minimal. Finally, some comments
on the 2010 Proposal espoused a view that conversion to fee-based accounts would be limited as
a result of limitations imposed by advisory firms on the types of accounts that they typically
serve on a fee basis. For all these reasons, the Department anticipates little conversion of
brokerage accounts to fee-based advisory accounts.
A client moving from a brokerage, commission-based account to a fee-based account as a
result of the rule could face higher fees in that account. At the same time, the fees associated
with fee-based accounts provide for additional services that have value for the client. To the
extent that conversion occurs, the net cost to the client would be any potential increase in fees
less the additional benefits of advisory services. However, as the selection of invest account
(e.g. brokerage versus advisory) is fiduciary advice, the Department expects that any customers
who do convert a brokerage account to a fee-based account as a result of the rule would be those
customers who place greatest value on the services such an account provides. The Department
thus concludes that the costs for any customers who do potentially convert to an advisory model
would therefore be minimal as well (as they would be offset by the benefits from receiving
additional unconflicted advice).
Additional discussion of impacts on current market practices is provided in Section 8.3.

5.4.3 Call Centers
The Department expects that the cost impact of this rule on financial services industry
call centers will be small. This rule and the associated PTEs will require call centers that
provide fiduciary advice to provide additional disclosures and different or additional training to
their staff, but it need not otherwise impair the existence of call centers. Concerns were raised
by commenters on the proposal that requiring advisers in call centers to sign contracts before
offering advice would be expensive and disruptive. In response to these comments changes
were made to the contract requirement that are expected to eliminate this concern. Also call
centers may currently refrain from offering specific recommendations, limiting their support of
IRA investors’ decisions to provision of education. In such circumstances, some training may
still be required to ensure that activities are limited to education as defined under this proposal,
but call center employees would not be fiduciary advisers and therefore not otherwise affected
by this proposal. The Department believes, however, that call centers can add value to the IRA
advice market by offering fiduciary advice.
The Department believes that each medium and large BD or RIA firm will have its own
call center, but it lacked sufficient data to confirm this assumption. Therefore, the Department
requested comment on the number of call centers and the number of call center staff in the
industry, but did not receive sufficient information.
Based on the Department’s experience in training highly skilled customer service staff on
new laws and regulations, the Department believes that additional training for existing call
center staff could add the equivalent of between one-half and 1 day of staff time in an online
conference call or classroom-style setting to existing training programs. The training would
likely be performed by in-house legal staff or similarly skilled outside contractors.
In addition to the conference call or classroom-style training, management and legal staff
would need to revise internal scripts or talking points. These talking points are likely to already
exist for call center staff, but would need to be revised to ensure they comply with the new
regulation. Management and legal staff would also need to draft answers to frequently asked
questions and to revise training protocols for new staff. Because most of these revisions and
trainings would be performed by in-house legal staff knowledgeable in the final rule and
243

exemptions,529 the additional incremental burden would likely be measured in hours or days,
rather than weeks or longer.

5.5

Indirect Cost

The Department expects that the final rule and exemptions, which seeks to improve
retirement security by mitigating conflicts of interest that currently reduce the quality of
investment advice, to have little effect on access to investment advice. The Best Interest
Contract Exemption extends substantial flexibility to advice providers to adopt the business
models that allow them to best serve their clients, including the ability to continue receiving
direct and indirect variable compensation, such as commissions and revenue-sharing payments.
In part as a result of this PTE, the Department anticipates that firms providing investment advice
today will continue to provide advice to similar clienteles as they do today and incur compliance
costs as discussed above. As the Department anticipates that firms providing advice will
continue to provide advice, it likewise expects minimal transition costs that could arise from
recipients of financial advice changing financial agents. A more extensive discussion of access
is provided in the uncertainty analysis in Section 8.4.

5.5.1 Impact on Financial Sector Labor Markets
The gains to investors discussed, and partially quantified, elsewhere in this analysis
consist of three parts: transfers of surplus to IRA investors from advisers and others in the supply
chain, reduction in underperformance from suboptimal allocation of capital (or in other words,
benefits to the overall economy from a shift in the allocation of investment dollars to projects
that have higher returns), and resource savings associated with reduced excessive trading and
reduced wasteful, unsuccessful efforts to outperform the market. Although resource savings
contribute to net benefits when considered in a standard cost-benefit analysis, transitional
frictions may introduce some social costs; for example, if the resource being saved is worker
labor, then there would be search and training costs associated with finding new employment
within or outside of the financial industry. These related costs have not been quantified due to a
lack of data, literature, or other evidence on either the portion of investor gains that consists of
resource savings (as opposed to transfers or improved capital allocation) or the amount of
transitional cost that would be incurred per unit of resource savings. A more extensive
discussion is therefore included in the uncertainty Section 8.4.2. The Department invited
comment and data that would inform the estimation of rule-induced transitional costs for the
labor market, but did not receive any additional data that would help quantify these costs.

5.5.2 Impact on Asset Providers
This chapter does not contain any cost estimates for asset providers, such as mutual
funds.530 These entities may incur indirect, frictional costs related to product re-design, training,
and information technology requirements. These firms may incur costs for legal review of the

529
530

The time necessary to familiarize legal staff with the new regulation is factored in elsewhere in this Regulatory Impact Analysis.
Fidelity did provide a comment containing costs for the annual disclosure, but it did not directly address costs for its mutual fund business.
The SIFMA comment also contained asset managers, but that comment seemed most directed at the BD part of the business, but it could
also address the mutual fund component as well.

244

final rule and exemptions, decisions of whether to change payment streams to advisers, and
actions to facilitate disclosure by their advisers recommending their products. These entities
could also incur costs associated with winding down poorly-performing funds that charge high
commissions; indeed, the incurring of such costs is consistent with the assumptions underlying
the quantification of potential gains-to investor that appears elsewhere in this regulatory impact
analysis. Additional discussion is included in Section 8.4.
The RDR report, for example, shows asset providers incurring substantial costs when the
United Kingdom implemented its financial advice regulations.531 On the other hand, the RDR
policies were much more restrictive than this proposed rule, so asset providers may not
experience large changes in the United States.

5.6

Additional Costs Related to PTEs and Exceptions to Fiduciary
Investment Advice

It was not clear if the FSI and SIFMA cost estimates included all the costs of compliance.
To be conservative, additional costs were estimated for complying with the new and amended
PTEs, and producing the disclosures to utilize fiduciary exceptions for transactions involving
fiduciaries with financial expertise, platform providers, and investment education providers. The
additional costs are summarized below. For a more detailed discussion of these costs, see the
PRA section of the rule, as well as the PRA sections of the specific exemptions, which are
published elsewhere in today’s issue of the Federal Register.
The new Best Interest Contract Exemption will result in 65.1 million contracts and
disclosures being distributed during the first year and 72.3 million contracts and disclosures
being distributed in subsequent years. These disclosures range from one page to fifteen pages
long. In addition to the costs discussed earlier in chapter 5, costs for printing and distributing the
disclosures and legal professionals’ fees to produce the disclosures under the streamlined option
for level-fee fiduciaries, will cost approximately $403.2 million during the first year and $404.7
million in subsequent years. 532
The new Principal Transactions PTE will result in 4.9 million contracts and disclosures
being distributed during the first year and 3.0 million contracts and disclosures being distributed
in subsequent years. These disclosures range from two pages to fifteen pages long. In addition
to the costs discussed previously, producing and distributing the disclosures will cost
approximately $7.2 million during the first year and $4.3 million in subsequent years.
The amended PTE 86-128 adds a recordkeeping provision for plans and managed IRAs,
and new disclosure requirements for managed IRAs. Complying with the new recordkeeping
provision will generate on average an additional $273,000 in annual costs for maintaining the
records. The new disclosure requirements will result in an additional 18,000 disclosures being
distributed during the first year and an additional 13,000 disclosures being distributed in
subsequent years. The disclosures range from two to seven pages long. In total, these new

531

532

Europe Economics, “Retail Distribution Review; Post Implementation Review,” Dec. 2014, 65, available at:
http://www.fca.org.uk/static/documents/research/rdr-post-implementation-review-europe-economics.pdf.
Costs are higher in subsequent years due to the phase-in of the rule.

245

recordkeeping and disclosure requirements will cost approximately $803,000 during the first
year and $395,000 in subsequent years.
The amended PTE 75-1 includes disclosure requirements consistent with disclosures
already required by the SEC. Therefore, the amended PTE 75-1 does not add any burden.
The amended PTE 84-24 continues to grant relief for insurance agents, insurance
brokers, and pension consultants to receive a commission in connection with the purchase by
ERISA plans and IRAs of fixed rate annuity contracts, and for mutual fund principal
underwriters to receive a commission in connection with the purchase of mutual fund shares in
transactions involving ERISA plans. In order for insurance agents, insurance brokers, and
pension consultants to receive a commission in connection with the purchase of all other
annuities by ERISA plans and IRAs, and in order for mutual fund principal underwriters to
receive a commission in connection with the purchase of mutual fund shares in transactions
involving IRAs, such investment advice fiduciaries would instead be required to rely on the Best
Interest Contract Exemption. The amended PTE 84-24 will result in 3.3 million disclosures
being distributed annually. Producing and distributing the disclosures, including staff time to
draft the disclosures, will cost approximately $19.5 million annually.
The fiduciary exception for fiduciaries with financial expertise will result in 36,000
disclosures being distributed annually. Producing and distributing the disclosures, including
staff time to draft the disclosures, will cost approximately $3.0 million annually.
The fiduciary exception for platform providers will result in 2,000 disclosures being
distributed annually. Producing and distributing the disclosures, including staff time to draft the
disclosures, will cost approximately $45,000 annually.
The fiduciary exception for investment education will result in 23,500 disclosures being
distributed annually. Producing and distributing the disclosures, including staff time to draft the
disclosures, will cost approximately $3.1 million annually.
The Department has not associated any burden with the swap transaction exception. Plan
fiduciaries covered by the swap transactions provision must already make the required
representation to the counterparty under the Dodd-Frank Act provisions governing cleared swap
transactions. This rule adds a requirement that the representation be made to the clearing
member and financial institution involved in the transaction. The Department believes that the
incremental burden of this additional requirement would be de minimis. Plan fiduciaries would
be required to add a few words to the representations required under the Dodd-Frank Act
provisions reflecting the additional recipients of the representation. Due to the sophisticated
nature of the entities engaging in swap transactions, the Department believes that all of these
representations are transmitted electronically; therefore, the incremental burden of transmitting
this representation to two additional parties is de minimis. Further, keeping records that the
representation had been received is a usual and customary business practice.
In total, the final rule and exceptions will result in the creation and distribution of 73.4
million additional disclosures during the first year and 78.7 million additional disclosures in
subsequent years. The additional costs of these exemptions and exceptions, excluding those
costs discussed previously, are approximately $436.9 million during the first year and $435.1
million in subsequent years.

5.7

Total Quantified Costs and Sensitivity Analysis

In the sections above the various quantified costs have been calculated and explained.
These estimates include costs using information in comments submitted by SIFMA and FSI for
246

broker-dealers. Costs have also been estimated for RIAs, both for those who decide to comply
with the Best Interest Contract Exemption to continue to receive certain payments and those
RIAs who will comply with the streamlined Best Interest Contract Exemption. Costs for
insurers to comply with exemptions have also been estimated. Costs have also been estimated
for other service providers to ERISA plans as well as costs for additional liability insurance.
Finally additional costs to comply with the disclosure requirements were calculated as well.
Several summary tables showing total quantified costs are included below. Figure 5-12 shows
the 10- and 20- year discounted costs for each cost category discussed above Figure 5-13
contains the sum of 10-year and 20 year discounted value of all of these costs. These discounted
estimates account for start-up costs in the first year following the final regulation’s and
exemptions’ initial applicability. The Department understands that in practice some portion of
these start-up costs may be incurred in advance of or after that year. SIFMA- and FSI-based
estimates are presented separately. Estimates are also broken-out by the assumptions of costreduction from changes made in the final rule and exemptions relative to the proposed
rulemaking; these categories were categorized as high, medium and low cost reduction
scenarios. Figure 5-16 shows the start-up and annual ongoing cost by exemption. This figure
uses the FSI estimates, using the medium cost reduction assumptions.
As discussed above in Section 5.3, the data used raises several concerns about data
quality, accuracy, and reliability. However, the data sources used are the best ones available.
One key point to note is that the SIFMA study, as discussed in Section 5.3, did not
incorporate small firm costs into its estimates of total costs as SIFMA did not find the small firm
costs estimates produced by its survey to be reliable. The methodology the Department used to
calculate total costs using the SIFMA report figures follows this standard and also does not
include costs for small firms. The estimates derived from the SIFMA report should therefore not
be thought of as including total costs, but only the part of the costs that could be quantified given
the study’s limitations and short comings.
Besides the variation in costs due to the uncertainty in the changes in the cost due to the
changes from the proposal to the final, two other sources of uncertainty in the estimates are
modeled below. As discussed in Section 5.2.7 not all BDs, RIAs, and insurers serve ERISA plan
or IRA investors. A survey found that 41 percent of RIAs advise ERISA plans or are pension
consultants, while 22 percent advise retail individuals. To provide a conservative estimate, the
Department assumes that there is no overlap and therefore 63 percent of firms advise either an
ERISA plan or IRA investors and will be affected by the rule. While the amount of overlap is
not known the true value of the overlap is likely to be nearly complete as the expertise to provide
service to IRAs would be similar to that needed to service plans. Figure 5-14 shows the results
if the overlap is complete and therefore only 41 percent of firms service ERISA Plans, IRAs or
both.
A third source of uncertainty is the number of RIAs that would need to use the Best
Interest Contract Exemption due to their receiving payments otherwise disallowed as a result of
the final rule. As discussed in Section 5.3.2.1, 13 percent of RIAs reported receiving
commissions and 8 percent of RIAs that could be affected by this rule reported receive
performance based fees. As with the percent of service providers servicing ERISA Plans and
IRAs no overlap was assumed between these two groups, and 21 percent of the RIAs were
estimated to incur costs to comply with the Best Interest Contract Exemption. This assumption
is also relaxed and in Figure 5-15 only 13 percent of RIAs are assumed to need to comply with
the Best Interest Contract Exemption to maintain current business practices of receiving
payments otherwise disallowed as a result of the final rule. The estimates in Figure 5-15 also
247

maintain the assumption that only 41 percent of RIAs, BDs, and insurers service ERISA Plans
and/or IRAs.
Due to the uncertainty in the assumptions used to quantify the costs of the final rule and
exemptions an array of estimates has been presented. Using the 10-year, three percent
discounted estimates DOL’s cost estimates contained in Figure 5-13, Figure 5-14, and Figure 515 range from $10.0 billion to $31.5 billion. A conservative but reasonable primary estimate of
quantified 10-year discounted costs using a three percent discount rate is $16.1 billion. This
estimate uses FSI cost estimates – which are most complete as they account for the full costs of
firms of all sizes – and assumes the medium cost reduction scenario. It also very conservatively
assumes there is no overlap between firms servicing ERISA plans and IRA investors and that
there is no overlap between RIAs receiving commissions and those receiving performance-based
fees.
Figure 5-12 Summary Figure of Costs (In Millions of Dollars)
10-Year Discounted Costs

20-Year Discounted Costs

3 Percent

7 Percent

3 Percent

7 Percent

$5,492

$4,737

$8,430

$6,390

Medium Reduction Scenario

$6,767

$5,836

$10,390

$7,874

Low Reduction Scenario

$8,043

$6,936

$12,349

$9,359

BD Firms-FSI
High Reduction Scenario

BD Firms-SIFMA
High Reduction Scenario

$3,974

$3,430

$6,087

$4,618

Medium Reduction Scenario

$4,930

$4,255

$7,551

$5,729

Low Reduction Scenario

$5,885

$5,079

$9,016

$6,840

RIA Firms using Exemption-SIFMA
High Reduction Scenario

$8,243

$7,119

$12,607

$9,573

Medium Reduction Scenario

$10,329

$8,919

$15,798

$11,996

Low Reduction Scenario

$12,414

$10,720

$18,988

$14,418

RIA Firms Using Exemption-FSI
High Reduction Scenario

$4,237

$3,659

$6,478

$4,919

Medium Reduction Scenario

$5,369

$4,637

$8,210

$6,235

Low Reduction Scenario

$6,502

$5,615

$9,942

$7,550

$4,336

$3,742

$6,641

$5,039

$5,417

$4,675

$8,298

$6,296

$6,498

$5,608

$9,955

$7,553

$1,034
$1,304

$892
$1,124

$1,588
$2,002

$1,203
$1,517

$1,573

$1,357

$2,417

$1,831

Insurers-SIFMA

High Reduction Scenario
Medium Reduction Scenario
Low Reduction Scenario
Insurers-FSI

High Reduction Scenario
Medium Reduction Scenario
Low Reduction Scenario

248

E&O Insurance

$620

$510

$1,081

$769

RIA Costs

$151

$135

$208

$167

Other Service Provider Costs

$57

$50

$81

$64

$3,713

$3,058

$6,475

$4,611

Other PRA Costs

Figure 5-13 Summary Figure of Quantified Costs (In Millions of Dollars)
10-Year Discounted Costs
3 Percent
7 Percent

20-Year Discounted Costs
3 Percent
7 Percent

High Reduction Scenario

$22,612

$19,351

$35,524

$26,614

Medium Reduction Scenario

$27,054

$23,184

$42,330

$31,777

Low Reduction Scenario

$31,496

$27,017

$49,137

$36,941

High Reduction Scenario

$13,785

$11,733

$21,997

$16,352

Medium Reduction Scenario

$16,143

$13,769

$25,608

$19,093

Low Reduction Scenario

$18,502

$15,805

$29,219

$21,833

SIFMA

FSI

Figure 5-14 Summary Figure of Quantified Costs, Complete Overlap of Service Providers
to ERISA Plans and IRAs (In Millions of Dollars)
10-Year Discounted Costs
3 Percent
7 Percent

20-Year Discounted Costs
3 Percent
7 Percent

High Reduction Scenario

$18,460

$15,768

$29,159

$21,786

Medium Reduction Scenario

$21,964

$18,791

$34,534

$25,862

Low Reduction Scenario

$25,469

$21,815

$39,910

$29,938

High Reduction Scenario

$11,074

$9,396

$17,833

$13,197

Medium Reduction Scenario

$12,803

$10,887

$20,483

$15,207

Low Reduction Scenario

$14,532

$12,379

$23,132

$17,217

SIFMA

FSI

249

Figure 5-15 Summary Figure of Quantified Costs, Complete Overlap of Service Providers to ERISA Plans
and IRAs and RIAs Needing to use Best Interest Contract Exemption (In Millions of Dollars)
10-Year Discounted Costs

20-Year Discounted Costs

3 Percent

7 Percent

3 Percent

7 Percent

High Reduction Scenario

$16,336

$13,934

$25,907

$19,318

Medium Reduction Scenario

$19,297

$16,488

$30,450

$22,762

Low Reduction Scenario

$22,258

$19,043

$34,993

$26,206

FSI
High Reduction Scenario

$9,997

$8,466

$16,184

$11,946

$11,434
$12,871

$9,706
$10,945

$18,387
$20,590

$13,617
$15,287

SIFMA

Medium Reduction Scenario
Low Reduction Scenario

Figure 5-16 Summary Figure of Costs by Provision (In Millions of Dollars)

Best Interest Contract Exemption (Discussed in Sections 5.2-5.5)
Best Interest Contract Exemption (Discussed in Section 5.6)
Best Interest Contract Exemption (Total)
Principal Transactions Exemption (Discussed in Sections 5.2-5.5)
Principal Transactions Exemption (Discussed in Section 5.6)
Principal Transactions Exemption (Total)
PTE 86-128 (Discussed in Section 5.6)
PTE 75-1 (Discussed in Section 5.6)
PTE 84-24 (Discussed in Section 5.6)
Disclosures for Transactions that are Not Fiduciary Advice (Discussed in Section 5.6)

250

First Year

Subsequent
Years

$2,440
$395

$540
$403

$2,834

$943

$1,954
$7

$430
$4

$1,961

$435

$0.8
$0
$20
$6

$0.4
$0
$20
$6

Figure 5-17 Comparison of Proposed and Final Best Interest Contract and Principal Transactions
Exemptions
DISCLOSURES
Proposed Exemption

Final Exemption

BICE Transaction Disclosure: Required retirement
investors to receive a chart illustrating the “total cost”
of recommended investments over 1-, 5- and 10- year
periods.

BICE Transaction Disclosure Modified: Requires more
general disclosures regarding best interest standard of care
and material conflicts of interest. Retirement investors
may obtain more specific disclosure of the costs, fees and
other compensation regarding recommended transactions
upon request.

The disclosure required an individually tailored
mathematical calculation for each recommendation,
including projections and accounting for the present
value of different costs paid at different times

Revisions reduce costs from proposal because requirement
for customized transaction disclosure based on each
recommended investment to each individual investor is
eliminated.

BICE Annual Disclosure: Required retirement
investors to receive an annual disclosure of
investments purchased, sold or held; the total dollar
amount of all fees and expenses paid by the plan,
participant or beneficiary account or IRA (directly and
indirectly) with respect to each investment; and the
total dollar amount of compensation received by the
adviser and financial institution as a result of each
investment.

Annual Disclosure Requirement Eliminated: The
Department is confident this results in a substantial cost
reduction. One the world’s largest financial services
providers commented that that total costs to implement
this requirement as proposed was more than $46 million
the first year and $18 million thereafter.

BICE Web Disclosure: Required financial institution to
disclose on a website freely accessible to the public in a
machine readable format all direct and indirect
material compensation payable to it, its Advisers and
affiliates in connection with each asset that could be
purchased by plans and IRAs or that were purchased
by plans and IRAs in the last 365 days. The disclosure
had to provide the source of the compensation, and
how the compensation varied within and among assets.
The compensation had to be expressed as a monetary
amount, formula or percentage of the assets involved
in the purchase, sale or holding.

BICE Web Disclosure Modified: The financial institution
must disclose information on web about its business model
and associated material conflicts of interest; a schedule of
typical account fees and charges; model contract
disclosures and a written description of its policies and
procedures. Also must include, to the extent applicable,
arrangements with parties providing third-party payments,
and the financial institution’s compensation and incentive
arrangements with advisers.
Reduces burden and costs from proposal by:
 Permitting of group disclosures based on
reasonably defined categories, and reducing asset
by asset granularity proposal.
 Making clear that individual adviser compensation
is not required to be disclosed.
 Allowing financial institutions to comply by relying
on other public disclosures, such as the SEC Form
ADV, and posting them on website.
 Not requiring information to be provided in a
machine readable format.
 Permit dollar amounts, percentages, and formulas.
 Providing correction procedure for good faith
errors.

251

DATA COLLECTION
Proposed Exemption

Final Exemption

Best Interest Contract Exemption Specific Data
Request: Required financial institutions to collect and
maintain detailed data relating to inflows, outflows,
holdings, and returns for retirement investments for
six years from the date of the applicable transactions
and to provide that data to the Department upon
request within six months. The Department reserved
the right to publicly disclose the information provided
on an aggregated basis. Under a general recordkeeping
requirement, financial institutions also were required
to maintain records necessary for the Department to
determine whether the exemption conditions had been
satisfied.

Specific Data Request Eliminated and Proposed General
Recordkeeping Requirement Retained: Reduces costs
from proposal as commenters suggested this requirement
would be extremely time-consuming and costly as it would
require extensive contracting and systems build-outs
among thousands of advisors and third-party custodians
or recordkeepers. Financial institutions should maintain
information to demonstrate compliance with exemption
conditions as part of their usual and customary business
practices.

IMPLEMENTING CONTRACTS
Proposed Exemption

Final Exemption

Best Interest Contract Exemption - Contract
Requirement: Required financial institutions and
advisors to enter into a written contract with all
retirement investors, including ERISA plan investors, IRA
owners and investors in non-ERISA plans before
recommending a purchase, sale or holding of an asset.
Principal Transactions Exemption – Contract
Requirement: Adviser and financial institution must
enter into a written contract with all retirement
investors, including ERISA plan investors, IRA owners,
and investors in non-ERISA plans before engaging in the
principal transaction.

Best Interest Contract Exemption and Principal
Transactions Exemption Contract Requirement Modified:
Contract only required for IRAs or non-ERISA plan. The
contract may be a master contract covering multiple
recommendations whose terms may appear in a standalone
document or be incorporated into an investment advisory
agreement, investment program agreement, account opening
agreement, insurance or annuity contract or application, or
similar document, or amendment thereto.
Revisions reduce cost and make exemption more workable
by:
 Not requiring individual Advisers to be a party to the
contract.
 Not requiring Financial Institutions to enter into
contracts with ERISA plan investors.
 Allowing contracts to be consummated before or at
the same time the recommended transaction is
executed and is part of contracts commonly used
already.
 Not requiring Financial Institution to sign contract
 Allowing retirement investors’ written or electronic
signature.
 Allowing negative consent procedure for existing
customers.

252

253

6. Regulatory Flexibility Act
The Regulatory Flexibility Act (5 U.S.C. § 601 et seq.) (RFA) imposes certain
requirements with respect to Federal rules that are subject to the notice and comment
requirements of Section 553(b) of the Administrative Procedure Act (5 U.S.C. § 551 et seq.) and
which are likely to have a significant economic impact on a substantial number of small entities.
Unless the head of an agency certifies that a final rule and exemptions are not likely to have a
significant economic impact on a substantial number of small entities, Section 604 of the RFA
requires that the agency present a final regulatory flexibility analysis (FRFA) describing the
rule’s impact on small entities and explaining how the agency made its decisions with respect to
the application of the rule to small entities. The Department’s FRFA of the final rule and
exemptions is provided below.

6.1

Need for and Objectives of the Rule

ERISA and the IRC together provide that anyone paid to provide advice on the
investment of plan or IRA assets is a fiduciary. As fiduciaries, they are subject to certain duties,
including the obligation to generally avoid conflicts of interest. However, a 1975 regulation
narrowly construed these ERISA and IRC provisions, thereby effectively relieving many
advisers of these duties.
The Department of Labor found that conflicted advice is widespread, causing serious
harm to plan participants and IRA investors, and that disclosing conflicts alone would fail to
adequately mitigate the conflicts or remedy the harm. The Department has determined that
regulatory action is necessary to extend fiduciary duty to more advisers in order to minimize the
harm’s impact of conflicts of interest, to raise standards of care, recognize the diverse and
complex fee practices that exist in today’s plan service provider market, to account for the shift
from DB to DC plans, to expand the scope of fiduciary protections for plans and their
participants and beneficiaries, and ensure compliance with basic fiduciary norms and ensure
redress whose norms are violated. By extending fiduciary status to more advisers and providing
new and amended prohibited transaction exemptions that will preserve a variety of current
business practices subject to protective conditions, the new proposal will mitigate conflicts,
support consumer choice, remedy significant market failure and thereby improve plan and IRA
investing to the benefit of retirement security.
As discussed in further detail in the Regulatory Impact Analysis above, the Department
believes that amending the current regulation by broadening the scope of service providers,
regardless of size, that would be considered fiduciaries would enhance the Department’s ability
to redress service provider abuses that currently exist in the ERISA plan and IRA service
provider market, such as imprudent and disloyal advice, undisclosed fees and misrepresentation
of compensation arrangements.

6.2

Affected Small Entities

The final rule and the accompanying new and amended prohibited transaction
exemptions (PTEs) will provide benefits to small plans, small plan sponsors, and IRA investors,
and impose costs on small service providers rendering investment advice to plan or IRA
investors. Small service providers affected by this rule include broker-dealers (BDs), Registered
Investment Advisers (RIAs), insurance companies and agents, pension consultants, and others
providing investment advice to plan and IRA investors.

254

For purposes of the RFA, the Department continues to consider an employee benefit plan
with fewer than 100 participants to be a small entity.533 Further, while some large employers
may have small plans, in general small employers maintain most small plans. Thus, the
Department believes that assessing the impact of this final rule and exemptions on small plans is
an appropriate substitute for evaluating the effect on small entities. The definition of small
entity considered appropriate for this purpose differs, however, from a definition of small
business that is based on size standards promulgated by the Small Business Administration
(SBA) (13 C.F.R. 121.201) pursuant to the Small Business Act (15 U.S.C. § 631 et seq.). The
2013 Form 5500 filings show nearly 595,000 ERISA-covered plans with fewer than 100
participants.
The SBA defines a small business in the Financial Investments and Related Activities
Sector as a business with up to $38.5 million in annual receipts. In response to a comment
received from the SBA’s Office of Advocacy, the Department contacted the SBA and received a
data set containing data on the number of firms by NAICS codes. The data set also contains the
number of firms in specific revenue categories. This data set would allow the estimation of the
number of firms with a given NAICS code that fall below the $38.5 million threshold and
therefore be considered small entities by the SBA.
This data set alone, however, does not provide a sufficient basis for the Department to
estimate the number of small entities affected by the rule. Not all firms within a given NAICS
code would be affected by this rule because being an ERISA fiduciary relies on a functional test
and is not based on industry status as defined by a NAICS code. Further, not all firms within a
given NAICS code work with ERISA-covered plans or IRAs.
To obtain an estimate of the number of affected entities that meet the SBA’s definition of
small entities, the Department has applied the dataset provided by SBA to the Department’s
previous estimates of the total number of affected entities. As discussed previously, the
Department believes that the universe of firms affected by the rule consists of 2,509 BDs, 16,983
RIAs, 398 Insurers, and 3,375 other ERISA Plan Service Providers. To estimate the number of
small firms as defined by the SBA, the Department identified the NAICS industries that each
firm type might fall within and then calculated the share of firms that met the SBA size standard
within those NAICS industries.534 Based on these calculations, the share of each type of firm
that meets the SBA definition of small entities is 96.2, 97.3, 99.3, and 99.5 percent for BDs,
RIAs, Insurers, and other service providers, respectively. Based on this methodology, the
Department estimates that the number of small entities affected by this rule is 2,414 BDs, 16,524
RIAs, 395 Insurers, and 3,358 other ERISA service providers.
The Department also considers ERISA-covered pension plans with less than 100
participants as a small entity for purposes of the RFA. These small pension plans will benefit
from the rule, because as a result of the rule, they will receive advice that meets fiduciary
standards of prudence and loyalty from their fiduciary service providers. The Department’s

533

534

The basis for this definition is found in §104(a)(2) of ERISA, which permits the Secretary of Labor to prescribe simplified annual reports
for pension plans that cover fewer than 100 participants.
The following NAICS codes were used for BDs; 522120, 523110, 523120, 523930, 523130, 523999; for RIAs: 523930, 523130, 523999;
for Insurers: 524113 and 524210; and for other ERISA plan service providers: 524292, 525910, 525930, 541110, 541211, and 523991.

255

estimates of the value small plans and their participants gain from non-conflicted advice are
discussed in Chapter 4, above.
Figure 6-1 Number of Entities Affected by Rulemaking Meeting SBA's Definition of a Small
Entity

BDs
RIAs
Insurers
Other Service
Providers
ERISA Pension Plans

6.3

Total Number of
Firms Affected

Percent of Firms Meeting
SBA Size Standard

Number of Small
Entities Affected

2,509
16,983
398

96.2%
97.3%
99.3%

2,414
16,524
395

3,375

99.5%

3,358

681,154

87.3%

594,939

Significant Issues Raised In Response to IRFA

In response to the proposed rulemaking the Department received over 3,000 comments.
Additionally, the Department held four days of public hearings with 25 panels to discuss the
proposed rule and PTEs. The Department held numerous meetings with interested parties to
discuss issues and concerns. Organizations representing small businesses submitted comments
expressing particular concern with three issues: the education carve-out, best-interest contract,
and the seller’s carve-out.
A discussion of comments on the rule and its investment education and seller’s carve-out
as well as changes that were made to these provisions in the final rule can be found at Section
2.9.1, above. A discussion of comments regarding the Best Interest Contract Exemption and
changes that were made in the final exemption may be found at 2.9.2.1, above.

6.4

Response to Comments Filed by the Chief Counsel for Advocacy
of the Small Business Administration

On July 17, 2015 the Office of Advocacy filed a comment with the Department of Labor
on the proposed rule. Its comment letter claimed that the Initial Regulatory Flexibility Analysis
was deficient as it did not adequately estimate the costs for small business or the number of
small entities that would be affected.
The claims regarding the number of effected entities were based on the perception that in
estimating the number of small entities “[DOL] divide firms into small, medium, and large size
categories based on an allocation methodology that is not fully explained” and that “[DOL’s]
methodology for eliminating non-ERISA/IRA firms from the affected firm count suffers from a
similar lack of transparency and clarity.”
In response, DOL contacted the SBA and received a data set from them containing data
on the number of firms by NAICS codes. The data set also provided the number of firms in
given revenue ranges. This data set allows the Department to estimate of the number of firms
with a given NAICS code that fall below the $38.5 million threshold and therefore are
considered small entities by the SBA, which is an improvement over the method the Department
used in the proposal. As discussed above, the Department has revised its estimates of the
number of affected entities that meet the SBA definition of small entities using the data supplied
by the SBA.

256

DOL disagrees with the SBA Office of Advocacy’s assertion that the Department’s
methodology to estimate which firms would provide fiduciary services to ERISA-covered plans
or IRAs was insufficiently transparent. Section 5.2.4 above contains nearly identical language to
the proposal in describing data sources used including citations describing the calculations.
Also, Figure 5-1 shows the results of each step of the calculation.
The SBA Office of Advocacy comment suggested that while DOL acknowledged the
uncertainty around its estimated costs, it did not include that uncertainty in the estimates
provided included in the Regulatory Impact Analysis. While it is true that the estimates included
in Chapter 6 of the Regulatory Impact Analysis for the proposal were point estimates, the
Regulatory Impact Analysis did present two possible estimates. Also in the Regulatory Impact
Analysis for the proposal, the Department requested data from all interested parties, including
small businesses which could be used to estimate the costs of the rulemaking. Few firms
provided such data, but those cost estimates that were submitted are discussed and/or included in
Chapter 5, above.
Significant changes were made to the rule and exemptions in response to information
received from public comments, the hearings, and discussions with the regulated community.
The Department extensively discusses those changes and responses in the Regulatory Impact
Analysis and in the preambles to the final regulation and the exemptions.

6.5

Impact of the Final Rule and Exemptions

The reporting, recordkeeping, and other compliance costs of the final rule and
exemptions and the accompanying new and amended PTEs are discussed in detail in Chapter 5,
above. To summarize, the final rule require three disclosures for scenarios where financial
institutions are not considered to provide investment advice that would make them fiduciaries
under the final rule: (1) a platform provider disclosure; (2) a disclosure for investment education
materials; (3) a disclosure for transactions involving independent plan fiduciaries with financial
expertise; and (4) a swap transaction disclosure. The Best Interest Contract Exemption requires
the following disclosures or other legal documents as conditions of the PTE: (1) a prerelationship disclosure for transactions involving plans; (2) a contract for transactions involving
IRAs; (3) written anti-conflict policies and procedures and a written description of them; (4)
pre-transaction disclosures; (5) more detailed information on demand supplementing the pretransaction disclosures; (6) a web-based disclosure; and (7) a disclosure to the Department. The
Best Interest Contract Exemption also requires financial institutions to maintain records
demonstrating compliance with the conditions of the exemption. Financial institutions that limit
recommendations in whole or in part to proprietary products or third-party payments will have to
prepare a written documentation regarding those limitations. Financial institutions that are
“level fee fiduciaries” will be required to make disclosures of their fiduciary status to retirement
investors, and document the reasons for the recommendation of a rollover from an ERISA plan
to an IRA, from another IRA or from a commission-based account to a fee-based account.
Finally, for the transition period between the Applicability Date and January 1, 2018, financial
institutions must make a transition disclosure. The Principal Transactions Exemption requires
the following disclosures or other legal documents as conditions of the PTE: (1) a prerelationship disclosure for transactions involving plans; (2) a contract for transactions involving
IRAs; (3) written anti-conflict policies and procedures and a written description of them; (4) pretransaction disclosures; (5) confirmations; (6) an annual statement; and (7) a web-based
disclosure. For the transition period between the Applicability Date and January 1, 2018,
financial institutions must make a transition disclosure. The Principal Transactions Exemption
257

also requires financial institutions to maintain records demonstrating compliance with the
conditions of the exemption.
The amended PTE 86-128 adds a recordkeeping provision for plans and managed IRAs,
and new disclosure requirements for managed IRAs. The amended PTE 75-1, Part V adds
disclosures consistent with those already required by the SEC. The amended PTE 84-24 does
not add any compliance costs, but narrows the scope of transactions permissible under the PTE.
In order to create and distribute the disclosures and other legal documents and maintain
records demonstrating compliance with the PTE conditions, affected entities are likely to employ
legal professionals and financial managers to draft and review the materials, financial managers
to organize the records, IT staff to update technology systems to produce the disclosures and
maintain the records, and clerical staff to print and distribute the disclosures and other legal
documents.
As described previously, the Department anticipates that BDs and insurers will be most
impacted by the final rule and exemptions, while RIAs and other ERISA plan service providers
will experience less of a burden. The Department assumes that firms will take advantage of
relief provided in the PTEs in order to maintain their current compensation structures – or
choose to forgo conflicted fee structures – based on the particular approach that will be most cost
effective for their business models. Regardless of which PTEs they use, small affected entities
will incur costs associated with developing and implementing new compliance policies and
procedures to minimize conflicts of interest; creating and distributing new disclosures;
maintaining additional compliance records; familiarizing and training staff on new requirements;
and obtaining additional liability insurance.
The final regulations and PTEs will also balance the playing field between firms that are
currently providing unconflicted advice that is in the best interest of their clients and those who
previously may have had an unfair advantage as they offered conflicted advice with a lower
standard of care even as they marketed themselves as trusted advisers or fiduciaries.
It is possible that some small service providers may find that the increased costs
associated with ERISA fiduciary status outweigh the benefit of continuing to service the ERISA
plan market or the IRA market. The Department does not believe that this outcome will be
widespread or that it will result in a diminution of the amount or quality of advice available to
small or other retirement savers, because some firms will fill the void and provide services to the
ERISA plan and IRA market. It is also possible that the economic impact of the rule on small
entities would not be as significant as it would be for large entities, because anecdotal evidence
indicates that small entities do not have as many business arrangements that give rise to conflicts
of interest. Therefore, they would not be confronted with the same costs to restructure
transactions that would be faced by large entities.
As discussed in Section 6.2, according to the SBA size standards most firms affected by
this rule are small firms. Instead of replicating the costs analysis here, interested parties are
directed to Chapter 5 for a discussion of costs. Readers should note that the analysis in Chapter

258

5 divided firms into small, medium and large size categories.535 The Department has used these
categories in reliance on the submissions of commenters, which similarly divide the universe and
presentation of cost data based on these three broad categories, and because of the limited
availability of data from other sources or based on other categories. For purposes of the FRFA
all firms in Chapter 5 that are considered small and medium sized firms have revenues under
$38.5 million annually and therefore meet the SBA’s definition of a small entity. In addition,
some of the firms in the large category meet the SBA’s definition of a small entity. Figure 6-1
shows the universe of affected firms, the percent of each type of firm that meet the SBA’s
definition of a small firm, and the total number of small firms impacted by the rulemaking.
Figure 6-2 and Figure 6-3 summarize the average total start-up costs and average total
annual ongoing costs. Readers are directed to Section 5.1 for a discussion of the reliability of
the underlying data. It should also be noted that when choices had to be made as part of the
cost-benefit analysis, the Department generally sought to err in favor of an overestimate of costs.
Figure 6-2 Start-Up Costs Per Firm by DOL Firm Size Category in Medium Cost Reduction Scenario

$556,301

FSI
Medium
Firm
$1,777,688

$7,366,036

Small
Firm
NA

$438,886

$1,324,428

$6,440,892

$5,420

$11,840

$463,508
$5,420

Small Firm
BDs
RIAs (Complying With
Exemption)
RIAs (Complying With
Streamlined
Exemption)
Insurers
Other Service
Providers

535

SIFMA
Large Firm

Medium Firm

Large Firm

$12,258,090

$17,253,532

NA

$9,132,626

$15,086,559

$49,600

$5,420

$11,840

$49,600

$1,411,545

$6,634,892

NA

$9,733,342

$15,540,968

$11,840

$49,600

$5,420

$11,840

$49,600

For these figures, the following was used to create the size categories. The size categories for BDs were based on net capital values with the
following ranges: small <$50 million; medium $50 million to $1 billion; large >$1 billion as revenue was not available. The size
categories for RIAs used assets under management with the following ranges (revenue data was not available): small <$1 billion; medium
$1 billion to $100 billion; large >$100 billion. The size categories for insurer were based on assets with the following ranges: small <$1
billion; medium $1 billion to $100 billion; large >$100 billion. Complete revenue data for other service providers to ERISA plans were not
available, a size distribution of 80 percent small, 15 percent medium, and 5 percent large was used. This distribution is more skewed to
large firms leading to a greater likelihood of an overestimate of costs because of the higher costs attributed to such firms.

259

Figure 6-3 Ongoing Costs Per Firm by DOL Firm Size Category in Medium Cost Reduction Scenario

$120,411

FSI
Medium
Firm
$384,781

$1,836,676

Small
Firm
NA

$94,997

$286,673

$1,605,997

$500

$1,500

$100,326
$500

Small Firm
BDs
RIAs (Complying
With Exemption)
RIAs (Complying
With Streamlined
Exemption)
Insurers
Other Service
Providers

6.6

SIFMA
Large Firm

Medium Firm

Large Firm

$2,653,266

$4,302,062

NA

$1,976,759

$3,761,740

$10,000

$500

$1,500

$10,000

$305,529

$1,654,369

NA

$2,106,784

$3,875,045

$1,500

$10,000

$500

$1,500

$10,000

Steps Agency Has Taken to Minimize Significant Economic Impact
on Small Entities

Section 5.3.1 of the Regulatory Impact Analysis includes a discussion of changes to the
proposed rulemaking that are intended to reduce costs affecting both large and small firms.

6.7

Related Rules

ERISA and the IRC together assign fiduciary status to any person who “renders
investment advice for a fee or other compensation, direct or indirect” with respect to plan or IRA
investments. Under ERISA, fiduciary advisers to plan investors owe undivided loyalty to plan
participants’ interests. In addition, ERISA and the IRC together forbid fiduciary advisers to both
plan and IRA investors from engaging in broadly-defined “prohibited transactions” in which the
advisers’ and investors’ interests might conflict. While fiduciary advisers generally must avoid
conflicts, ERISA and the IRC provide certain parallel statutory “prohibited transaction
exemptions” that allow some transactions that involve conflicts of interest. The Department also
has authority to issue rules under both ERISA and the IRC that determine when persons
rendering advice on the investment of plan or IRA assets must act as fiduciaries. The
Department issued the current rule in 1975, and imposed a five-part regulatory test for
determining fiduciary status, which was much narrower than the broad statutory language.
The ERISA and IRC rules governing advice on the investment of plan and IRA assets
overlap with the separate and somewhat different provisions of federal securities laws such as
the Exchange Act and the Advisers Act and rules issued by the SEC that govern the conduct of
RIAs and BDs who advise retail investors. Congress, as part of the Dodd-Frank Act, directed
the SEC to consider a uniform fiduciary standard for RIAs and BDs who advise retail customers.
The Department consulted closely with the IRS/Treasury, SEC staff, and FINRA in developing
the new proposal. For a complete and thorough description of the interaction of these federal
laws, please see Chapter 3, above.
The Department strives to ensure consistency with regulations issued by other
government agencies. In response to comments received on the proposal, changes have been
made and language added to help ensure there is no conflict with other federal regulations. The
Department does not believe that the final rule and exemptions will conflict with any relevant
laws or regulations.
260

261

7. Regulatory Alternatives
In conformance with Executive Order 12866, the Department considered several
alternatives in finalizing its rule to more broadly define the circumstances under which a person
is considered to be a “fiduciary” by reason of providing investment advice regarding plan or IRA
assets and providing new and amended exemptions from the ERISA and IRC prohibited
transactions provisions that would preserve beneficial business models for advisers delivering
investment advice. These alternatives were informed by public comments, hearing testimony,
meetings with stakeholders, consultations with other financial regulators, and suggestions from
Congress.
The Department considered the alternative of issuing the 2015 Proposal and
corresponding exemptions as final without modification, but as discussed in Chapter 2 and 5
above, the Department made significant revisions to make the final rule and exemptions more
workable in response to public comments. For example, many commenters stated that the
proposed Best Interest Contract Exemption is unworkable, due to the cost that would be required
to comply with its requirements. In response to these comments, as discussed in Section 7.7
below, the Department has made several revisions to the exemption to make it less burdensome
and costly for financial institutions to comply with, such as eliminating (1) the proposed annual
disclosure, (2) the requirement for financial institutions to collect and maintain data relating to
inflows, outflows, holdings, and returns for retirement investments for six years from the date of
the applicable transactions and to provide that data to the Department upon request within six
months,536 and (3) the contract requirement for ERISA-covered plans.
The Department substantially revised the pre-transaction disclosure in the final
exemption by making it more general and requiring financial institutions to provide more
specific information only upon request, and providing financial institutions with considerable
flexibility regarding how best to present the information in the website disclosure by adopting a
more principal-based approach.
The Department also made several burden reducing changes to the contract requirement
in the proposed Best Interest Contract Exemption and proposed Principal Transactions
Exemption. For example, for “new contracts,” the final exemption provide flexibility for the
contract terms to occur in a standalone document or in an investment advisory agreement,
investment program agreement, account opening agreement, insurance or annuity contract or
application, or similar document, or amendment thereto. Also, existing customers may assent to
the contract either by affirmative consent or by negative consent procedure. Also, the
Department has clarified the required timing of the contract by deleting the specific requirement
that the contract be entered into before the advice recommended transaction. Relative to the
alternative of leaving these conditions the same as the proposal, the changes to the final rule and
exemptions reduce estimated costs by $13.2 billion to $24.8 billion of quantified costs over 10
years.
Also, as discussed in Section 7.10 below, the Department considered continuing, as
proposed, to allow financial institutions to receive relief under PTE 84-24 for transactions

536

The Department reserved the right to publicly disclose the information provided on an aggregated basis, although it made clear it would not
disclose any individually identifiable financial information regarding retirement investor accounts.

262

involving fixed-indexed annuities. After taking into account the risks and complexities of these
investments, however, the Department has determined that indexed annuities are appropriately
subject to the same protective conditions of the Best Interest Contract Exemption that apply to
variable annuities. The Department estimates that providing relief for fixed-index annuities
under PTE 84-24 instead of under the Best Interest Contract Exemption would have reduced
compliance costs by between $34.0 million and $37.8 million over ten years.
All of the alternatives are discussed in detail below.

7.1

Populating Asset Allocation Models and Interactive Investment
Materials with Designated Investment Alternatives

The 2015 Proposal excepted investment education from the definition of investment
advice. In doing so, the Department incorporated much of the Department’s earlier 1996
Interpretive Bulletin (IB 96-1), 537 but with the important exceptions that asset allocation models
and interactive investment materials could not include or identify any specific investment
product or specific investment alternative available under the plan or IRA. The Department
understood that not incorporating these provisions of IB 96–1 into the proposal represented a
significant change in the information and materials that may constitute investment education.
Accordingly, the Department specifically invited comments on whether the change was
appropriate.
A few commenters supported this change. They argued that participants are highly
vulnerable to subtle, but powerful, influences by advisers when they receive asset allocation
information. They believe that ordinary participants may view these models, including specific
investments included therein, as specific investment recommendations even if the provider does
not intend it as advice and even if the provider includes caveats or statements about the
availability of other products. In contrast, other commenters argued – particularly with respect
to ERISA-covered plans - that it is a mistake to prohibit the use of specific investment options in
asset allocation models used for educational purposes. They said this information is a critical
step to “connect the dots” for retirement investors in understanding how to apply educational
tools to the specific options or options available in their plan or IRA. With the change, the
commenters asserted that the Department had effectively shifted the obligation to populate asset
allocation models to plan participants, who for a variety of reasons are unlikely to do so, thereby
significantly undermining what has become a valuable tool for participants.
After evaluating the comments and considerations above, the Department has determined
that asset allocation models and interactive investment materials can identify a specific
investment product or specific alternative available under ERISA-covered plans if, in addition to
meeting the general conditions for asset allocation models and interactive investment tools; (1)
the alternative is a designated investment alternative (DIA) under an employee benefit plan; (2)
the alternative is subject to fiduciary oversight by a plan fiduciary independent from the person
who developed or markets the investment alternative or distribution option; (3) the asset
allocation models and interactive investment materials identify all the other DIAs available
under the plan that have similar risk and return characteristics, if any; and (4) the asset allocation

537

29 C.F.R. 2509.96-1 (IB 96-1),

263

models and interactive investment materials are accompanied by a statement that identifies
where information on those DIAs may be obtained. When these conditions are satisfied with
respect to asset allocation or interactive investment materials, the communications can be
appropriately treated as non-fiduciary “education” rather than fiduciary investment
recommendations, and the interests of plan participants are protected by fiduciary oversight and
monitoring of the DIAs.
The Department does not agree that the same conclusion applies in the case of
presentations of specific investments to IRA owners because of the lack of review and prudent
selection of the presented options by an independent plan fiduciary, and because of the
likelihood that such “guidance” or “education” would often amount to specific investment
recommendations in the IRA context. The Department was not able to reach the conclusion that
it should create a broad safe harbor from fiduciary status for circumstances in which the IRA
provider effectively narrows the entire universe of investment alternatives available to IRA
owners to just a few coupled with asset allocation models or interactive materials. Nor could the
Department readily import the conditions applicable to such plan communications to IRA
communications.
Similarly, because the provision is limited to DIAs available under employee benefit
plans, the use of asset allocation models and interactive materials with specific investment
alternatives available through a self-directed brokerage account would not be covered by the
“education” provision in the final rule. Such communications lack the safeguards associated
with DIAs, and pose many of the same problems and dangers as identified with respect to IRAs.

7.2

Extending Counterparty Exception to Include Smaller Plans,
Participants and Beneficiaries

The 2015 Proposal provides a carve-out from the general investment advice definition for
incidental advice provided in connection with an arm’s length sale, purchase, loan, or bilateral
contract between an expert plan investor and the adviser. The carve-out also applied in
connection with an offer to enter into such a transaction or when the person providing the advice
is acting as a representative, such as an agent, for the plan’s counterparty. The proxies for
financial expertise were that the person be a fiduciary of a plan with 100 participants or more or
that the plan fiduciary has responsibility for managing at least $100 million in plan assets.
Additional conditions applied that were intended to ensure the parties understood the nonfiduciary nature of the relationship.
The Department explained in the 2015 Proposal that the overall purpose of the proposed
carve-out was to avoid imposing ERISA fiduciary obligations on sales pitches that are part of
arm’s length transactions where neither side assumes that the counterparty to the plan is acting
as an impartial trusted adviser, but the seller is making representations about the value and
benefits of proposed deals. The Department believed that, under appropriate circumstances,
counterparties to the plan do not suggest that they are an impartial fiduciary and the plans do not
expect a relationship of undivided loyalty or trust. The premise of the proposed carve-out was
that both sides of such transactions understand that they are acting at arm’s length, and neither
party expects that recommendations will necessarily be based on the buyer’s best interests.
Consumer advocates generally agreed with the Department’s views expressed in the
preamble that it was appropriate to limit the carve-out to large plans and sophisticated asset
managers, and disagreed with commenters that urged that the carve-out be broadened. They
expressed concern that allowing investment advisers to claim “seller” status across the entire
retail market would effectively reinstate the five-part test from the 1975 regulation by allowing
264

brokers and other advisers to use disclosures in account opening agreements, investor
communications, advertisements, and marketing materials to evade fiduciary responsibility and
accountability for their investment recommendations.
On the other hand, many commenters argued for extending the “counterparty” or
“seller’s” carve-out to include the retail market of smaller plans and individual participants,
beneficiaries and IRA owners. The commenters contended that the lines drawn in the proposal
were based on a flawed assumption that small plans and individuals cannot understand the
difference between a sales pitch and advice. They also argued that there is no statutory basis for
distinguishing the fiduciary duties based on plan size. Some asserted that the Department should
consider a sophisticated investor test even accepting the assumption that some retail investors
may be confused about when a person presenting financial advice was acting as a seller and not
as an adviser. Others alternatively argued that the carve-out should be expanded to fiduciaries of
participant-directed plans regardless of size.
In evaluating the alternatives regarding this provision, the Department concluded that it
would be inconsistent with the language or intent of ERISA Section 3(21) to extend this
exclusion to small retail employee benefit plan investors or IRA owners. The Department
explained its rationale in the preamble to the proposal. In summary, the Department did not
include retail investors in this carve-out because the Department did not believe (1) the
relationships fit the arm’s length characteristics that the seller’s carve-out was designed to
preserve; (2) disclaimers of adviser status were effective in alerting retail investors to the nature
and consequences of conflicting financial interests; (3) IRA investors in particular do not have
the benefit of a menu selected or monitored by an independent plan fiduciary; (4) small business
sponsors of small plans are more like retail investors compared to large companies that often
have financial departments and staff dedicated to running the companies’ employee benefit
plans; and (5) there were other more appropriate ways to ensure such retail investors had access
to investment advice, such as prohibited transaction exemptions, and investment education.
The Department continues to believe for all of those reasons that it would be
inappropriate to provide a broad “seller’s” exemption for investment advice in the retail market.
Recommendations to retail investors and small plan providers are routinely presented as advice,
consulting, or financial planning services. In fact, in the securities markets, brokers’ suitability
obligations generally require a significant degree of individualization. As discussed in Chapter 3
above, most retail investors and many small plan sponsors are not financial experts, are unaware
of the magnitude and impact of conflicts of interest, and are unable to effectively assess the
quality of the advice they receive. IRA owners are especially at risk because they lack the
protection of having a menu of investment options chosen by an independent plan fiduciary that
is responsible for protecting their interests. Similarly, small plan sponsors are typically experts
in the day-to-day business of running an operating company, not in managing financial
investments for others.
Further, the Department is not prepared to adopt the approach suggested by some
commenters that the provision be expanded to include individual retail investors through an
accredited or sophisticated investor test that uses wealth as a proxy for the type of investor
sophistication that was the basis for the Department proposing some relationships as non-

265

fiduciary. The Department agrees with the commenters that pointed out that merely concluding
someone may be wealthy enough to be able to afford to lose money by reason of bad advice
should not be a reason for treating advice given to that person as non-fiduciary.538
In assessing the economic impact of expanding the provision to include smaller plans and
individual retail investors, the Department concluded that such a decision would likely result in
creating a loophole that would cause the rule to fail to make any real improvement in consumer
protections because it could be used by financial service providers to evade fiduciary
responsibility for their advice through the same type of boilerplate disclaimers that some
advisers use to avoid fiduciary status under the 1975 “five-part test” regulation. The Department
estimates that all of the over 23,000 financial institutions and service providers working in the
retirement market would have in-house resources spend one hour of clerical time, one hour of
information technology time, and 25 minutes of legal time539 creating and updating disclaimers
to include on all materials annually in an effort to avoid fiduciary responsibility. This would
produce an annual burden of $5.1 million in added compliance costs540 with a ten year cost of
$38.1 million to $44.6 million. In such a scenario, much of the costs and much of the gains to
investors discussed previously would be negated. Compared to the chosen regulatory option,
extending the counterparty exception to include smaller plans, participants, and beneficiaries
could potentially cause the regulated community to incur relatively small added costs, but at the
price of eliminating much or even nearly all of the consumer protections associated with this
project, as firms and advisers would capitalize on the provision as a means of avoiding
responsibility to adhere of fiduciary norms. Moreover, although the seller’s carve-out may not
be available in the retail market for communications directly to retail investors, the final rule
includes other provisions that are more appropriate ways to address some concerns raised by
commenters and ensure that small plan fiduciaries, plan participants, beneficiaries, and IRA
owners would be able to obtain essential information regarding important decisions they make
regarding their investments without the providers of that information crossing the line into
providing recommendations that would be fiduciary in nature. Under paragraph (b)(2) of the
final rule, platform providers (i.e., persons that provide access to securities or other property
through a platform or similar mechanism) and persons that help plan fiduciaries select or
monitor investment alternatives for their plans can perform those services without those services

538

539

540

The Department continues to believe that a broad based “seller’s” exception for retail investors is inconsistent with recent congressional
action, the Pension Protection Act of 2006 (PPA). Specifically, the PPA created a new statutory exemption that allows fiduciaries giving
investment advice to individuals (pension plan participants, beneficiaries, and IRA owners) to receive compensation from investment
vehicles that they recommend in certain circumstances. 29 U.S.C. §1108(b)(14); 26 U.S.C. § 4975(d)(17). Recognizing the risks presented
when advisers receive fees from the investments they recommend to individuals, Congress placed important constraints on such advice
arrangements that are calculated to limit the potential for abuse and self-dealing, including requirements for fee-leveling or the use of
independently certified computer models. The Department has issued regulations implementing this provision at 29 C.F.R. 2550.408g-1
and 408g-2. Thus, the PPA statutory exemption remains available to parties that would become investment advice fiduciaries because of
the broader definition in this final rule and the new and amended administrative exemptions published with this final rule and exemptions
(detailed elsewhere) provide alternative approaches to allow beneficial investment advice practices that are similarly designed to meet the
statutory requirement that exemptions must be protective of the interests of retirement plan investors.
As discussed in the Paperwork Reduction Act section of the preamble to this rule, the Department tasked several attorneys with drafting
sample legal documents in an attempt to determine the hour burden associated with complying with the information collections in the
counterparty exception in response to a recommendation made during the Department’s August 11, 2015 public hearing on the proposed
rule and exemptions. Commenters did not provide time or cost estimates needed to drafted this disclosure; the legal burden estimates in
this analysis, therefore, use the data generated by the Department to estimate the time required to create the sample disclosure.
23,265 financial institutions x ((1 hour of clerical time x $55.21 per hour) + (1 hour of IT time x $107.07 per hour) + (25 minutes x $133.61
per hour)) = $5.1 million.

266

being labeled recommendations of investment advice. Similarly, as discussed above, under
paragraph (b)(2) of the final rule, general plan information, financial, investment, and retirement
information, and information and education regarding asset allocation models would all be
available to a plan, plan fiduciary, participant, beneficiary or IRA owner and would not
constitute the provision of an investment recommendation, irrespective of who receives that
information.

7.3

Treating Appraisals, Fairness Opinions, or Similar Statements as
Fiduciary Investment Advice

Similar to the 1975 regulation, which includes “advice as to the value of securities or
other property,” the 2010 Proposal covered certain appraisals and valuation reports. The 2015
Proposal, however, was considerably more focused than the 2010 Proposal. Responding to
comments to the 2010 Proposal, the 2015 Proposal only covered appraisals, fairness opinions, or
similar statements that relate to a particular transaction. The 2015 Proposal also expanded the
2010 Proposal’s carve-out for general reports or statements of value provided to satisfy required
reporting and disclosure rules under ERISA or the Code. In this manner, the 2015 Proposal
focused on instances where the plan or IRA owner was looking to the appraiser for advice on the
market value of an asset that the investor was considering to acquire, dispose, or exchange. The
proposal also contained a carve-out specifically addressing valuations or appraisals provided to
an investment fund (e.g., a collective investment fund or pooled separate account) holding assets
of various investors in addition to at least one plan or IRA. In addition, the Department decided
not to extend fiduciary coverage to valuations or appraisals for ESOPs relating to employer
securities because it concluded that its concerns in this space raise unique issues that are more
appropriately addressed in a separate regulatory initiative.
Many commenters requested that the Department expand the valuation carve-outs to
clarify that routine or ministerial valuation functions necessary and appropriate to plan
administrative functions or integral to the offering and reporting of investment products are not
fiduciary advice. Commenters also wanted an explanation of what was meant by “in connection
with a specific transaction” and argued that many appraisals merely support fairness opinions
that fiduciary investment managers render in connection with specific transactions. Other
commenters asked the Department to remove valuations of all types from the definition of
investment advice or reserve the issue of valuations pending further study. Other commenters
suggested that the Department make certain exceptions for all valuations provided to ESOPs
regardless of whether they are done on a transactional basis or for independent plan fiduciaries
who engaged the valuation provider. Some others suggested that the current professional
standards for appraisers are sufficient or that the Department should develop its own.
Other commenters agreed with the Department that appraisal and valuation information
is extremely important to plans when acquiring or disposing of assets. Some also expressed
concern that valuations can steer participants toward riskier assets at the point of distribution.
After considering the comments, the Department has chosen the alternative of not
treating appraisals, fairness opinions, or similar statements to be fiduciary investment advice
under this regulatory package. The Department has concluded that ESOP valuations present
special issues that should be the focus of a separate project. The Department also believes that
piecemeal determinations as to inclusions or exclusions of particular valuations may produce
unfair or inconsistent results. Accordingly, rather than single out ESOP appraisers for special
treatment under the final rule, the Department has concluded that it is preferable to broadly
address appraisal issues generally in a separate project so that the Department can ensure
267

consistent treatment of appraisers under ERISA’s fiduciary provisions. Given the common
issues and problems appraisers face, it is quite likely that the comments and issues presented to
the Department by ESOP appraisers will be relevant to other appraisers as well.

7.4

Basing Exemptive Relief on Disclosure Alone

Some commenters argued that disclosure of potential adviser conflicts is, by itself,
sufficiently protective of plan and IRA investors’ interests. According to these comments, if
conflicts are transparent, then investors can choose optimally between more and less conflicted
advisers. Therefore, the commenters advocate that the Department should issue broad PTEs that
exempt all or almost all existing and potential adviser business models and compensation
arrangements on the sole condition that material conflicts be disclosed. The Department
estimates that relying on disclosure alone could reduce compliance costs by between $9.8 billion
and $17.5 billion over ten years.541
This argument necessarily presumes that investors will adequately understand the
implications of disclosed conflicts and factor that understanding into their choice of adviser and
investments. This presumption is highly questionable. As the Department discusses at length in
Chapter 3, above, available academic and empirical evidence strongly suggest that disclosure
alone will be ineffective at mitigating conflicts in financial advice.
As noted above, IRA investors are often older, and therefore, particularly vulnerable to
and targeted for abuse. Available evidence suggests that IRA owners may be poorly equipped to
act as a check on adviser misbehavior. Most are not financially sophisticated, and even those
that are might find it difficult to accurately detect lapses in the quality of advice. Nor are IRA
owners likely to understand advisers’ conflicts. Many ignore disclosures. Some others may
react to disclosures in ways that exacerbate the problem. Plan sponsors, especially small plan
sponsors, likewise may lack financial expertise, and/or be unclear about the fiduciary status of
their advisers.542
It appears that disclosures often fail to make investors aware of their advisers’ conflicts,
let alone understand their nature and potential implications. Representatives of all brokerage
firms interviewed by Hung et al. (2008), reported extensive efforts to clearly disclose conflicts,
but several acknowledged that “investors rarely read these disclosures … [F]or many investors,
the fact that they were given disclosures was seen as meaningless.” Burke et al. (2015)
summarize additional literature suggesting that retirement investors often fail to devote
meaningful attention to relevant disclosures.
Haziza and Kalay (2014) study investor behavior following a ruling by the Israeli
Securities Authority that portfolio managers must obtain written investor consent before
receiving part of the commission investors paid their broker to execute a trade. According to the
authors, “One would expect an overwhelming opposition to the kickback as consenting investors
are exposed to avoidable losses due to (moral hazard) access trading.” Yet 89 percent of

541

542

Estimates are obtained by setting all cost categories except the “disclosure” category to 100 percent cost reduction in the model. The range
is obtained by comparing the SIFMA and FSI medium cost reduction scenario. In addition, we assumed, for these purposes, that 63 percent
of all service providers would be affected by the rule, and applied a three percent discount rate.
GAO, GAO-08-774, “Private Pensions: Fulfilling Fiduciary Obligations Can Present Challenges for 401(k) Plan Sponsors” (2008);
available at: http://gaonet.gov/assets/280/278247.pdf.

268

investors affirmatively consented. The authors characterize this finding as “quite remarkable
considering that not responding is taken as a prohibition.” They find that more sophisticated
investors are less likely to consent, consenting investors underperform in the following year, and
consent is not a reward for past success.
If IRA investors could somehow be prompted to pay attention to and understand adviser
conflicts, would that serve to mitigate the conflicts? Available evidence suggests that investor
reaction to clear disclosure of conflicts is difficult to predict and not always beneficial to their
own interests (Robertson 2011; Cain, Loewenstein, and Moore 2005).
Loewenstein, Cain, and Sah (2011) describe how disclosure of adviser conflicts can
“backfire.” Once conflicts are disclosed, advisers might employ “strategic exaggeration” of their
own bias to offset customers’ discounting of their recommendations. They might feel “morally
licensed” to pursue their own interests over that of customers who have been duly warned.
Advice recipients might “anchor” to the advice and then adjust insufficiently for bias. They
might interpret disclosure as a sign of honesty and/or believe that payments that cause conflicts
signal high professional standing. They might follow biased advice because they feel socially
constrained from questioning their advisers’ integrity or threatening their livelihood. The
authors go on to review some experimental evidence for these phenomena. Also according to
the authors, “discounting advice appropriately for a disclosed conflict of interest requires a
mental model of advisor behavior to predict the impact of the conflict – let alone the disclosure
of that conflict – on the advice. Lacking such a model, advice recipients will not know what to
do with the disclosed information…”543
In a study of actual retirement investment advice interactions in Australia, investors
“were rarely able to tell whether or not the advice they received had a reasonable basis.” In most
cases where the Australian authority found “major shortcomings in the advice,” the investors
“thought the advice was satisfactory and said they intended to follow it.”544 Chater, Huck, and
Inderst (2010) find (with respect to an experimental setting in the European Union) that
disclosure of conflicts can have a different negative effect. Even investors who see conflicts as a
red flag may respond poorly.
In this study, the authors found that “[d]isclosing conflicts of interest elicits a ‘knee jerk’
reaction that can be harmful as well as helpful. Subjects exhibited ‘contrarian’ behavior in their
investment choices when biased incentives were disclosed. This led to better decisions when the
advisor’s and advisee’s interests were adversely aligned but worse decisions when their interests

543

544

Similar conclusions have been reached in connection with other types of professional advisory relationships. In one example, Malmendier
and Shanthikumar (2007) find that small investors fail to account for the positive bias associated with recommendations from analysts (or
brokers) who are affiliated with the underwriters of the firms they are reporting on, and consequently their returns suffer. In a second
example, as summarized by Burke et al. (2015), potential conflicts of interest are also a concern in the medical field: doctors and dentists
are often compensated directly or indirectly for recommending particular treatments, in ways that may conflict with the interest of patients;
the position of trust that medical professionals attain may exacerbate the problems of these conflicts. Schwartz, Luce, and Ariely (2011) use
health care claims data to show that dental patients in long-standing relationships with their dentists are likely to choose more expensive
(but not necessarily clinically superior) procedures; in experiments, they go on to show that patients are reluctant to seek second opinions
for fear of damaging their dentist-patient relationship, and that clinically-irrelevant social behavior (a dentist granting or refusing to grant a
personal favor to the patient) affects propensity to seek second opinions. Using survey data, Schwartz, Gino, and Ariely (2011) show that
patients recognize conflicts of interest for other people’s doctors but focus less on their own doctors; that patients are not concerned by
“indirect” conflicts of interest (e.g., doctors recommending a drug on which they are doing research); and that trust developed over time
reduces the willingness of patients to discount potentially conflicted advice.
Australian Securities and Investment Commission, “Shadow Shopping Survey on Superannuation Advice,” ASIC.

269

were aligned. Subjects lost trust even when an advisor with misaligned incentives was not
actually able to deceive them, showing that their reaction is reflexive” (Chater, Huck and Inderst
2010, 9-10).
Prentice (2011) provides a thorough discussion of “stock brokers and the limits of
disclosure.” He argues that “investors who receive a document indicating that their stock broker
owes them no fiduciary duty often (a) will not even read it and therefore go on assuming that
they are, in fact, owed a fiduciary duty by their brokers, as most investors currently believe, or
(b) if they do read it, they will not go to the trouble of figuring out what…it means…” (internal
citations omitted). He surveys a number of well-documented human cognitive behavioral biases,
and explains how each is likely to render disclosures of an adviser’s non-fiduciary status
ineffective or harmful. These include “over optimism” (other people might be misled by their
advisers but not me), “overconfidence” (I can tell whether my adviser is honest and his or her
advice is impartial), “illusion of control” (I can look out for myself), “false consensus effect”
(I’m honest, like most people, so my adviser must be too), “personal-positivity bias” (perhaps
many advisers are dishonest, but not mine), “insensitivity to the source of information” (whether
the advice is trustworthy is unrelated to whether the person giving it is conflicted), a tendency to
discount low-probability risks (I will not be defrauded), and “social proof” (lots of people trust
my adviser, or advisers generally, so I should too). He further observes that an adviser’s oral
communications with his or her client can color the client’s perception of any written disclosure,
and that an adviser’s “likeability” can cement trust. “Cognitive dissonance” (in this case, not
wanting to admit error), confirmation bias (focusing mostly on information that affirms initial
belief), and a tendency to anchor to initial intuitive judgments545 – can prevent an investor from
adjusting his or her opinion of an adviser based on new information. Finally, even an investor
who wished to adjust his or her level of trust to account for an adviser’s conflicts cannot know
how or how much to do so.
According to Prentice, advisers, in turn, exhibit behavioral cognitive biases that make it
more likely that their conflicts will bias their advice at their customers’ expense, sometimes
without realizing that they are doing so. They may suffer from “bounded ethicality” (it seemed
like the right thing to do based on my incomplete information), “self-serving bias” and a
tendency to anchor on first intuition (if it benefits me/feels good, I will focus most on
information that makes it also look like the right thing to do). They may be wedded to a selfimage that is moral, competent and deserving (my advice is never influenced by my conflicts,
and any related gains to me are merited), and/or exhibit “ethical fading” (moral considerations
wane as personal interests wax). All of these tendencies can lead advisers to calibrate advice to
profit themselves at their customers’ expense, even as they believe that they are acting honestly
in their customers’ interest. The more an adviser stands to gain, the more she will be
unconsciously pressed to convince herself that the most profitable path is also the right one.
Prentice concludes that “disclosure alone seems a frail tool with which to attack the many ills
that arise from blatant conflicts of interest in the financial industry.”
A broader look at transparency as a policy tool further suggests that disclosure alone
would be ineffective at mitigating conflicts in financial advice. There is evidence that disclosure

545

See Bubb (2015), however, who suggests that disclosures that act on an investor’s initial intuitive judgment might be effective.

270

alone fails to ensure efficiency in the U.S. mutual funds market (Palmiter and Taha 2008).
According to Ben-Shahar and Schneider (2011), advice must be “genuinely in the service of the
client, free of conflicts and influences,” and “experts are needed in the first place because people
cannot rely solely on disclosures.” Weil et al. (2006, 158, 161) argue that disclosure is effective
only where it provides “pertinent information that enables users to substantially improve their
decisions with acceptable [user] costs” and is comprehensible. Given that investors are hard
pressed to understand advisers’ conflicts or their implications, this suggests that disclosure of
conflicts will be ineffective. Finally, Willis casts doubt on the potential for financial education
to improve consumers’ financial decisions and use of financial advisers (Willis 2008).546
Overseas experience also suggests that disclosure alone does not effectively mitigate
conflicts in investment advice. For example, regulators in the UK found that requiring
disclosure and demanding that advisers act in their clients’ interest together did not sufficiently
discourage mis-selling of retail investment products largely due to commission-driven sales
practices and low qualification standards for advisers.547 The UK has since implemented
regulations in which financial advisers are no longer permitted to earn commissions from fund
companies in return for selling or recommending their investment products. These rules are
more stringent than those included in the Department’s new rule and exemptions.548
For all of these reasons the Department believes that a rule that relies on disclosure alone
to mitigate adviser conflicts would be ineffective and would therefore yield little or no investor
gains and fail to justify its compliance cost. The Department therefore attaches additional
investor protections to its PTEs.

7.5

Not Providing a Best Interest Contract Exemption

A few commenters regarding the 2015 Proposal advocated that the Department should
not issue any exemption from the prohibited transactions provisions of ERISA and the Code for
investment advice fiduciaries effectively asserting that all investment advice should be
unconflicted. Although the Department shares these commenters’ desire to promote
unconflicted advice, it did not adopt this alternative due to concerns about the disruptive impact
that barring all conflicts would have on the availability of and access to advice - especially for
small investors. Of course, the Department will continue to monitor this marketplace and, if

546

547
548

According to Willis (2008, 247-248), “Unfortunately, consumers have difficulty selecting advisers who possess sufficient expertise and
incentives to act in the consumers’ best interests. Once a consumer has selected an adviser, reliance on the advisor can become another
form of passivity in that consumers do not always sufficiently monitor the adviser’s performance.”
“[B]efore implementing an expert’s advice, a consumer has little means to determine whether the benefits of the advice outweigh the costs
of obtaining it. Without independent advice, consumers tend to rely on the advice dispensed by the ‘expert’ closest at hand, the seller.
Even with substantial literacy gleaned from financial education, the consumer rarely is as familiar as a salesperson with the latest financial
products. This ‘free’ advice can have a price. Among other things, yield-spread premiums for selling borrowers higher-cost mortgages than
those for which they qualify, and soft-dollar payments to investment brokers for favoring particular funds, can place the financial interests
of mortgage and investment brokers at odds with their clients.”
“Financial-product salespeople can take advantage of the ‘reciprocity effect’ invoked by ‘befriending’ the consumer, who then reciprocates
the seller’s ‘kindness’ with trust and business. Social mores inhibit customers from challenging the credibility of this new ‘friend.’
Linguistic conventions contribute to role confusion: the broker, officer, or agent is ‘my broker,’ ‘my loan officer,’ or ‘my agent,’ even
without any fiduciary duty to the consumer, and the insurer or lender ‘gives’ the coverage or credit, rather than ‘selling’ the financial
product. Once trusted, sellers have broad opportunities to influence consumer financial decisions.” (Internal citations omitted.)
FSA, “A Review of Retail Distribution” (2007), paragraph 3.
Letter from David Geale, FCA. UK, to Joseph Piacentini, U.S. Department of Labor, (2014).

271

necessary, adjust the conditions of the exemptions in the future if there are indications that the
exemptions appear to be abused by advisers.

7.6

ERISA-Covered Plans Have to Enter into the Best Interest Contract

The proposed Best Interest Contract Exemption permits investment advice fiduciaries
and certain related entities to receive compensation for services provided in connection with the
purchase, sale, or holding by plan participants, beneficiaries, IRAs and small employee benefit
plans of certain assets as a result of the investment advice. Specifically, the exemption would
permit fiduciary advisers and their firms to receive fees such as commissions, 12b-1 fees, and
revenue sharing in connection with investment transactions by the plan participants,
beneficiaries, IRAs and small plans, thus preserving many current fee practices.
The proposed exemption required fiduciary advisers and their firms to enter into a
written contract with the plan/IRA investor. The Department’s imposition of the contract
requirement is intended, in particular, to protect IRA investors with protections they do not
otherwise have under ERISA or the IRC. The contract between the IRA or non-ERISA plan,
and the financial institution, forms the basis of IRA/non-ERISA plans’ enforcement rights. The
Department intends that all the contractual obligations imposed on the financial institution (the
impartial conduct standards and the warranty) will be actionable by IRA/non-ERISA plans.
Because these standards are contractually imposed, an IRA or non-ERISA plan has a contractual
claim if, for example, its adviser recommends an investment product that is not in the investor’s
best interest.
In the Department’s view, these contractual rights serve a critical function for IRA
owners and participants and beneficiaries of non-ERISA plans. Unlike participants and
beneficiaries in plans covered by Title I of ERISA, IRA owners and participants and
beneficiaries in non-ERISA plans do not have an independent statutory right to bring suit against
fiduciaries for violation of the prohibited transaction rules. Nor can the Secretary of Labor bring
suit to enforce the prohibited transactions rules on their behalf.549 Thus, for investors in IRAs
and plans not covered by Title I of ERISA, the contractual requirement creates a mechanism for
investors to enforce their rights and ensures that they will have a remedy for misconduct. In this
way, the exemption creates a powerful incentive for financial institutions and advisers alike to
oversee and adhere to basic fiduciary standards, without requiring the imposition of unduly rigid
and prescriptive rules and conditions.
A number of commenters indicated that the contract requirement was unnecessary for
ERISA plans, because the ERISA statutory framework already provides enforcement rights to
such plans, their participants and beneficiaries, and the Secretary of Labor. Some commenters
additionally questioned the extent to which the contract provided additional rights or remedies,
and whether state-law contract claims would be pre-empted under ERISA’s pre-emption
provisions. The Department eliminated the proposed contract requirement with respect to
ERISA plans in this final exemption in response to public comment on this issue. In the

549

An excise tax does apply in the case of a violation of the prohibited transaction provisions of the Code, generally equal to 15% of the
amount involved. The excise tax is generally self-enforced; requiring parties not only to realize that they've engaged in a prohibited
transaction but also to report it and pay the tax. Parties who have participated in a prohibited transaction for which an exemption is not
available must pay the excise tax and file Form 5330 with the Internal Revenue Service.

272

Department’s view, the requirement that a financial institution provide written acknowledgement
of fiduciary status for itself and its advisers provides protections in the ERISA plan context that
are comparable to the contract requirement for IRAs and non-ERISA plans. As a result of the
written acknowledgment of fiduciary status, the fiduciary nature of the relationship will be clear
to the parties both at the time of the investment transaction, and in the event of subsequent
disputes over the conduct of the advisers or financial institutions. There will be far less cause for
the parties to litigate disputes over fiduciary status, as opposed to the substance of the
fiduciaries’ recommendations and conduct.

7.7

Leave Best Interest Contract Exemption Unchanged from 2015
Proposal

The Department retained the contract requirement with respect to IRAs and non-ERISA
plans in the final Best Interest Contract Exemption. In the Department’s view, these contractual
rights serve a critical function for IRA owners and participants and beneficiaries of non-ERISA
plans. Unlike participants and beneficiaries in plans covered by Title I of ERISA, IRA owners
and participants and beneficiaries in non-ERISA plans do not have a statutory right to bring suit
against fiduciaries for violation of the prohibited transaction rules, and fiduciaries are not
personally liable to them for the losses caused by their misconduct. Nor can the Secretary of
Labor bring suit to enforce the prohibited transactions rules on their behalf. Thus, the
contractual requirement creates a mechanism for investors to enforce their rights and ensures
that they will have a remedy for misconduct. It also creates a powerful incentive for fiduciary
advisers to oversee and adhere to basic fiduciary standards.
Many commenters stated that the proposed Best Interest Contract Exemption is
unworkable due to the cost that would be required to comply with its requirements. In response
to these comments, the Department has made several revisions to the exemption in addition to
limiting the contract requirement to non-ERISA plans and IRAs. The Department believes that
these changes make the exemption less burdensome for financial institutions to comply with.
Some of these changes are discussed below.
Data Collection — The proposed Best Interest Contract Exemption would have required
Financial Institutions to collect and maintain data relating to inflows, outflows, holdings, and
returns for retirement investments for six years from the date of the applicable transactions and
to provide that data to the Department upon request within six months. The Department reserved
the right to publicly disclose the information provided on an aggregated basis, although it made
clear it would not disclose any individually identifiable financial information regarding
retirement investor accounts.
Commenters expressed concern about the burden and costs of obtaining and maintaining
the necessary data and responding to the Department within the timeframe set forth in the
proposal. In response to the comments, the Department eliminated the data request in its
entirety.

273

While the proposed data collection requirement was not adopted as part of the final
exemption, the Department adopted the separate proposed general recordkeeping requirement,
which is discussed below and which was also included in the 2015 Proposal.
In addition, the Department notes that the final exemption did not adopt the proposed
annual disclosure, and the pre-transaction and website disclosures were revised in ways that
eliminate much of the need for data collection by the firm. The cost reduction of these changes
is estimated to be between $2.5 billion and $3.3 billion over ten years.550
Disclosure Requirements — The Department has substantially revised the disclosure
requirements that were in the 2015 proposed Best Interest Contract Exemption to make
compliance less burdensome and costly. The changes made to each disclosure are discussed
below. The cost reduction of these changes is estimated to be between $3.8 billion and $9.3
billion over ten years.551
Pre-transaction Disclosure — The proposed pre-transaction disclosure would have
required a retirement investor to receive a chart setting forth the “total cost” of the recommended
investment for 1-, 5- and 10- year periods, expressed as a dollar amount, assuming an investment
of the dollar amount recommended by the adviser and reasonable assumptions about investment
performance. A number of commenters raised significant objections to the proposed transaction
disclosure. They generally indicated the disclosure would be costly to implement and an
extensive transition period would be required for financial institutions to comply. In this vein,
several commenters stated that financial institutions do not currently assemble or maintain all of
the required information and that current systems could not deliver the disclosures. Thus, costly
system modifications would be necessary for financial institutions to obtain the data necessary to
comply with the requirement.
In response to the commenters, the Department has significantly revised the transaction
disclosure requirement in the final exemption to reduce burden, focus on the most salient
information about the contractual relationship and material conflicts of interest, and require more
detailed disclosures to be available upon request to retirement investors who are interested in
receiving them. The final exemption requires the transaction disclosure to:

550

551

552

•

State the best interest standard of care owed by the adviser and financial institution to
the retirement investor, disclose any material conflicts of interest;

•

Inform the retirement investor that the Investor has the right to obtain copies of the
financial institution’s written description of its required written policies and
procedures, as well as specific disclosure of costs, fees and other compensation
including third-party payments regarding the recommended transaction.552 The

Estimates are obtained by setting the data collection category to 0 percent cost reduction in the model. The range is obtained by comparing
the SIFMA and FSI medium cost reduction scenario. For these purposes, the Department additionally assumed that 63 percent of all service
providers are affected by the rule, and applied a three percent discount rate.
Estimates are obtained by setting disclosure cost category to 0 percent cost reduction in the model. The range is obtained by comparing the
SIFMA and FSI medium cost reduction scenario. For these purposes, the Department additionally assumed that 63 percent of all service
providers are affected by the rule, and applied a three percent discount rate.
The costs, fees, and other compensation may be described in dollar amounts, percentages, formulas, or other means reasonably designed to
present materially accurate disclosure of their scope, magnitude, and nature in sufficient detail to permit the retirement investor to make an
informed judgment about the costs of the transaction and about the significance and severity of the material conflicts of interest.

274

information required under this section must be provided to the retirement investor
before the transaction, if requested prior to the transaction, and if the request occurs
after the transaction, the information must be provided within 30 business days after
the request; and
•

Include a link to the financial institution’s website required by the exemption and its
address, and inform the retirement investor that: (i) model contract disclosures
updated as necessary on a quarterly basis are maintained on the website, and (ii) the
financial institution’s written description of its policies and procedures are available
free of charge on the website.

These disclosures do not have to be repeated for subsequent recommendations by the
adviser and financial institution of the same investment product within one year, unless there are
material changes in the subject of the disclosure.
As revised, the pre-transaction disclosure adopted in the final exemption involves
significant reductions in burdens and costs, compared with the proposed pre-transaction
disclosure. The proposal would have required a customized disclosure for each recommended
investment and the adviser and financial institution would have been required to calculate cost
projections based on the retirement investor’s dollar amount, and convert the costs into dollar
figures over the three holding periods. In comparison, the final pre-transaction disclosure is
more general and requires more specific information only to be provided upon request. Even if
more specific information is requested, the final exemption does not require calculation of a
specific amount expressed in dollars, but rather allows the information to be disclosed in dollar
amounts, percentages, formulas, or other means reasonably designed to present materially
accurate disclosure of their scope, magnitude, and nature in sufficient detail to permit the
retirement investor to make an informed judgment about the costs of the transaction and about
the significance and severity of the material conflicts of interest.
Annual Disclosure — In addition to the pre-transaction disclosure, the proposed
exemption would have required the Adviser or Financial Institution to provide the following
written information to the retirement investor, annually, within 45 days of the end of the
applicable year, in a succinct single disclosure:
(1) A list identifying each asset purchased or sold during the applicable period and
the price at which the asset was purchased or sold;
(2) A statement of the total dollar amount of all fees and expenses paid by the plan,
participant or beneficiary account, or IRA (directly and indirectly) with respect to
each asset purchased, held, or sold during the applicable period; and
(3) A statement of the total dollar amount of all compensation received by the adviser
and financial institution, directly or indirectly, from any party, as a result of each
asset sold, purchased or held by the plan, participant or beneficiary account, or
IRA during the applicable period.
The Department received numerous comments expressing concerns about the burden,
cost, and utility of the annual disclosure requirement. In response to such comments, the
Department did not adopt the annual disclosure requirement in the final exemption. The
Department is confident that the elimination of the annual disclosure results in a substantial cost
reduction from the proposed annual disclosure. The potential magnitude of the reduction was
illustrated in a comment from one of the world’s largest financial services providers. The
commenter estimated that it would incur total costs to implement the annual disclosure
275

requirement of more than $46 million the first year and more than $18 million dollars annually
thereafter based on a detailed cost assessment process for each affected area of its business.553
The Department notes that these cost estimates exceed the anticipated costs for large firms to
comply with the rule and exemptions that were contained in the Deloitte report.554 The
commenter noted that some of the magnitude of the expense is related to the firm’s large size,
but many of the same expenses would be incurred by smaller firms as well.
Website Disclosure — The proposed exemption would have required a financial
institution to maintain a website, freely accessible to the public, which shows the following
information:
(A) The direct and indirect material compensation payable to the adviser, financial
institution and any affiliate for services provided in connection with each asset
(or, if uniform across a class of assets, the class of assets) that a plan, participant
or beneficiary account, or an IRA is able to purchase, hold, or sell through the
adviser or financial institution, and that a plan, participant or beneficiary account,
or an IRA has purchased, held, or sold within the last 365 days. The compensation
may be expressed as a monetary amount, formula, or percentage of the assets
involved in the purchase, sale, or holding; and
(B) The source of the compensation, and how the compensation varies within and
among assets.
The financial institution's website would have been required to provide access to the
information described in (A) and (B) above in a machine readable format.
The Department’s intent in proposing the website disclosure was to provide broad
transparency about the pricing and compensation structures adopted by financial institutions and
advisers. The Department contemplated that the data could be used by financial information
companies to analyze and provide information comparing the practices of different advisers and
financial institutions. This information would allow retirement investors to evaluate and
compare the practices of particular advisers and financial institutions.
A number of commenters viewed the proposed website disclosure as too costly,
burdensome, and unlikely to be used by IRA investors, or expressed confidentiality and privacy
concerns. In particular, commenters opposed disclosure of adviser-level compensation. A few
commenters misinterpreted the proposal to require disclosure of the precise total compensation
amounts earned by each individual adviser, and strongly opposed such disclosure.
The Department has reworked the final website disclosure requirement to be based on a
more principles-based approach to address commenters’ concerns. The Department accepted the
suggestion of a commenter that the website disclosure should contain: a schedule of typical
account or contract fees and service charges, and a list of product manufacturers with whom the
financial institution maintains arrangements that provide payments to the adviser and financial

553

554

The comment states that many employees participated in the cost assessment process including, among others, personnel, finance,
technology, risk and compliance, product management, analytics, digital communications, distribution services, and platform support.
Deloitte report on behalf of SIFMA, “Report on the Anticipated Operational Impacts to Broker-Dealers of the Department of Labor’s
Proposed Conflicts of Interest Rule Package” (July 17, 2015).

276

institution, including whether the arrangements impact adviser compensation. Another
commenter suggested that the Department require disclosure of the financial institution’s
business model and the material conflicts of interest associated with the model. The commenter
further suggested the Department should require disclosure of the financial institution’s
compensation practices with respect to advisers, including payout grids and non-cash
compensation and rewards. The Department has adopted these suggestions as well.
With respect to the level of detail required, the Department has modified the website
disclosure requirement by providing financial institutions with considerable flexibility regarding
how best to present the information subject to the following principle: the website must “fairly
disclose the scope, magnitude, and nature of the compensation arrangements and material
conflicts of interest in sufficient detail to permit visitors to the website to make an informed
judgment about the significance of the compensation practices and material conflicts of interest
with respect to transactions recommended by the financial institution and its advisers.”
In response to comments, the final website disclosure requirement also reduces cost and
burden by permitting financial institutions to rely on other public disclosures, including those
required by the SEC and/or the Department to provide information required by the exemption by
posting them to its website.555
The Department is confident that the revision to the website disclosure requirement in the
final exemption will result in significant cost savings. The proposed website disclosure required
the financial institution to calculate and disclose compensation payable to itself, its advisers, and
its affiliates with respect to each asset recommended or a class of assets. In the final exemption,
the Department reduced this burden by reducing the specificity of the information provided. The
financial institution must disclose its third-party compensation arrangements with investment
product providers, and its compensation and incentive arrangements with advisers. The final
exemption allows such disclosures to be grouped together based on reasonably-defined
categories of investment products or classes, product manufacturers, advisers, and arrangements,
and financial institutions may disclose reasonable ranges of values, rather than specific values, as
appropriate. The final exemption also makes clear that individual adviser compensation is not
required to be disclosed. The final exemption also did not adopt the requirement that the
information in the website disclosure be made available in machine readable format.
Recordkeeping — This provision of the Best Interest Contract Exemption requires the
financial institution to maintain for six years records necessary for the Department and certain
other entities, including plan fiduciaries, participants, beneficiaries, and IRA owners, to
determine whether the conditions of the exemption have been satisfied. These records include,
for example, records concerning the financial institution’s incentive and compensation practices
for its advisers, the financial institution’s policies and procedures, any documentation governing
the application of the policies and procedures, the documents required to be prepared for
proprietary products and third-party payments, contracts entered into with retirement investors

555

These commenters argued that the information required to be disclosed as part of the exemption may already be part of other existing
disclosures, such as those provided pursuant to ERISA Sections 404(a)(5) and 408(b)(2) and the SEC’s required mutual fund summary
prospectuses and Form ADV. The Department has accepted these comments insofar as the information required disclosed pursuant to other
requirements also satisfies the conditions of the exemption, and so long as the Financial Institution provides an explanation that the
information can be found in the disclosures and a link to where it can be found.

277

that are IRAs and non-ERISA plans, and disclosure documentation. The Department believes
that it will not be costly for firms to comply with this requirement, because they already
maintain records similar to those required under the final exemption as part of their usual and
customary business practices.
The requirement to maintain the records necessary to determine compliance with the
exemption both encourages thoughtful compliance and provides an important means for the
Department and retirement investors to assess whether financial institutions and their advisers
are, in fact, complying with the exemption’s conditions and fiduciary standards. Although the
requirement does not lend itself to the same sorts of statistical and quantitative analyses that
would have been promoted by the data collection requirement, it too assists the Department and
retirement investors in evaluating compliance with the exemption, but at substantially less cost.
Implementing Contracts — The Department has made several burden reducing changes
to the contract requirement in the proposed Best Interest Contract Exemption and proposed
Principal Transactions Exemption. For example, for “new contracts,” the final exemptions
provide flexibility for the contract terms to occur in a standalone document or in an investment
advisory agreement, investment program agreement, account opening agreement, insurance or
annuity contract or application, or similar document, or amendment thereto. For retirement
investors with “existing contracts,” the final exemptions permit assent to be evidenced either by
affirmative consent, as described above, or by a negative consent procedure in which the
financial institution delivers a proposed contract amendment to the retirement investor prior to
January 1, 2018, and deems the retirement investor’s failure to terminate the amended contract
within 30 days as assent. An existing contract is defined in the exemptions as “an investment
advisory agreement, investment program agreement, account opening agreement, insurance
contract, annuity contract, or similar agreement or contract that was executed before January 1,
2018, and remains in effect.” If the financial institution elects to use the negative consent
procedure, it may deliver the proposed amendment by mail or electronically, but it may not
impose any new contractual obligations, restrictions, or liabilities on the retirement investor by
negative consent.
Moreover, the Department has clarified the required timing of the contract. As proposed,
the exemptions generally required that, “[p]rior to recommending that the plan, participant or
beneficiary account, or IRA purchase, sell or hold the asset, the adviser and financial institution
enter into a written contract with the retirement investor that incorporates the terms required by
[the exemption]….” A large number of commenters responded to various aspects of this
proposed requirement. Many commenters objected to the timing of the contract requirement.
They said the timing “prior to” any recommendations would be contrary to existing industry
practices. The commenters indicated that preliminary discussions may evolve into
recommendations before a retirement investor has decided to work with a particular adviser and
financial institution. Requiring a contract upfront would chill such preliminary discussions,
marketing, and an investor’s ability to shop around, commenters said. Many commenters on this
subject suggested that the proper timing of the contract would be before execution of the
investment transaction at issue or even later. Several commenters that strongly supported the
contract requirement nevertheless agreed that the timing could be adjusted without loss of
protection to the retirement investors.
In the Department’s view, the critical aspect of the requirement is that all instances of
advice be covered by an enforceable contract. Therefore, the Department adjusted the
exemptions by deleting the specific requirement that the contract be entered into before the
advice recommendation. Instead, the exemptions provide that the advice must be subject to an
278

enforceable written contract entered into before or at the same time as the recommended
transaction. The final exemptions also allow the contract to be incorporated into other
documents to the extent desired by the financial institution. Additionally, as requested by
commenters, the Department confirmed in the final exemptions that the contract requirement
may be satisfied through a master contract covering multiple recommendations and does not
require execution prior to each additional recommendation.
The Department eliminated the proposed contract requirement with respect to ERISA
plans in the final exemptions in response to public comments on this issue. A number of
commenters indicated that the contract requirement was unnecessary for ERISA plans due to the
statutory framework that already provides enforcement rights to such Plans, their participants
and beneficiaries, and the Secretary of Labor. Some commenters questioned the extent to which
the contract provided additional rights or remedies, and whether it would be preempted under
ERISA’s preemption provisions. The changes made in the final rule and exemptions reduce
these estimated costs by $5.9 billion to $6.6 billion over ten years.556
With the above changes, the Department is confident that the Best Interest Contract
Exemption now provides a flexible and workable mechanism for advisers to preserve their fee
practices while aligning their interests with those of their clients. This should ensure access to
investment advice for plan and IRA investors with small account balances is not impaired.
Relative to the alternative of leaving these conditions the same as the proposal, the changes to
the final rule and exemptions reduce estimated costs by $13.2 billion to $24.8 billion of
quantified costs over 10 years.

7.8

Arbitration

The proposed Best Interest Contract Exemption provides that individual claims may be
the subject of contractual pre-dispute binding arbitration but class claims must be permitted to
proceed in court. Commenters on the proposed exemption were divided on the approach taken
in the proposal. Some commenters objected to limiting retirement investors’ right to sue in court
on individual claims and specifically focused on FINRA’s arbitration procedures. These
commenters described FINRA’s arbitration as an unequal playing field with insufficient
protections for individual investors. They asserted that arbitrators are not required to follow
federal or state laws, and so would not be required to enforce the terms of the contract. In
addition, commenters complained that the decision of an arbitrator generally is not subject to
appeal and cannot be overturned by any court. According to these commenters, even when the
arbitrators find in favor of the consumer, the consumers often receive significantly smaller
recoveries than they deserve. Moreover, some asserted that binding pre-dispute arbitration may
be contrary to the legislative intent of ERISA, which provides for “ready access to federal
courts.”
Some commenters opposed to arbitration indicated that preserving the right to bring or
participate in class actions in court would not give retirement investors sufficient access to

556

Estimates are obtained by setting the cost category of implementing the contract requirement under the Best Interest Contract Exemption to
0 percent cost reduction in the model. The range is obtained by comparing the SIFMA and FSI medium cost reduction scenario. For these
purposes, the Department additionally assumed that 63 percent of all service providers are affected by the rule and applied a three percent
discount rate.

279

courts. According to these commenters, allowing financial institutions to require resolution of
individual claims by arbitration would impose additional and unnecessary hurdles on investors
seeking to enforce the best interest standard. One commenter warned that the regulation would
make it more difficult for retirement investors to pursue class actions because the individualized
requirements for proving fiduciary status could undermine any claims about commonality.
Commenters said that class action lawsuits tend to be expensive and protracted, and even where
successful, investors often recover only a small portion of their losses.
Other commenters just as forcefully supported pre-dispute binding arbitration
agreements. Some asserted that arbitration is generally quicker and less costly than judicial
proceedings. They argued that FINRA has well-developed protections in place to protect the
interests of aggrieved investors. One commenter pointed out that FINRA requires that the
arbitration provisions of a contract be highlighted and disclosed to the customer, and that
customers be allowed to choose an “all-public” panel of arbitrators. FINRA rules also impose
larger filing fees on the industry party than on the investor. Commenters also cited evidence that
investors are as likely to prevail in arbitration proceedings as they are in court. Thus, these
commenters argued that permitting mandatory arbitration for all disputes would be in investors’
best interest.
A number of commenters argued that arbitration should be available for all disputes that
may arise under the exemption, including class or representative claims. Some of these
commenters favored arbitration of class claims due to concerns about costs and potentially
greater liability associated with class actions brought in court. Some commenters took the
position that the ability of the retirement investor to participate in class actions could deter
Financial Institutions from relying on the exemption at all.
After consideration of the comments on this subject, the Department has decided to adopt
the general approach taken in the proposal. Accordingly, contracts with retirement investors
may require pre-dispute binding arbitration of individual disputes with the adviser or financial
institution. The contract, however, must preserve the retirement investor’s right to bring or
participate in a class action or other representative action in court in such a dispute. The final
exemption also provides that contract provisions may not set an amount representing liquidated
damages for breach of the contract. However, the final exemption expressly permits retirement
investors to knowingly waive their rights to obtain punitive damages or rescission of
recommended transactions to the extent such waivers are permitted under applicable law.
The Department recognizes that for many claims, arbitration can be more cost-effective
than litigation in court. Moreover, the exemption’s requirement that financial institutions
acknowledge their own and their advisers’ fiduciary status should eliminate an issue that
frequently arises in disputes over investment advice. In addition, permitting individual matters
to be resolved through arbitration tempers the litigation expense for financial institutions,
without sacrificing retirement investors’ ability to secure judicial relief for systemic violations
that affect numerous investors through class actions.
On the other hand, the option to pursue class actions in court is an important enforcement
mechanism for retirement investors. Class actions address systemic violations affecting many
different investors. Often the monetary effect on a particular investor is too small to justify an
individual claim, even in arbitration. Exposure to class claims creates a powerful incentive for
financial institutions to carefully supervise individual advisers, and ensure adherence to the
impartial conduct standards. This incentive is enhanced by the transparent and public nature of
class proceedings and judicial opinions, as opposed to arbitration decisions, which are less
280

visible and pose less reputational risk to firms or advisers found to have violated their
obligations.
The ability to bar investors from bringing or participating in such claims would
undermine important investor rights and incentives for advisers to act in accordance with the
best interest standard. As one commenter asserted, courts impose significant hurdles for
bringing class actions, but where investors can surmount these hurdles, class actions are
particularly well suited for addressing systemic breaches. Although by definition
communications to a specific investor generally must have a degree of specificity in order to
constitute fiduciary advice, a class of investors should be able to satisfy the requirements of
commonality, typicality and numerosity where there is a widespread problem, such as the
adoption or implementation of non-compliant policies and procedures applicable to numerous
retirement investors, the systematic use of prohibited or misaligned financial incentives, or other
violations affecting numerous retirement investors in a similar way. Moreover, the judicial
system ensures that disputes involving numerous retirement investors and systemic issues will be
resolved through a well-established framework characterized by impartiality, transparency, and
adherence to precedent. The results and reasoning of court decisions serve as a guide for the
consistent application of the law in future cases involving other retirement investors and
financial institutions.

7.9

Assets Covered Under Best Interest Contract Exemption

The proposed Best Interest Contract Exemption provided that the exemption would
permit an adviser, financial institution and their affiliates and related entities to receive
compensation for services provided in connection with the purchase, sale, or holding of an asset
by a plan, participant, beneficiary, or IRA owner, as a result of an adviser’s or financial
institution’s investment advice to a retirement investor. As proposed, the exemption was limited
to otherwise prohibited compensation generated by investments that are commonly purchased by
plans, participant and beneficiary accounts, and IRAs. Accordingly, the exemption defined the
“assets” that could be sold under the exemption as bank deposits, CDs, shares or interests in
registered investment companies, bank collective funds, insurance company separate accounts,
exchange-traded REITs, exchange-traded funds, corporate bonds offered pursuant to a
registration statement under the Securities Act of 1933, agency debt securities as defined in
FINRA Rule 6710(l) or its successor, U.S. Treasury securities as defined in FINRA Rule
6710(p) or its successor, insurance and annuity contracts (both securities and non-securities),
guaranteed investment contracts, and equity securities within the meaning of 17 C.F.R. 230.405
that are exchange-traded securities within the meaning of 17 C.F.R. 242.600. The Department’s
intent in proposing a limited definition of asset was to permit investment advice on the types of
investments that retirement investors typically rely on to build a basic diversified portfolio under
a uniform set of protective conditions.
Commenters representing the industry strenuously objected to the limited definition of
“asset.” Commenters took the position that the limited definition would be inconsistent with the
Department’s historical approach of declining to create a “legal list” of investments for plan
fiduciaries. Some commenters argued that Congress imposed only very narrow limits on the
types of investments IRAs may make, and therefore the Department should not impose other
limitations in an exemption.
Many commenters viewed the limited definition of asset as the Department substituting
its judgment for that of the adviser and stating which investments are permissible or “worthy.”
Some commenters believed that the best interest standard alone should guide the
281

recommendations of specific investments. Some asserted that the limitations could undermine
advisers’ obligation to act in the best interest of retirement investors. However, the Department
also received comments supporting the proposed definition of asset as an appropriate safeguard
of the exemption. These commenters expressed the view that the list was sufficiently broad to
allow an adviser to meet a retirement investor’s needs, while limiting the risks of other types of
investments. Retirement investors would still have access to these products under either a
pooled investment vehicle such as mutual funds, or a compensation model that does not involve
conflicted advice. Some commenters expressed support for exclusion of specific investment
products, such as non-traded REITs, private placements, and other complex products, indicating
these investments may be associated with extremely high fees. A commenter asserted that there
have been significant problems with recommendations of non-traded REITs and private
placements in recent years. Another commenter urged that the exemption not provide relief for
the recommendation of variable annuity contracts, although they were in the proposed definition
of asset.
After careful consideration of these comments, the Department eliminated the definition
of asset in the final exemption. In this regard, the Department ultimately determined that the
other safeguards adopted in the final exemption – in particular, the requirement that advisers and
financial institutions provide investment advice in accordance with the impartial conduct
standards, the requirement that financial institutions adopt anti-conflict policies and procedures;
and the requirement that financial institutions disclose their material conflicts of interest - were
sufficiently protective to allow the exemption to apply more broadly to all securities and other
investment property. If adhered to, these conditions should be protective with respect to all
investments. However, the Department cautioned in the preamble to the Best Interest Contract
Exemption that it expects advisers and financial institutions to exercise special care in
complying with their obligations under the exemption when recommending investment products
with attributes such as unusual complexity, illiquidity, risk, lack of transparency, high fees or
commissions, or tax benefits that are generally unnecessary in these tax preferred accounts.

7.10

Allowing Fixed-Indexed Annuities to be Covered under PTE 84-24

In the proposed amendment to PTE 84-24, the Department took the approach that
generally focused on revoking relief under PTE 84-24 for transactions involving IRAs and
variable annuities. The Department believed that the conditions included in the proposed Best
Interest Contract Exemption were more appropriate and protective of IRA investors purchasing
variable annuities. On the other hand, the Department was not certain that the conditions
proposed for the Best Interest Contract Exemption were readily applicable to insurance and
annuity contracts that are exempt securities.
The Department received many comments on the proposed Best Interest Contract
Exemption and the amendment to PTE 84-24 regarding the appropriate approach to annuity
contracts. Some commenters, expressing concern about the risks associated with variable
annuities, commended the Department for proposing that they should be recommended under the
conditions of this exemption rather than PTE 84-24. One commenter cited the provision of
FINRA’s Investor Alert, “Variable Annuities: Beyond the Hard Sell,” which says:

282

“Investing in a variable annuity within a tax-deferred account, such as an individual
retirement account (IRA) may not be a good idea. Since IRAs are already tax-advantaged, a
variable annuity will provide no additional tax savings. It will, however, increase the expense of
the IRA, while generating fees and commissions for the broker or salesperson.”557
Other commenters wrote that fixed annuities, particularly indexed annuities, should also
be subject to the requirements of the Best Interest Contract Exemption rather than PTE 84-24.
One commenter indicated that indexed and variable annuities raise similar issues with respect to
conflicted compensation, and that different treatment of the two would create incentives to sell
more indexed annuities subject to the less restrictive regulation.
Other commenters wrote that advisers and financial institutions should be able to rely on
PTE 84-24 for all insurance products, rather than bifurcating relief between two exemptions.
Commenters emphasized the benefit, for compliance purposes, of one exemption for all
insurance products. These commenters highlighted the importance of lifetime income options,
and the ways the Department, the Treasury Department and the IRS have worked to make
annuities more accessible to retirement investors. They expressed concern that the approach to
annuity contracts in the proposals could undermine those efforts.
In this regard, many commenters expressed concern that the disclosure requirements
proposed in the Best Interest Contract Exemption were inapplicable to insurance products and
that they would not be able to satisfy the best interest and other impartial conduct standards, or
provide a sufficiently broad range of assets to satisfy the conditions of Section IV of the
exemption, as proposed. Several raised questions about how the proposed definition of
“Financial Institution” would apply to insurance companies. According to these commenters,
the conditions proposed for this exemption would be so difficult and costly that broker-dealers
would stop selling variable annuities to certain IRA customers and retirement plans rather than
comply.
As discussed in detail above, in response to these comments, the final Best Interest
Contract Exemption has been revised so that the conditions identified by commenters are less
burdensome and more readily complied with by all financial institutions, including insurance
companies and distributors of insurance products. In particular, the Department has revised the
pre-transaction disclosure so that it does not require a projection of the total cost of the
recommended investment, which commenters indicated would be difficult to provide in the
insurance context. The Department also did not adopt the proposed data collection requirement,
which also posed problems for insurance products, according to commenters.
Further, the Department adjusted the language of the exemption in other places and
addressed interpretive issues in the preamble to address the particular questions and concerns
raised by the insurance industry. For example, the Department revised the “reasonable

557

“Variable Annuities: Beyond the Hard Sell;” available at: http://www.finra.org/sites/default/files/InvestorDocument/p125846.pdf. FINRA
also has special suitability rules for certain investment products, including variable annuities. See FINRA Rule 2330 (imposing heightened
suitability, disclosure, supervision and training obligations regarding variable annuities); see also FINRA rule 2360 (options) and FINRA
rule 2370 (securities futures).

283

compensation” standard throughout the exemption to address comments from the insurance
industry regarding the application of the standard to insurance transactions. Additionally,
guidance is provided regarding the treatment of insurers as financial institutions, within the
meaning of the exemption. Finally, the Department provided specific guidance in Section IV of
the exemption on satisfaction of the best interest standard by proprietary product providers.
Indexed annuities are complex products that can be confusing for consumers to
understand. Both the SEC staff and FINRA have issued guidance on indexed annuities. In its
2010 Investor Alert, “Equity-Indexed Annuities: A Complex Choice,” FINRA explained the
need for an Alert, as follows:
“Sales of equity-indexed annuities (EIAs) … have grown considerably in recent years.
Although one insurance company at one time included the word ‘simple’ in the name of its
product, EIAs are anything but easy to understand. One of the most confusing features of an
EIA is the method used to calculate the gain in the index to which the annuity is linked. To
make matters worse, there is not one, but several different indexing methods. Because of the
variety and complexity of the methods used to credit interest, investors will find it difficult to
compare one EIA to another.”
FINRA also explained that equity-indexed annuities “give you more risk (but more
potential return) than a fixed annuity but less risk (and less potential return) than a variable
annuity.”
Similarly, in its 2011 “Investor Bulletin: Indexed Annuities,” the SEC staff stated, “You
can lose money buying an indexed annuity. If you need to cancel your annuity early, you may
have to pay a significant surrender charge and tax penalties. A surrender charge may result in a
loss of principal, so that an investor may receive less than his original purchase payments. Thus,
even with a specified minimum value from the insurance company, it can take several years for
an investment in an indexed annuity to ‘break even.’”
Given the risks and complexities of these investments, the Department has determined
that indexed annuities are appropriately subject to the same protective conditions of the Best
Interest Contract Exemption as apply to variable annuities. These are complex products
requiring careful consideration of their terms and risks. Assessing the prudence of a particular
indexed annuity requires an understanding, inter alia, of surrender terms and charges; interest
rate caps; the particular market index or indices to which the annuity is linked; the scope of any
downside risk; associated administrative and other charges; the insurer’s authority to revise
terms and charges over the life of the annuity; and the specific methodology used to compute the
index-linked interest rate; and any optional benefits that may be offered, such as living benefits
and death benefits. In operation, the index-linked interest rate can be affected by participation
rates; spread, margin or asset fees; interest rate caps; the particular method for determining the
change in the relevant index over the annuity’s period (annual, high water mark, or point-topoint); and the method for calculating interest earned during the annuity’s term (e.g., simple or
compounded interest). Investors can all too easily overestimate the value of these contracts,
misunderstand the linkage between the contract value and the index performance, underestimate
the costs of the contract, and overestimate the scope of their protection from downside risk (or
wrongly believe they have no risk of loss). As a result, retirement investors are acutely
dependent on sound advice that is untainted by the conflicts of interest posed by advisers’
incentives to secure the annuity purchase, which can be quite substantial. Both categories of
annuities, variable and indexed annuities are susceptible to abuse, and retirement investors
would equally benefit in both cases from the protections of the Best Interest Contract
284

Exemption, including the conditions that clearly establish the enforceable standards of fiduciary
conduct and fair dealing as applicable to advisers and financial institutions.
Moreover, as discussed in Chapter 2 above, there is uncertainty regarding the regulation of
indexed annuities that creates a complicated and confusing regulatory landscape for consumers.
The SEC attempted to regulate such annuities under Federal securities law in a 2009
rulemaking.558 However, the rule was challenged shortly after it was issued and overturned on
procedural grounds by the U.S. Court of Appeals for the D.C. Circuit.559 As part of the DoddFrank Act, Congress directed the SEC to treat insurance policies or annuity contracts (which
could include indexed annuity contracts) as exempt securities as long as, among other
requirements, the product is issued in a state that has adopted suitability requirements which
substantially meet or exceed the minimum requirements established by the NAIC Suitability in
Annuity Transactions Model Regulation or by an insurer that adopts and implements practices on
a nationwide basis for the sale of annuity contracts that meet or exceed the Model Suitability
Regulation.560 However, a recent report indicates that only approximately two-thirds of the
states have adopted some version of the Model Suitability Regulation.561 This has created
uneven protections and confusion for consumers, because individual state insurance laws are
neither uniform nor consistent.
The Department notes that many insurance industry commenters stressed a desire for one
exemption covering all insurance and annuity products. The Department agrees that efficient
compliance with fiduciary norms is promoted by a common set of requirements, but concluded
that the Best Interest Contract Exemption is best suited to address the conflicts of interest
associated with variable annuities, indexed annuities, and similar investments, rather than the
less stringent PTE 84-24. Accordingly, the Department has limited the availability of PTE 84-24
to “fixed rate annuity contracts,” as defined in the exemption, while requiring advisers
recommending variable and indexed annuities to rely on the Best Interest Contract Exemption,
which is broadly available for any kind of annuity or asset, subject to its specific conditions. In
this manner, the final exemption creates a level playing field for variable annuities, indexed
annuities, and mutual funds under a common set of requirements, and avoids creating a
regulatory incentive to preferentially recommend indexed annuities.
The Department conservatively assumes in Chapter 5 that all insurance companies would
incur compliance costs associated with the Best Interest Contract Exemption; however, it
estimates that providing relief for fixed-indexed annuities under PTE 84-24 instead of under the
Best Interest Contract Exemption would have reduced its cost estimate by between $34.0 million
and $37.8 million over ten years. The largest costs associated with the Best Interest Contract
Exemption are fixed costs that are triggered during the first instance that a financial institution
uses the exemption. These costs are borne by financial institutions whether they use the
exemption once or regularly. Therefore, the financial institutions that would be most likely to
realize significant cost savings from providing relief for fixed-indexed annuities under PTE 84-

558
559
560
561

17 C.F.R. 230.151(b).
Am Equity Inv. Life Ins. Co. v. SEC, 613 F.3d 166 (D.C. Cir. 2010)
Dodd-Frank Act, Pub L. No. 111-203,Ssection 989J, 124 Stat. 136, 1949-50 (2010).
See Nat’l Ins. Producer Registry, Annual Report 9 (2013), available at: http://www.nipr.com/docs/annual-reports/2013-nipr-annualreport.pdf.

285

24 instead of under the Best Interest Contract Exemption are those financial institutions that
would not sell any other products requiring relief under the Best Interest Contract Exemption.
The Department has identified 12 life insurance companies that sell fixed-indexed
annuities, but whose total reported sales volume does not include any variable annuity sales.
The Department believes that these companies are the ones most likely to use the Best Interest
Contract Exemption to sell indexed annuities but no other products.562 These insurance
companies are small relative to the annuity provider market at large. The Department
conservatively assumes that all of these insurance companies service the ERISA plan and/or IRA
markets. Additionally, the Department has information on the companies’ size relative to other
life insurers in the annuity market, and it accordingly placed the companies in the appropriate
size categories. This produced a Departmental estimate of no large insurance companies, nine
medium insurance companies and three small insurance companies who the Department is most
certain would be impacted by these policy decisions.563 As discussed above, the per-firm costs
for insurance companies to comply with the Best Interest Contract Exemption are $6.6 million
for large, $1.4 million for medium, and $464,000 for small firms for start-up costs and $1.7
million for large, $306,000 for medium, and $100,000 for small firms for ongoing costs.
Therefore, the reduction to the cost estimate resulting from allowing fixed-index annuities to be
covered under PTE 84-24 would have been $14.1 million during the first year, $3.1 million
during subsequent years, and $34.0 million to $37.8 million over ten years.

7.11

Wait for SEC Action

Some commenters advised the Department to postpone its regulatory initiative until the
SEC has completed its rulemaking activities under the Dodd-Frank Act with respect to
establishing a uniform fiduciary standard of conduct for BDs and RIAs under the federal
securities law.564 The commenters asserted that if the Department moves forward with its
regulatory initiative, inconsistent rules and increased costs and complexities would result for
participants, beneficiaries, and IRA investors who have different types of accounts at the same
financial institution.
A number of consumer groups, in contrast, have jointly taken a strong position against
delay of Departmental action. The groups argue that ERISA and the IRC do not conflict with
other laws governing financial advice, and that retirement accounts merit special protection.565
After carefully considering this issue, the Department selected the alternative to proceed
without waiting for the SEC to act. The Department has consulted with SEC staff over the past
three years when it was developing the 2015 Proposal to avoid conflicts between the final rule

562

563

564
565

To identify the companies that sell only fixed income annuities, the Department compared Tables B-1 and B-3 of U.S. Individual Annuity
Yearbook — 2014 Loma Secure Retirement Institute (LIMRA). Of those companies who reported positive fixed indexed annuity sales
volume, the Department concluded that the insurance company only sells fixed indexed annuities if its total reported annuity sales volume
matches its total reported fixed annuity sales volume.
The U.S. Individual Annuity Yearbook Tables B-1 and B-3 contain individual data on the largest 76-79 insurance companies in the US
market and then total data for the remaining companies. Therefore, due to data constraints, it is possible that this measure undercounts
small insurance companies that sell indexed annuities and do not sell variable annuities.
The SEC’s Semiannual Regulatory Agenda states that the SEC will issue a proposed rule in October 2016.
October 18, 2013 letter from Fund Democracy, Consumer Federation of America, AARP, Americans for Financial Reform, and Public
Citizen’s Congress Watch to The Honorable Sylvia Matthews Burwell, Director, Office of Management and Budget, Executive Office of
the President. Available at: http://www.consumerfed.org/pdfs/Fiduciary-Sign-On-Letter-10-18-13.pdf.

286

and exemptions and federal securities law. Although the Department and the SEC have different
statutory responsibilities, both agencies recognize the importance of working together on
regulatory issues in which our interests overlap, particularly where action by one agency may
affect the community regulated by the other agency. To that end, the Department has sought
technical assistance from the SEC on the development of this rule. The technical assistance that
the SEC staff has provided has helped the Department in its efforts to ensure that the rule
achieves the goal of striking a balance between protecting individuals looking to build their
savings and minimizing disruptions to the many good practices and good advice that the
financial services industry provides today.
The SEC staff provided technical assistance on all aspects of the Department’s Proposal,
including the regulatory impact analysis. The Department has made numerous changes in
response to observations and issues raised by the SEC staff and is grateful for the staff’s
technical assistance.
The fiduciary duties included in ERISA and the IRC are different from those applicable
to RIAs under the Advisers Act. The duties would continue to differ even if both regimes were
interpreted to attach fiduciary status to exactly the same parties and activities. When Congress
enacted ERISA, it provided that all investment advisers to plan and IRA investors would be
subject to the ERISA and/or IRC fiduciary regime. Importantly, compared with securities laws,
ERISA and the IRC are generally less tolerant of fiduciary conflicts of interest, and in that
respect provide a higher level of protection to plan participants and IRA investors, reflecting the
importance of plans and IRAs to retirement security, and the tax subsidies they enjoy. The
ERISA and IRC standards applicable to fiduciary investment advisers will continue to overlap
with, and differ from, those applicable under securities laws, irrespective of whatever regulatory
action the Department or SEC takes. Indeed, BDs and RIAs who provide investment advice
with respect to plan or IRA assets, and satisfy the 1975 regulation’s five-part test, are already
subject to the fiduciary provisions of ERISA and/or the IRC.
The Department understands the roles of the SEC and other federal and state agencies in
regulation of financial advice provided to retail investors. At the same time, however, the IRC
prohibited transactions provisions, as enacted by Congress as part of ERISA in 1974,
specifically apply to IRA investment advice, and the Department is solely responsible for
interpreting these provisions.566 It is thus incumbent on the Department to protect IRA investors
from harmful adviser conflicts. An examination of trends and evidence accumulated since 1974
suggests that such special protections, if anything, are even more critical today than when
Congress first enacted ERISA more than 40 years ago. The Department’s role in applying these
protections is well established under law and in practice.
IRAs’ important role in retirement security, which warrants special protections against
conflicts in advice, underscores the need for the final rule and exemptions to ensure the broad
application of these protections.
IRAs were established in 1974 as a vehicle to promote retirement savings. In supporting
IRAs, lawmakers pointed to the need to provide tax preferences similar to those applicable to

566

See Reorganization Plan No. 4 1978, 5 U.S.C. § App. (2010).

287

job-based pensions to workers who did not have access to such pensions. They also pointed to
rollover IRAs’ potential to make job-based pensions more portable.
The special protections for IRAs embodied in the IRC prohibited transactions provisions
are mirrored by the large tax subsidies IRAs enjoy under other IRC provisions. These subsidies
are estimated to amount to $17 billion in 2016 alone.567 This figure dramatically understates the
degree to which current IRA savings have been subsidized by taxpayers, however. Most of the
savings flowing into IRAs come not from direct contributions but from rollovers primarily from
job-based retirement plans, (mostly from DC plans, including 401(k)s)568 – and much of the
savings currently in these plans may eventually be rolled over into IRAs. The tax preference for
DC plans is estimated to amount to $65 billion in 2016.569 Moreover, these IRA and DC figures
vastly understate the accumulated taxpayer subsidy in DC and IRA savings, reflecting only oneyear’s subsidy.
Delay associated with waiting for SEC action would impose substantial costs on plan
participants and IRA investors, as current harms from conflicted advice would continue.
Likewise, delay would defer, but do little to reduce, compliance costs. Prompt action is
warranted. The Department’s issuance of the final rule and exemptions before the SEC issues a
rule is no less conducive to harmonization than would be the reverse order. The Department
therefore has elected to proceed without undue delay.
The Department agrees with commenters, however, that overlapping rules should not
conflict with or impair other applicable rules. The Department has taken care to adhere to
ERISA’s and the IRC’s specific text and purposes. At the same time, however, the Department
has sought to understand the impact of the final rule and exemptions on firms subject to the
securities laws and other federal or state laws, and to take the effects into account by
appropriately calibrating the impact of the rule on those firms. In the Department’s view, the
final regulation neither undermines nor contradicts the provisions or purposes of the securities
laws or other federal or state laws. Instead, the regulation has been designed to work in harmony
with other federal laws, and the Department has consulted – and will continue to consult – with
other federal agencies to ensure that, to the extent possible, the various legal regimes are
appropriately harmonized.

7.12

Alternative “Best Interest” Conduct Standard Formulations

A number of commenters suggested alternatives to the Department’s approach taken in
the proposed Best Interest Contract Exemption to formulate a “best interest” conduct standard.
Some commenters suggested alternative approaches that included a standard characterized as a
“best interest” standard of conduct, combined with certain of the other safeguards that the
Department had proposed, including reasonable compensation, disclosures of material conflicts,
or anti-conflict policies and procedures.
In some instances, commenters indicated that a different best interest standard would be
appropriate but failed to provide an alternative to the Department’s definition. Others suggested

567

568
569

Office of Management and Budget, “Fiscal Year 2017 Analytical Perspectives” (2016), 231. Available at:
https://www.whitehouse.gov/sites/default/files/omb/budget/fy2017/assets/spec.pdf.
ICI, “U.S. Retirement Market, Third Quarter 2015,” 2015.
Office of Management and Budget, “Fiscal Year 2017 Analytical Perspectives,” 2016.

288

a definition of “best interest” that did not include a duty of loyalty constraining advisers from
making recommendations based on their own financial interests, or from subordinating the
interests of retirement investors to their own financial interests. Some of these definitions
focused exclusively on the fiduciary obligation of prudence, while excluding the equally
fundamental fiduciary duty of loyalty.
As a general matter, the Department adopted its best interest formulation because none of
the suggested alternative approaches incorporated all the components that the Department views
as essential to making the required findings for granting an exemption, or provided alternatives
that included conditions that would appropriately safeguard the interests of retirement investors
in light of the exemption’s broad relief from the conflicts of interest and self-dealing prohibitions
under ERISA and the Code. The Department remains convinced of the critical importance of the
core requirements of the exemption, including an up-front commitment to act as a fiduciary,
enforceable adherence to the impartial conduct standards, the adoption of policies and
procedures reasonably designed to assure compliance with the impartial conduct standards, a
prohibition on incentives to violate the best interest standard, and fair disclosure of fees,
conflicts of interest, and material conflicts of interest. In addition, in contrast to many of the
proposed alternatives, the Department’s approach generally prevents firms and advisers from
contracting out of these basic obligations or waiving them through disclosure. As discussed
above and in the preambles to the Rule and Best Interest Contract Exemption, the Department
has concluded that the ability to disclaim these obligations would result in a large loophole that
would largely negate the consumer-protection purposes of this regulatory initiative.
The Impartial Conduct Standards require adherence to basic fiduciary norms and
standards of fair dealing -- rendering prudent and loyal advice that is in the best interest of the
customer, receiving no more than reasonable compensation, and refraining from making
misleading statements. These fundamental standards enable the Department to grant an
exemption that flexibly covers a broad range of compensation structures and business models,
while safeguarding the interest of retirement investors against dangerous conflicts of interest.
The conditions were critical to the Secretary of Labor’s ability to make the required findings
under ERISA Section 408(a) and Code Section 4975(c)(2) that the associated exemptions,
including the Best Interest Contract Exemption, were administratively feasible, in the interests of
plans, participants, beneficiaries, and IRA owners and protective of their interests.
In addition, most of the best interest standard alternatives suggested by commenters
appeared to lack a clear means of enforcement. A number of commenters suggested they could
abide by a best interest standard but at the same time objected to the enforcement mechanisms
that the Department proposed, particularly in the IRA market. As stated in the Section 7.12
above, the Department does not believe that these alternatives will adequately protect retirement
investors, particularly those in the IRA market, from harmful conflicts of interest, or that
financial institutions and their advisers will be properly incentivized to comply with a best
interest standard, if there is no enforceable mechanism for retirement investors to enforce
adherence to that standard or to obtain redress when they’ve been injured by violation of the
standards. From the perspective of retirement investors, a right without a remedy is scarcely a
right at all.
A number of commenters expressed particular concern about the application of the
Department’s Best Interest requirement that the recommendation be made “without regard to the
financial or other interests of the Adviser, Financial Institution” or other parties. Some of these
commenters suggested that the Department use different formulations that were similar to the
Department’s, but might be construed to less stringently forbid the consideration of the financial
289

interests of persons other than the retirement investor. In response to commenter concerns, the
Department created a specific “Best Interest” test for advisers and financial institutions that
make recommendations from a restricted range of investments based on proprietary products or
investments that generate third-party payments. The test makes clear how such firms can satisfy
the “best interest” standard by satisfying specific conditions, even as they limit the universe of
recommended products. Outside of this context, the Department has retained the “without
regard to” language as best capturing the exemption’s intent that the Adviser’s recommendations
be based on the Investor’s best interest. This approach also accords with ERISA’s
requirement570 that plan fiduciaries act “solely in the interest” of plan participants and
beneficiaries.

7.13

Issue a Streamlined, “Low-Cost Safe Harbor” PTE

The Department considered issuing a PTE that would effectively provide relief from the
relevant ERISA and IRC PT provisions based on recommendations of high quality low-fee
investments that appear unlikely to cause harm associated with conflicted advice. Such a PTE
might establish thresholds for “low cost” investments that advisers could recommend. The PTE
would allow the adviser and firm to receive a wide range of variable and third-party
compensation. The PTE might attach no or few additional conditions.
The aims of such a PTE might include:


Providing a targeted reduction in regulatory burdens where market failures are absent
or very small by enabling fiduciary advisers to accept variable, third-party
compensation without having to satisfy the conditions of the Best Interest Contract
Exemption;



Rewarding and encouraging best-practices with respect to optimizing the quality,
amount, and combined, all-in cost of recommended financial products, financial
advice, and any other bundled services; and



Ensuring small savers’ access to quality, affordable financial products and advice.

If these aims could be achieved, the PTE might make it possible to secure some investor
gains at lower cost than otherwise achievable under the new rule and exemptions. However, the
Department identified a number of practical challenges in designing and implementing a PTE
that would achieve these aims, and therefore did not include one in the proposal. The
Department invited comments on the feasibility and desirability of such a PTE, whether such a
PTE would achieve its intended aims or other beneficial aims (or have unintended negative
consequences), as well as comments that identify practical design and implementation
challenges, and offer suggestions to overcome such challenges.
In the proposal, the Department noted that although it finds the idea attractive, it has been
unable to operationalize the high quality low-fee streamlined exemption and therefore did not
propose text for such an exemption. Instead, the Department sought public input to assist its
consideration and design of the exemption. The Department asked a number of specific
questions, including which products should be included, how the fee calculations should be

570

ERISA Section 404(a)(1).

290

established, performed, communicated, and updated, what, if any additional conditions should
apply, and how a streamlined exemption would affect the marketplace for investment products.
The vast majority of commenters were opposed to creating a streamlined exemption for
low-fee products. They expressed the view that this would over-emphasize the importance of
fees, despite prior Department guidance noting that fees are just one of many factors to consider.
Commenters also raised many of the same operational concerns the Department had raised in the
preamble, such as identifying the appropriate fee cut off, as well as the potential for undermining
suitability and fiduciary obligations under securities laws, with a sole focus on products with low
fees.
The Department did receive a few comments in support of a low-fee streamlined
exemption. These commenters generally recommended that the exemption be limited to certain
investments, most commonly mutual funds, and perhaps just those with fees in the bottom five
or ten percent. One commenter requested a carve-out from the Regulation’s definition of
“fiduciary,” or a streamlined exemption, for retirement investments in high-quality, low-cost
financial institutions savings products, like CDs, when a direct fee is not charged and a
commission is not earned by the bank employee. Other commenters were willing to consider a
low fee streamlined exemption, but argued that more information was necessary and any such
exemption would need to be proposed separately.
The commenters’ concerns described above echoed the Department’s concerns regarding
the low-fee streamlined exemption. Therefore, despite some limited support, the Department
chose the alternative of not proceeding with a low-fee streamlined exemption. In particular, the
Department did not receive enough information in the comments to address the significant
conceptual and operational concerns associated with the approach. For example, after
consideration of the comments, the Department was unable to conclude that the streamlined
exemption would result in meaningful cost savings. Most financial institutions and advisers
would likely only be able to rely on such a streamlined exemption in part. The adviser would
still need to comply with this exemption for many of the investments he or she would
recommend outside of the streamlined exemption. Many of the costs associated with this
exemption are upfront costs (e.g., policies and procedures, contracts), that the financial
institution would have to incur whether or not it used the streamlined exemption. As a result, the
streamlined exemption may not have resulted in significant cost savings. In addition, the e
Department was unable to overcome the challenges it saw in using a low-fee threshold as a
mechanism to jointly optimize quality, quantity, and cost. Fundamentally, it is unclear how to
set a “low-fee” threshold that achieves all of these aims. A single threshold could be too low for
some investors’ needs and too high for others’. Further, any threshold might encourage the
lowest existing prices to rise to its level, potentially harming investors.

7.14

Delaying Applicability Date

The 2015 proposal stated that the final rule would be effective 60 days after publication
in the Federal Register and the requirements of the final rule and exemptions would generally
become applicable eight months after publication of a final rule and related PTEs. Many
commenters expressed concern with the applicability date and asserted that eight months were
wholly inadequate due to the time and budget requirements necessary to make changes to
information systems, compliance processes, and compensation structures. Some commenters
suggested that the applicability date should be extended to as much as 18 to 36 months (and
some suggested even longer, e.g., five years) following publication to allow service providers
sufficient time to make changes necessary to comply with the new rule and exemptions.
291

Many other commenters asked the Department to provide a grandfather rule or similar
rule for existing contracts or arrangements or a temporary exemption permitting all currently
permissible transactions to continue for a certain period of time. As part of these concerns, some
commenters highlighted possible challenges with compliance, asking the Department to provide
that good faith efforts with reasonable diligence to comply with the terms of the rule and related
exemptions are sufficient for compliance and one requested a stay on enforcement of the rule for
36 months. Some commenters suggested that the Department delay the application of the rule
until such time as financial institutions could make the changes to their practices and
compensation structures necessary to comply with the conditions of the Best Interest Contract
Exemption or re-propose or issue the rule and exemptions as interim final rules with requests for
additional comment Other commenters thought that the effective and applicability dates in the
proposal were reasonable and asked that the final rule go into effect promptly in order to reduce
ongoing harms to savers that result from conflicts of interest.
The Department, after carefully considering all of these alternatives, has chosen the
alternative of retaining the 2016 Final Rule 60-day effective date, (and making the exemptions’
effective date the same), while delaying the applicability date of both for one year after
publication in the Federal Register. Sixty days is the earliest point that major rules can become
final under the Congressional Review Act. The Department adopted this alternative to provide
certainty to plans, plan fiduciaries, plan participants and beneficiaries, IRAs, and IRA owners
that the new protections afforded by the final rule and new and amended PTEs are now officially
part of the law and regulations governing their investment advice providers. Similarly, the
financial services providers and other affected service providers will also have certainty that the
2016 Final Rule and associated PTEs are final. The Department expects that retaining the
effective date will remove uncertainty as an obstacle to regulated firms allocating resources
toward transition and longer term modifications to compliance systems and business practices.
After considering comments, the Department lengthened the initial applicability date to
one-year and provided transition relief for certain PTE conditions discussed below but did not
provide transition relief for the final rule itself of for the obligation of fiduciaries to adhere to the
impartial conduct standards. The public comments requesting such relief generally expressed
concerns with meeting the conditions of the new and amended prohibited transaction exemptions
by the applicability date rather than with the final rule itself. The Department believes that an
applicability date that is 12 months after the date of publication, provides adequate time for plans
and their affected financial services and other service providers to adjust to the basic change
from non-fiduciary to fiduciary status. The Department believes that delaying the application of
the new fiduciary rule and impartial conduct standards would inordinately delay the basic
protections of loyalty and prudence that the rule provides. Moreover, a long period of delay
could incentivize financial institutions to increase efforts to provide conflicted advice to
retirement investors before they become subject to the new rule.
The Department agrees that it is appropriate to provide transition relief for satisfaction of
the full exemption conditions to ensure that financial institutions and advisers have sufficient
time to prepare for full compliance. Therefore, the Department has chosen the alternative of
providing relief from the ERISA and IRC prohibited transactions rules for financial institutions
and advisers using the Best Interest Contract and Principal Transactions Exemptions during the
period between the Applicability Date and January 1, 2018 (the “transition period”). During the
transition period, full relief under the exemption will be available for financial institutions and
advisers subject to more limited conditions that provide significant safeguards to mitigate the
harmful effects of conflicts of interest by requiring prompt implementation of certain core
292

protections of the exemption. Providing this relief will have the effect of making compliance
with the exemption more workable and efficient and spreading out or reducing upfront costs.
Specifically, during the transition period, the financial institution and its advisers must
comply with the impartial conduct standards contained in the exemptions when making
recommendations to retirement investors. In addition, the financial institution must designate a
person or persons, identified by name, title or function, responsible for addressing material
conflicts of interest and monitoring advisers’ adherence to the impartial conduct standards.
During the transition period, the financial institution must also provide a written notice to the
retirement investor acknowledging its and its adviser(s)' fiduciary status with respect to the
transaction. The financial institution must state in writing that it and its advisers will comply
with the impartial conduct standards and describe its material conflicts of interest. Further, the
financial institution’s notice must disclose whether it recommends proprietary products or
investments that generate third-party payments; and, to the extent the financial institution or
adviser limits investment recommendations, in whole or part, to proprietary products or
investments that generate third-party payments, the financial institution must notify the
retirement investor of the limitations placed on the universe of investment recommendations.
Finally, the financial institution must comply with the recordkeeping provision of the exemptions
regarding transactions entered into during the transition period.
After the transition period, financial institutions and advisers must satisfy all of the
applicable conditions of the exemption for any prohibited transactions occurring after that date.
This includes the requirement in the Best Interest Contract Exemption and Principal Transactions
Exemption to enter into a contract with a retirement investor, where required, affirmatively
warrant that the financial institution has adopted and will comply with written policies and
procedures reasonably and prudently designed to ensure that its individual advisers adhere to the
impartial conduct standards and provide the required disclosures.
While the exemption for the transition period does not require a warranty or disclosure of
anti-conflict policies and procedures, the Department envisions that financial institutions will
nevertheless adopt prudent supervisory mechanisms to ensure that advisers comply with the
impartial conduct standards during the transition period. And additionally, while a contract is not
required, the acknowledgment of fiduciary status provides an important piece of evidence for any
retirement investor wishing to bring a claim against a financial institution or adviser as a result of
advice provided during the transition period.
The final rule and exemptions have the potential to provide significant gains to retirement
investors by requiring fiduciary IRA advisers to forgo conflicted fee structures when providing
fiduciary advice to IRA investors or to provide advice that is in their clients’ best interest or
satisfy the protective conditions of the Best Interest Contract Exemption. As discussed in detail
in Chapter 3, even taking into account the gradual movement of IRA assets into more optimal
investments, and backing out improvements in cost-effectiveness that might be expected without
the new rule and exemptions, the Department expects that the new rule and exemptions have the
potential to restore to IRA investors approximately $33 billion to $36 billion over 10 years and

293

$66 billion to $76 billion over 20 years, even if one just considers the type of investment product
and typical conflict of interest involving front-end-load mutual funds.
By choosing the alternative of providing transition relief for the conditions of the Best
Interest Contract Exemption, the rule’s short-term effectiveness may be below 100 percent.
However, the potential gains to investors remain significant.571 For example, if the rule and
exemptions are only 50% effective in the first year (when several of the Best Interest Contract
Exemption and Principal Transactions Exemption provisions are not yet in effect), the quantified
subset of gains – specific to the front-end-load mutual fund segment of the IRA market – would
amount to between $30 billion and $33 billion over 10 years.572

7.15

Providing Streamlined Conditions in Best Interest Contract
Exemption for “Level-Fee Fiduciaries”

Several commenters asked whether a fiduciary investment adviser would need to
utilize the Best Interest Contract Exemption if the only compensation the adviser receives is a
fixed percentage of the value of assets under management. Whether a particular relationship
or compensation structure would result in an adviser having an interest that may affect the
exercise of its best judgment as a fiduciary when providing a recommendation, in violation of
the self-dealing provisions of prohibited transaction rules under Section 406(b) of ERISA,
depends on the surrounding facts and circumstances. The Department believes that, by itself,
the ongoing receipt of a level-fee, such as a fixed percentage of the value of a customer’s
assets under management, where such values are determined by readily available independent
sources or independent valuations, typically would not raise prohibited transaction concerns
for the adviser or financial institution. Under these circumstances, the compensation amount
depends solely on the value of the investments in a client account, and ordinarily the interests
of the adviser in making prudent investment recommendations, which could have an effect on
compensation received, are aligned with the investor’s interests in growing and protecting
account investments.
However, there is a clear and substantial conflict of interest when an adviser
recommends that a participant roll money out of a plan into a fee-based account that will
generate ongoing fees for the adviser that he would not otherwise receive, even if the fees
going forward do not vary with the assets recommended or invested. Similarly, the prohibited
transactions rules could be implicated by a recommendation to switch from a low activity
commission-based account to an account that charges a fixed percent of assets under
management on an ongoing basis.
Although the Department considered requiring advisers in these circumstances to
comply with all provisions of the Best Interest Contract Exemption, the Department decided
to streamline the provisions applicable to such “level fee fiduciaries” in Section II(h) of the

571

572

According to the Department’s estimates, gains-to-investors begin accruing once new advice is given following the applicability date of the
final rule.
These gain estimates exclude additional potential gains to investors For example, these potential additional gains include improvements in
the performance of IRA investments other than front-end-load mutual funds and potential reductions in excessive trading and associated
transaction costs, and timing errors (such as might be associated with return chasing), as well as additional potential gains attributable to the
application of fiduciary standard of care.

294

exemption, because the prohibited transaction in these examples is relatively discrete and the
provision of advice thereafter generally does not involve prohibited transactions.573
A financial institution and its advisers are level fee fiduciaries if they and their affiliates
only receive a “level fee” as defined in the exemption that is disclosed in advance to the
retirement investor, for the provision of advisory or investment management services regarding
the plan or IRA assets. This occurs most frequently in the case of a recommendation to rollover
assets from an ERISA plan to an IRA, where the adviser is going to manage or provide
investment advice on an ongoing basis regarding the IRA’s assets. However, the streamlined
conditions applicable to level fee fiduciaries are not limited to rollover recommendations
scenario and would apply in any case in which only a fixed percentage of the value of the assets
under management will be paid. Although “robo-advice providers”574 are generally carved out
of the Best Interest Contract exemption, this streamlined exemption is available to them too to
the extent they satisfy the definition of level fee fiduciary and comply with the exemption’s
conditions.
The following streamlined conditions apply to such level-fee fiduciaries: (i) Prior to the
recommended transaction, the financial institution must provide the retirement investor with a
written statement of the financial institution’s and its advisers’ fiduciary status; (ii) The financial
institution and adviser must comply with the exemption’s impartial conduct standards of Section
II(c); and (iii) certain documentation must be created. Specifically, in the case of a
recommendation to roll over from an ERISA plan to an IRA, a rollover from another IRA, or a
switch from a commission-based account to a fee-based account, the financial institution must
document the specific reasons why the recommendation was considered to be in the best interest
of the retirement investor. In the case of a recommended rollover from an ERISA plan, this
documentation must include consideration of the retirement investor’s alternatives to a rollover,
including leaving the money in his or her current employer’s plan, if permitted, and must take
into account the fees and expenses associated with both the plan and the IRA; whether the
employer pays for some or all of the plan’s administrative expenses; and the different levels of
services and investments available under each option. In the case of a recommendation to
rollover from another IRA or switch from a commission-based account to a level fee
arrangement, the level fee fiduciary must document the reasons that the level fee arrangement is
considered to be in the best interest of the retirement investor, including, specifically, the
services that will be provided for the level fee.

573

574

In general, after the rollover, the ongoing receipt of compensation based on a fixed percentage of the value of assets under management
does not require a prohibited transaction exemption. However, certain practices involve violations that would not be eligible for the relief
granted in the Best Interest Contract Exemption. For instance, if an adviser compensated in this manner engaged in “reverse churning,” or
recommended holding an asset solely to generate more fees for the adviser, the adviser’s behavior would constitute a violation of section
406(b)(1) of ERISA that is not covered by the Best Interest Contract Exemption or its Level Fee provisions. In its “Report on Conflicts of
Interest” (Oct. 2013) , p. 29, FINRA suggests a number of circumstances in which advisers may recommend inappropriate commission- or
fee-based accounts as means of promoting the adviser’s compensation at the expense of the customer (e.g., recommending a fee-based
account to an investor with low trading activity and no need for ongoing monitoring or advice; or first recommending a mutual fund with a
front-end-sales-load, and shortly later, recommending that the customer move the shares into an advisory account subject to asset-based
fees). Such abusive conduct, which is designed to enhance the adviser’s compensation at the retirement investor’s expense, would violate
the prohibition on self-dealing in ERISA section 406(b)(1) and Code section 4795(c)(1)(E), and fall short of meeting the impartial conduct
standards required for reliance on the Best Interest Contract Exemption and other exemptions.
Robo-advice providers, as defined in the exemption, furnish investment advice to a retirement investor generated solely by an interactive
web site in which computer software-based models or applications make investment recommendations based on personal information each
investor supplies through the web site without any personal interaction or advice from an individual adviser.

295

Including this alternative in the Best Interest Contract Exemption results in savings of
$20.1 billion to $38.7 billion in quantified cost over ten years relative to if the provision were not
included, because the written contract, policies and procedures, warranty, website disclosure and
pre-transactions disclosure requirements do not apply to level fee fiduciaries.575

7.16

Conclusion

In conclusion, the Department considered a large variety of important regulatory
alternatives, in finalizing the rule and new amended PTEs including several identified by
commenters. The qualitative and, where possible, quantitative assessments of these alternatives
(detailed immediately above) suggest that none would protect plan and IRA investors as
effectively as the Department’s chosen alternatives. At the same time, the Department has made
many changes in the final rule and exemptions that make them more workable and less
burdensome than the 2015 Proposal to ensure that their protections are effective without
imposing unreasonable costs.

575

Estimates are obtained by comparing costs accruing to RIAs as estimated under the Final rule and when the number of RIAs using the Best
Interest Contract Exemption is set to 100% in the model. The range is obtained by comparing the SIFMA and FSI medium cost reduction
scenario. For these purposes, the Department additionally assumed that 63 percent of all service providers are affected by the rule and
applied a three percent discount rate.

296

297

8. Uncertainty
As detailed in Chapters 3 and 4 above, this regulatory analysis finds that conflicted
advice is widespread, causing serious, avoidable harm to plan and IRA investors. By extending
fiduciary status to more advice and providing flexible and protective PTEs that apply to a broad
array of compensation arrangements, the final rule and exemptions aim to mitigate conflicts and
deliver substantial gains for retirement investors and other economic benefits that more than
justify its costs. The Department’s analysis (in Chapter 3) quantifies only a subset of gains that
the Department estimates will accrue to retirement investors.
This chapter details uncertainties attendant to the Department’s analysis. Sections 8.1
and 8.2 respectively examine uncertainty associated with the magnitude of the harm attributable
to advisory conflicts and the magnitude of investor gains that are anticipated under this final rule
and exemptions. These sections consider uncertainties attendant to both the partial, quantitative
estimates of harm and gains provided in this analysis and the additional, large but unquantified
harms and gains documented herein. Section 8.3 examines uncertainty related to compliance
costs.
Section 8.4 addresses uncertainties associated with the final rule and exemptions’
potential secondary market effects. Anticipated secondary effects include changes in the nature,
delivery and compensation of plan and IRA advisory services, and consequent changes in plan
and IRA investments and associated mix of financial products and deployment of capital.
Comments on the 2015 Proposal expressed divergent views with respect to these secondary
effects. Some predicted major negative effects, arguing that the proposal would sharply increase
the price of advice and thereby erode small investors’ (and small plans’) access to beneficial
advice and with it, their prospects for retirement security. Others disagreed, predicting
improvements in the transparency and efficiency of markets for investment advice and financial
products that would benefit consumers and society at large. This analysis concludes that
structural secondary effects will be highly positive. Negative effects generally will be confined
to short-term frictions.

8.1

Magnitude of Harm

This analysis documents large negative economic effects of advisory conflicts. The
direct negative effects manifest as underperformance576 suffered by plan and IRA investors
exposed to advisory conflicts. This analysis quantifies a subset of these effects, and documents
but does not quantify additional such effects. The Department is confident that this harm is
more than large enough to warrant the regulatory intervention embodied in the final rule and
exemptions. The exact magnitude of this harm is subject to uncertainty, however, with respect
to both the precision of the Department’s estimates of the quantified subset of the harm and the
magnitude of the additional, unquantified harm.

576

Underperformance in this context broadly includes any avoidable decrement in the risk-adjusted net value delivered to a consumer by a
financial product or advisory service below that which would be delivered in an efficient market. Some examples of such
underperformance include systematically sub-par returns to a mutual fund, transaction costs that exceed related benefits (such as costs
associated with excessive turnover of portfolios), inefficiently high prices paid for annuity-related guarantees, losses (or opportunity costs)
from avoidable timing errors, and mismatching of investments with investors’ preferences.

298

8.1.1 Quantified Harm
As detailed in Chapter 3, because of data limitations, none of the available empirical
studies is able to examine all relevant harm. Consequently, this analysis quantifies only a subset
of such harm.577 There is strong and deep empirical evidence that conflicted advisers often
recommend more expensive and poorer performing mutual funds to retail customers. This
evidence, taken together, suggests that IRA holders receiving conflicted investment advice can
expect the mutual funds in which they invest to underperform by an average of 50 to 100 basis
points per year. The source of underperformance alone could cost IRA investors $95 billion to
$189 billion over the next 10 years and between $202 billion and $404 billion over the next 20
years.
These estimates are uncertain in part due to limitations in available data and statistical
methods. The estimates are grounded largely in research that compares investment results
between two groups of mutual funds: those distributed by full-service brokers, and those
distributed directly to consumers. As detailed in Chapter 3, this research documents relative
underperformance in broker-distributed funds. The gap is a strong indicator of the magnitude of
this quantified subset of harm from adviser conflicts, but it is not a clean measure of such harm,
for several reasons.

577



The gap might overstate or understate this harm insofar as it does not take into
account the commissions paid to distributing full-service brokers, nor fees that some
direct investors pay directly for advice (others are unadvised), nor the value of the
advisory services. Some comments on the 2015 Proposal provided data suggesting
that commissions are smaller than the fees, and argued that this demonstrates that the
gap overstates the harm. But many of those same comments reported that fee-based
advice arrangements usually provide more service than commission-based advice
arrangements, so in fact there is no basis for the comment’s assertion that the
omission of fees and commissions from the comparison would cause the observed
gap to be overstated.



The gap might understate the quantified subset of harm insofar as available data do
not facilitate a complete and clean classification of mutual funds by exposure to
advisory conflicts. Any misclassification would dilute the decrement in investment
results that is a consequence of advisory conflicts making the observed gap smaller.



The gap might overstate this harm if some underperformance reflects unobserved,
indirect, fair compensation for advice (generally in the form of revenue shared by the
mutual fund’s asset manager with the distributing broker). However, as documented
in Chapter 3 and elsewhere in this analysis, retail investors generally are unaware of
such indirect compensation (and of 12b-1 fees), so it is implausible that they reflect
efficient market pricing of advisory services.

Also see the qualitative discussion of investor gains in Section 3.3.2.

299



The gap might overstate or understate future harm insofar as the gap varies widely in
both magnitude and even direction across years, so its observed magnitude is
sensitive to the period studied.578



Uncertainty about the future growth and composition of the IRA market and the
market for IRA investment advice introduce still more uncertainty into the estimated
magnitude of this quantified subset of the harm.

Two comments on the 2015 Proposal offered their own estimates of this gap. In contrast
to other available evidence and the conclusion of this analysis, these comments generally
purported to show that there is no evidence of a gap and therefore no evidence of this particular
subset of harm from advisory conflicts.
In the first such comment, ICI examines recent data covering essentially the entire
universe of U.S. mutual funds.579 ICI’s comment compares performance of front-end-load funds
(distributed by BDs compensated by commission) and no-load funds (distributed directly with
little or no exposure to advisory conflicts). The comparison yields a performance gap smaller
than that evident in other research. As elaborated in Chapter 3, this analysis takes into account
ICI’s findings, but for several reasons rejects ICI’s interpretation, which asserts that the findings
contradict the conclusions reached in the Department’s 2015 NPRM analysis. The Department’s
conclusions in both its 2015 analysis and this current analysis rely on a broad body of evidence,
importantly including a study from CEM that, by studying the effect of the magnitude of a
particular advisory conflict within a distribution channel, measures the effect of conflicts more
directly and reliably than the approach adopted by ICI, Bergstresser et al. (2009) and others
which rely on comparisons between different distribution channels. Moreover, by combining
data on domestic and international equity funds, ICI’s analysis appears to have obscured material
differences between the two. And the time horizon studied by ICI is too short to meaningfully
examine the volatile performance gap (See Section 3.2.4).
In the second such comment, NERA Economic Consulting (in an analysis commissioned
and submitted by SIFMA), assesses the performance of commission-based and fee-based
accounts within a confidential data set consisting of 63,000 IRA accounts with returns data
spanning from 2012 through the first quarter of 2015.580 The Department found this analysis
flawed and its conclusions unsupported and therefore unreliable.
NERA’s analysis offers little detail about its data source, and it is unclear whether the
data are representative of financial institutions or IRA accounts in the United States.
NERA argues that commission-based accounts incur lower fees than fee-based accounts.
The magnitude of the difference ranges from about 57 basis points for relatively small accounts
with balances below $25,000 up to about 100 basis points for accounts with balances from
$100,000 to $250,000. However, its comparison excludes important fee components that often
affect commission-based accounts, such as markup/markdown revenue and 12b-1 fees. NERA

578

579
580

As detailed in Chapter 3 and Appendix A, this wide variability is reported in at least one comment on the 2015 Proposal, and is
corroborated over a longer period by Morningstar data. Also as detailed there, notwithstanding this variability, available evidence strongly
supports the conclusion that broker-distributed funds underperform on average over long periods.
ICI comment letter, (July 21, 2015); available at: http://www.dol.gov/ebsa/pdf/1210-AB32-2-00749.pdf.
SIFMA, “NERA Analysis: Comment on the Department of Labor Proposal and Regulatory Impact Analysis” (July 20, 2015); available at:
http://www.sifma.org/issues/item.aspx?id=8589955443.

300

acknowledges this bias but does not do anything to mitigate it. NERA also points out that
today’s fee-based accounts trade more frequently. This strongly suggests that today’s fee-based
accounts require more service than commission-based accounts, so it is neither surprising nor
relevant that fees are larger than commissions. If markets are efficient, prices investors pay will
depend on levels of service, not the form of payment.
NERA purports to show that median fees are higher in fee-based accounts at all account
sizes. However, losses from conflicted advice are likely to manifest themselves mostly away
from the median: if even 49 percent of commission-based investors were paying exorbitant fees,
this pattern would be invisible at the median. Commissions may be excessive for a minority of
accounts, excessive trading may be found in a minority of accounts, underperformance may be
serious for a minority of accounts, etc. Median statistics cannot show any such pattern. Because
NERA analyzed account-level data, it is unclear why it did not address this possibility.
NERA also compares commission- and fee-based account returns, showing the former to
be about equal to those in fee-based accounts. This finding conflicts with the substantial body of
peer-reviewed academic research reviewed in Chapter 3, which documents a performance gap.
Yet NERA devotes merely one page to the analysis and presents quarterly differences in median
returns only (Table 4, page 10).
The analysis of rates of return also falls short in several respects. First, the comment fails
to adjust for differences in riskiness (volatility) of account portfolios. This is important if assets
in fee-based and commission-based accounts differ in the average level of risk. For example, a
portfolio invested only in stocks that make up the S&P 500 index would have realized
compound annual growth rate of approximately 19 percent over the period of the study, much
higher than the historical average rate of return on stocks. But of course investing in stocks only
will not be suitable for all investors, particularly not for many of those nearing retirement. The
NERA memorandum (page 4) shows that account holders of fee-based accounts tend to be older
than commission-based account holders. Roughly 58 percent of fee-based account holders are
age 60 or older, compared with roughly 48 percent of commission-based account holders. Based
on their higher ages, fee-based account holders probably invest in less risky assets than
commission-based account holders. In fact, data in the NERA memorandum itself confirm that
commission-based accounts are invested in riskier assets than fee-based accounts.581
Second, the NERA comment based its underperformance analysis entirely on median
quarterly rates of return. At best, such data support a conclusion about underperformance at the
median; they do not support any conclusion about accounts above or below the median. For
example, the median would be the same if 49 percent of commission-based accounts performed
extremely poorly. The median measure in effect would hide any meaningful result that NERA
could have derived from its data. Based solely on the median, NERA’s conclusion that
underperformance is not an issue is unconvincing.
Third, NERA’s analysis spans an unreasonably short time horizon. The volatility
reported by NERA itself demonstrates that any given year might produce a result opposite of the
true, long-run effect on performance. NERA’s analysis of returns data, spanning less than three
years, fails to provide a reliable estimate of the true difference in returns across commission- and

581

See AACG’s review of the NERA analysis for more details (Padmanabhan, Panis and Tardiff 2016).

301

fee-based accounts. Available peer-reviewed academic studies together span longer periods and
find a performance gap that evidences harm from advisory conflicts.582

8.1.2 Unquantified Harm
As detailed in Chapters 3 and 4, there is strong evidence that advisory conflicts inflict
more types of harm than are quantified in this analysis. This analysis quantifies only a portion of
the underperformance of mutual funds held by IRA investors. In addition to this harm, there is
evidence that advisory conflicts lead to losses of other kinds in connection with IRA investments
in mutual funds, losses in connection with other IRA investments, and losses to plan investors.
Data limitations preclude quantification of these losses, but there is ample evidence that they are
of an order of magnitude similar to the quantified harm. Put differently, the quantified harm
alone dramatically understates the total harm known to result from advisory conflicts – albeit by
an uncertain amount.
With respect to IRA mutual fund investments, adviser conflicts (in particular conflicts
arising from commission compensation) sometimes lead to excessive trading, avoidable timing
errors, and poor matching of investors with funds. IRA investors’ associated losses are
sometimes large and large on aggregate. Commissions, such as those associated with mutual
funds’ front-end-load sharing, directly reward advisers for recommending more frequent trades.
Actively-managed funds tend to pay higher commissions. Some retail investors are predisposed
to chase returns (and thereby suffer avoidable timing losses), and variability in active funds’
returns relative to benchmarks can feed both trading frequency and return chasing. Advisers in
effect often are rewarded for exploiting rather than mitigating investors’ harmful behavioral
biases. Some comments on the 2015 Proposal argue that advisers add important value by
discouraging panic selling and thereby averting potentially large timing errors. At the same
time, however, as detailed in Chapter 3, this analysis concludes that conflicts can and sometimes
do increase excessive trading costs and avoidable timing errors. Comments on the 2015
Proposal document how various commission structures can sometimes produce very large
financial incentives for advisers to sell certain products at certain times.583 This type of
incentive is especially likely to sometimes result in poor matching of products with consumers’
needs.
With respect to other IRA investments, and as detailed in Chapter 3, there is evidence
that similar conflicts exist and result in similar harms. Variable annuities share certain features
and characteristics with mutual funds. Fixed-indexed annuities have certain similarities with
variable annuities from the consumer’s perspective. However, in contrast with mutual funds,
both variable and fixed-indexed annuity products generally provide consumers with a variety of
insurance protections (often subject to limiting conditions), carry higher fees, and pay higher
commissions. Higher fees and commissions often might be justified as compensation for
advisers’ effort to recommend these more complex products and for insurers’ assumption of
various risks. However, the products’ complexity amplifies both consumers’ need for advice
and the potential for abuse. Advisers are often rewarded for distributing higher-priced variable
annuities or certain products that may not be the best match for consumers’ needs. Still other

582
583

Both the volatility of the gap and its average direction are evident over an even longer period in Morningstar data, as shown in Appendix A.
See Mercer Bullard’s written testimony, (August 13); available at: http://www.dol.gov/ebsa/pdf/1210-AB32-2-WrittenTestimony28.pdf.

302

types of investments expose IRA investors to potential harm from advisory conflicts. For
example, variations in spreads in principal transactions may encourage advisers to recommend
trades and products that are not optimally aligned with investors’ interests.
Finally, with respect to plans, as detailed in Chapter 4, adviser conflicts can cause similar
harm. There is evidence that plan advisers sometimes promote the inclusion of affiliated funds
in participant-directed DC plan menus, and that DB plans that rely on consultants with
undisclosed conflicts suffer large underperformance. In addition, advisers often have incentives
to encourage investors to imprudently roll over account balances from ERISA plans to IRAs
even though the ERISA plans offer fiduciary protection and often have lower cost and higher
performance. The rule and related exemptions offer additional protections in this context, but
these benefits too have not been quantified.
It should be noted that variation in adviser incentives to recommend different products
sometimes can result in misalignment between investors’ interests and products both within and
among product types. For example, while one adviser might be rewarded more for
recommending one mutual fund rather than another, another adviser might be rewarded more for
recommending a fixed-indexed annuity rather than a mutual fund. All such misalignment entails
IRA or plan investor losses beyond those quantified in this analysis.
In summary, there is ample evidence that unquantified harm from advisory conflicts,
separate from and additional to the harm quantified in this analysis, is of large, albeit uncertain,
magnitude. While the scale of such harm is well documented, data limitations preclude its
quantification.

8.2

Investor Gains

This analysis (in Chapters 3 and 4) documents both quantified and additional
unquantified gains to IRA and plan investors expected under the final rule and exemptions.
However, as elaborated in this section, the magnitude of the gains is uncertain.

8.2.1 Quantified Investors Gains
The precise magnitude of the final rule and exemptions’ quantified subset of anticipated
investor gains is uncertain. As discussed above, (see Section 3.3.1) the final rule’s and
exemptions’ mitigation of adviser conflicts associated with variation in mutual funds’ front-endload sharing alone is expected to produce gains for IRA mutual fund investors that will total $33
billion to $36 billion over 10 years. This quantified gain (which is only a subset of the much
larger, total expected gain) is subject to uncertainty, for several reasons. First, the final rule and
exemptions may not fully mitigate adviser conflicts associated with variation in mutual funds’
front-end-load sharing. Second, the parameters and data underlying the estimates are uncertain.
Sensitivity analyses on the assumptions used to generate these estimates are provided in
Appendix B, Section B.3.
The effectiveness of the final rule and exemptions’ protections is uncertain. The final
amended and new PTEs would allow fiduciary advisers (including both existing fiduciary
advisers under the 1975 regulation and new fiduciary advisers added by the final rule and
exemptions) and affected financial institutions to receive certain variable fees and compensation
resulting from the provision of fiduciary advice, subject to protective conditions intended to
ensure the advice’s impartiality. If advisers were to fail to adhere to the protective conditions,
including conditions that prohibit compensation practices and employment incentives that are
intended or would reasonably be expected to compromise advisers’ impartiality, then anticipated
gains to retirement investors and other economic benefits would be reduced.
303

The Department is not aware of any direct empirical evidence that would support
estimation of the degree to which the final rule and exemptions might fail to ensure adviser
impartiality. As noted in Chapter 2, recent reforms in the UK to eliminate adviser conflicts
substantially increased the sale of low-fee funds. This suggests that effective reforms to combat
conflicts can indeed promote impartiality. But this falls short of direct evidence as to whether
the Department’s final rule and exemptions, which differ from the UK reforms, will be fully
successful at ensuring impartiality.
The Department believes that the final PTEs’ scope and conditions, once fully phased in,
will be effective at ensuring that investment advisers act in investors’ best interest. Both
advisers and relevant financial institutions will have strong incentives to ensure that advice
honors investors’ best interest. In order to protect the interests of the plan participants and
beneficiaries, IRA owners, and plan fiduciaries, the final Best Interest Contract Exemption
allows investment advisers to receive variable compensation and indirect compensation from
third parties in connection with providing investment advice only if they adhere to basic
standards of impartial conduct, adopt policies and procedures reasonably designed to mitigate
any harmful impact of conflicts of interest, and disclose basic information on their conflicts of
interest and on the cost of their advice. It prohibits compensation practices and employment
incentives that are intended or would reasonably be expected to compromise advisers’
impartiality. In addition, firms and advisers providing investment advice to IRA investors and
other plans not covered by Title I of ERISA must agree to adhere to the final exemption’s
impartial conduct standards in a written contract that is enforceable by the investors.584 Pursuant
to these provisions, advisers will no longer be encouraged to subordinate investors’ interests.
Advisers and financial institutions that fail to satisfy impartial conduct standards will be exposed
to class action litigation where costs to make a class of investors whole for their relevant losses
can exceed any profit from noncompliance. Finally, advisers and financial institutions failing to
satisfy relevant exemption conditions will potentially face steep excise taxes. The Department
will monitor compliance and market developments under the new rule to assess whether it is
achieving its intended goals and inform possible future changes to the regulation and/or the
PTEs’ scope or conditions.
In addition, the Department notes that as impartiality increases, products that are not in
the best interest of the investor will see a net outflow of funds while products that are in the best
interest of the investors could experience an inflow of assets. As discussed earlier in the
Regulatory Impact Analysis, providers who sell their products by incentivizing advisers to
recommend the products will find that those incentives have been mitigated. As a result, any
movement by advisers toward more impartiality is expected to reduce the propensity of those
who compensate them to use variable payments to induce advisers to recommend preferred
products. This in turn will lead to still more impartiality on the part of advisers.
Lacking empirical evidence as to how successful the final rule and PTEs will be in
removing all the adverse impacts of conflicts of interest, the Department has considered the
possibility that the industry may not fully comply with the rule and exemption conditions, and
that the combined rule and exemption conditions may not be fully effective at ensuring advisers’

584

See Section 2.9.2 for more information on the Best Interest Contract Exemption.

304

impartiality as anticipated by the Department. If financial institutions and advisers identify ways
to circumvent the protections in the final rule and exemptions, they might continue to impose
costs on their customers and — because of their ability to continue subordinating their clients’
interests to their own — the anticipated gains to investors would be reduced. If only 75 percent
of anticipated gains were realized the quantified subset of such gains – specific to the front-endload mutual fund segment of the IRA market – would amount to between $24 billion and $27
billion over 10 years. If only 50 percent were realized, this subset of gains would total between
$16 billion and $18 billion. In addition, some gaps in effectiveness are likely during the months
after the final rule and exemptions become applicable but before some of the protective PTE
conditions take effect. If the rule and exemptions are only 50% effective in the first year
following the initial applicability date (which includes an approximately nine-month transition
period when some Best Interest Contract Exemption provisions are not yet in effect), the
quantified subset of gains – specific to the front-load mutual fund segment of the IRA market –
would amount to between $30 billion and $33 billion over 10 years. These gains estimates
include only the subset of gains to investors that this analysis quantifies, omitting additional
unquantified gains and other economic benefits.
With respect to parameters, as detailed in Chapter 3, different available research provides
somewhat different estimates of the size of conflicts’ negative effects on investment results. In
addition, data on future growth and mix of IRA assets are projected and therefore uncertain.

8.2.2 Unquantified Investor Gain
Because of the limited scope of the research paper and analysis underlying the investor
gains quantified by the Department, the estimates themselves are limited in their scope and omit
other large but unquantified investor gains expected under the final rule and exemptions. The
estimates capture only the potential for improved performance and reduced loads in IRA mutual
fund investments. They do not capture the potential for improved consumer value from financial
products other than mutual funds. They do not capture the potential for transaction cost savings
from longer holding periods, or the potential for reductions in return-chasing and related timing
errors. They do not capture the potential for narrower spreads in principal transactions. They do
not capture the potential for better alignment of financial products with consumers’ needs. They
do not capture similar investment gains by plan investors. As discussed earlier in this chapter
(and elaborated in Chapters 3 and 4), there is evidence that the unquantified harm from advisory
conflicts is large. The large magnitude of that harm demonstrates that the unquantified gains
investors will realize from the final rule and exemptions’ mitigation of advisory conflicts is
similarly large. The quantified investor gains alone dramatically understate the total gains
expected under the final rule and exemptions, albeit by an uncertain amount.
In addition to the quantified and unquantified improvements in IRA and plan investors’
investment results, there are likely to be additional unquantified investor gains, discussed
qualitatively herein, also of uncertain magnitude. The IRA and plan markets for fiduciary advice
and other services may become more efficient as a result of more transparent pricing and greater
certainty about the fiduciary status of advisers and about the impartiality of their advice. There
may be gains from the increased flexibility that the final rule will provide with respect to
fiduciary investment advice currently falling within the ambit of the 1975 regulation. The final
rule defined boundaries between fiduciary advice and education may improve access to plan
participant and IRA investor educational services. Innovation in new advice business models,
including technology-driven models, may be beneficially accelerated, and nudged away from
conflicts and toward transparency, thereby promoting healthy competition in the fiduciary advice
market. In the plan market, clarity about advisers’ fiduciary status under ERISA will strengthen
305

the Department’s enforcement efforts and deter abuse. Additionally, in both the plan market and
the ERISA market, the imposition of a prudent expert standard would also be expected to raise
the level of quality advice, even for many advisers who are not conflicted or who are less
responsive to conflicts of interest. All of these may yield additional gains, of uncertain
magnitude, for plan and IRA investors.

8.3

Compliance Costs

Based on industry estimates of costs to comply with the 2015 Proposal, the final rule and
exemptions’ ten-year compliance cost is estimated to be between $10.0 billion and $31.5 billion,
or less if, as expected, more cost effective business models gain market share. This estimate is
uncertain, reflecting uncertainty about the number and mix of affected advisers, the time and
effort required to review practices for compliance, the degree of change in such practices
necessary to achieve compliance (including the degree to which such practices fit within the
scope of existing and proposed PTEs), the cost associated with such change, and the potential
magnitude and speed of improvements in cost-effectiveness. Uncertainty also arises from the
lack of granularity in available industry estimates, and the process of adjusting these estimates to
reflect changes to the 2015 Proposal included in the final rule and exemptions. The price and
loss ratios associated with errors and omissions insurance is also uncertain.
Much of the estimated compliance cost is associated with satisfaction of PTE conditions.
The number of advisers who will take advantage of the relevant PTEs is uncertain, however.
Some advisers may find it more advantageous to simply avoid prohibited transactions. As
elaborated in Chapter 5, in considering this uncertainty, the Department has aimed to err on the
side of overestimating the compliance costs by assuming wide use of PTEs.
Compliance costs, and consumer costs generally, will be less than estimated in this
analysis if, as expected under the final rule and exemptions, more efficient advisory models and
financial products gain market share. With respect to advisory models, for example, simplified
and less variable compensation arrangements may help ease compliance with the Best Interest
Contract Exemption, and direct, flat-fee arrangements can obviate any need for exemptive relief.
Algorithms that generate impartial recommendations may further ease compliance, and may
sometimes increase advisers’ productivity. With respect to financial products, for example, more
consumers, such as small IRA investors, may migrate to relatively simple, inexpensive solutions
that are calibrated to suit common circumstances and adjust automatically as time passes and
circumstances change. These might include passively managed target date funds or similar ETFs
that automatically rebalance portfolios and adjust risk exposure as investors age. They may mix
such strategies with other innovative products recently gaining favor, such as deferred annuities
that insure them against exhausting resources at very advanced ages, sometimes referred to as
“longevity insurance.” Target date funds and deferred annuities may require relatively
infrequent, streamlined, and minor monitoring and adjustments. Consequently, following such
simple and inexpensive yet potentially effective strategies can reduce the need for complicated
and expensive advice. Efficient advisory models and financial products may act as complements
and their efficiencies may be additive.
The Regulatory Impact Analysis estimates compliance costs based on comments from
major securities industry trade groups. Such estimates are best understood as industry’s
proffered assessment of its own cost to adapt its existing business structure, practices, and
product/service offerings to the changed regulatory environment. This largely static view is
illustrated by many industry comments’ contention that all advice must continue to be provided
in one of the two modes that it says currently dominate the market: a commission model that
306

currently serves nearly all small and many large advised accounts, and an asset-based fee model
that currently serves almost exclusively larger accounts. These comments appear to presume
that only these compensation structures are viable, and that their pricing is largely fixed and
therefore will determine their customer base rather than the reverse. For purposes of this
Regulatory Impact Analysis, this view neglects competitive market forces and innovations that
are expected to result in important, cost-saving market adjustments. 585
If new rules add expense to service delivery models that involve myriad, complex
relationships and revenue streams that can cause adviser conflicts, models that lack this baggage
and associated regulatory expense will gain market share. The latter historically have suffered
competitive disadvantage due to the opacity of prices in the former. Financial products that have
been favored in commission-driven, sales-oriented consumer markets will be displaced by
products that are favored by impartial expert advisers. Again, the latter historically have
suffered competitive disadvantage due to the opacity of prices in the former. Advisory fee
structures and service bundles likewise will evolve in response to market demands and
regulatory incentives. Simpler fee structures could ease compliance with Best Interest Contract
Exemption conditions including disclosure provisions and adherence to required warranties.
As elaborated later in this chapter, the relatively static view of compliance costs espoused
by many industry comments on the 2015 Proposal is belied by innovation already underway in
the market for advisory services, and by comments submitted by businesses engaged in such
innovation. These innovations benefit greatly from advances in technology, including advances
in relevant information technology, consumer interface technology, and financial analytic
technology. Innovations to date include the availability of highly affordable “robo-advice
provider” services – but also include hybrid models where consumers face live personal advisers
who are backed by new, cost-reducing technologies.
Regulators will also incur some cost to implement and enforce the final rule and
exemptions. The net amount of such cost is highly uncertain as it depends on compliance levels
and efficacy of policies and procedures of the financial institutions. High compliance and
effective policies and procedures would limit regulators’ cost. Regulators’ costs will also be
limited to the extent that less conflicted advisory services, which are likely to be less prone to
abuse, gain market share under the final rule and exemptions.

8.4

Secondary Market Effects

This section assesses the potential for secondary market effects, concluding that,
notwithstanding some near-term frictional costs, such effects will consist mainly of
improvements in economic efficiency. Negative effects predicted by some commenters, such as
erosion of small investors’ access to beneficial advice and their prospects for retirement security,
are not anticipated.

585

Failure to anticipate cost-saving innovations has sometimes led federal agencies to overestimate the future cost of regulations. For example
see Eban Goodstein and Hart Hodges, “Behind the Numbers: Polluted data,” November-December 1997; available at:
http://prospect.org/article/behind-numbers-polluted-data; and Harrington, Morgenstern and Nelson (2000). Although Harrington et al.
identify instances of cost-saving innovations, they find that the agencies in their sample do not systematically overestimate regulatory costs
on a per-unit basis, with the reason for overestimation of costs at the aggregate level instead being a combination of difficulty with
establishing baseline conditions and noncompliance.

307

While investor gains and associated compliance costs constitute the direct effects of this
final rule and exemptions, a variety of secondary effects are also anticipated, as the markets for
financial services and products adjust to the new regulatory imperative to honor plan and IRA
investors’ best interests. These effects generally include longer-term improvements in economic
efficiency that will advance social welfare, and some temporary frictional costs which may erode
social welfare in the near term. Some comments on the 2015 Proposal predicted that these and
perhaps some additional costs will be persistent and structural rather than temporary and
frictional. At the broadest level, these comments predicted sharp and lasting increases in the
cost of investment advice and consequent erosion in the amount of advice delivered and
investment results, particularly for smaller investors. However, careful review of these
comments revealed important flaws in the evidence and analysis they presented. In light of these
flaws, the evidence presented in the preceding chapters, and improvement to the 2015 Proposal
that are included in this final rule and exemptions, the Department believes there is minimal risk
of such negative effects.
The final rule and exemptions will have a variety of indirect effects on existing markets
for financial products and services. The character and magnitude of these effects are uncertain.
The Department believes these effects are likely to tend toward greater long-term economic
efficiency and thereby improve overall social welfare. However, transitional frictions may
introduce some social costs, and the long-term distributional impacts are uncertain. The
discussion that immediately follows explores qualitatively some potentially important indirect
effects and their potential social welfare implications, with an eye toward the starting point of
historical market conditions. It is not intended to be exhaustive.
The final rule and exemptions aim to mitigate harms from advisory conflicts without
unduly advantaging or disadvantaging any business model. Rather, the rule aims to level the
field on which different models compete for plan and IRA advisory customers. Historically,
investors’ high information costs combined with the absence of fiduciary conduct standards have
tilted the market in favor of advisory business models (and financial products) whose price is
shrouded, and business practices that divert resources from enriching investors to rewarding
advisers for promoting the products that profit financial firms most. In that environment,
advisory services that appear to be impartial and free, but are neither, enjoy an inefficiently large
market share, at the expense of services that are truly impartial and competitively priced, and
investors consequently suffer. Financial products promoted by `conflicted advisers likewise
enjoy an inefficiently large market share, diverting resources to the products’ manufacturers (and
possibly to manufacturers’ profit586) from uses more beneficial to investors. The final rule and
exemptions are intended and expected to mitigate these economic inefficiencies and to move
markets toward a more optimal mix of advisory services and financial products.
While the Department expects the frictions associated with market adjustments to be
temporary, it understands that they may be significant and may pose a particular challenge to
some parties in the near term, especially financial businesses whose practices historically have

586

One industry observer predicted that under the 2015 Proposal industry “operating margins on IRA assets could contract up to 30 percent.”
Morningstar, “Financial Services Observer: The U.S. Department of Labor’s Fiduciary Rule for Advisors Could Reshape the Financial
Sector” (Oct. 2015); available at:
file:///C:/Users/Bloom.Teresa/Downloads/The%20U%20S%20%20Department%20of%20Labor's%20Fiduciary%20Rule%20for%20Advis
ors%20Could%20Reshape%20%20%20.pdf.

308

deviated most from fiduciary standards. Such businesses may need to undertake major changes
to adviser incentive structures and loyalties, and/or lose market shares to businesses more
prepared or willing to align adviser and investor interests and honor fiduciary norms. Such
businesses may include some independent securities broker dealers and insurance agencies
whose commission and other compensation structures have been highly variable and laden with
more acute conflicts of interest. They may also include some insurers and mutual fund
companies who historically have relied heavily on highly variable compensation to promote
sales of the products that profit them more over alternative products that might better serve their
customers. Investors whose advisers and product providers are so affected also may experience
some amount of disruption as markets adjust, and may incur some costs to find, acquire, and
adjust to new services and products from the same or different vendors. These same investors,
however, absent this final regulation and exemptions, would likely have been the most adversely
affected by adviser conflicts, and therefore may stand to gain the most from reform,
notwithstanding near-term disruptions. Finally, it should be noted that the same frictions that
present challenges for some businesses may enhance opportunities for others, namely those
businesses most prepared or willing to align their interests with their customers and adhere to
fiduciary standards of care and loyalty.
The Department notes that the markets for financial advice, financial products and other
financial services are highly dynamic. They are characterized by innovation in both product
lines and business models, and by large ongoing shifts in labor and other resources across
product and service vendors and business models. These dynamics often involve large
transactions, including recruiting bonuses, client account transfers and other asset flows, all of
which may entail substantial frictional costs. The Department believes it likely that any
frictional cost associated with this final rule and exemptions will be justified by the rule’s
intended long-term effects of greater market efficiency and a distributional outcome that favors
retirement investors over the financial industry.

8.4.1 Advisory Firms’ Responses
The Department’s assessment of the financial industry’s responses to the final rule and
exemptions is subject to uncertainty. It is likely, however, that those responses will add to the
plan and IRA investor gains predicted by this analysis.
Industry responses are likely to vary across market segments, across business models, and
across firms. Many firms will face choices, and presumably will pursue whatever path is
expected to be most advantageous. For example, consider a large financial services firm that
advises IRA investors as both a BD paid by commission and RIA paid by asset-based fees. With
respect to the clients served under the BD model, the firm may respond to the rule in any of the
following ways (and possibly in other ways not listed):
•

The firm may elect to rely on the Best Interest Contract Exemption (and possibly
other PTEs) in order to minimize changes to its compensation arrangements and other
relevant practices. The firm might absorb or pass on to customers the cost of
satisfying applicable PTE conditions. This response may be most advantageous for
many firms, especially in light of the cost-reducing improvements to the 2015
Proposal that are reflected in the final rule and exemptions (and detailed in Chapters 2
and 5). This response might be least disruptive of existing adviser/IRA investor
relationships, but would involve frictional costs in the form of start-up costs to satisfy
PTE conditions. The compliance cost and (partial) investor gains estimates presented
in this analysis generally reflect this response. This response generally is more static
309

than alternatives so this analysis might understate the net investor gains that will be
achieved following beneficial market adjustments.
•

The firm may elect to avoid potential conflicts and minimize the need for exemptive
relief by moving IRA investors who are BD clients to fee-based RIA arrangements.
If the firm’s RIA fee structure remains fixed, this action might increase the price
many investors pay for advice and thereby increase the firms’ revenue, and possibly
its profits. Over time prices would adjust to reflect the marginal cost of services
provided and profits would stabilize at competitive levels. This response is likely to
be advantageous in instances where the cost to provide prohibited transactions-free
RIA services is less than the cost of BD services that satisfy applicable PTE
conditions. Therefore this response generally would result in larger aggregate net
gains for IRA investors than the preceding response (after transitional costs are
defrayed), and this analysis might understate net investor gains. As discussed later in
this section, there is evidence that prohibited transactions-free RIA services are
becoming more affordable, making it more likely that this analysis overstates
compliance costs and understates net IRA investor gains. This response would
reshuffle adviser/IRA investor relationships, producing frictional costs for both
parties.

•

The firm might cease to advise some or all heretofore BD clients on IRA investments.
This response might be likely in instances where the firm’s cost to satisfy PTE
conditions or to convert BD to RIA arrangements both exceed IRA investors’
willingness to pay, as might be the case where existing business practices are deeply
laden with (and reliant on) myriad, serious advisory conflicts. This response could
generate relatively large frictional costs for some advisers, and for affected IRA
investors as they select a new source (and possibly new level) of (now fiduciary)
advice. However IRA investors affected by such disruptions often will benefit most
in the long run, as absent reform, they are most likely to have suffered large losses
which they are likely to be unaware of.

Other types of firms that advise plan and IRA investors, such as small and large
independent BDs, RIAs, insurers using captive agents, independent insurance agents,
independent marketing organizations, and comprehensive financial planners, likewise will face
choices about how to respond to this final rule and exemptions. As with large BD/RIA dual
registrants, other advisory firms will gravitate toward structures and practices that efficiently
avoid or manage conflicts to deliver impartial advice consistent with fiduciary conduct
standards. Firms that achieve these ends most efficiently will gain market share. One industry
observer predicted that the 2015 Proposal would produce gains for discount brokerage, fee-based
advisers, and so-called robo-advice providers, while effects on full-service wealth managers
might be mixed.587 However a particular firm responds, its plan and IRA customers and
competitors will also respond to favor the most economically efficient results. The trajectory of
responses by different types of firms is uncertain, but resulting shifts in market share are
expected to reduce compliance costs and thereby amplify net plan and investor gains.

587

Morningstar, “Financial Services Observer: The U.S. Department of Labor’s Fiduciary Rule for Advisors Could Reshape the Financial
Sector” (Oct. 2015).

310

Advisory firms’ responses to the final rule and exemptions (and to related changes in
consumer demand and competition) will impact the labor market for advisers. These dynamics
may involve frictional costs and have distributional effects. For example, advisers may migrate
from advisory firms where conflicts had been most deeply embedded to firms that are wellsituated to efficiently provide impartial advice compliant with the final rule and exemptions.
The overall movement is likely to be toward greater long-term efficiency, with a more efficient
allocation of labor and other resources to investment advice and other productive enterprises.

8.4.2 Product Manufacturers’ Responses
As with advisers’ responses, financial product manufacturers’ responses to the final rule
and exemptions are uncertain, likely to improve social welfare, but carry some frictional costs.
The final rule and exemptions, by regulating advice to retirement investors, directly affect the
distribution of financial products, such as mutual funds and insurance products, but do not
directly affect their manufacture. However, adviser and investor responses to the rule will affect
investors’ relative demand for different products, as well as advisers’ relative demand for
different compensation arrangements, support services and business relationships. Adjusting to
these changes in demand will entail frictional costs. Compensation arrangements will be
adjusted to reduce and mitigate advisory conflicts. Adviser support services will evolve to ease
compliance. Business relationships and structure might evolve to advance these same ends.
Financial products that are relatively expensive, underperforming, and/or not optimally aligned
with affected IRA and plan investors’ interests, and that are currently relying on sales incentives
that can bias advice to keep their net flows competitive, are likely to lose market share to more
consumer-friendly products. One industry observer predicts that under the 2015 Proposal
passive investment products would gain market share, in part due to “increased usage … from
financial advisers that formerly may have been swayed by distribution payments.” Effects on
active asset managers could be mixed, while some insurers may be challenged.588

8.4.3 Investors’ Responses
The final rule and exemptions will affect demand for financial advice in multiple ways.
They will make retirement investment advice more impartial, and the prices of advice, investing,
and investments more transparent. They will improve many IRA investors’ heretofore poor
understanding of how advisers are regulated and paid, and raise awareness that they are
obligated to honor retirement investors’ interest. All of this may increase investors’ trust in
advisers and increase their demand for advice. At the same time, the price of some advice might
rise to reflect advisers’ compliance costs. As a result, the amount of advice provided might rise
or fall and the mix of kinds of advice may change.
The final rule and exemptions will affect the demand for financial products and financial
services beyond advice, and supply will adjust in response. Passive and other lower cost
investments and consumer-friendly insurance products may gain market share. As discussed
earlier in the Regulatory Impact Analysis, providers who sell their products by incentivizing
advisers to recommend the products will have to take a different approach, as misaligned

588

Morningstar, “Financial Services Observer: The U.S. Department of Labor’s Fiduciary Rule for Advisors Could Reshape the Financial
Sector” (Oct. 2015).

311

incentives become prohibited. As a result, products that are not in the best interest of the
investor will see a net outflow of funds while assets that are in the best interest of the investors
could experience an inflow of assets. This flow of assets will cause shifts in the asset provider
market, with associated frictional costs and distributional effects. As with the advice market,
these markets are likely to move toward greater efficiency, with a more optimal allocation of
resources dedicated to producing a more optimal mix of financial products and services.
Small plan and IRA customers in particular will respond not only to any increases in the
price of affected financial products and services, but also to increased transparency of such
prices. Transparency will increase for several reasons. First, the final rule and exemptions and
PTEs include provisions to improve transparency. Second, advisory conflicts entail
complicated, indirect adviser compensation arrangements that frustrate transparency. As less
conflicted business practices gain market share, transparency will therefore improve. Third,
advisers adhering to fiduciary conduct standards will do more to educate plan and IRA investors.
Plan and IRA investors therefore will be equipped to make more optimal choices with respect to
advisory services, as well as with respect to financial products.

8.4.4 Impact on Small Plan and IRA Investors
Comments on the 2015 Proposal expressed divergent views with respect to its potential
effects on small IRA and plan investors. Some predicted major negative effects, arguing that the
proposal would sharply increase the price of advice and thereby erode small IRA and plan
investors’ access to beneficial advice and with it, their prospects for retirement security. Others
disagreed, arguing that smaller investors are more vulnerable to advisory conflicts and therefore
likely to benefit more from the proposed reform, and that impartial advice compliant with the
proposal can be supplied affordably to small plan and IRA investors.
This analysis concludes that the final rule and exemptions will benefit small plan589 and
IRA investors. While the exact trajectory and future shape of advisory markets is uncertain, and
some frictions can be expected in the near term, the Department believes that quality, affordable
advisory services will be amply available to small plans and investors under the final rule and
exemptions. Moreover, innovations currently underway in the market for advisory services
identified in many comments on the 2015 Proposal and discussed later in this chapter, are likely
to be accelerated by this final rule and exemptions, which the Department expects will increase
the availability of quality, affordable advisory services for small plans and IRA investors in the
future.
Many comments predicting negative effects relied on analyses that do not withstand
scrutiny. After close review, much of the analysis set forth in these comments proved to be
flawed or otherwise unpersuasive. Nonetheless, as detailed below, the Department took these
comments into careful consideration in developing its final rule and exemptions and this
associated Regulatory Impact Analysis. The final rule and exemptions reflect responsive
changes to the 2015 NPRM that reduce compliance costs and disruptions to existing

589

The benefits that will be derived from the final rule and exemptions by plans, participants, and beneficiaries are discussed in Section 4.3.
The Department expects these benefits to accrue to large and small plans and their participants and beneficiaries.

312

arrangements and practices that some comments suggest would adversely affect small plan and
IRA investors.
Many negative comments on the 2015 Proposal argue that compliance costs, particularly
costs incurred to satisfy conditions of the proposed Best Interest Contract Exemption or avoid
prohibited transactions, would make advice unaffordable for small IRA investors. 590 However,
as detailed in Chapter 5, revisions to the 2015 Proposal reflected in the final Best Interest
Contract Exemption will reduce compliance costs and thereby help make advice affordable to
small investors. Moreover, the Department believes it is likely that unconflicted advisory
service models that avoid or minimize prohibited transactions and therefore require little or no
exemptive relief also will be affordable for small investors under the final rule and exemptions.
In an efficient market, the price of advisory services should reflect the level and quality
of the service, and not vary with the mode of payment. Put differently, the price of advice
should not be higher merely because an adviser charges direct fees and avoids prohibited
transactions. IRA advisers avoiding prohibited transactions are unaffected by the final rule and
exemptions, so investors should be able to obtain advice at the same price under the final rule
and exemptions as absent the rule. Of course markets generally are not perfectly efficient. For
example, some advisory firms may be able to reduce transaction costs by recommending only
proprietary products, or by taking compensation from a small number of product providers rather
than directly from a much larger number of investors. But the Department believes that the
difference is likely to be small.
Some of the comments appear to neglect the role of fixed costs that advisory firms must
incur in order to serve large investors. In some instances, the firm’s fixed costs of regulatory
compliance (such as satisfaction of certain applicable PTE conditions for which costs are mostly
fixed) may be high, but their marginal cost to provide advice to one additional customer,
particularly one who requires minimal advisory services, may be low enough to make the smallbalance account profitable.
Some comments appear to overlook instances where firms may provide advice to smallbalance savers at a loss because they expect those accounts to grow and become profitable in the
future. Some accounts may grow quickly if they receive rollovers, and savers may be more
likely to roll over their accounts to a firm where they have an established account and are
receiving advice.

590

For example, consider a comment letter from Davis & Harman LLP, which included a study by Quantria Strategies LLC. On page 21 of
the study, Quantria states that: “Since the brokerage account fee arrangements violate the re-proposed regulations with respect to IRA
investors, advisers will shift to advisory accounts or require larger minimum account balances. […] The fixed costs associated with all
retirement plans (e.g. research, investment, or administrative costs) indicate that small account balances require comparable effort and input
as an account with larger account balance. Therefore, broker-dealers and advisers would be unable to charge fees sufficient to cover their
costs, and this would reduce the availability or intensity of investment advice for IRA account holders.” The Davis & Harman LLP
comments, including the Quantria study, are available at http://www.dol.gov/ebsa/pdf/1210-AB32-2-00746.pdf.

313

A number of the negative comments directly criticized the Department’s 2015 NPRM
analyses, and/or provided alternative analysis that reached different conclusions. 591 The
Department carefully reviewed these analyses.592 Its review revealed analytic flaws that render
the comments’ conclusions unsupported and unreliable.
Some comments that predict small investor losses from reduced access to advice appear
to presume that many small investors benefit from such advice today, and will in the future
absent reform. However, few households with small IRAs or modest means currently report
receiving professional advice (see Appendix C). It is less clear to what degree small savers have
access to affordable advice. Comments on the 2015 Proposal and available empirical and
anecdotal evidence together suggest that some advisory firms offer little or no service to small
investors, while others offer more. One comment reports that 9 percent of fee-based accounts
have balances below $25,000 suggesting that fee-based advice is sometimes available to smaller
accounts.593 However, the comment appears to leave open the possibility that some of these
small accounts belong to customers who own additional, possibly larger, accounts. A mystery
shopper study finds that brokers in New York and Boston routinely turn down individuals with
less than $100,000 in savings; these individuals found it very difficult to even get a first
appointment with a broker.594 Some advisory firms that do not offer small investors face-to-face
advice or telephone access to an assigned personal adviser nonetheless appear to offer some type
of advisory or educational decision-making support, such as access to call centers and on-line
tools. As elaborated later in this chapter, a variety of low-cost advisory services for small
investors are emerging in the market today.
Some comments fail to take into account statutory and administrative PTEs – other than
the Best Interest Contract Exemption – and guidance issued by the Department available to
fiduciary advisers that allow them to maintain their current business models when receiving
indirect payments while ensuring that investors are protected from conflicts of interest. For
example, in Advisory Opinion Nos. 97-15A and 2005-10A, the Department explained that a
fiduciary investment adviser could provide investment advice to a plan with respect to
investment funds that pay it or an affiliate additional fees without engaging in a prohibited
transaction if those fees are offset against fees that the plan otherwise is obligated to pay to the
fiduciary. Also, a statutory exemption created by Congress allows financial advisers to receive

591

592
593

594

See, for example, Compass Lexecon comment (http://www.dol.gov/ebsa/pdf/1210-AB32-2-00795.pdf), ICI comment
(http://www.dol.gov/ebsa/pdf/1210-AB32-2-00749.pdf); plus two follow-up comments (http://www.dol.gov/ebsa/pdf/1210-AB32-203056.pdf; and comment received December 2, 2015), Litan-Singer comment (http://www.dol.gov/ebsa/pdf/1210-AB32-2-00517.pdf) plus
two follow-up comments (http://www.dol.gov/ebsa/pdf/1210-AB32-2-02967.pdf and http://www.dol.gov/ebsa/pdf/1210-AB32-203075.pdf), NERA comment (http://www.dol.gov/ebsa/pdf/1210-AB32-2-00506.pdf), Oliver Wyman comment
(http://www.dol.gov/ebsa/pdf/1210-AB32-2-00515.pdf), Oxford Economics report (available at http://www.oxfordeconomics.com/recentreleases/the-economic-consequences-of-the-us-department-of-labor-s-proposed-new-fiduciary-standard) and three reports from Quantria
Strategies (http://www.dol.gov/ebsa/pdf/1210-AB32-2-00746.pdf).
Also see AACG’s assessment of the validity of these claims (Padmanabhan, Panis and Tardiff 2016).
See SIFMA, “NERA Analysis: Comment on the Department of Labor Proposal and Regulatory Impact Analysis” (July 20, 2015); available
at: http://www.sifma.org/issues/item.aspx?id=8589955443; and AACG’s review (Padmanabhan, Panis and Tardiff 2016).
This issue also was addressed in Professor Antoinette Schoar’s testimony. She stated the following: “[T]he actual argument must be that
the industry fears that the only way people are willing to pay for advice is through conflicted payments, where the full costs of the advice
are hidden from the customer. In other words, the industry must be implicitly asking whether customers would be willing to pay for this
advice if they realized how much they actually are being charged. But it definitely cannot be in the interest of customers to only be exposed
to these conflicted payments. If the fear of providing non-conflicted advice leads to some brokers to drop out of the market, the proposed
rule would actually be doing exactly its job by screening out advisors who are not planning to act in the best interests of their clients.”
Testimony of Antoinette Schoar from August 11, 2015, Department of Labor. Pages 376-385 of hearing transcript; available at:
http://www.dol.gov/ebsa/pdf/1210-AB32-2-HearingTranscript2.pdf.

314

indirect compensation if conditions are met that limit the potential for abuse and self-dealing,
including requirements for fee-leveling or the use of independently certified computer models.595
Comments suggesting that small-balance IRA holders who currently receive advice
would no longer purchase advice under the 2015 Proposal generally fail to address the separate
roles played by price levels and price transparency. Many small investors who currently
purchase professional advice fail to understand its cost and so may be buying more advice than
would be optimal.596 It is possible that some investors also fail to fully appreciate the benefits of
advice. Therefore a theoretical case can be made that shrouded prices sometimes will lead to
more optimal levels of advice.
Several comments that emphasize the value of professional investment advice fail to
distinguish between conflicted investment advice and impartial investment advice and thereby
overstate the value of conflicted investment advice. Research demonstrates that conflicted
advisers help investors overcome biases when it is in the adviser’s own best interest to do so, but
they also reinforce investor biases when those biases are beneficial to the adviser.597 For
example, conflicted advisers overwhelmingly push clients toward higher-fee actively-managed
funds and away from lower-fee index funds when the adviser can generate more revenue by
selling an actively-managed fund. In contrast, impartial fiduciary investment advice helps
investors overcome all biases. The relative value of conflicted and fiduciary investment advice
is demonstrated in Chapter 3 of this Regulatory Impact Analysis.
One group of the comments that fails to distinguish between conflicted investment advice
and impartial advice refers to the Department’s analysis of its 2011 PPA advice PTE regulation.
That analysis found that DC and IRA investors’ errors cost these investors $114 billion annually,
and the increased access to advice pursuant to the PPA PTE would restore $7 billion to $18
billion annually to those investors. Importantly, the advice projected to be extended was
fiduciary investment advice. Were the Department to project that the affected investors would
receive conflicted investment advice the forecasted benefits of the PTE would be much lower
and possibly negative. The Department consistently aims to promote impartial investment
advice for IRA holders and plan participants who seek help.
Many comments on the 2015 NPRM Regulatory Impact Analysis claim to find a strong
causal link between investment advice and savings. These comments typically demonstrate a
correlation – that advised individuals have more financial assets than non-advised individuals –
but interpret it as a causal result – that financial advisers cause consumers to accumulate assets.
However, there is little evidence that financial advisers improve retirement savings. A RAND
study commissioned by the Department provides a thorough assessment of the literature on the

595

596

597

ERISA section 408(b)(14) and (g), The Department has issued regulations implementing this provision at 29 C.F.R. 2550.408g-1 and 408g2.
Testimony of Antoinette Schoar from August 11, 2015, Department of Labor, Page 35 of hearing transcript available at:
http://www.dol.gov/ebsa/pdf/1210-AB32-2-HearingTranscript2.pdf.
See Mullainathan et al. (2012) and Antoinette Schoar’s discussion of newer research in her oral testimony at the Aug. 11 DOL hearing;
available at: http://www.dol.gov/ebsa/pdf/1210-AB32-2-HearingTranscript2.pdf.

315

adviser-savings relationship.598 While the correlation between advice seeking and savings is
well established, there is limited and contradictory evidence on causation. In theory, three
different causal relationships could generate the advice-savings correlation: First, as suggested in
many industry comments, financial advice may in fact improve savings rates. Second, savings
may cause advice seeking behavior. This relationship should be intuitive; people with little or
no savings have few reasons to seek advice about how to invest those savings. Finally, some
other characteristic of an individual – perhaps a personality trait or an outside influence – may
make a person more likely to seek financial advice and more likely to save for retirement. Most
of the available literature on advice and savings is not able to distinguish between these three
possible effects. According to RAND, the study most able to determine the degree of causality
between advice and savings found that the use of a financial adviser does not appear to increase
savings.599
The Department does not expect the final rule and exemptions to negatively affect
savings rates. Nonetheless, it should be noted that welfare is highest when savings are
optimized, not when it is maximized as some comments seem to imply. Many negative
comments incorrectly characterize any decrease in retirement savings as a cost. Yet this “cost”
must be netted against the benefit derived from the alternate use of these funds. For example, if
money is used to retire expensive debt rather than contributed to an IRA, net worth at retirement
might be higher.
Several of the negative comments on the 2015 NPRM Regulatory Impact Analysis argue
that brokers add value by preventing clients from selling funds after a downturn. For example,
see Economists Incorporated’s review of four studies commissioned by the Department.600
Though the comment offers no empirical evidence of brokers’ value in this area, it points out
that brokers’ incentives are consistent with encouraging clients to remain in the market
following a downturn. The Department agrees that broker incentives are in some ways
consistent with encouraging clients to remain in the market following a downturn. However,
broker incentives generally favor more frequent and larger trades, raising the possibility that
some might exploit investors’ panic for profit. Moreover, broker incentives are also consistent
with encouraging more market involvement following an upturn. Just as remaining in the
market following a downturn may be helpful to an investor, becoming more heavily invested in
the market following an upturn may be harmful to an investor. Therefore, it is unclear,
theoretically, whether these broker incentives benefit the client on net over the course of a
market cycle.
Economists Incorporated estimate that “just these two of broker-provided benefits –
coaching to stay invested through market downturns, and assistance in portfolio rebalancing –
conservatively total 44.5 basis points annually” and assert that these outweigh the purported

598

599
600

See Burke and Hung, (2015). A comment from Economists Incorporated (see Robert Litan and Hal Singer comment on the Best Interest
Contract Exemption submitted September 24, 2015; available at: http://www.dol.gov/ebsa/pdf/1210-AB32-2-03075.pdf.) reviews the
RAND study and generally agrees with RAND’s analysis of the literature pertaining to evidence on the relationship between advice and
savings. The Economists Incorporated comment emphasizes theoretical possibilities that have not yet been ruled out by the literature, but
offers no new evidence of a causal effect of advice on savings.
See Marsden, Zick, and Mayer (2011).
See Robert Litan and Hal Singer comment submitted September 24, 2015; available at: http://www.dol.gov/ebsa/pdf/1210-AB32-203075.pdf.

316

benefits of the rule.601 This estimate is based on a publication from Vanguard.602 This 44.5
basis point estimate comes from two sources: avoidance of timing and rebalancing.603 The
benefit from avoidance of timing is estimated by estimating the underperformance of accounts
with exchanges during and after the Great Recession compared to the Vanguard target-date fund.
The benefit from rebalancing is measured by estimating the differences between the returns from
a rebalanced portfolio and hypothetical non-rebalanced portfolio. Therefore, the Vanguard
publication does not attempt to assess the value of advice from brokers. Instead, the paper is
intended to promote a Vanguard “concept” called “Vanguard Advisor’s Alpha” by
demonstrating the concept’s potential to add value relative to alternatives. This distinction in the
purpose of Vanguard’s analysis is emphasized in the Vanguard paper:
“Paying a fee for advice and guidance to a professional who uses the tools and tactics
described here can add meaningful value compared to the average investor experience,
currently advised or not. We are in no way suggesting that every advisor—charging any
fee—can add value, but merely that advisors can add value if they understand how they
can best help investors” (Emphasis is original).
Quantitative analyses in the Vanguard paper measure the “Potential value-add” of the
Vanguard Advisors Alpha concept. The actual behavior of self-directed investors is compared to
an ideal strategy. There is no effort made by the researchers to compare self-directed investor
behavior to the actual behavior exhibited by investors who receive advice from Vanguard
representatives. The authors are unable to say anything about the actual benefit of Vanguard
advisers, much less the benefit of advice from broker-dealers who operate under a different,
conflicted business model and are likely to recommend different products.
Thus, an inspection of the single source relied upon by Economists Incorporated reveals
that Economists Incorporated’s own quantitative estimates generally are hypothetical.
Economists Incorporated provides no empirical evidence to back their claim of “Yet-To-BeRecognized Costs” of the rule.
Furthermore, the Vanguard quantitative analysis itself is problematic in the following
ways. The quantitative analysis of rebalancing fails to take into account the associated costs.
Vanguard recognizes the existence of costs associated with rebalancing (“There are costs
associated with any rebalancing strategy, including taxes and transaction costs, as well as time
and labor on the part of advisors.”), yet it continues to emphasize the gross benefit of
rebalancing and fails to quantify the net benefit – after taking into account those costs.
Vanguard’s market timing findings are based on a hypothetical analysis – how much better off
investors would be if they had made better choices, not how much better off they were as a
consequence of receiving financial advice. The authors examine Vanguard clients who had
deviated from their initial investments and shows how much better off they would have been if
they had instead invested consistently in a target date fund benchmark. However, the benefits
that would have accrued to this group of clients, selected using the criteria that they were

601

602

603

Robert Litan and Hal Singer, “Good Intentions Gone Wrong: The Yet-To-Be-Recognized Costs of the Department of Labor’s Proposed
Fiduciary Rule” (July 2015).
Francis M. Kinniry Jr., Colleen M. Jaconetti, Michael A. DiJoseph, and Yan Zilbering, “Putting a value on your value: Quantifying
Vanguard Advisor’s Alpha,” Vanguard, (March 2014).
According to the publication by Vanguard, the estimated benefits from avoidance of timing – not making exchanges during market
downturn - are 27 basis points, while the estimated benefits from rebalancing are 17.5 basis points.

317

deviating from their initial investment, is not necessarily the same as the benefit for all investors.
Secondly, it does not provide evidence on the effectiveness of financial advice in practice at
convincing investors to stay disciplined and not attempt to time the market. Thirdly, it is unclear
how often financial advisers throughout the current marketplace are actually providing advice
that discourages market timing.
Previous studies have shown that investors’ market timing generally is poor and that
investors who receive advice from a broker exhibit worse market timing than those who don’t.
Using data that cover both a market upturn and a downturn, Friesen and Sapp (2007) and
Bullard, Friesen and Sapp (2008) find that investors in load funds exhibit worse market timing
than investors in no-load funds.604 The implication of this result is that the brokers’ incentive to
encourage excess market participation at just the wrong time – when the market has seen
substantial recent increases – does more harm than their incentive to encourage sustained market
participation at the right time – following a downturn. In other words, while one can point out
specific market conditions where broker incentives are aligned with what is in the client’s best
interest, perpetuating a system where brokers provide advice that is in their own interest rather
than being focused on the interests of the retirement saver tends overall to harm to the saver.
For all of these reasons, the Department does not anticipate that the final rule and
exemptions will have substantial unintended negative effects on the availability or affordability
of advice that are predicted by NERA and others. Rather, it is more likely that the new rule and
exemptions will nudge the investment advice market’s evolution toward greater efficiency and
better results for plan and IRA investors.
Nonetheless, the possibility for some negative consequences for some plans, plan
participants or IRA investors cannot be ruled out. At a minimum some might experience shortterm disruptions in service as their existing advisers make changes in response to the new rule.
There may be a period of increased rates of switching to new advisers, with associated transition
costs, although this would most likely lead to a more efficient market equilibrium, reflecting
better informed matches between customers and advisers.

8.4.5 Innovation
The Department is confident that the final rule and exemptions will accelerate positive
innovations in the market that will serve small savers particularly well. Enabled by new
technologies, new business models already are delivering inexpensive, quality advice to small
investors. The final rule and exemptions will promote the availability of such advisory services,
because the business models’ technologies can help efficiently ensure the impartiality they
demand and because increased price transparency will favor such cost-saving innovations.
The Department believes the final rule and exemptions will promote healthy competition
in the market for advice on the investment of IRA assets, to the advantage of IRA investors.
This analysis has presented evidence that consumers currently mistake biased advice for
impartial advice, and are unaware of some of the fees they pay for that advice. This indicates an

604

Also see the discussion of these studies in Chapter 3. Some comments raised the possibility that brokers helped prevent panic selling
specifically with respect to the most recent market downturn in 2007 and 2008. This most recent downturn is outside of the sample used in
the Friesen and Sapp (2007) and Bullard et al. (2008) papers.

318

inefficient market where consumers spend too much and get too little. Imperfect information is
causing the market for IRA advice to fail. Under the new rule and exemptions, IRA investors
will expect and get impartial advice and the price will be more transparent, so the market will be
more efficient.
Under the final rule and exemptions, more efficient models will gain market share. More
consumers, such as small IRA investors, may migrate to inexpensive solutions such as passively
managed target date funds or similar ETFs. They may mix such strategies with other innovative
products recently gaining favor, such as deferred annuities that insure them against exhausting
resources at very advanced ages, sometimes referred to as “longevity insurance.” Following
such simple and inexpensive yet potentially effective strategies can reduce the need for
complicated and expensive advice.
Technological innovation likely will be harnessed to ease compliance with the Best
Interest Contract Exemption. Possible applications include automated monitoring of trades for
patterns of inappropriate recommendations, and automated generation of sound, impartial
recommendations for consideration by consumer-facing advisers.
Equally important, technology can directly lower the cost to deliver beneficial advice.
The market is already beginning to serve small accounts with quality, impartial, affordable
advice or other effective support for sound saving and investing decisions. The final rule and
exemptions are likely to promote healthier development of emerging business models that rely
heavily on technology to generate and deliver advice and/or that build advice into financial
products themselves, as is the case with target date funds. Such new technologies and
innovations in financial products already appear to be making advice and other potentially
effective investment support more affordable and available to many consumers. For example,
technologies make it possible to automatically pull data from customers’ accounts with nearly all
financial institutions, with customer consent. Computer algorithms can compare their financial
information with data on the price and performance of a wide universe of investment
alternatives, generating options for consideration, or even recommendations, very inexpensively.
Some of these newer business models lean toward independence in advice, but absent policy
changes such as those included in the new rule and exemptions, they may face the same
competitive pressures that have led more conflicted models to prevail so far. Conflicted models
currently can prevail even with inferior value because their price and quality are shrouded.
So-called “robo-advice providers” and products (such as target date funds) that reduce
the need for complex advice are already rapidly gaining market share. Going forward, they
promise to make advice far more affordable for small investors, especially young investors who
generally are more accustomed to technology-based tools.605 More traditional advisory firms are
scrambling to develop, partner with, or acquire such innovative tools, and to combine these with
more traditional services to deliver tailored services to more market segments at far lower cost
than that historically associated with traditional approaches alone.
Robo-advice providers can have various business models in terms of the amount of
hands-on assistance and the types of services offered. Despite this variation, they share a

605

For example, AT Kearney (2015) projects that “robo” advisers will manage roughly $2.2 trillion by 2020.

319

common characteristic - they utilize technology to meet the core portfolio management needs of
mass retail investors.
Because the core portfolio management is captured in a computer algorithm, roboadviser services generally can be scaled up more easily than traditional advisory services. The
marginal cost incurred by a robo-adviser to service additional customers is very small relative to
that incurred by traditional advisers. Consequently robo-advice providers can profitably service
small investors – and even bring new investors into the market – at low prices. Robo-adviser
firms often serve investors with assets under $500,000.606 Historically, small investors
sometimes have been underserved by traditional investment firms because it has not been
economical to serve them one at a time. However, this advantage in scale comes at a price – a
somewhat limited range of services. Services provided are typically comprised of asset
allocation with passively-managed ETFs or mutual funds only. Some may believe that this is
not a huge limitation because small investors are the main client base. The financial needs of
small investors often can be easily met by basic services and, given the low balance of such
accounts it is probably not worth paying for a more intensively personalized strategy.
Robo-advice providers and other inexpensive investment firms have grown quickly over
a short period of time. Although their market share is still small – about 0.1 percent of total
household investible assets – they can influence the market significantly by means of their low
fees and cost-efficient business models. Some would predict that this new type of firm will
replace traditional firms; however, robo-advice providers and traditional firms are not
necessarily substitutes for several reasons.
First, robo- and traditional advice providers serve different populations. Robo-advice
providers or other low-cost investment firms often attract young technology-savvy investors with
low balances, whereas traditional advisers often target older investors with high net worth.
Because robo-advice providers’ client bases are relatively young, robo-advice providers are well
positioned for future growth. If a firm provides a simple technology-only platform for young
investors with low-balances and helps such investors accumulate more wealth over time, later
those investors can be easily brought into a full service program within a more traditional firm.
Due to this generational component, it is not surprising that some traditional firms have begun
acquiring or partnering with robo-advice provider firms.
Second, robo-advice providers and traditional advisers have advantages in different tasks
related to investment services. Beyond portfolio management, human advisers provide a wide
range of services such as tax and estate planning. In contrast, robo-advice providers currently
offer a somewhat narrow range of services within portfolio management. Using computer
algorithms, robo-advice providers automate a few elements of investment services, such as
portfolio rebalancing and tax-loss harvesting. These are typically time-consuming and errorprone tasks if done manually. If human advisers automate these tasks using technology, human
advisers can more efficiently allocate their time to the tasks that can bring more revenue. This
difference in comparative advantage makes robo-advice providers and traditional adviser firms
complement rather than substitute each other.

606

Providers of computerized advice have recently lowered their minimum account balances. As reported in the Wall Street Journal, Personal
Capital Corp. requires $25,000, Wealthfront Inc. requires $500, and Betterment LLC has no minimum but requires a $100 automatic
monthly deposit for accounts below $10,000.

320

Third, robo-advice providers have not been tested in a bear market. If or when the
economy next slows down and the market goes through a correction, relying on a computer
algorithm only may be inadequate to avoid panic selling and the need for a human adviser may
increase. To prepare for a potential downturn, some robo-advice providers reportedly diversify
their revenue streams. Some firms may choose to reach out to financial firms and build
partnership with them. Others may offer more contacts with human advisers. Although it is not
clear how the investment market will shape up during a future market correction, the
partnerships between robo-advice provider firms and traditional firms may become prevalent.
Because of the complementary nature of these two types of business models, robo-advice
providers and traditional firms may merge into one integrated business model. In just one recent
example, a large asset manager firm purchased a small robo-advice provider firm.607 Whether
they are merged or remain separate, unbundling of services may be accelerated due to the
presence of robo-advice providers. Investment firms may be more willing to differentiate the
levels of services and charge fees accordingly. This differentiation is likely to increase profit for
the firms, as it would bring more investors to the market. This is likely to positively affect
investors, as well, because investors can have more choices on the level of services and fees
based on their needs.
Although it is too early to precisely predict how the investment market will evolve over
time, the Department is confident that the number of low-cost automated investment service
firms will increase and their presence will accelerate changes in the market. Therefore, it is
critical to create a regulatory environment where these new innovative firms can grow free from
conflicts. Currently, some robo-advice providers are offering services mostly free from conflicts:
some claim no commission, no performance fees and/or no compensation from third parties, and
others claim to serve investors as fiduciaries. However, this conflict-free model may not last
long as more robo-advice providers expand their businesses through partnering with traditional
firms. For example, one robo-adviser firm assists financial advisers in working with their clients
using their automated account system and they work with other brokerages and custodians, as
well.
The Department expects that the final rule and exemptions will help ensure that these
new approaches evolve toward less conflicted and more innately impartial business models,
rather than succumbing to the competitive pressures that have led more conflicted models to
dominate today’s highly imperfect marketplace. In addition, the new rule and exemptions will,
in the Department’s view, promote the availability of such advisory services, both because the
business models’ technologies can help efficiently ensure the impartiality the rule demands, and
because the new exemptions’ public fee disclosure provisions will help the business models
compete for clients of all sizes by highlighting the now transparent higher price of what had
appeared to be “free” advice provided by full service BDs and others.
Also, financial services firms already are moving toward more fee-based advice models,
considering flatter compensation models, and integrating technology. A growing number of
advisers appear also to be favoring broader application of fiduciary standards. And there is

607

The Economist, “Robo-Advisers Does not compute: The growth of firms selling computer generated financial advice is slowing” (Oct. 31,
2015); available at: http://www.economist.com/news/finance-and-economics/21677245-growth-firms-selling-computer-generated-financialadvice-slowing-does-not.

321

evidence that holding BD representatives to fiduciary standards at the state level does not impair
access to their advice services (Finke and Langdon 2012).
With respect to fees, there is ample room for innovation and market adaptation on the
way advisers are compensated. As consumers gain awareness that advice was never “free,”
demand is likely to grow not only for asset-based fee arrangements, but also for hourly or flat fee
arrangements. Advisory firms may compensate advisers less by commission and more by salary
or via rewards tied to customer acquisition or satisfaction.
The increased use of the electronic delivery of financial advice and investment
management is a powerful trend in the investment industry to improve the retirement savings of
millions of Americans. Technology may actually boost the demand for personal advice and help
firms deliver their services in new, innovative ways. Low cost models are emerging that
integrate technology-based, automated advisory services that include simple user interfaces with
technology-supported, live advisory services. These models have the potential to tailor service
levels to consumers’ needs and preferred price points. Thus, there are new opportunities for
expansion of direct-to-consumer models while that may challenge some of aspects of the
traditional advice model. Some firms have already started to redefine the wealth management
landscape, inventing alternative business models and expanding the boundaries of the wealth
management client base.
In the advisory spectrum there are hybrid financial advisers and wealth managers who
provide an online financial planning structure that is partly automated (robo) and part human
(online financial and investment managers). Generally, this model combines the digital client
portal and investment automation with a virtual financial adviser to deliver personal advice.
These companies are registered investment advisers providing simplified financial solutions
through sophisticated online platforms, eliminating or reducing the need for human interaction.
Firms like Wealthfront Inc. and Betterment LLC offer a direct-to-consumer business model to
offer fully automated investment services without assistance from a financial adviser. One
example of an online hybrid financial adviser is Personal Capital Advisor Corporation offering
registered investment advisers, as well as providing automated investing and financial planning
tools with 900,000 registered users and $1.9 billion of assets under management. Personal
Capital Advisor Corporation offers free electronic gathering of the data from a consumer’s
financial accounts and advice-generating tools projecting long-term forecasts using Monte Carlo
simulations and customization to the needs of investors. SigFig Wealth Management LLC is
another example of a hybrid robo and online financial adviser that applies automated investing
strategies to managing investor portfolios, with a personal investment adviser that is available on
a virtual basis. The Vanguard Group, Inc. also combines digital products with human advisers
creating Personal Advisor Services, which charges 0.30 percent of assets a year,608 enlists a
human financial adviser at the very beginning to set goals and risk tolerance for the investor, and
then uses digital services to manage the account during the accumulation phase. While most of
the established firms are still charging above 1 percent on assets under management (AUM),

608

The 0.30% charge is in addition to the expense ratio of its mutual funds and ETFs.

322

digital entrants are leveraging low-cost managed ETF and single-stock investment portfolios that
provide asset diversification with much lower pricing (i.e., less than 30 basis points).609
Recently many robo and hybrid financial advisers have lowered investment minimums in
order to expand access to investment advice and to increase assets under management. Personal
Capital, for example, has reduced the minimum requirement for opening an account to $25,000
from $100,000, Vanguard’s Personal Advisor Services has reduced its minimum to $50,000
from $100,000, and Wealthfront Inc. reduced its minimum investment to $500 from $5,000. In
the U.S., 76 million households have less than $50,000 of investible assets610 and many in this
category have been underserved in the financial advice market.
Technological advances have been making the creation and delivery of investment
advice easier and less costly in other countries as well. For example, the UK has been studying
the emergence of new advice models that allow consumers - particularly small savers and
consumers who may never have accessed advice or who may have chosen to exit the advice
market after being exposed to the true cost of the advice they had been purchasing and deciding
it was not worth the benefits derived from it - to access guidance and investment advice.
Demand for low-cost online investment propositions is being driven by a number of factors,
including the greater scrutiny of costs with better disclosure of fees and the fact that consumers
have become increasingly comfortable managing their financial affairs online. The FCA has
issued guidance to support firms in the UK wishing to develop new ways of accessing guidance
and advice. This included providing examples using online technology which may help firms
develop cost-efficient new models.
One example of innovation in the UK is provided by Nutmeg, an online discretionary
investment manager that allows customers to invest small amounts of money (£100 per month or
an initial deposit of £1000) into portfolios of assets that are often comprised primarily of ETFs
but which could also include other equities, bonds or commodities. Initially, Nutmeg decides
how to invest its customers’ money based on personal profiles, taking into consideration
investments’ risk and customers’ timelines for investing. Another example of a simplified
advice model from the UK that is targeting smaller savers is Wealth Horizon, which was
launched in 2014, and which offers a hybrid model combining an automated advice platform and
front-end with human advisers behind the scenes to help investors through the process of setting
up portfolios.611 Customers can sign up with as little as £1,000 and advice will be delivered
online and over-the-phone by registered advisers eliminating the face-to-face element. Lastly,
Money on Toast in the UK, one of the first robo-advice companies which appeared as a response
to the RDR, offers advice on investments and creates tailored reports with investment
recommendations. Unlike some of its competitors, Money on Toast goes as far as to offer
restricted personal advice and discretionary management and uses both active and passive funds.
Although the market share of pure robo-adviser model is currently small compared to the
traditional adviser model, it nonetheless may help move the entire market toward more

609

610

611

Ernst & Young (EY), “Advice Goes Virtual: How New Digital Investment Services are Changing the Wealth Management Landscape”
(2015); available at: http://www.ey.com/Publication/vwLUAssets/Advice-goes-virtual/$FILE/EY-Digital-investment-services-Canada.pdf.
Anna Tergesen, “Automated ‘Robo’ Adviser Lowers its Investment Minimum” (Nov. 20, 2015), Wall Street Journal; available at:
http://www.wsj.com/articles/automated-robo-adviser-lowers-its-investment-minimum-1447855380.
EY, “Advice Goes Virtual: How New Digital Investment Services are Changing the Wealth Management Landscape”(2015).

323

affordable models that take advantage of new technologies. This influence of the robo-adviser
model on the entire market is illustrated by the birth of these hybrid models. The robo-adviser
model intensifies competition in the market as traditional adviser model firms either launch their
own robo-adviser services or purchase small robo-adviser firms.612 These types of technologyenhanced adviser models - whether pure robo-adviser or hybrid models – will contain the overall
costs associated with providing investment advice and strategies through this competition.
Furthermore, it will help low-balance account holders obtain the investment advice at an
affordable cost. This new technology-driven model may accelerate the already existing market
trend toward low-cost products such as target-date funds and index funds. Investors have
increasingly directed their investment dollars toward passive investment products in recent
years. The share of all products that were actively managed products was close to 100 percent in
1993. However, the forecasts estimate that this share of active funds will decline to a little over
60 percent by 2020, whereas the share of index funds and ETFs’ share will increase to slightly
less than 40 percent.613 These passively-managed funds, ETFs and index funds often have lower
costs than traditional mutual funds. New technologies similarly enable ETFs and index funds to
keep the overall costs low. Because the market is already heading toward low-cost models, the
estimated compliance costs based on the current data are likely overestimates.
In summary, while the specific future trajectory of innovation is uncertain, the
Department is confident that the final rule and exemptions will accelerate innovation and help
ensure that innovations deliver the largest possible benefit to plan and IRA investors of all sizes.

8.5

Net Welfare Gains Considered Separately

Circular A-4 suggests that agencies should estimate benefits, costs and transfers
separately. Benefits and cost estimates should reflect real resource use, whereas transfers
involve gains to one person or group that are entirely offset (in monetary terms, though not
necessarily as regards utility) by losses to another person or group. As such, a comparison of
benefits and costs yields an estimate of net social welfare gains, so defined. Transfers, while an
important component of a regulatory impact analysis and often a source of distributive effects
that may be grounds for regulatory interventions, are not directly relevant for this specific
purpose.
Data, methodology or other limitations can preclude the estimation of benefits as defined
by Circular A-4, as well as the quantitative comparison of benefits and cost. In the particular
case of this rule and its accompanying exemptions, data constraints that limit the estimation of
investor gains to a subset of those gains, uncertainty about the share of the estimated and
additional unquantified gains to investors consisting of social benefits, and uncertainty about the
effectiveness of the protective provisions of the exemptions in delivering those gains, renders an
estimate of benefits unavailable.614 In such cases, Circular A-4 suggests agencies conduct a

612

613

614

The Economist, “Robo-Advisers Does not compute: The growth of firms selling computer generated financial advice is slowing” (Oct. 31,
2015); available at: http://www.economist.com/news/finance-and-economics/21677245-growth-firms-selling-computer-generated-financialadvice-slowing-does-not.
Morningstar, “The U.S. Department of Labor’s Fiduciary Rule for Advisors Could Reshape the Financial Sector,” Financial Services
Observer (Oct. 2015).
As an example of how the reduction in underperformance may be incomplete consider the possibility that litigation may sometimes lead to a
finding that an advisory firm’s policies and procedures (as set forth in the best interest contract exemption) are protective when, in fact, they
are insufficient.

324

“break-even” analysis, which identifies the threshold level of the uncertain parameter where the
rule’s benefits (defined above as excluding the distributive effects of transfers) would at least
equal its costs. The Department is convinced that the evidence presented in the Regulatory
Impact Analysis on the benefits, costs, and distributional impacts (including transfers to
retirement investors from primarily the financial industry) is sufficient justification for the rule.
In this section it also conducts a break-even analysis of costs and benefits alone, excluding
retirement investor gains attributable to transfers.
The Department’s break-even analysis is complicated due to the number of uncertain
parameters. For illustrative purposes, the Department uses its primary annualized costs ($1.9
billion) and partial gains estimates ($4.2 billion) despite uncertainties in these figures and the
omission from the latter figure of unquantified gains. This allows the Department to solve for the
uncertain parameters of effectiveness and the share of the gains estimate consisting of benefits
(as defined above), which are both theoretically bounded between 0 and 1. For the rule and
accompanying exemptions to at least break even in terms of social welfare as defined by
Circular A-4 alone (excluding gains attributable to transfers), the product of these two
parameters must be greater than or equal to the ratio of costs to the sum of estimated and
unquantified gains.615 Using the primary estimates (and thereby leaving out unquantified gains),
the break-even product of effectiveness and the share of gains consisting of benefits is roughly
0.45. In other words, if the gains to investors entirely consisted of benefits, then the rule and
exemptions must be at least 45% effective to break even. Likewise, if the rule and exemptions
were 100% effective at eliminating conflict-driven underperformance, the share of gains
consisting of benefits must be at least 45%. If unquantified gains were included to facilitate a
more complete comparison that captures all of the gains, rather than only partial gains related to
the reduction of conflicts of interest with respect to front-end-load mutual funds in the IRA
market, this break-even ratio would be smaller.

615

To see why the product of the effectiveness rate and the share of gains due to benefits must be equal to the ratio of costs to gains in order to
break even, consider the following formula: G x S x E – C = W.
Where G is the gains to investors, S is the share of the gains consisting of benefits, E is the effectiveness rate, C is costs, and W represents
net benefits. The rule “breaks even” where W=0. Replacing W with 0 and rearranging algebraically, S x E = C/G.

325

9. Conclusion
This document has presented the Department’s regulatory impact analysis of its final rule
and exemptions. The analysis finds that conflicted advice is widespread, causing serious harm to
plan and IRA investors, and that disclosing conflicts alone would fail to adequately mitigate the
conflicts or remedy the harm. By extending fiduciary status to more advice and providing
flexible and protective PTEs that apply to a broad array of compensation arrangements, the final
rule and exemptions will mitigate conflicts, support consumer choice, and deliver substantial
gains for retirement investors, compromising both social welfare gains and transfers to investors
from the financial industry, and other worthwhile economic effects that together more than
justify their costs.
A wide body of economic evidence supports a finding that the impact of these conflicts
of interest on investment outcomes is large and negative. The supporting evidence, reviewed in
Chapter 3, includes, among other things, statistical analyses of conflicted investment channels,
experimental studies, government reports documenting abuse, and economic theory on the
dangers posed by conflicts of interest and by the asymmetries of information and expertise that
characterize interactions between ordinary retirement investors and conflicted advisers. A
careful review of this data, which consistently point to a substantial failure of the market for
retirement advice, suggests that IRA holders receiving conflicted investment advice can expect
their investments to underperform by an average of 50 to 100 basis points per year over the next
20 years. The underperformance associated with conflicts of interest – in the mutual funds
segment alone – could cost IRA investors between $95 billion and $189 billion over the next 10
years and between $202 billion and $404 billion over the next 20 years. While these expected
losses are large, they represent only a portion of what retirement investors stand to lose as a
result of adviser conflicts. Data limitations impede quantification of all of these losses, but there
is ample qualitative, anecdotal, and in some cases empirical evidence that they occur and are
large both in instance and on aggregate.
The Department expects the final rule and exemptions to deliver large gains for
retirement investors by reducing, over time, the losses identified above. Because of data
limitations, as with the losses themselves, only a portion of the expected gains are quantified in
this analysis. The Department’s quantitative estimate of investor gains from the final rule and
exemptions takes into account only one type of adviser conflict: the conflict that arises from
variation in the share of front-end-loads that advisers receive when selling different mutual funds
that charge such loads to IRA investors. Published research, reviewed in Chapter 3, provides
evidence that this conflict erodes investors’ returns. The Department estimates that the final rule
and exemptions, by mitigating this particular type of adviser conflict, have the potential to
produce gains for IRA investors worth between $33 billion and $36 billion over 10 years and
between $66 and $76 billion over 20 years.
These quantified potential gains do not include additional potentially large, expected
gains to IRA investors resulting from reducing or eliminating the effects of conflicts in IRA
advice on financial products other than front-end-load mutual funds, or the effect of conflicts on
advice to plan investors on any financial products. Moreover, in addition to mitigating adviser
conflicts, the final rule and exemptions raise adviser conduct standards, potentially yielding
additional gains for both IRA and plan investors. The total gains to retirement investors thus
have the potential to be substantially larger than these particular, quantified gains alone.

326

The final exemptions include strong protections calibrated to ensure that adviser conflicts
are fully mitigated such that advice is impartial. If, however, advisers’ impartiality is sometimes
compromised, gains to retirement investors consequently will be reduced correspondingly.
The Department estimates that the cost to comply with the final rule and exemptions will
be between $10.0 billion and $31.5 billion over 10 years with a primary estimate of $16.1
billion, mostly reflecting the cost incurred by affected fiduciary advisers to satisfy relevant
consumer-protective PTE conditions. These cost estimates may be overstated insofar as they
generally do not take into account potential cost savings from technological innovations and
market adjustments that favor lower-cost models. They may be understated insofar as they do
not account for frictions that may be associated with such innovations and adjustments.
Just as with IRAs, there is evidence that conflicts of interest in the investment advice
market also erode the retirement savings of plan participants and beneficiaries. For example,
and discussed in Chapter 4, the U.S. Government Accountability Office (GAO) found that
defined benefit pension plans using consultants with undisclosed conflicts of interest earned 1.3
percentage points per year less than other plans.616 Other GAO reports have found that adviser
conflicts may cause plan participants to roll plan assets into IRAs that charge high fees or 401(k)
plan officials to include expensive or underperforming funds in investment menus.617 A number
of academic studies find that 401(k) plan investment options underperform the market, and at
least one study attributes such underperformance to excessive reliance on funds that are
proprietary to plan service providers who may be providing investment advice to plan officials
that choose the investment options.
The final rule and exemptions’ positive effects are expected to extend well beyond
improved investment results for retirement investors. The IRA and plan markets for fiduciary
advice and other services may become more efficient as a result of more transparent pricing and
greater certainty about the fiduciary status of advisers and about the impartiality of their advice.
There may be benefits from the increased flexibility that the final rule and related exemptions
will provide with respect to fiduciary investment advice currently falling within the ambit of the
1975 regulation. The final rule’s defined boundaries between fiduciary advice, education, and
sales activity directed at independent fiduciaries with financial expertise may bring greater
clarity to the IRA and plan services markets. Innovation in new advice business models,
including technology-driven models, may be accelerated, and nudged away from conflicts and
toward transparency, thereby promoting healthy competition in the fiduciary advice market.
A major expected positive effect of the final rule and exemptions in the plan advice
market is improved compliance and the associated improved security of ERISA plan assets and
benefits. Clarity about advisers’ fiduciary status will strengthen the Department’s ability to
quickly and fully correct ERISA violations, while strengthening deterrence.
A part of retirement investors’ gains from the final rule and exemptions represents
improvements in overall social welfare, as some resources heretofore consumed inefficiently in
the provision of financial products and services are freed for more valuable uses. The remainder

616

617

U.S. Government Accountability Office, GAO-09-503T, Private Pensions: Conflicts of Interest Can Affect Defined Benefit and Defined
Contribution Plans (2009) available at: http://www.gao.gov/new.items/d09503t.pdf.
GAO Publication No. GAO-11-119, 36.

327

of the projected gains reflects transfers of existing economic surplus to retirement investors,
primarily from the financial industry. Both the social welfare gains and the distributional effects
can promote retirement security, and the distributional effects more fairly (in the Department’s
view) allocate a larger portion of the returns on retirement investors’ capital to the investors
themselves. Because quantified and additional unquantified investor gains from the final rule
and exemptions comprise both welfare gains and transfers, they cannot be netted against
estimated compliance costs to produce an estimate of net social welfare gains. Rather, in this
case, the Department concludes that the final rule and exemptions’ positive social welfare and
distributional effects together justify their cost.
In conclusion, the Department’s analysis indicates that the final rule and exemptions will
mitigate adviser conflicts and thereby improve plan and IRA investment results, while avoiding
greater than necessary disruption of existing business practices. The final rule and exemptions
will deliver large gains to retirement investors and a variety of other economic benefits, which,
in the Department’s view, will more than justify its costs.

328

329

Appendix A: Analysis of Broker-Sold Mutual Fund Performance
Using Data from Morningstar
A.1

Introduction

In Chapter 3 of this document, the Department concludes that conflicts of interest in
investment advice cause substantial harm to IRA holders. This conclusion is based on a
preponderance of evidence presented in the 2015 NPRM Regulatory Impact Analysis and
additional evidence provided by commenters on the 2015 NPRM Regulatory Impact Analysis
and testimony at the DOL hearing on conflicts of interest in investment advice in August 2015.
Evidence presented in the 2015 NPRM Regulatory Impact Analysis demonstrates that conflicts
of interest are prevalent in the retail market for investment advice and that advisers frequently
act on those conflicts of interest to the detriment of the customer. Studies demonstrate that
domestic equity mutual funds sold by brokers – in a market where conflicts of interest are
pervasive – consistently underperform direct-sold domestic equity mutual funds.618 These
studies suggest that the harm from conflicts due to the measured underperformance is between
50 and 100 basis points per year and that if other manifestations of harm from conflicts are taken
into account (unmeasured by most of the studies), the total harm from conflicts could be greater
than 100 basis points per year. The Department estimates that the gains to IRA front-end-load
mutual fund investors alone have the potential to be worth between $33 billion and $36 billion
over 10 years and between $66 and $76 billion over 20 years. New data and testimony provided
as comments on the 2015 NPRM Regulatory Impact Analysis and at the DOL hearing in August
2015 demonstrate that conflicts of interest continue to be present in the retail market for
investment advice and continue to harm IRA holders.
Some comment letters on the 2015 NPRM Regulatory Impact Analysis criticize the
Department for relying on “old data” – data presented in peer-reviewed academic papers – in
developing estimates of the harm caused by conflicted advice and gains-to-investors of the
proposed rule. One comment in particular, from ICI, claims that the nature of competition in the
mutual fund market has fundamentally changed, rendering the “old data” irrelevant.619 ICI
criticizes the Department for failing to utilize recent, publicly available mutual fund performance
data to independently verify and update the findings in the academic studies. Several other
comments point to ICI’s criticisms, analysis, and conclusions.
While the evidence in the public record is more than sufficient to demonstrate that ICI’s
claims about market changes are false,620 the Department has conducted its own analysis to
supplement the evidence. Just as ICI used Morningstar data on mutual fund performance for its
analysis, the Department has acquired and analyzed mutual fund performance data from
Morningstar. The Department finds that broker-sold domestic equity mutual funds
underperformed direct-sold domestic equity mutual funds, on average, by 59 to 85 basis points
per year over the entire sample period, 1980-2015, and underperformed by 81 to 101 basis points
over the recent period, 2008-2015. These results affirm several conclusions that the Department

618

619
620

As noted in the 2015 NPRM Regulatory Impact Analysis, the opposite appears to be true for foreign equity mutual funds when performance
is aggregated on an asset-weighted basis.
See ICI comment letter (July 21, 2015); available at: http://www.dol.gov/ebsa/pdf/1210-AB32-2-00506.pdf.
See Section 3.2.4.

330

has drawn from the academic literature and other data in the public record. First, the analysis
confirms the Department’s finding that, contrary to ICI’s claim, the nature of competition in the
mutual fund market has not fundamentally changed in recent years. Second, the analysis
confirms the Department’s characterization of the academic literature presented in the 2015
NPRM Regulatory Impact Analysis. Third, the relative-performance data presented in ICI’s
comment letter obscure stark differences in the relative performance of broker-sold and directsold funds for domestic equity funds versus international equity funds.
The remainder of this appendix details the Department’s analysis.

A.2

Data

Using a Morningstar Direct subscription, DOL downloaded data on 19,079 domestic
equity mutual funds, 7,160 international equity mutual funds, and 3,237 sector equity mutual
funds covering the years 1980 to 2015. Since some funds were removed for various reasons,
DOL’s final analysis included a subset of these 29,476 funds.621 The year 1980 was chosen as
a start date because it appears to be the beginning of the modern era in mutual fund distribution,
and so that the sample would span back at least as far as each of the studies included in Figure 317. Along with the mutual fund name and ticker symbol, key data elements downloaded include
annual and monthly returns, monthly assets, Morningstar-calculated 1-year 1-factor alphas, and
share class identified by Morningstar.
Sector and domestic equity mutual funds were aggregated and will henceforth be referred
to simply as domestic equity mutual funds. Sector mutual funds make up a small fraction of all
equity mutual fund assets (just under 10 percent in 2015) and appear to consist primarily of
domestic assets. For ease of illustration, the Department presents the combined performance of
domestic and sector funds as one group.
Three measures of risk-adjusted returns are considered for equity funds: Morningstar’s
1-factor alpha, a DOL calculated 1-factor alpha, and a DOL calculated 3-factor alpha. These
risk-adjusted alphas measure the performance of a fund after accounting for volatility that is
correlated with market risk.
The DOL- and Morningstar-calculated 1-year 1-factor alphas differ in several important
ways. First, the Morningstar 1-factor alpha appears to use the S&P 500 index as its benchmark
for domestic equity mutual funds whereas the DOL alphas use the Fama-French factors available
on Kenneth French’s Dartmouth website.622 Second, Morningstar appears to multiply by 12 in
order to convert a monthly alpha to an annual alpha, whereas the DOL monthly alphas are
converted to annual alphas using a methodology that takes into account compounding.623

621
622

623

For details on the mutual funds included in this analysis, see AACG’s Morningstar analysis report (Padmanabhan, Panis and Tardiff 2016b).
Home Page of Kenneth R. French, Tuck School of Business, Dartmouth College; available at:
http://mba.tuck.dartmouth.edu/pages/faculty/ken.french/index.html.
The DOL monthly alphas are converted to annual alphas as follows. Monthly alphas are used in conjunction with monthly returns to
determine the mutual fund’s monthly benchmark returns. Monthly returns and monthly benchmark returns are then converted to annual
returns and annual benchmark returns, respectively, by compounding gross returns. A mutual fund’s annual alpha is the difference between
the annual returns and the annual benchmark returns. This method of converting monthly alphas to annual alphas has little impact on the
overall underperformance estimates. Across all sample years, the DOL conversion methodology decreases average estimated
underperformance by only 1 to 2 basis points relative to the Morningstar methodology.

331

Measures of aggregate broker-sold mutual fund underperformance may differ across the DOL
and Morningstar alphas for one additional reason. For international equity mutual funds,
Morningstar 1-factor alphas only go back as far as 1999, whereas Fama-French international
equity factors, and likewise the associated DOL alphas, go back to 1990. Where an alpha is
missing, the fund is not included in the asset-weighted average alpha calculation for that year.

A.3

Broker-Sold Domestic Equity Mutual Fund Underperformance

The Department’s analysis of mutual fund performance data affirms its previous
conclusion that broker-sold domestic equity mutual funds consistently underperform direct-sold
domestic equity mutual funds. Figure A-1 presents the average Morningstar 1-factor alpha,
DOL 1-factor alpha, and DOL 3-factor alpha over the full sample period, by distribution
channel, as well as the difference in the metrics across the distribution channels.
Figure A-1 Risk-adjusted Domestic Equity Mutual Fund Performance by Distribution
Channel, 1980-2015 (Percentage Points, Annual)
Direct-Sold

BrokerSold

Difference

Morningstar 1-Factor Alpha

0.427

-0.425

0.852**

DOL 1-Factor Alpha

0.472

-0.207

0.680*

DOL 3-Factor Alpha

0.214

-0.381

0.594**

Domestic Equity Mutual Funds

** Indicates that the difference is statistically significant at the 1% level.
* Indicates that the difference is statistically significant at the 5% level.

These averages are calculated using a two-step process. First, the risk-adjusted
performance metrics (alphas) are averaged on an asset-weighted basis within each year,
generating an average for each year. In order to take into account the distribution of assets in the
market, funds are weighted by assets held within years. Next, the yearly averages are averaged
together across years on an equal-weighted basis. Unlike within individual years, across years it
is less clear whether asset-weighting is appropriate. In samples that span a much shorter number
of years, the difference may be negligible. But in the current sample, asset-weighting across
years would cause the earlier years to carry far less weight in the average.
The average Morningstar 1-factor alpha for direct-sold domestic equity mutual funds
between 1980 and 2015 is 0.427 percent, while the average Morningstar 1-factor alpha for
broker-sold equity mutual funds over the same time period is -0.425 percent. The difference
represents broker-sold equity mutual fund underperformance of about 85 basis points. The
performance differences are somewhat smaller, but still very economically meaningful when
performance is measured by DOL-calculated 1-factor and 3-factor alphas. Using 1-factor alphas
calculated by DOL, the broker-sold domestic equity mutual fund underperformance was about
68 basis points per year. Broker-sold equity mutual funds underperform by an average of 59
basis points per year between 1980 and 2015 according to the DOL calculated 3-factor alphas.
DOL’s 3-factor alpha metric is clearly preferred to its 1-factor alpha, but the relative merit of
DOL’s 3-factor alpha and Morningstar’s 1-factor alpha is less clear. The discussion in Section
A.2 above includes a discussion of how they differ.
Broker-sold mutual fund underperformance in the domestic equity space is evident over
the entire sample period 1980-2015. Figure A-2 presents the average underperformance of
broker-sold domestic equity funds for various time periods within the 1980-2015 sample.
332

Figure A-2 Broker-Sold Domestic Equity Mutual Fund Underperformance over Various Time Periods
(Percentage Points, Annual)
1980-2015

1980-1987

1988-1997

1998-2007

2008-2015

Morningstar 1-Factor Alpha

0.852

0.368

1.045

0.918

1.012

DOL 1-Factor Alpha

0.680

0.265

0.882

0.677

0.844

DOL 3-Factor Alpha

0.594

0.149

0.565

0.806

0.813

In the most recent period, 2008-2015, broker-sold domestic equity mutual fund
underperformance remains substantial. Using the Morningstar 1-factor alpha metric, the brokersold underperformance for the 2008-2015 period is just over 100 basis points. The DOL 1-factor
and 3-factor alphas indicate that broker-sold domestic equity mutual fund underperformance in
that period are between 80 and 85 basis points.
ICI claims that a fundamental change has occurred in the mutual fund market such that
broker-sold funds now compete directly with direct-sold funds. In support of this claim, ICI
presents data showing that the share of mutual funds with a front-end-load-share class that also
have a no-load-share class has increased between 2000 and 2010. While this shift may have the
potential to change mutual fund incentives, the data do not bear it out. As demonstrated in
Figure A-2, recent broker-sold domestic equity mutual fund underperformance is just as large or
larger than it has been in the past. 624

A.3.1

Hypothesis Tests

While Figure A-2 demonstrates that broker-sold domestic equity mutual funds continue
to underperform direct-sold domestic equity mutual funds, this section presents a more formal
analysis of ICI’s claims.
What is the most appropriate test to determine whether the underperformance of brokersold mutual funds has changed over time? ICI’s comment letter suggests that the change in the
mutual fund market has been occurring gradually for at least the last 20 years. Other literature
suggests that this gradual change began as early as 1980.625 One can test this hypothesis by
looking for a time trend in the underperformance of broker-sold mutual funds.

624

625

An alternative empirical investigation of whether the nature of competition in the mutual fund market has changed would replicate the CEM
study using newer data. ICI, in fact, performed this investigation subsequent to their initial comment letter as reported in a letter to the
Department dated December 1, 2015. Available at: http://www.dol.gov/ebsa/pdf/ ici-letter-to-supplement-comment-12-01-2015.pdf. CEM
studied broker incentives and subsequent mutual fund returns between 1993 and 2009 and found that performance decreases as the loadshare paid to the broker increases. ICI reports in footnote 18 (page 9) of their December 1st letter that they have affirmed the CEM finding
using data from 2010 through 2013. Mutual funds still appear to have less incentive to invest in performance when they are able to sell their
product by incentivizing brokers instead. This result stands in direct contrast to ICI’s earlier claim that the nature of competition in the
mutual fund market has changed.
See ICI, “Mutual Fund Distribution Channels and Distribution Costs,” ICI Perspective, Vol.9, No. 3, July 2003.

333

Hypothesis 1: Within the period 1980-2015, the underperformance of broker-sold
domestic equity mutual funds is decreasing in time.
Null Hypothesis: Within the period 1980-2015, the underperformance of broker-sold
domestic equity mutual funds remains constant over time.
Figure A-3 Results of Three Regressions of Domestic Equity Performance Difference (Direct-Sold Performance
Minus Broker-Sold Performance) on Year
Intercept (1997)

Year

Performance Measure

n

Coefficient

t-statistic

p-value

Coefficient

t-statistic

p-value

Morningstar 1-Factor Alpha

36

0.84

3.12

0.004

0.0161

0.62

0.540

DOL 1-Factor Alpha

36

0.67

2.28

0.029

0.0146

0.52

0.608

DOL 3-Factor Alpha

36

0.58

2.91

0.006

0.0295

1.54

0.133

Figure A-3 presents the results of three regressions that test Hypothesis 1. In each of the
regressions, the independent variable is the broker-sold mutual fund underperformance – the
asset-weighted difference between direct-sold and broker-sold mutual fund performance – in a
given year. The underperformance is regressed on year in a univariate regression. One
regression is run for each performance metric – Morningstar 1-factor alpha, DOL 1-factor alpha,
and DOL 3-factor alpha. The measure of interest in the regression is the coefficient on the
variable ‘Year.’ A positive value would indicate that the underperformance of broker-sold
mutual funds is increasing over time, while a negative value would indicate that the
underperformance of broker-sold mutual funds is getting less severe.
The regression results in Figure A-3 indicate no long-term trend in broker-sold mutual
fund underperformance. If anything, it appears that the underperformance of broker-sold
domestic equity funds may be increasing over time (the point estimate of the coefficient on Year
is positive in all three regressions), though the results are not statistically significant. No
evidence exists to support the hypothesis proposed in ICI’s comment letter that the
underperformance of broker-sold mutual funds has gradually decreased over time.
This non-result can be seen more clearly in graphical form. Figure A-4 plots the
underperformance of domestic equity broker-sold mutual funds in a given year for two of the
three risk-adjusted performance metrics – Morningstar 1-factor alpha and DOL 3-factor alpha.626
Each point in the scatterplot represents the average difference in performance (direct-sold
domestic equity mutual fund performance minus broker-sold domestic equity mutual fund
performance) for a given risk-adjusted performance metric in a given year. The light blue
diamonds represent underperformance estimated by the Morningstar 1-factor alpha while the
dark blue circles represent underperformance estimated by the DOL 3-factor alpha. A trend line
is drawn for each performance metric reflecting the regression results presented in Figure A-3.
In both cases, the trend line is slightly upward sloping, but the time-trend is not statistically
significant.

626

The DOL 1-factor alpha is not included in order to maintain readability. Note, though, that the trend line for the DOL 1-factor alpha
appears extremely similar to those shown in this graph.

334

In contrast to the hypothesis that the market has changed gradually over time, some
commenters simply criticize the 2015 NPRM Regulatory Impact Analysis for relying on “old
data.” These criticisms suggest that a fundamental market change may have occurred at a single
point in time such that returns data observed before the market change could be considered “old”
and returns data observed after the market change could be considered “new.” It is not clear at
precisely what time these commenters claim the market change occurred, but many of the
commenters cite one or more ICI analyses of returns data beginning in 2008. Therefore, it seems
reasonable to test the hypothesis that a fundamental market change occurred on or around
Figure A-4 -- Broker-sold Domestic Equity Mutual Fund
Underperformance, 1980-2015
Direct-sold domestic equity mutual fund performance minus broker-sold domestic equity
mutual fund performance (asset-weighted, percentage points)

6

4

Domestic
Equity Morningstar 1factor Alpha

2

Domestic
Equity - DOL 3factor Alpha

0
Linear
(Domestic
Equity Morningstar 1factor Alpha)
Linear
(Domestic
Equity - DOL 3factor Alpha)

-2

-4

-6
1980

1985

1990

1995

2000

2005

2010

2015

January 1, 2008.
Hypothesis 2: The underperformance of domestic equity broker-sold funds during the
period 2008-2015 is less than the underperformance of domestic equity broker-sold
funds during the period 1980-2007.
Null Hypothesis: The underperformance of domestic equity broker-sold funds during
the period 2008-2015 is equal to the underperformance of domestic equity broker-sold
funds during the period 1980-2007.
Figure A-5 presents the results of three statistical t-tests of Hypothesis 2, one each for the
risk-adjusted performance metrics: Morningstar 1-factor alpha, DOL 1-factor alpha, and DOL 3factor alpha. In each test, the mean of broker-sold domestic equity mutual fund
underperformance – the asset-weighted difference between direct-sold and broker-sold mutual
fund performance – in the 1980-2007 was compared to the mean of underperformance in the
years 2008-2015. A t-test was performed, allowing for unequal variances, to assess whether any
observed change in the underperformance of broker-sold mutual funds is statistically significant.
335

Figure A-5 Results of Three Statistical t-tests on Average Broker-Sold Domestic Equity Mutual Fund
Underperformance Before and After January 1, 2008
1980-2007

2008-2015

Change in Underperformance

Performance measure

n

Mean

Variance

n

Mean

Variance

Change

t-statistic

p-value
(two-tailed test)

Morningstar 1-Factor
Alpha

28

0.81

3.23

8

1.01

0.42

0.21

0.50

0.619

DOL 1-Factor Alpha

28

0.63

3.86

8

0.84

0.28

0.21

0.51

0.615

DOL 3-Factor Alpha

28

0.53

1.86

8

0.81

0.15

0.28

0.96

0.343

The statistical tests presented in Figure A-5 indicate no statistically significant change in
the underperformance of broker-sold mutual funds from the “old” time period, 1980-2007, to the
“new” time period, 2008-2015. Once again, the point estimate for the change goes in the
opposite direction of the change suggested by ICI. The data show that, according to the
Morningstar 1-factor alpha, the average underperformance of broker-sold domestic equity
mutual funds increased from 81 basis points in the 1980-2007 period to 101 basis points in the
2008-2015 period. Likewise, according to the DOL 1-factor and 3-factor alphas, the average
underperformance of broker-sold domestic equity mutual funds increased from 63 and 53 basis
points in the 1980-2007 period to 84 and 81 basis points in the 2008-2015 period. These
differences are not statistically significant, so the changes may not be meaningful in terms of
predicting future underperformance (using an average over a longer time horizon is likely to be a
better predictor than an average of just the most recent values). But there is certainly no
evidence here supporting ICI’s conjecture that the market has fundamentally changed.
Finally, it is possible that ICI intended to claim that the recent market change
differentiates the recent period (2008-2015) from the period containing the data that the 2015
NPRM Regulatory Impact Analysis relied upon (generally 1993-2009)627, but that the recent
market change does not necessarily differentiate the recent period from any earlier period (e.g.
1980-1992).
Hypothesis 3: The underperformance of domestic equity broker-sold funds during the
period 2008-2015 is less than the underperformance of domestic equity broker-sold
funds during the period 1993-2007.
Null Hypothesis: The underperformance of domestic equity broker-sold funds during
the period 2008-2015 is equal to the underperformance of domestic equity broker-sold
funds during the period 1993-2007.
This claim also appears to have no merit in the data. Figure A-6 demonstrates that the
broker-sold domestic equity mutual fund underperformance is virtually identical in the 19932007 and 2008-2015 periods. In each case, the absolute value of the difference in average
underperformance across the two time periods is 12 basis points or less, and the change is not

627

The 2015 NPRM Regulatory Impact Analysis relied upon data between 1993 and 2009. However, for the purpose of applying a proper
statistical test on non-overlapping samples, Hypothesis 3 shortens this period to 1993-2007.

336

uniform across the performance metrics. None of the differences in Figure A-6 are statistically
significant.
Figure A-6 Results of Three Statistical t-tests on Average Broker-Sold Domestic Equity Mutual Fund
Underperformance, 2008-2015 versus 1993-2007
1993-2007

2008-2015

Change in Underperformance

Performance measure

n

Mean

Variance

n

Mean

Variance

Change

t-statistic

p-value
(two-tailed test)

Morningstar 1-Factor
Alpha

15

0.99

2.42

8

1.01

0.42

0.02

0.05

0.960

DOL 1-Factor Alpha

15

0.72

3.34

8

0.84

0.28

0.12

0.24

0.813

DOL 3-Factor Alpha

15

0.90

0.97

8

0.81

0.15

-0.09

-0.32

0.755

A.4

Aggregated Performance

ICI’s results obscure the underperformance of broker-sold domestic equity mutual funds
by averaging mutual fund performance across all types of equity mutual funds – domestic and
foreign. The academic literature presented in the 2015 NPRM Regulatory Impact Analysis
recognizes that when measuring average performance on an asset-weighted basis, broker-sold
foreign equity mutual funds overperform direct-sold foreign equity mutual funds.628 Figure A-7
demonstrates that when broker-sold domestic equity mutual fund underperformance is combined
with broker-sold foreign equity overperformance, the aggregated results appear more mixed. In
addition to the risk-adjusted alphas, Figure A-7 presents aggregated performance differences
using raw, unadjusted returns.

628

While the result is demonstrated in the literature cited in the 2015 NPRM Regulatory Impact Analysis and affirmed here, it is not clear why
broker-sold foreign equity mutual funds consistently outperform direct-sold foreign equity mutual funds when averaged on an assetweighted basis. Bergstresser et al. (2009) points to a single, high-performing fund family that makes up a large plurality of broker-sold
foreign equity funds (about 40 percent in the DOL Morningstar sample). The Department’s analysis finds that this fund family explains a
large portion, but not all of, the overperformance of foreign equity mutual funds. A second and related possible explanation points to the
potential for active management to provide more value in international markets than in domestic markets, especially the domestic large-cap
mutual fund market. In the domestic large-cap mutual fund market, asset managers generally have immediate access to nearly full
information. Consequently, there is little mispricing to exploit and little opportunity for particular asset managers to add net value. In this
space, any negative effect of conflicts of interest (reduced performance) is less likely to be obscured by other factors affecting performance.
In contrast to the domestic market, there is greater opportunity for asset managers in the foreign equity market to add value if they can
obtain superior information. To the extent that foreign equities are less likely to be efficiently priced, these markets may provide
opportunities for active managers to exploit inefficiencies and generate above-market returns. It’s possible that a particular asset managers
or managers whose funds provide superior performance and also happen to be sold through the broker channel provide enough value to
overwhelm any potentially negative effects of conflicts of interest in the foreign equity performance data. This theory is not inconsistent
with results from CEM showing that, across all types of mutual funds, brokers steer clients to funds that pay the broker more and result in
poorer performance for the client.

337

Figure A-7 Risk-adjusted Equity Mutual Fund Performance by Investment Location and
Distribution Channel, 1980-2015 (Percentage Points, Annual)
Domestic Equity Mutual Funds
Assets: $2.73 trillion)

(2015

Direct-Sold

Broker-Sold

Difference

Morningstar 1-Factor Alpha

0.427

-0.425

0.852**

DOL 1-Factor Alpha

0.472

-0.207

0.680*

DOL 3-Factor Alpha

0.214

-0.381

0.594**

Direct-Sold

Broker-Sold

Difference

Morningstar 1-Factor Alpha

0.778

2.018

-1.226*

DOL 1-Factor Alpha

1.350

3.070

-1.720*

DOL 3-Factor Alpha

1.399

3.013

-1.614*

Direct-Sold

Broker-Sold

Difference

Returns

12.839

11.899

0.941**

Morningstar 1-Factor Alpha

0.404

-0.175

0.579*

DOL 1-Factor Alpha

0.506

0.354

0.151

DOL 3-Factor Alpha

0.304

0.245

0.060

Foreign Equity Mutual Funds
Assets: $0.85 trillion)

Aggregated Domestic and Foreign
Equity Mutual Funds

(2015

** Indicates that the difference is statistically significant at the 1% level.
* Indicates that the difference is statistically significant at the 5% level.

The aggregated performance differences presented in Figure A-7 are comparable to the
performance difference presented in Figure 4 of ICI’s comment letter.629 ICI finds that the frontend-load-share classes underperform retail no-load-share classed by 43 basis points on average
when performance is measured relative to a Morningstar category average. While important
distinctions exist between the two methodologies – timeframe, risk-adjustment technique,
inclusion/exclusion of bond mutual funds, and examination of front-end-load funds vs. all funds
sold by brokers to name a few – the ICI result does fall within the range of the risk-adjusted
aggregated equity mutual fund performance differences presented in Figure A-7 (6 to 58 basis
points).

A.5

Conclusion

Based on evidence in the record – academic studies cited in the 2015 NPRM Regulatory
Impact Analysis, comments on the 2015 NPRM Regulatory Impact Analysis, and testimony at

629

See page 20 of ICI comment letter, (July 21, 2015); available at: available at: http://www.dol.gov/ebsa/pdf/1210-AB32-2-00749.pdf.

338

the DOL hearing in August 2015 – the Department has concluded that conflicts of interest
continue to harm IRA retirement investors. As a supplemental analysis, the Department
examined mutual fund performance data from Morningstar. The results of the Department’s
analysis affirm that 1) contrary to ICI’s claim, the nature of competition in the mutual fund
market has not fundamentally changed in recent years; broker-sold domestic equity mutual funds
continue to dramatically underperform direct-sold domestic equity mutual funds, 2) the
Department’s characterization of the academic literature presented in the 2015 NPRM
Regulatory Impact Analysis remains valid, and 3) the relative-performance data presented in
ICI’s comment letter obscures stark and meaningful differences in the relative performance of
broker-sold and direct-sold funds for domestic equity funds versus international equity funds.

339

Appendix B: Bases for Estimates of Harm and Subsets of Gain to
Investors
An overview of the methodology used to estimate a small subset of the final rule and
exemptions’ expected gains to investors appears in Section 3.3.1 above. This section provides
more detail on the calculations. Section B.1 presents the projection of investment performance
over the projection period for the baseline and alternative scenarios, while Section B.2 presents
the methodology and calculation of the estimates. Section B.3 discusses each of the assumptions
used in the projections in detail and offers a sensitivity analysis for many of the assumptions.
Section B.4 presents underperformance estimates and the assumptions required to generate those
estimates.

B.1

Investment Performance

As discussed in Section 3.3.1, the alternative scenarios diverge from the baseline scenario
solely on investment performance net of loads. The baseline projection assumes that investment
returns are equal to 6 percent minus two values: 1) any front-end-loads paid during the year, and
2) the effect of current and past loads on performance.

340

341

(A)
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
2028
2029
2030
2031
2032
2033
2034
2035
2036

Year

102
99
96
93
90
87
84
81
79
76
74
72
69
67

126

122

118

114

110

107

103

100

97

94

91

88

85

82

124

153

106

128

158

130

133

163

109

137

169

134

142

174

113

146

180

139

151

186

117

156

192

120

161

198

143

167

205

148

(C)

(B)

42

45

48

51

55

59

63

67

71

76

82

87

93

99

106

113

121

130

138

148

---

---

---

---

---

---

---

---

---

(D)

Alternative
scenario 2
Baseline
Baseline
average load average load- average load
paid by IRA share paid to paid by IRA
holder (basis broker (basis holder (basis
points)
points)
points)

-0.14%

-0.14%

-0.15%

-0.15%

-0.16%

-0.16%

-0.17%

-0.17%

-0.18%

-0.19%

-0.19%

-0.20%

-0.20%

-0.21%

-0.22%

-0.23%

-0.23%

-0.24%

-0.25%

-0.26%

---

---

---

---

---

---

---

---

-0.07%

-0.08%

-0.08%

-0.09%

-0.09%

-0.10%

-0.11%

-0.11%

-0.12%

-0.13%

-0.14%

-0.15%

-0.16%

-0.17%

-0.18%

-0.19%

-0.20%

-0.22%

-0.23%

-0.25%

---

---

---

---

---

---

---

---

---

(F)
= (D) * 16.8%

(E)
= (B) * 16.8%
---

Alternative
scenario 2
direct effect
of loads on
average
performance
net of loads

Baseline
direct effect
of loads on
average
performance
net of loads

-0.30%

-0.31%

-0.32%

-0.33%

-0.34%

-0.35%

-0.37%

-0.38%

-0.39%

-0.40%

-0.42%

-0.43%

-0.45%

-0.46%

-0.48%

-0.49%

-0.51%

-0.52%

-0.54%

-0.56%

-0.58%

-0.60%

-0.62%

-0.64%

-0.66%

-0.68%

-0.70%

-0.73%

-0.75%

(G)
= (C) * 0.4494

Effect of load
paid in a
given year on
current and
future
performance

-0.33%

-0.34%

-0.36%

-0.37%

-0.38%

-0.39%

-0.41%

-0.42%

-0.43%

-0.45%

-0.46%

-0.48%

-0.49%

-0.51%

-0.53%

-0.54%

-0.56%

-0.58%

-0.60%

-0.62%

---

---

---

---

---

---

---

---

---

(H)

Baseline
effect of
current and
past loads on
performance
in a given
year

Figure B-1 -- Calculation of Investment Performance Net of Loads for Baseline and Alternative Scenarios

0.00%

0.00%

0.00%

0.00%

0.00%

0.00%

0.00%

0.00%

0.00%

0.00%

0.00%

-0.01%

-0.02%

-0.06%

-0.11%

-0.17%

-0.25%

-0.33%

-0.42%

-0.53%

---

---

---

---

---

---

---

---

---

(I)

Alternative
scenarios
effect of past
loads on
performance
in a given
year
(K)
=6% + (E) + (I)

------------------5.22%
5.33%
5.43%
5.52%
5.61%
5.67%
5.73%
5.77%
5.80%
5.81%
5.81%
5.82%
5.83%
5.83%
5.84%
5.84%
5.85%
5.85%
5.86%
5.86%

------------------5.12%
5.15%
5.18%
5.21%
5.23%
5.26%
5.28%
5.30%
5.32%
5.35%
5.37%
5.39%
5.41%
5.43%
5.44%
5.46%
5.48%
5.50%
5.51%
5.53%

Alternative
scenario 1
investment
performance
net of loads

(J)
=6% + (E) + (H)

Baseline
investment
performance
net of loads

------------------5.23%
5.35%
5.45%
5.55%
5.64%
5.71%
5.77%
5.82%
5.85%
5.86%
5.87%
5.88%
5.89%
5.89%
5.90%
5.91%
5.91%
5.92%
5.92%
5.93%

(L)
=6% + (F) + (I)

Alternative
scenario 2
investment
performance
net of loads

------------------5.47%
5.58%
5.67%
5.75%
5.83%
5.89%
5.94%
5.98%
5.99%
6.00%
6.00%
6.00%
6.00%
6.00%
6.00%
6.00%
6.00%
6.00%
6.00%
6.00%

(M)
=6% + (I)

Alternative
scenario 3
investment
performance
net of loads

Figure B-1 displays the projected average front-end-loads paid by IRA investors (column
B) over the projection period. These load projections are generated using the average load
observed in the CEM data and then scaling downward based on the trend in front-end-load size
observed by the ICI (see Sections B.3.2.1 and B.3.2.3).630 Also displayed are the average frontend-load-shares projected to be received by brokers (column C). These are calculated by
multiplying the average front-end-loads paid by IRA investors by 81.4 percent, which is the
assumed share of the total load that will be shared with the broker (see Section B.3.2.2).
In a given year, only a fraction of IRA front-end-load mutual fund assets will turn over
and incur a front-end-load. Data from CEM suggest that this fraction is approximately 16.8
percent (see Section B.3.2.4).631 The direct effect of those loads on average investment
performance (column E) is estimated to be the average front-end-load paid by IRA investors in
that year (column B) multiplied by 16.8 percent.
The indirect effect of front-end-loads on performance in a given year is somewhat more
complicated to calculate. For mutual funds that pay load-shares exclusively to unaffiliated
brokers, an estimate from CEM suggests that for every 100 basis points of the load that go
toward the broker’s load-share, an IRA investor can expect to experience a decrease in
performance of 49.7 basis points. In the case of mutual funds that pay load-shares exclusively to
captive brokers, for every 100 basis points of the load that go toward the broker’s load-share, an
IRA investor can expect to experience a decrease in performance of 14.5 basis points.
Unaffiliated brokers constitute a large fraction of the advice market, so an asset weighted
average of these effects results in an average decrease in performance of 44.9 basis points for
every 100 basis points in load-share (see Section
B.3.1). Column (G) estimates this effect by
Figure B-2 Distribution of Purchase Dates
multiplying the average load-share (Column C) by
for Front-end-load Assets Owned in a
0.449. But this is not the average effect on
Given Year
performance in a given year. For that, a distribution
of purchase dates for front-end-load funds owned in a
% of year t front-end-load assets
given year is needed. The estimate that 16.8 percent
Year
purchased in each year
of load funds turn over in a given year is a good place
t-9
to start. Figure B-2 provides a distribution of
1.0%
purchase dates for IRA front-end-load funds owned
t-8
3.0%
in a given year t (see Section B.3.2.5). The baseline
t-7
6.0%
effect of current and past loads on performance
t-6
8.0%
(Column H) is calculated by applying the purchase
t-5
10.0%
date distribution (Figure B-2) to the current year and
t-4
12.4%
previous 9 years in Column (G).
Finally, the baseline investment performance
net of loads (Column J) is calculated as 6 percent
minus the direct effect of loads (Column E) minus
the effect of current and past loads on performance
(Column H).

630
631

t-3
t-2
t-1
t
Total

13.2%
14.2%
15.4%
16.8%
100.0%

There is more discussion of this projection and other assumptions in Section B-3: Assumptions and Uncertainty.
DOL calculation uses CEM, Table 1. Inflows subject to load = 1.40 percent of TNA per month or 16.8 percent of TNA per year.

342

Alternative scenario 1 differs from the baseline only in the effect of loads on current and
future performance. The effect on current and future performance is the same as the baseline for
years 2008-2016 – the years before the requirements of the rule become applicable – but it drops
to zero once the requirements take effect. As discussed previously, this downward pull on
performance is eliminated because affected advisers are no longer incentivized to recommend
particular funds. Column (I) estimates the effect of past loads on performance in a given year
using the load effect in Column (G) and the purchase date distribution in Figure B-2. The
alternative scenario load effect on performance remains fairly high (-0.53 percent) in 2017
because most of the assets owned in 2017 were purchased prior to the effective date of the rule.
Gradually, as time moves further from the effective date of the rule, the alternative scenario load
effect on performance goes to zero. The alternative scenario 1 investment performance net of
loads (Column K) is calculated by subtracting the baseline direct effect of loads (Column E) and
the alternative scenario effect of loads on performance (Column I) from 6 percent.
The effect of past loads on performance is the same across all three alternative scenarios;
alternative scenario 2 differs from alternative scenario 1 because loads fall faster under the 2nd
alternative. The average load projection for alternative scenario 2 is displayed in Column (D).
As in the baseline scenario, the alternative scenario direct effect of loads on average
performance net of loads (Column F) is the average load (Column D) multiplied by 16.8 percent,
the estimate of front-end-load assets turnover per year. The alternative scenario 2 investment
returns net of loads (Column L) is 6 percent minus the alternative scenario 2 direct effect of
loads minus the alternative scenario effect of past loads on performance.
Under alternative scenario 3, loads immediately go to zero when the proposal becomes
effective, so there is no direct effect of loads on investment performance. As in the other
alternative scenarios, there is still an indirect effect of past loads on performance. The
alternative scenario 3 investment performance (Column M) is 6 percent minus the alternative
scenario effect of past loads on performance (Column I).
The investment performance estimates in Columns (J), (K), (L), and (M) are the key
factors in the 10- and 20-year quantified subset of IRA investors’ expected gains causing asset
accumulation to diverge across the scenarios.

B.2

Front-End-Load-Mutual-Fund-Gains-to-Investors Estimates

The Department estimates the quantified subset of IRA investors’ expected gains over
10- and 20-year periods beginning in April 2017, one year after the rule and exemptions are
finalized.632 The estimates are derived by comparing alternative scenarios, under the rule, to the
baseline scenario where no rule is finalized. Figure B-3 walks through the 10-year front-endload-mutual-funds-gains-to-investors calculation for alternative scenario 1. Under all of the
scenarios, IRA assets in April 2017 are projected to total about $8.7 trillion.633 Of the $8.7
trillion, approximately 43.9 percent (about $3.8 trillion) are invested in mutual funds, and of

632

633

The initial applicability date for the rule is set for one year after publication of the final rule in the Federal Register. In contrast, the frontend-load-mutual-funds-gains-to-investors estimates are discounted back to the present, April 2016. See Section 2.9.1 and 2.9.2 for
discussion of applicability dates of the rule and exemptions.
Cerulli Associates, “Retirement Markets 2015.” Cerulli’s projection of year-end 2016 IRA assets is used as an approximation for projected
IRA assets in April 2017. Similar approximations are used in other years.

343

those IRA assets invested in mutual funds, approximately 27 percent incur a front-end-load. The
resulting pool of assets under consideration is $1.043 trillion in 2017 (Figure B-3, Row B).
Figure B-3 Calculation of 10-year Front-load-mutual-fund-gains-to-investors Under Alternative Scenario 1 ($ billion)
Projection Year*
Baseline total IRA year-begin assets
Baseline year-begin front-load mutual fund
assets
Scenario 1 year-begin front-load mutual
fund assets
Baseline year-end front-load mutual fund
assets
Scenario 1 year-end front-load mutual fund
assets

2017

2018

2019

2020

2021

2022

2023

2024

2025

2026

(A)

8,652

9,308

10,000

10,729

11,497

12,302

13,151

14,032

14,958

15,930

(B)

1,043

1,099

1,157

1,217

1,278

1,340

1,404

1,468

1,533

1,600

(C)

1,043

1,100

1,160

1,222

1,287

1,354

1,423

1,494

1,566

1,640

(D)

1,096

1,156

1,217

1,280

1,345

1,410

1,478

1,546

1,615

1,686

(E)

1,097

1,159

1,223

1,290

1,359

1,431

1,505

1,580

1,657

1,736

Asset differential

(F)

1.0

2.9

5.9

9.8

14.6

20.4

27.0

34.2

41.8

49.6

Asset differential withdrawn

(G)

0.04

0.13

0.26

0.43

0.65

0.90

1.19

1.51

1.85

2.19

Asset differential carry-over

(H)

0.9

2.8

5.6

9.3

14.0

19.5

25.8

32.6

39.9

47.4

Discounted asset differential withdrawn

(I)

0.04

0.11

0.21

0.33

0.47

0.63

0.78

0.94

1.09

1.23

Discounted asset final asset differential

(J)

---

---

---

---

---

---

---

---

---

26.7

Total discounted (April 2016 dollars) frontload-mutual-funds-gains-to-investors

32.5

* Projection year 2017 begins in April 2017 and ends in April 2018. Other projection years similarly begin in April of the named year and end in April of the following year.

The alternative scenarios diverge from the baseline scenario over the course of
projection year 2017 due to improved investment performance.634 Beginning-of-year front-endload mutual fund assets are in Rows (B) and (C) of Figure B-3 for the baseline and alternative 1
scenarios, respectively. Investment performance estimates from Columns (J) and (K) of Figure
B-1 are then applied to generate end-of-year front-end-load mutual fund asset estimates of
$1.096 trillion under the baseline scenario and $1.097 trillion under alternative scenario 1 (Rows
D and E of Figure B-3). The “Asset Differential” (Row F) is simply the difference between
these two amounts – $1.0 billion in 2017. A small fraction, 4.42 percent of this differential is
withdrawn, while most of the differential is carried over to 2018. 635
Each projection year starts with a fresh baseline projection of total IRA assets (Row A).
Baseline IRA front-end-load mutual fund assets are calculated each year using the estimate that
43.9 percent of IRA assets are in mutual funds. The share of IRA mutual funds incurring a frontend-load is assumed to decline over the projection period from 27 percent in 2017 to 23 percent
in 2026 and 19 percent in 2036. Unlike 2017, in 2018 and in subsequent years, the alternative
scenarios begin the year with higher IRA front-end-load mutual fund assets than the baseline
scenario. The differential that is carried over from the previous year is added to the baseline
IRA front-end-load mutual fund assets to estimate the alternative scenario IRA front-end-load
mutual fund assets. At the end of 2017, the asset differential carry-over was $0.9 billion (Row
H), so the 2018 alternative scenario IRA front-end-load mutual fund asset amount is $1.100

634

635

Projection year 2017 begins in April 2017 and ends in April 2018. Other projection years similar begin in April of the named year and end
in April of the following year.
For simplicity, the calculations assume that all contributions and withdrawals occur at the very end of each calendar year.

344

trillion (Row C), $0.9 billion more than the baseline amount of $1.099 trillion (Row B). The
alternative scenarios diverge further from the baseline scenario in 2018 and in later years when
end-of-year assets are again calculated. The carry-over of the asset differential from year to year
ensures that the calculations take into account the compound nature of improved returns over
multiple years. Each year a small fraction of the asset differential is withdrawn, while the
majority of the asset differential is carried over to the following year.
The asset differential at the end of the 10-year period (2025, Row H) together with the
portion of the asset differential withdrawn in each year (Row G) makes up the 10-year quantified
subset of IRA investors’ expected gains under alternative scenarios 1. However, before those
numbers are summed, they are each discounted by the appropriate number of years at a rate of
5.4 percent (Rows I and J) so that the 10-year front-end-load-mutual-fund-gain-to-investors is
expressed in April 2016 dollars.
The 10- and 20-year quantified subset of IRA investors’ expected gains under each of the
alternative scenarios are estimated using a method identical to that presented in Figure B-3.

B.3

Assumptions and Uncertainty

Uncertainty is inherent in any forward-looking projection, and the Department’s 10- and
20-year estimates of the quantified subset of IRA investors’ expected gains are no exception.
Every assumption that is required to generate the estimates adds another layer of uncertainty to
the projections. The Department has investigated all of the assumptions used herein. This
section discusses each of the assumptions, the assignments that the Department has chosen for
each assumption, alternative assignments, and related evidence. Figure B-4 presents a summary
of the assumptions and assignments.
Figure B-4 Assignments used in the Department's 10- and 20-year Front-load-mutual-fundgains-to-investors Estimates
Assumption

Assignment
44.9 basis point decline in
performance per 100 basis points in
load paid to broker

Effect of loads on returns
2014 average loads
- front-end-load
- front load paid to broker
- broker share of load

169 basis points
137 basis points
81.4%

Annual rate of decline in average load

3.2%

Annual decline in average load - alternative scenario 2

6.4%
Figure B-2

Purchase date distribution of load assets

16.8%

Average annual inflow subject to load

Figure B-3, Row (A)

Total IRA assets
% IRA assets in equity, bond, and hybrid mutual funds

43.9%

% 2014 IRA mutual fund assets incurring front-end-load

29.2%

Annual decline in % assets incurring front load

2.0%

Discount rate

5.4%

Average investment return excluding load effects

6.0%

Fraction of year-end assets withdrawn

4.4%

345

B.3.1

Effect of Loads on Returns

CEM investigate whether mutual funds that pay a higher load-share to brokers perform
better or worse than those that pay a lower load-share. Using U.S. mutual fund data from 19932009, the authors compare the average load-share that a mutual fund pays to brokers to the future
performance to the same mutual fund. In a series of regressions, the authors also control for a
number of other variables that are predictive of performance, including inflows, redemptions,
fund size, and fund family size.636 The fourth regression treats unaffiliated and captive brokers
separately. This regression may be the most appropriate to use because it allows for the data to
demonstrate the effects of different incentives across the two main types of broker arrangements.
CEM’s excess-load-paid-to-broker variable can be thought of as a proxy for the severity of
conflicts of interest. The results from the fourth regression strongly suggest that the severity of
the conflict of interest differs depending on whether the broker to whom the load-share is paid is
captive or unaffiliated. In other words, $X of load-share paid to an unaffiliated broker creates a
different severity of conflict of interest from $X of load-share paid to a captive broker.
Furthermore, the fourth regression allows the projections to take account of the different
size of load-shares paid to captive and unaffiliated brokers. Load-shares paid to unaffiliated
brokers are larger on average (see CEM Table 1) and the effect of a dollar of load-share paid to
unaffiliated brokers on performance is larger.
In the fourth regression specification, the data reveal that investment returns in the year
following an inflow decrease by 49.7 basis points for every 100 basis points of load-share for
mutual funds that pay load-shares exclusively to unaffiliated brokers and by 14.5 basis points for
every 100 basis points of load-share for mutual funds that pay load-shares exclusively to captive
brokers.637 The fourth regression does not include mutual funds that pay load-shares to both
unaffiliated and captive brokers. These mutual funds that pay load-shares to both unaffiliated
and captive brokers constitute about 23 percent of the mutual funds included in CEM Table V.
Mutual funds paying load-shares exclusively to unaffiliated brokers constitute about 83
percent of the set of mutual funds that pay load-shares exclusively to either unaffiliated or
captive brokers. Mutual funds that pay load-shares exclusively to captive brokers constitute the
remaining 17 percent. The average front-end-load paid to unaffiliated brokers in the CEM data
(2.30 percent) is higher than the average front-end-load paid to captive brokers in the CEM data
(1.73 percent). Weighting by both of these factors generates an average loss in performance of
44.9 basis points for every 100 basis points in load-share paid to brokers across all payments
recorded in the data. The Department relies on these estimates to forecast future returns. In
particular, the baseline scenario projection assumes that, in the future, investment returns will
suffer by 44.9 basis points for every 100 basis points paid in load-share. There is uncertainty
surrounding this projection for four reasons:

636

637

As the authors used data at the mutual fund level, they were unable to explicitly control for consumer sophistication by adding a variable to
the regressions; however, the entire analysis was conducted within the broker-sold segment of the market, thereby controlling for
differences in consumer sophistication across the broker-sold and direct-sold segments.
CEM, Table V, Column 4.

346

(1) The CEM regression results estimate the impact of load-sharing on investment returns
during the 12 months following the inflow, while the Department’s related
assumption applies to investment returns for the life of the investment.
(2) The CEM results are only a single data point. The results may be specific to the
period of the data, 1993-2009, and may not hold in the future.
(3) The Department’s methodology assumes that the best estimate of the effect of loadsharing on performance for the 23% of mutual funds that pay load-shares to both
captive and unaffiliated brokers is a weighted average of the effect of load-sharing on
performance for mutual funds that pay load-shares exclusively to captive brokers and
the effect of load-sharing on performance for mutual funds that pay load-shares
exclusively to unaffiliated brokers. Whether this assumption could be improved upon
is unknown.
(4) Three other regression specifications, none of which distinguish between captive and
unaffiliated broker arrangements, estimate a decrease in performance of 33-35 basis
points for every 100 basis points of load-share paid to the broker. Were the
Department to assume a reduction of 34.0 basis points rather than 44.9 basis points,
the estimated quantified subset of IRA investors’ expected gains would decrease by
16 percent to 24 percent.
Each of these concerns suggest that the Department should be cautious in applying the
assumption that investment returns will decrease by about 44.9 basis points for every 100 basis
points paid in load-share. The Department has conducted further analysis of the related literature
in order to test the assumption. Two themes emerged from this research:
(1) The finance literature supports the hypothesis that investment returns suffer following
load-sharing, in part, because the mutual fund lacks an incentive to invest in
performance. This incentive is lacking, not only in the first year, but in all of the
subsequent years that the IRA investor remains in the mutual fund. The literature
suggests that the CEM results should hold for the life of the fund, not just the first
year following an inflow.
(2) Across a broad array of studies, broker-sold mutual funds underperform direct-sold
mutual funds. Both the direction and the magnitude of the results are consistent with
the CEM results, suggesting that the CEM results are not an outlier and are not
dependent on the particular data or methodologies used.
In addition, the Department recently received supplemental data in a letter from ICI. The
letter indicates that ICI replicated the analysis contained in Table V of CEM using data from
2010 to 2013.638 ICI finds the effect of load-shares on performance to be even more severe than
indicated in CEM for both load-shares paid to unaffiliated brokers and load-shares paid to all
brokers:
The coefficient estimates and fit of the first-stage model, though based on data for the
years 2010 to 2013, are in all their important aspects similar to those reported in CEM. Like

638

See ICI Comment Letter (July 21, 2015); available at: http://www.dol.gov/ebsa/pdf/1210-AB32-2-00749.pdf

347

CEM, we use this first-stage regression to create an “residual load fee” variable, which is the
residual from the first-stage regression. In the second stage, we regress the excess return of each
front-load fund — measured as that fund’s return in the coming 12 months relative to the return
on that fund’s Morningstar category in the coming 12 months—against the fund’s residual load
fee and an array of other independent variable similar to those used in CEM. The results in the
second-stage regression are also in all their important elements very similar to those reported in
CEM. For example, we find a coefficient estimate on the residual load fee paid to brokers (both
affiliated and unaffiliated) of –0.48 percent, a bit larger in absolute size than their reported
estimate of –0.34 percent (see Table V on page 226 of the CEM study). We find a coefficient
estimate on the residual load fee paid to unaffiliated brokers of –0.64 percent, which implies an
even larger effect than the –0.4972 coefficient reported in CEM.”
These ICI results affirm the Department’s choice for its assumption regarding the effect
of load-sharing on performance.
In a comment on the 2015 NPRM Regulatory Impact Analysis, ICI suggests that it may
be inappropriate to use any estimated relationship between load-shares and performance derived
from the CEM results because the CEM regressions are not asset-weighted.639 However, two of
the authors of the CEM paper also provided a comment demonstrating that the variation in fund
size is handled appropriately:640
“All the regressions in the paper use robust standard errors which control for the
heteroscedasticity in variance often associated with funds of different size. This implies that for
each observation the variance estimate is allowed to vary in proportion to the independent
predictors of the regression (which include log asset size) so effectively the variances in the
regression are asset-weighted. We also cluster standard errors by each fund so standard errors
are allowed to vary fund-by-fund. Lastly, log of asset size is included as a control variable in all
the regressions. Thus, the variation in fund assets is addressed thoroughly in our analysis.”
Further discussion of points raised by ICI appears in Chapter 3.641

B.3.1.1 Broker-Sold Underperformance as a Result of
Mutual Fund Incentives
Del Guercio and Reuter (2014) (DGR) provide a helpful framework for interpreting the
CEM results. The authors hypothesize that broker-sold, actively-managed funds underperform
direct-sold, actively-managed funds because the direct-sold funds invest more in performance.
In fact, the authors find that, in the direct-sold segment, active funds achieve high enough returns
to make up for their higher costs. However, broker-sold actively-managed funds underperform
direct-sold actively managed funds by over 100 basis points per year.
DGR provide several pieces of evidence suggesting a lack of investment in alpha
(superior performance above and beyond that of the market) contributes to broker-sold mutual
fund underperformance. Small cap stocks, by their nature, require more resources to investigate
than large cap stocks, per investment dollar. If broker-sold mutual funds fail to invest in alpha,

639
640
641

See ICI’s comment letter (July 21, 2015); available at: http://www.dol.gov/ebsa/pdf/1210-AB32-2-00749.pdf.
See Christoffersen and Evans’ comment letter, (Sept. 10, 2015); available at: http://www.dol.gov/ebsa/pdf/1210-AB32-2-02766.pdf.
For additional discussion of ICI’s critique of the Department’s use of CEM results, see Padmanabhan, Panis and Tardiff (2016).

348

broker-sold fund underperformance will show up to a greater degree in the small-cap space,
where more resources are required. The data reveal that the underperformance of broker-sold
funds is more dramatic when focusing on small-cap growth funds (Table III, Panel B). The
authors find direct “evidence that direct-sold funds are more actively managed than broker-sold
funds,” by showing that actively-managed broker-sold funds track closer to indices than directsold funds. Broker-sold funds are also more likely to expose investors to higher systemic risk
(beta), more likely to outsource portfolio management and less likely to hire asset managers with
superior educational backgrounds. All of these pieces of evidence suggest that broker-sold funds
invest less in alpha than direct-sold funds, but why?
All mutual funds want to attract inflows because revenues rise with assets under
management. DGR find that inflows…
“[I]n the direct-sold segment are significantly more sensitive to risk-adjusted returns than
fund flows in the broker-sold segment. Specifically, while the estimated coefficients on lagged
alpha are positive in both segments, the estimated coefficient for the direct-sold segment is
larger (0.176 versus 0.021), significantly different from zero (p-value of 0.000), and significantly
different from the coefficient for the broker-sold segment (p-value of 0.001). These coefficients
imply that a one standard deviation increase in alpha will increase fund size over the next 12
months by approximately 6.18% in the direct-sold and 0.59% in the broker-sold segments, or in
dollar terms, by $86.9 million and $5.0 million, respectively. Since the typical actively managed
fund’s management fee is approximately 75 basis points, this implies incremental annual
revenue to the fund of $651,660 for the average direct-sold fund and only $37,445 for the
average broker-sold fund. Thus, if families in the direct-sold segment could invest in the
managers, analysts, or trading infrastructure that would generate this increase in alpha at lower
annual cost than $651,660, they would presumably do so, whereas families in the broker-sold
segment have a much weaker incentive to make alpha-generating investments.”
Conversely, broker-sold mutual fund inflows respond more to raw (non-risk-adjusted)
returns which can be increased without investing in performance – by taking on more systemic
risk.
Recognizing the broker-sold mutual funds’ lack of incentive to invest in alpha is helpful
in interpreting the CEM results. CEM find that investment returns in the year following an
inflow decrease by 49.7 basis points for every 100 basis points of load-share for mutual funds
that pay load-shares exclusively to unaffiliated brokers. Importantly, this result arises in spite of
the regulations in place at the time the study was conducted to protect investors from conflicts of
interest. The DGR results suggest that a lack of incentive to invest in alpha contributes to the
underperformance. The lack of incentive for broker-sold mutual funds to invest in alpha persists
for the life of the fund, well beyond the first 12 months following the inflow; funds that pay high
load-shares to brokers need not invest in performance to attract new inflows. Therefore, the
underperformance in the first 12 months following the inflow (as estimated by CEM) can be
expected to continue for as long as the IRA client holds the fund.
The DGR results and the subsequent interpretation of the CEM results are consistent with
the theoretical financial economics literature on mutual funds. Berk and Green (2004) present a
compelling model of mutual fund performance where fund managers exhibit high average skill,
yet competitive allocation of investor assets reduces investor returns in actively managed funds
to the level of returns in index funds. The construction and results of this model are consistent
with the direct-sold segment of the market. Similarly, models that account for the conflicts of
interest inherent in the broker-sold segment of the market produce theoretical predictions that are
349

consistent with the underperformance of the broker-sold segment observed by DGR (Inderst and
Ottaviani 2009; 2012).

B.3.1.2 Robust Evidence of Broker-Sold Mutual Fund
Underperformance
A deep pool of academic studies confirms the robustness of the result that broker-sold
mutual funds underperform direct-sold mutual funds. Both the direction and the magnitude of
the results in these studies are consistent with the CEM results, providing assurance that the
CEM results accurately reflect the condition of the market. The literature comparing the
performance of broker-sold and direct-sold mutual funds is discussed in Section 3.2.4 and
summarized in Figure 3-17. The CEM results imply that front-end-load mutual funds in the
broker-sold segment underperform no-load funds by approximately 1 percent per year, a
magnitude typical of the results in Figure 3-17.642

B.3.2

Load and Performance Projections

The baseline load projections, in Column (B) of Figure B-1, begin with an assumption
regarding average loads in the year 2014 – the most recent year for which strong mutual fund
load data can be obtained. The estimate for average front-end-loads paid by IRA investors in
2014 is generated by taking the average front-end-load paid by investors in the CEM sample
(1993-2009) and scaling it down based on the rate of decrease in loads in the data presented in
the ICI Fact Book 2015.643

B.3.2.1 Average Load Paid by Investors in 2014
The average front-end-load paid by investors in the CEM sample is relatively
straightforward. CEM Table 1 presents the average load paid by investors who receive advice
from captive brokers (2.40 percent) and unaffiliated brokers (2.77 percent). The Department
calculated average for the entire sample (2.71 percent) is simply the weighted average of these
two numbers, weighted by the number of observations for captive (25,807) and unaffiliated
(123,824) brokers.
While the CEM sample includes funds from all years between 1993 and 2009, loads in
the population have been shown to be decreasing over time. The ICI Fact Book 2015 lists
average front-end-sales-loads actually paid for Equity, Hybrid, and Bond mutual funds for the
years 1990, 1995, 2000, 2005, 2010, and 2014 (Figure 5.8, page 105). Meanwhile, the ICI U.S.
Retirement Markets quarterly Excel spreadsheets list the amount of IRA assets in Domestic
Equity, World Equity, Hybrid, and Bond mutual funds each year going back to 1990 (Table
16).644 By imputing the front-end-loads paid for all sample years, one can estimate the assetweighted average front-end-load paid by IRA investors during the 1993-2009 period. This
comes out to138.8 basis points. Similarly, one can estimate the asset-weighted front-end-load
paid in 2014 – 86.4 basis points – by the same method. The ratio of these two numbers –

642

643
644

A sample average load-share of 220, multiplied by a 44.9 basis point decrease in performance for every 100 basis point increase in loads
(see top of Section B.3.1 above) results in average underperformance of 99 basis points (220 * 0.449 = 99).
ICI “2015 Investment Company Fact Book.”
ICI, “The U.S. Retirement Market, Third Quarter, 2015,” 2015.

350

86.4/138.8 = 62.3 percent represents the estimated decline in loads from the sample period,
1993-2009, to 2014. Applying this ratio to the average load paid by investors in the CEM
sample generates an estimated 2014 average load paid by IRA investors of 169 basis points (271
basis points * 62.3 percent = 169 basis points).
This raises two questions: Why is the asset-weighted average load paid, calculated using
the ICI data (138.8 basis points), different from the average load paid by investors in the CEM
sample (271 basis points)? Similarly, why not simply use the asset-weighted average front-endload for 2014 (86.4 basis points, calculated from the ICI data) in estimating the quantified subset
of IRA investors’ expected gains, rather than using the ICI data to scale down the CEM average
load? It is not entirely clear why the CEM and ICI average loads differ to the extent that they do,
but a portion of the difference may be attributable to the manner in which the averages are
calculated. CEM averages appear to be equal-weighted averages, while ICI calculates their
reported numbers on an asset-weighted basis.645 This means that investors with large accounts –
and small loads (load charges typically decrease as investment size increases) – have greater
influence in the ICI averages. For example, consider two investors. One investor invests
$400,000 in a mutual fund and is charged a 1 percent sales load. The other (IRA) investor
invests $150,000 in the same mutual fund and is charged a sales load of 2.5 percent. The assetweighted average sales load actually paid is approximately 1.4 percent (1 percent /
(400,000/550,000) + 2.5 percent / (150,000/400,000) = 1.4 percent), whereas the equal-weighted
average sales load would be 1.75 percent ((1 percent + 2.5 percent) / 2 = 1.75 percent).
Under most circumstances, the asset-weighted average load would be the appropriate
average to use in calculating an aggregate gains estimate; however, in this case it is problematic.
The Department’s estimates concern only loads paid by IRA investors for IRA investments.
IRA accounts tend to have much smaller account balances, on average, than non-IRA accounts,
including non-IRA mutual fund accounts. According to tabulations from the Survey of
Consumer Finances, the average IRA account balance is approximately $149,000, compared to
an average non-IRA mutual fund balance of $387,000 and an average non-IRA taxable
investment account balance of $412,000 (see Panis and Brien (2016) presenting 2013 household
survey data on the IRA marketplace). Super-large accounts are generally limited to non-IRAs as
Internal Revenue Service regulations place limits on both IRA and defined contribution
retirement plan contributions (a primary source of IRA assets through rollovers). Investors with
very large accounts can pay very low front-end-loads to invest in mutual funds. These investors,
which are exclusively non-IRA, can dramatically skew the average load when it is calculated on
an asset-weighted basis.
In addition, the ICI front-end-load averages appear to include some institutional funds.
Page 104 of the Fact Book describes how the decline in loads partly “reflects the increasing role
of mutual funds in helping investors save for retirement. Funds that normally charge front-end
load fees often waive load fees on purchases made through defined contribution (DC) plans,
such as 401(k) plans.” So the averages include large discounts enjoyed, not only by large
taxable accounts, but also by large institutional investors.

645

DOL staff conversation with ICI analysts, Aug. 22, 2014.

351

The above discussion is not to say that the CEM sample average load is necessarily an
accurate estimate of the asset-weighted average load paid by IRA mutual fund investors; there is
uncertainty here. However, the CEM sample average load seems to be a more plausible estimate
of the asset-weighted average load paid by IRA mutual fund investors than the ICI weighted
average load. The ICI numbers clearly underestimate the asset-weighted average load paid by
IRA investors, given the size of IRA accounts relative to taxable accounts. While the CEM
sample average load may also include discounted loads paid by taxable account holders and
institutional investors, the averages are calculated on an equal-weighted basis, and therefore are
not skewed by these discounted, non-IRA loads. Using the less-relevant average 2014 frontend-load of 86.4 basis points (the ICI asset-weighted average) instead of the more-applicable
169 basis points (the CEM sample average scaled to 2014) would cause the estimated quantified
subset of IRA investors’ expected gains to decrease by 48 percent to 49 percent.

B.3.2.2 Average Load Paid to Brokers in 2014
Similar to the average front-end-load paid by investors, the average front-end-load paid
to brokers in the CEM sample (load-share, 2.20 percent) is the weighted average of the average
front-end-load paid to captive brokers (1.73 percent) and the average front-end-load paid to
unaffiliated brokers (2.30 percent) weighted by the number of observations for each.
The ratio of the broker load-share to the total front-end-load paid (2.20/2.71 = 81.4
percent) is assumed to remain constant throughout the projection period. Therefore, the average
load paid to brokers in 2014 is assigned a value of 137 basis points (169 * 81.4 percent = 137).
This assumption may work to understate the estimated gains of the rule to investors. As
previously discussed in Section 3.2.4, CEM finds that higher loads tend to decrease inflows
while higher load-shares tend to increase inflows. These two forces acting in tandem would tend
to increase the broker share of the load. It is quite possible – and in fact suggested by the
evidence – that the broker share of the load is lower in the early part of the CEM sample and
higher in the later part of the sample.646 It is further probable that the broker share of the load
has continued to increase since 2009 and will continue to increase throughout the projection
period in the absence of regulatory intervention. Because the underperformance of load funds is
tied to the load paid to the broker, a higher broker share of the load would increase the
underperformance in the baseline scenario, and, in turn, increase the estimated quantified subset
of IRA investors’ expected gains.
However, the extent to which the gains to investors are underestimated as a result of this
assumption is limited. The broker share of the load has a theoretical maximum of 100 percent.
If the broker share of load assumption were changed to 100 percent for the entirety of the
projection period, estimated quantified subset of IRA investors’ expected gains would increase
by approximately 14 percent to 23 percent.

646

Unfortunately, there are no reported results in the paper that identify whether this is true.

352

B.3.2.3 Decline in Loads
As mentioned above, data from ICI clearly show that front-end-loads have declined over
the historical period 1990-2014.647 The Department estimates that these loads have decreased at
a rate of approximately 3.2 percent per year between 2000 and 2014. The estimated subset of
IRA investors’ expected gains assume that, under the baseline scenario, average loads will
continue to decrease at a rate of 3.2 percent per year. The baseline scenario load projection
(Figure B-1, Column B), begins with the assigned 2014 average load paid by investors (Section
B-3.2.1 above), and projects future average loads by decreasing that value by 3.2 percent per
year. In 2017, the start of the 10- and 20-year projection periods, the average load paid by IRA
investors is projected to be 153 basis points, under the baseline scenario. In 2026 and 2036, the
ends of the 10- and 20-year projection periods, baseline average loads paid by IRA investors are
projected to be 114 and 82 basis points, respectively.
Alternative scenario 2 projects an increase in the rate at which average loads decline.
The rule may put downward pressure on the size of loads that IRA investors pay to the extent
that paying high fees is not in the best interest of the IRA holder. It is unknown how quickly
average IRA loads may decline as a result of the rule. Alternative scenario 2 uses the
assumption that average loads for IRA holders decline at twice the rate that they otherwise
would have declined under the baseline scenario.
Alternative scenario 3 presents the extreme where all loads paid by IRA holders go to
zero immediately. While this is not an expected outcome, the scenario quantifies the potential
for gains if loads paid by IRA holders fall even more quickly than assumed in alternative
scenario 2.
While average loads have clearly trended downward historically – and that trend is
reflected in the baseline scenario – there may be reason to believe that, in the absence of
regulation, this trend will slow down or even stop in the near future. Front-end-loads are
ostensibly for the purpose of compensating brokers for the services that they provide.648 If level
of service is in fact the primary determinant of loads, a significant decline in loads (as projected
under the baseline scenario) would require a decrease in either the level of service provided per
customer or the cost of providing those services. As such, some may not find the baseline
projected decline in the average IRA load, from 169 basis points in 2014 to 82 basis points in
2036, to be realistic. Cutting the assigned baseline rate of decline in the average load in half (1.6
percent) increases the estimated subset of IRA investors’ expected gains by 11 percent to 19
percent. Eliminating the baseline scenario decline in loads (assuming average IRA loads of 169
basis points throughout the projection period), increases the estimated subset of IRA investors’
expected gains by 22 percent to 42 percent. In both cases, there is a larger impact on the 20-year
estimated subset of IRA investors’ expected gains (relative to the 10-year estimates).

B.3.2.4 Average Annual Inflow Subject to Load
Columns (E) and (F) of Figure B-1 estimate the average direct effect of loads on
performance for IRA investors, under the baseline and alternative 2 scenarios, respectively.

647
648

ICI 2015 Fact Book, Figure 5.8, 1059.
See, for example, ICI Fact Book 2015, 104.

353

Loads decrease performance because the load is taken out of the IRA investor’s pool of money
available for investment. The aggregate direct effect of loads on performance depends heavily
on the frequency with which front-end-load assets are turned over. Front-end-loads are, of
course, paid only when front-end-load mutual funds are purchased. The more IRA investors buy
and sell front-end-load mutual funds, the more load charges they will incur.
In the CEM sample, Inflows Subject to Load averaged 1.4 percent of total net assets
(TNA) each month (Table I), or about 16.8 percent per year. Assuming a net flow of zero, this
inflow rate corresponds to front-end-load assets being bought and sold approximately once every
6 years on average. The Department has not located any alternative data sources to estimate
load fund turnover.649 The projections assume that 16.8 percent of front-end-load assets are
turned over each year throughout the projection period, under all scenarios. Columns (E) and
(F) of Figure B-1 are the product of the turnover assumption (16.8 percent) and the average load
paid under the given scenario, Columns (B) and (C), respectively.
The assumption regarding the average annual inflow subject to load is closely tied to the
assumed distribution of purchase dates for assets which incurred a front-end-load. Sensitivity
analyses for these assumptions are presented at the end of the following Section B.3.2.5.

B.3.2.5 Distribution of Purchase Dates for Assets which
Incurred a Front-End-Load
The average effect of current and past loads on performance (Columns H and I of Figure
B-1) depends on when the front-end-load mutual funds were purchased. In the absence of a rule
(both under the baseline scenario and under the alternative scenarios for years prior to the
finalization of the rule), loads decrease future performance by 44.9 basis points for every 100
basis points of load-share paid to the broker (see Section B.3.1 above). However, loads are
projected to decline over the projection period which implies a corresponding decline in the
effect on performance. With the rule, the load effect on performance goes to zero because the
conflicts of interest associated with the load are mitigated. All of these factors necessitate an
assumption for the distribution of purchase dates for front-end-load assets that are owned in a
given year. The assumption impacts the estimated subset of IRA investors’ expected gains by
determining how quickly assets are moved into better performing mutual funds once the
requirements of the rule become applicable.
The construction of Figure B-2 begins with the assumption that 16.8 percent of IRA
front-end-load mutual fund assets are turned over each year. Therefore, 16.8 percent of frontend-load mutual fund assets owned in a given year, t, were purchased in that same year. It is
unclear what percentage of assets owned in a given year, t, would have been purchased in the
previous year, t-1. If assets purchased in year t-1 were equally likely to be sold again in year t,
then one could assume that 14.0 percent of assets owned in year t were purchased in year t-1. In
this case, 16.8 percent of assets were purchased in year t-1, but then 16.8 percent of those were
re-purchased in year t, so 16.8 percent – (16.8 percent * 16.8 percent) = 14.0 percent of assets
year t assets were last purchased in year t-1. However, it seems unlikely that load-incurring
assets purchased in year t-1 would be equally likely to be sold again in year t.

649

ICI indicated they would send inflow data underlying Figure 5.10 (page 98) of ICI Fact Book 2014, but it has not yet been received.

354

Why would one want to incur a load two years in a row? If no assets purchased in year t-1 were
turned over in year t, the percentage of year t assets originally purchased in year t-1 would be
16.8 percent. This extreme again seems unlikely. The projections split the difference and
assume that 15.4 percent of IRA front-end-load mutual fund assets owned in a given year, t, were
purchased in the preceding year, t-1. The rest of Figure B-2 is constructed to ensure that the
percentage of assets originating in a given year and prior years adds up to 100 percent and
decreases in a somewhat smooth manner as one gets further and further removed from the given
year, t.
The projections are modestly sensitive to the assumed average annual inflow subject to
load and the assumed distribution
Figure B-5 Alternative Assignments for the Distribution of
of purchase dates for assets that
Purchase Dates for Front-end-load Assets Owned in a Given
incurred a front-end loach.
Year
Decreasing the assumed load-asset
turnover rate to 12.6 percent and
(1)
(2)
using the first set of alternative
% of year t front-end-load % of year t front-end-load
assignments for the distribution of
assets purchased in each
assets purchased in each
purchase dates in Figure B-5
Year year
year
(middle column), decreases the
t-9
7.6%
0.0%
estimated subset of IRA investors’
t-8
8.1%
1.4%
expected gains by 4 percent to 14
t-7
8.6%
3.8%
percent. Conversely, increasing the
t-6
9.1%
6.2%
assumed load-asset turnover rate to
t-5
9.6%
8.6%
21.0 percent and using the second
t-4
10.2%
11.2%
set of alternative assignments for
t-3
the distribution of purchase dates in
10.8%
13.6%
Figure B-5 (rightmost column),
t-2
11.4%
16.0%
increases the estimated subset of
t-1
12.0%
18.2%
IRA investors’ expected gains by 2
t
12.6%
21.0%
percent to 7 percent. The impact of Total
100.0%
100.0%
these assumptions on the 10-year
estimated subset of IRA investors’ expected gains is larger than the impact on the 20-year
estimates.

B.3.2.6 Average Investment Returns Excluding Load
Effects
Columns (E) through (I) of Figure B-1 estimate the impacts of loads both directly and
through load-sharing, which can remove the mutual fund’s incentive to invest in performance.
In order to generate estimated subset of IRA investors’ expected gains from these impacts, an
investment return assumption is required. The projections assume that, before applying any load
effects, nominal investment returns are equal to 6 percent for all scenarios throughout the
projection period.650 This assumption has a small impact on estimated subset of IRA investors’

650

The 2015 NPRM Regulatory Impact Analysis also included an assumption of 6 percent for non-conflicted, nominal investment returns. The
Department did not receive any comments objecting to this assumption.

355

expected gains. Decreasing the investment returns assumption to 4 percent decreases the
estimated subset of IRA investors’ expected gains by 6 percent to 13 percent. Conversely,
increasing the investment returns assumption to 8 percent increases the by 6 percent to 16
percent. The impact of the investment returns assumption is larger with respect to the 20-year
estimated subset of IRA investors’ expected gains relative to the 10-year estimated subset of
IRA investors’ expected gains.

B.3.3

Front-End-Load Mutual Fund Assets

Until this point in Section B.3, all of the assumptions discussed have contributed to
projecting the performance of assets under a baseline and several alternative scenarios. The
aggregate estimated subset of IRA investors’ expected
gains also depends on the amount of front-end-load
Figure B-6 Projected Total IRA
Assets at Beginning of Year
mutual fund assets.

Cerulli Associates projects IRA year-begin
asset levels for 2017-2021 of $8.7 trillion, $9.3
trillion, $10.0 trillion, $10.7 trillion, and $11.5 trillion,
respectively.651 The Department projects IRA assets
beyond year-begin 2021 by continuing Cerulli’s
projected growth trend. Figure B-6 displays the
projected asset levels and growth rates.

651

Growth Rate

8,652
9,308
10,000
10,729

7.6%
7.4%
7.3%
7.2%

11,497

7.0%

12,302

6.9%

13,151
14,032
14,958
15,930
16,934
17,984
19,063
20,187
21,358
22,554
23,795
25,056
26,359
27,703

6.7%
6.6%
6.5%
6.3%
6.2%
6.0%
5.9%
5.8%
5.6%
5.5%
5.3%
5.2%
5.1%
4.9%

DOL

B.3.3.1 Total IRA Assets

Year
2017
2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
2028
2029
2030
2031
2032
2033
2034
2035
2036

Assets
($ billions)

Cerulli

The dollar amount of front-end-load mutual
fund assets depends primarily on three factors: total
IRA assets, the percentage of IRA assets invested in
mutual funds, and the percentage of IRA mutual fund
assets incurring a front-end load. The second factor,
the percentage of IRA assets invested in mutual funds,
appears to be relatively stable over the last 15 years, so
a single assumption (43.9 percent) is used for all years
in the projection period. Trends are observed in the
other two factors. Total IRA assets appear to be
increasing while the percentage of IRA mutual fund
assets incurring a front-end-load is likely decreasing.
For these factors, assumptions are generated for each
year of the 10- and 20-year projections periods.

Cerulli Associates, “Retirement Market 2015.” Asset amounts presented in the source document are year-end, whereas amounts presented
here are beginning of year assets.

356

B.3.3.2 Percentage of IRA Assets in Mutual Funds
While the Department is concerned about the impact of conflicts on all IRA assets, the
subset of IRA investors’ expected gains projections focus on IRA assets in domestic equity,
world equity, hybrid, and bond mutual funds. The projections take a narrower focus because this
subset of IRA assets may be the best space to generate a reliable, quantitative estimate of a
portion of the rule’s gains to investors. CEM find that load sharing is associated with decreased
future performance in equity, international, fixed income, balanced, and municipal mutual funds.
Money-market mutual funds do not appear to be part of the CEM sample (Table II). In order to
appropriately apply the performance projections derived from the CEM estimates, moneymarket mutual fund and non-mutual fund IRA assets must be excluded. This exclusion should
not imply that conflicts are not a problem in these areas nor that the rule will not produce
substantial gains to individuals who hold these excluded assets in IRA accounts. It simply
means that a quantitative estimate of the gains to investors in these areas is not available.
The ICI’s U.S. Retirement Market quarterly spreadsheets list asset amounts for Total
IRA Assets (Table 7) and domestic equity, world equity, hybrid, and bond mutual funds in IRAs
(Table 16).652 Since 1999, the share of total IRA assets invested in mutual funds (excluding
money-market funds) has fluctuated between 37 percent and 48 percent, with an average of 43.3
percent (DOL calculations). The fluctuation in the mutual fund share of IRA assets seems to
correlate with movement in equity markets, but lacks a trend upward or down.
Because there is no discernable recent trend in mutual fund share of IRA assets, the
estimates utilize the most recent data point (43.9 percent in third quarter 2015) and project the
mutual fund share of IRA assets to remain constant throughout the projection period. Using the
long-term average of 43.3 percent for the share of IRA assets in mutual funds decreases the
estimated subset of IRA investors’ expected gains by less than 2 percent.

B.3.3.3 Percentage of IRA Mutual Fund Assets Incurring
a Front-End-Load
Ideally, the projections would utilize a data source that looks within IRA accounts and
measures the percentage of IRA mutual fund assets incurring a front-end-load. This factor,
combined with the two listed above, would generate an estimate for the dollar amount of IRA
mutual fund assets incurring a front-end-load in a given year. In the absence of this ideal data,
another data source must be used to anchor the assumed percentage of IRA mutual fund assets
incurring a front-end-load over the course of the projection period.

652

ICI, “The U.S. Retirement Market, Third Quarter, 2015,” 2015.

357

The ICI’s 2014 and 2015 Fact Books collectively present the total net assets of mutual
funds by share class from 2004 to 2014 (Figure 5.11, page 99 of the 2014 Fact Book and Figure
5.10, page 107 of the 2015 Fact Book). In 2014, front-end-load assets were 29.2 percent of all
non-annuity mutual fund assets excluding Institutional no-load funds (DOL calculation).
Institutional no-load assets are
excluded from the denominator
Figure B-7 Projected Percentage of IRA Mutual Fund Assets
of the calculation because IRA
Incurring a Front-end-load by Year
investors do not have access to
institutional share classes.
Percentage of IRA mutual

IRA mutual fund assets

The data show a decline
fund assets incurring a
incurring a front-end-load as a
in the percentage of mutual
Year front-end-load
percentage of all IRA assets
fund assets incurring a front2017
27.4%
12.0%
end-load. In 2004, the earliest
2018
26.9%
11.8%
year in which data are available 2019
26.3%
11.6%
from these sources, front-end2020
25.8%
11.3%
load assets totaled 35.7 percent
2021
25.3%
11.1%
of all non-annuity mutual fund
2022
24.8%
10.9%
assets excluding institutional
2023
24.3%
10.7%
no-load funds. The 2014 share
2024
23.8%
10.5%
of 29.2 percent signals a
2025
23.3%
10.3%
decline of approximately 2.0
2026
percent per year. The 2014
22.9%
10.0%
2027
percentage of IRA mutual fund
22.4%
9.8%
assets incurring a front-end2028
22.0%
9.6%
load is assigned a value of 29.2
2029
21.5%
9.5%
percent and that percentage is
2030
21.1%
9.3%
projected to decline by 2.0
2031
20.7%
9.1%
percent per year through the 10- 2032
20.3%
8.9%
and 20-year projection periods.
2033
19.9%
8.7%
Figure B-7 presents the
2034
19.5%
8.5%
assigned percentage of IRA
2035
19.1%
8.4%
mutual fund assets incurring a
2036
18.7%
8.2%
front-end-load in each year of
the projection period (middle column). The historical data used to produce these assumptions
are not specific to IRA mutual funds. Instead, an implicit assumption is made; the percentage of
IRA mutual fund assets incurring a front-end-load is assumed to be comparable to the percentage
of all retail mutual fund assets incurring a front-end-load. If IRA investors are especially
unsophisticated, is it likely that the percentage of IRA mutual fund assets incurring a front-endload is higher than the percentage across all retail mutual fund assets; however, the Department
has not found any data to verify this hypothesis.
There also remains uncertainty in the projections surrounding the rate of decline in the
percentage of mutual fund assets incurring a front-end-load. It is unclear what factors are
driving the decline in front-end-loads. If investor demand is driving the decline in front-endloads, the movement away from front-end-loads could be limited to the more sophisticated,
highly financially literate consumer base. In this case, the decline in front-end-loads might
quickly slow or stop once that segment of investors has moved on. On the other hand, if
advisers are driving the movement away from front-end-loads, the decline could continue, or
even accelerate. In fact, the rate of decline in front-end-load-share of mutual fund assets has
358

been even larger – 4.2 percent – in more recent years (2009-2014). While maintaining the
baseline assumption that the 2014 front-end-load percentage of IRA mutual fund assets is 29.2
percent, eliminating the projected decline in this percentage increases the estimated subset of
IRA investors’ expected gains by 18 percent to 30 percent. Conversely, increasing the projected
decline in front-end-load percentage of mutual fund assets to 4.0 percent decreases the estimated
subset of IRA investors’ expected gains by 15 percent to 23 percent.
The rightmost column of Figure B-7 displays the projected IRA mutual fund assets
incurring a front-end-load as a percentage of all IRA assets. This column is the product of the
center column and the assumed mutual fund share of IRA assets (43.9 percent, see Section
B.3.3.2 above). The baseline scenario IRA front-end-load mutual fund assets (Row B of Figure
B-3) are calculated by multiplying the baseline total IRA assets (Row A of Figure B-3) by the
projected IRA mutual fund assets incurring a front-end-load (rightmost column of Figure B-7).

B.3.4

Aggregation of Yearly Gains to Investors

The aggregation of yearly front-end-load-mutual-fund benefits depends primarily on two
variables: the discount rate, and the fraction of assets withdrawn from IRA accounts each year.

B.3.4.1 Discount Rate
The estimated subset of IRA investors’ expected gains are weighted more heavily toward
the ends of the 10- and 20-year projection periods because the effects of the rule will take time to
filter through IRA front-end-load-mutual-fund assets. When the requirements of the rule
become applicable brokers who previously advised IRA accounts will generally not be required
to review past advice. IRA assets in underperforming funds will only be affected when the IRA
investor receives new advice. Therefore, the estimated subset of IRA investors’ expected gains
will grow as more and more time passes following the finalization of the rule.
At least one comment on the 2015 NPRM Regulatory Impact Analysis claimed to notice
a mismatch between the benefit and cost estimates presented in the Regulatory Impact Analysis.
That comment stated that the cost estimates were discounted using both a 3 percent and 7
percent real discount rate while the benefit estimates were discounted using only a 3 percent real
discount rate.653 In fact, the 2015 NPRM Regulatory Impact Analysis did present partial gainsto-investor estimates discounted using a 7 percent real discount rate. Readers interested in
gains-to-investor estimates aggregated using the 7 percent real discount rate can find those
estimates in this section.
Because the estimated subset of IRA investors’ expected gains tends to be back-loaded,
the assigned discount rate has a significant effect on the estimates. Circular A-4 states that real
discount rates of 3 percent and 7 percent should be used. 654 Inflation (change in CPI-U)
averages 2.3 percent for calendar years 2019-2026 in the assumptions underlying the

653
654

See FSI comment, (July 21, 2015); available at: http://www.dol.gov/ebsa/pdf/1210-AB32-2-00724.pdf.
Circular A-4, “Regulatory Impact Analysis: A Primer” (Sept. 2003); available at:
https://www.whitehouse.gov/sites/default/files/omb/inforeg/regpol/circular-a-4_regulatory-impact-analysis-a-primer.pdf.

359

Administration’s FY2017 budget. 655 Combining the inflation projection and the 3 percent real
discount rate implies a nominal projected discount rate of 5.369 percent throughout the
projection period (1.03 * 1.023 = 1.05369). Increasing the discount rate to 9.461 percent
(reflecting a 7 percent real discount rate, 1.07 * 1.023 = 1.09461) throughout the projection
period decreases the estimated subset of IRA investors’ expected gains by 33 percent to 51
percent. Using the 9.461 percent nominal discount rate an estimated subset of IRA investors’
expected gains is $22 billion under Scenario 1. As expected, the impact of the discount rate
assumption on the 20-year estimated subset of IRA investors’ expected gains is larger than the
impact on the 10-year estimated subset of IRA investors’ expected gains.

B.3.4.2 Fraction of Year-End Assets Withdrawn
Compound interest is a well-established financial principle and adds to the estimated
subset of IRA investors’ expected gains of a rule that helps IRA investors over time. Gains that
accrue in one year can be carried over to the next and accrue additional benefits through
reinvestment. The estimated subset of IRA investors’ expected gains reflect this compounding
effect. Most of the asset differential (difference between end-of-year assets under the baseline
and alternative scenarios) is carried over to the following year. However, a portion of IRA
assets are also withdrawn each year.
The ICI U.S. Retirement Market quarterly spreadsheets present the withdrawals from
IRAs between 2000 and 2012.656 Withdrawals average 4.42 percent of total IRA assets each
year. The subset of IRA investors’ expected gains projections assume that withdrawals will
equal 4.42 percent of the asset differential in each year of the projection period. This projection
also relies on the additional assumption that IRA mutual fund assets incurring a front-end-load
are withdrawn at the same rate as the overall population of IRA assets.
The withdrawal rate assumption has relatively little impact on estimated subset of IRA
investors’ expected gains. If the assigned withdrawal rate were increased to 7 percent for the
entirety of the projection period (the highest withdrawal rate in the 2000-2012 historical period
was 6.5 percent in 2008), the estimated subset of IRA investors’ expected gains would be
reduced by less than 1 percent.

B.3.5

Effects of Proposed Regulatory Action on Market Trends

The Department has estimated the subset of IRA investors’ expected gains at $33 billion
to $36 billion over 10 years and $66 billion to $76 billion over 20 years relative to a baseline
where the size of front-end-loads and the proportion of assets subjects to front-end-loads are
projected to substantially decrease.657 Removing both of these projected declines – assuming
load sizes and assets subject to loads remain at their 2014 levels – increases the projected subset
of IRA investors’ expected gains to $47 billion to $52 billion over 10 years and $122 billion to

655

656

657

See Table 2-1, “Analytical Perspectives: Budget of the U.S. Government,” (2016); available at:
https://www.whitehouse.gov/sites/default/files/omb/budget/fy2017/assets/spec.pdf
See ICI, “The U.S. Retirement Market, Third Quarter, 2015,” 2015. According to this spreadsheet, withdrawal rates for 2013 and 2014 are
unavailable.
All estimated subset of IRA investors’ expected gains in this section refer to alternative scenarios 1 and 2 only.

360

$141 billion over 20 years. Which is the most appropriate baseline – one where loads maintain
current levels or one where they decrease at a rate similar to the recent past?
If markets exist in a vacuum, decreasing loads would clearly be the most applicable
baseline; however, regulatory action, and even the expectation of regulatory action, can have
significant impacts on markets. The Department has been working on the project that has
culminated in the re-proposal of this rule since 2008. The public has been aware of the project
since work began, and the project has received widespread public and industry attention since
the original proposal in October 2010. These dates line up quite well with the accelerated drop
in mutual fund assets incurring a front-end-load, though not as well with the decrease in the size
of front-end-loads.
To the extent that the Department’s work on this project has generated downward
pressure on the size and frequency of front-end-loads, the failure of the Department to finalize a
rule could have the expected effect of a rebound in those trends. It is unclear to what extent the
recent decline in loads (sizes and rates) can be attributed to action on the part of the Department,
but evidence on the impact of expected regulatory action suggests that the appropriate baseline
scenario may be one where the decline in loads decelerates or even disappears.

B.4

Estimates of the Harm Due to Conflicted Advice

Two sections of this Regulatory Impact Analysis present projections that were produced
using a similar methodology to that used to estimate the quantified subset of IRA investors’
expected gains. (Section 3.2.4 considers the gap in performance between broker-sold and directsold IRA assets as well as the underperformance which results from conflicts of interest. Section
4.2.2.2 contemplates the performance difference between accounts in employer-based retirement
plans and IRAs that are subject to conflicted advice.) In all cases, the performance measures are
presented as market aggregate dollar values over 1, 10, and 20 years. This section details the
assumptions required to generate the projections in these sections and the values assigned to
those assumptions.
Section 3.2.4 presents several estimates for the amount of loads and underperformance
that could result from conflicts of interest. Those estimates are summarized in Figure B-8. The
assumptions used to generate all of the estimates in Figure B-8 differ from the above estimated
subset of IRA investors’ expected gains assumptions in the following ways:
1. The percentage of IRA mutual fund assets incurring a front-end-load does

not decline over the projection period.
2. The size of front-end-loads does not decline over the projection period.
3. There is no phase in effect. The calculations count all underperformance

due to conflicts of interest, including conflicted advice that occurs both
before and after the requirements of the rule become applicable.
Additional variation in the assumptions used is detailed in Figure B-8.

361

Figure B-8 Underperformance and Load Projections
Projected Loss from
Underperformance and Loads
($ billion)
Row
(1)
(2)
(3)
(4)

1-year

10-year

20-year

6.1
9.0
16.9
24.7

67.3
99.3
184.3
273.4

142.4
211.8
387.8
583.2

Assignments
Percentage of IRA
assets in affected
asset categories

43.9%
43.9%
43.9%
43.9%

Percentage of affected asset
category IRA assets sold
through conflicted channels

Average size of
front-end-load
(2014 or historical)

29.2%
29.2%
50.0%
50.0%

169
169
271
271

Row (1) of Figure B-8 reflects an estimate of the underperformance directly attributable
to loads as applied to the current IRA front-end-load mutual funds market. This projection uses
the 2014 estimates for percentage of IRA assets in mutual funds (43.9 percent), percentage of
IRA mutual fund assets incurring a front-end-load (29.2 percent) and average size of front-endload (169 basis points). Row (2) adds in the direct cost of loads, analogous to Alternative
Scenario 3 from the estimated subset of IRA investors’ expected gains.
Row (3) acknowledges the fact that front-end-loads have decreased over time and
contemplates the possibility that, as front-end-loads have disappeared, harms from conflicts of
interest have been shifted to other revenue streams rather than eliminated. In order to estimate
harms in this situation, Row (3) projects the underperformance that would occur if all brokersold mutual funds (50 percent658) incurred a front-end-load and the average size of front-endloads was that of the CEM sample period (270.61 basis points). Row (4) adds in the direct cost
of loads that would occur, given the assumptions.
Section 3.2.4 also presents estimates of the underperformance of broker-sold mutual
funds relative to those on the direct-sold side of the market. These projections use some of the
same methodology as the front-end-load-mutual-fund-gain-to-investors estimates. However,
rather than projecting loads across the projections periods, these estimates simply take the
assumed performance gap and apply it to the projected IRA assets, similar to Figure B-3.
Figure B-9 presents the assigned values for underperformance and projections over 1, 10,
and 20 years. Row (1) assigns 50 basis points to the underperformance of broker-sold funds and
Row (2) assigns 100 basis points. All rows assume that 43.9 percent of IRA assets are in nonmoney-market mutual funds and that 50 percent of mutual funds are broker-sold throughout the
projection period.

658

Bergstresser et al. (2009) categorizes approximately 50 percent of retail mutual funds as broker-sold.

362

Figure B-9 Broker-sold Mutual Fund Underperformance
Projected Underperformance ($ billion)
Row
(1)
(2)

Assignment

1-year

10-year

20-year

Underperformance of
broker-sold funds

8.6
17.1

94.7
189.3

201.9
403.9

50 basis points
100 basis points

In addition to the market impact of conflicted-advice-related underperformance, Section
3.2.4 presents estimates for the effect of 50, 100, or 200 basis point underperformance on an
individual investor’s retirement savings. The section states that an ERISA plan investor who
rolls her retirement savings into an IRA could lose 6 to 23 percent of the value of her savings
over 30 years of retirement by accepting advice from a conflicted financial adviser. The
estimates rely on three assumptions: a nominal return on investment rate of 6 percent; an
inflation rate of 2.3 percent; and that the retiree will consume all of her savings in exactly 30
years. The resulting percentage-loss estimates do not depend on the amount of savings rolled
into an IRA, but, for purposes of illustration, consider an investor who rolls over $200,000. For
each of the 30 years, the investor experiences a real return rate of 3.62 percent (1.06/1.023 =
1.0362). After experiencing those returns, the investor withdraws $X where X is determined
such that the retiree consumes all of her savings in exactly 30 years. In the case of the retiree
who experiences no underperformance, she consumes $11,034 in real dollars each year and has a
balance of $0 at the end of the 30 year period.
However, if the retiree accepts investment advice that results in underperformance of 0.5
percent per year, the consumption possibilities are reduced. For each of the 30 years, the
investor now experiences a real return rate of 3.12 percent (3.62 percent minus 0.5 percent). The
retiree is able to consume only $10,359 per year while depleting her savings in exactly 30 years.
The reduction in consumption due to the underperformance is 6 percent: ($11,034 - $10,359) /
$11,034 = 6 percent.
If the investment advice results in underperformance of 1 percent per year, the
consumption potential is more severely diminished. The retiree now is able to consume only
$9,705 per year for each of the 30 years. The reduction in consumption due to 1 percent
underperformance is 12 percent: ($11,034 - $9,705) / $11,034 = 12 percent.
Finally, if the investment advice results in underperformance of 2 percent per year, the
consumption potential is further diminished. The retiree now is able to consume only $8,466 per
year for each of the 30 years. The reduction in consumption due to 2 percent underperformance
is 23 percent: ($11,034 - $8,466) / $11,034 = 23 percent.
Section 4.2.2.2 presents estimates for the market impact of the underperformance of IRA
rollover assets, in cases where advice is given. The projections assume that 50 percent of all
IRA rollovers come from employer sponsored retirement plans and involve conflicted advice
from a broker or other individual. The IRA rollover underperformance estimates are
methodologically similar to the broker-sold mutual fund underperformance estimates presented
in Figure B-9. The primary difference is the asset base.

363

Figure B-10 presents projections for
IRA assets over the 1-, 10-, and 20-year
projection periods. IRA rollover and IRA
rollover growth rate projections for 20172020 come from The Cerulli Associates’
“Retirement Markets 2015” report.659 For the
remainder of the 10- and 20-year projection
period, IRA rollovers are projected by
extending the trend in the IRA rollover
growth rate.

Figure B-10 Projected IRA Rollovers

Year
2017
2018
2019
2020

IRA Rollover
Growth
Rate
1.047
1.046
1.045
1.045

453
473
495
517

Cumulative
Rollover Assets
($ billion)
453
926
1,421
1,938

2021
2022

1.044
1.043

540
563

2,477
3,040

2023

1.042

586

3,626

IRA Rollovers
($ billion)

Figure B-11 presents the IRA
2024
1.042
611
4,238
rollover broker-sold mutual fund
2025
1.041
636
4,874
underperformance projections. These
2026
1.040
662
5,535
projections only consider rollovers that occur
2027
1.039
687
6,223
within the projection period. As a result, the
2028
1.039
714
6,937
10-year underperformance is much more
2029
1.038
741
7,678
than 10 times the size of the 1-year
2030
1.037
769
8,447
underperformance, and the 20-year
2031
1.036
796
9,244
underperformance is significantly more than
2032
1.036
825
10,069
twice the size of the 10-year
2033
1.035
854
10,923
underperformance. Each year additional
2034
1.034
883
11,806
assets are rolled over into broker-sold mutual 2035
1.033
912
12,718
funds adding to the cumulative rollover total
2036
1.033
942
13,661
and rollovers that occurred previously during
the projection period experience an additional year of underperformance. Row (1) of Figure B11 assigns 50 basis points to the underperformance of broker-sold funds relative to the
performance of employer-sponsored retirement plan assets and Row (2) assigns 100 basis points
to the underperformance.
Figure B-11 IRA Rollover Broker-sold Mutual Fund Underperformance

Row
(1)
(2)

659

Projected
Underperformance
($ billion)
11020year year
year
1.0 47.8 152.5
2.0 95.6 304.9

Assignment
Underperformance of IRA rollover broker-sold
funds
50 basis points
100 basis points

Cerulli Associates, “Retirement Markets 2015.”

364

365

Appendix C: Small Saver Market
The Department believes that “small savers” (that is, those individuals or households
with low account balances and/or of modest means) are most negatively impacted by the
detrimental effects of conflicted advice. With fewer economic resources, small savers are
particularly vulnerable to any practices that diminish their resources by extracting unnecessary
fees or by yielding lower returns. These savers cannot afford to lose any of their retirement
savings.
Comments concerning small savers appear to frequently misconstrue the nature of the
market. First, when establishing who small savers are, the comments often conflate all savers
with small IRA account balances with all savers of modest means. While there is often overlap
between the two groups, there are also very important differences between them, and failing to
recognize these differences can result in making predictions based on faulty assumptions.
For example, “small savers” are sometimes assumed to be those who hold small
retirement savings accounts, such as total IRA balances of less than $25,000 or some other
threshold. But defining small savers solely in this way can be misleading. Many small IRA
accounts are held by individuals or households with high incomes and/or wealth. In fact, nearly
70 percent of all households headed by individuals under age 65 with IRA savings less than
$25,000 are in the top half of the U.S. income distribution (Figure C-1). An overwhelming
majority of households owning such small IRAs also own their homes (75 percent), and many
own one or more financial assets outside of IRAs, including job-based defined contribution (DC)
accounts (59 percent), stocks (21 percent), savings bonds (18 percent), mutual funds (12
percent), and CDs (7 percent). That these households hold other investable assets likely makes
them attractive as prospective advisory clients.

Altogether, 11 percent of all households headed by individuals under age 65 own IRAs
valued at less than $25,000. The share owning small IRAs increases with both income and net
worth, and wealthier households are far more likely to own small IRAs (Figure C-2). Just 3
percent of households in the bottom income quartile and 5 percent of households in the bottom
net worth quartile own such small IRAs.
366

A substantial number of people with modest means do save for retirement, but these
savings are held mostly in job-based plans, not IRAs. Job-based plans are significantly more
common than IRAs among all but the most well-off households (Figure C-3). Looking
specifically at the bottom one-fourth of households by income and by net worth, while the

majority of both groups do not have either a job-based plan or an IRA, 11 percent and 20 percent
respectively have job-based plans, compared to just 4 percent and 5 percent respectively that
own IRAs (Figure C-4). Very small percentages have both.

367

A similar picture emerges if we consider the aggregate amount that households have
saved for retirement: Job-based DC plan savings outweigh IRA savings on aggregate among
those with low or moderate net worth (Figure C-5). This is the case even though low-paying
jobs often do not come with access to job-based retirement benefits, and many families of
modest means consequently do not have the opportunity to save in such plans. When looking

just at families of modest means who are offered job-based DC plans, it can be seen that DC
take-up rates far surpass IRA ownership rates (Figure C-6).

368

While these statistics illustrate that, for most moderate income families, job-based DC
plans play a far larger role in retirement savings than IRAs, they actually understate role of jobbased plans. Most of the assets held by IRAs originated as rollovers from job-based plans. In
2012, inflows to traditional IRAs from rollovers totaled an estimated $301 billion, while direct
contributions totaled only an estimated $14.1 billion. According to one survey,660 of all new
traditional IRAs in 2013, two-thirds were funded solely by rollovers, and another 22 percent
were funded by transfers from an IRA that was held at a different financial firm. Just 11 percent
were funded in whole (8.6 percent) or part (2.5 percent) by direct contributions. Among
investors who held traditional IRAs in 2013, 45 percent had made a rollover between 2007 and
2013. The industry today competes fiercely to capture these rollovers, but, as described in
Section 4.2.2.2, investors are particularly vulnerable to adviser conflicts of interest during the
rollover process. This means that many small savers whose primary engagement with the IRA
market comes when they roll their job-based DC account balances into IRAs could see their
retirement security significantly undermined if rollover advice is not made to comply with a
fiduciary standard.
Some comments on the 2015 Proposal also stated that advisers play an important role in
promoting saving among small savers and contended that the Proposal might undermine this
promotion of savings. While Section 8.4.4 addresses these claims specifically, it is also
important to note that not all advice to save/invest more is good advice and that conflicts of

660

ICI, “The IRA Investor Profile: Traditional IRA Investors’ Activity, 2007-2013,” July 2015.

369

interest can sometimes result in such advice undermining small savers’ overall financial security.
Commissions and revenue sharing might encourage brokers to sell savings products (such as
mutual funds) to families of modest means, and this might encourage them to save more.661
However, this might not always be in the family’s best interest. For example, some of these
families hold expensive debt. Ten percent of households under age 65 that fall in the bottom
net-worth quartile face debt payments in excess of 40 percent of their income, 16 percent are
more than 60 days behind in debt payments, and 35 percent maintain a credit card balance
(Figure C-7). For households saddled with expensive debt, buying mutual funds instead of
retiring debt is likely to reduce net worth and financial security, not increase it. Brokers’
compensation arrangements, however, typically reward them for recommending high-fee funds
over potentially better-performing, lower-fee funds, and for recommending that families invest
as much as possible, even if paying down debt would be a far better choice. Their financial
interests therefore often conflict acutely with that of the small savers they might advise.

Finally, some comments on the 2015 Proposal also appeared to exaggerate the extent to
which advisers, especially brokers, currently advise small IRA investors and thereby increase
their savings. Most small IRA investors today are served not by brokers, but by banks and other
service providers. In fact, small savers turn most often to friends and family for investment
advice. By increasing trust in professional advisers, small savers may actually seek out more
professional advice.

661

Although, as discussed above, the evidence that advisers have a positive impact on savings is extremely limited.

370

The Department notes that small savers appear to hold their IRAs at banks more often
than with brokers, and rarely receive financial advice from brokers. Among households under
age 65 that own IRAs worth less than $25,000, just 30 percent of those in the bottom income
quartile, and one third of those in the bottom net worth quartile, held their IRAs at brokerages
(Figure C-8). Likewise, very few households with modest means or small IRAs receive advice
from brokers, while much larger proportions seek advice on-line or from bankers. This is based
on data from a household survey asking respondent which sources of information they used to
make decisions about savings and investments. Therefore the advice that households reported to
obtain from banks may cover a wide range of topics such as college savings or emergency funds,
not specifically about retirement savings.
Among households headed by individuals under age 65 who are in the bottom net worth
quartile, just 4 percent report turning to brokers for financial advice, whereas 33 percent turn to
online sources, 30 percent to bankers, 16 percent to financial planners, 8 percent to magazines,
books, and/or newspapers, and 7 percent to TV and/or radio (Figure C-9). More report relying
on accountants or even lawyers than report consulting brokers. Just 6 percent of households
headed by Hispanics and 5 percent of those headed by African Americans report getting
financial advice from brokers, as do just 10 percent of those headed by Caucasians. Even among
households holding IRAs in brokerage accounts, less than one-fourth report receiving financial
advice from brokers, while 52 percent get advice online, 44 percent from financial planners, and

% of All HHs Age <65 with IRA < $25K

Figure C-8
More Small-Saver IRAs at Banks than at Brokerages
70%

Commercial Banks

60%

Brokerages

Other

50%
40%
30%
20%
10%
0%

Income %ile

Net worth %ile

22 percent from magazines, books and/or newspapers. As many rely on bankers as on brokers
for financial advice.

371

372

373

Appendix D: Partial Gains to Investors and Compliance Costs
Accounting Table
Figure D-1 Partial Gains to Investors and Compliance Costs Accounting Table
Category

Primary
Estimate

Low
High
Year
Estimate
Estimate
Dollar
Partial Gains to Investors
(Includes Benefits and Transfers)

Discount
Rate

Period
Covered

Annualized

$3,420

$3,105

2016

7%

April 2017April 2027

Monetized
($millions/year)

$4,203

$3,814

2016

3%

April 2017April 2027

Gains to Investors Notes: The Department expects the final rulemaking to deliver large gains for retirement
investors. Because of limitations of the literature and other available evidence, only some of these gains can be
quantified: up to $3.1 or $3.4 billion (annualized over April 2017 – April 2027 with a 7 percent discount rate) or up
to $3.8 or $4.2 billion (annualized over April 2017- April 2027 with a 3 percent discount rate). These estimates
focus only on how load-shares paid to brokers affect the size of loads IRA investors holding load funds pay and the
returns they achieve. These estimates assume that the rule will eliminate (rather than just reduce)
underperformance associated with the practice of incentivizing broker recommendations through variable frontend-load sharing. If, however, the rule’s effectiveness in reducing underperformance is substantially below 100
percent, these estimates may overstate these particular gains to investors in the front-end-load mutual fund
segment of the IRA market. However, these estimates account for only a fraction of potential conflicts, associated
losses, and affected retirement assets. The total gains to IRA investors attributable to the rule may be higher than
the quantified gains alone for several reasons. For example, the proposal is expected to yield additional gains for
IRA investors, including potential reductions in excessive trading and associated transaction costs and timing
errors (such as might be associated with return chasing), improvements in the performance of IRA investments
other than front-load mutual funds, and improvements in the performance of ERISA plan investments.
The partial-gains-to-investors estimates include both economic efficiency benefits and transfers from the financial
services industry to IRA holders.
The partial gains estimates are discounted to April, 2016.

Compliance Costs
Annualized

$1,960

$1,205

$3,847

2016

7%

April 2017
–April 2027

Monetized
($millions/year)

$1,893

$1,172

$3,692

2016

3%

April 2017April 2027

Notes: The compliance costs of the final include the cost to BDs, Registered Investment Advisers, insurers, and
other ERISA plan service providers for compliance reviews, comprehensive compliance and supervisory system
changes, policies and procedures and training programs updates, insurance increases, disclosure preparation and
distribution to comply with exemptions, and some costs of changes in other business practices. Compliance costs
incurred by mutual funds or other asset providers have not been estimated.

374

Insurance Premium Transfers
Annualized
Monetized
($millions/year)
From/To

$73

2016

7%

$73

2016

3%

From: Insured service providers without
claims

375

April 2017
–April 2027
April 2017April 2027

To: Insured service providers with claims
– funded from a portion of the increased
insurance premiums

Bibliographyi
Agarwal, Sumit, John C. Driscoll, Xavier Gabhaix, and David Laibson. "The Age of Reason:
Financial Decisions over the Life Cycle and Implications for Regulation." Brookings
Papers on Economic Activity, 2009.
Agnew, Julie, Hazel Bateman, Christine Eckert, Fedor Iskhakov, Jordan Louviere, and Susan
Thorp. "Individual Judgment and Trust Formation: An Experimental Investigation of
Online Financial Advice." Australian School of Business Research Paper No.
2013ACTL21, 2014.
Anagol, Santosh, Shawn Cole, and Shayak Sarkar. Understanding the Advice of CommissionsMotivated Agents: Evidence from the Indian Life Insurance Market". Harvard Business
School Working Paper, 2013.
Anguelov, Chris E., Howard M. Iams, and Patrick J. Purcell. "Shifting Income Sources of the
Aged." Social Security Bulletin 72, no. 3 (2012).
Bakos, Yannis, Henry Lucas, Wonseok Oh, Sivakumar Viswanathan, Gary Simon, and Bruce
Weber. Electronic Commerce in the Retail Brokerage Industry: Trading Costs of Internet
Versus Full Service Firms. Center for Digital Economy Research, Stern School of
Business, New York University, Working Paper IS-99-014, 1999.
Barber, Brad, Terrance Odean, and Lu Zheng. "Out of Sight, Out of Mind: The Effects of
Expenses on Mutual Fund Flows." Journal of Business 78, no. 6 (2005): 2095-2119.
Beh, Hazel, and Amanda M. Willis. "Insurance Intermediaries." Connecticut Insurance Law
Journal 15, no. 2 (2009): 571-98.
Ben-Shahar, Omri, and Carl E. Schneider. "The Failure of Mandated Disclosure." University of
Pennsylvania Law Review 159 (2011): 647-749.
Bergstresser, Daniel, John Chalmers, and Peter Tufano. "Assessing the Costs and Benefits of
Brokers in the Mutual Fund Industry." Review of Financial Studies 22, no. 10 (2009):
4129-4156.
Berk, Jonathan, and Richard Green. "Mutual Fund Flows and Performance in Rational Markets."
Journal of Political Economy 112, no. 6 (2004): 1269-1295.
Bluethgen, Ralph, Steffen Meyer, and Andreas Hackethal. High-Quality Financial Advice
Wanted! European Business School, Working Paper, 2008.
Bolton, Patrick, Xavier Freixas, and Joel Shapiro. "Conflicts of Interest, Information Provision,
and Competition in the Financial Services Industry." Journal of Financial Economics 85
(2007): 297-330.
Brien, Michael, and Constantijn Panis. Classifying Portfolio Volatility. Report Prepared for the
U.S. Department of Labor, Deloitte & Advanced Analytical Consulting Group, 2013.
Bubb, Ryan. "TMI? Why the Optimal Architecture of Disclosure Remains TBD." Michigan Law
Reivew 113, no. 6 (2015): 1021-1042.
Bullard, Mercer, Geoffrey C Friesen, and Travis Sapp. "Investor Timing and Fund Distribution
Channels." Social Science Research Network. 2008.
Burke, Jeremy, and Angela Hung. Do Financial Savers Influence Savings Behavior? RAND
Corporation, 2015.
376

Burke, Jeremy, and Angela Hung. Trust and Financial Advice. RAND Corporation, 2015.
Burke, Jeremy, Angela Hung, Jack Clift, Steve Garber, and Joanne Yoong. Impact of Conflicts of
Interest in the Financial Services Industry. RAND Corporation, 2015.
Cain, Daylian M., George Loewenstein, and Don A. Moore. "The Dirt on Coming Clean:
Perverse Effects of Disclosing Conflicts of Interest." Journal of Legal Studies 34 (2005):
1-25.
Chalmers, John, and Jonathan Reuter. What is the Impact of Financial Advisors on Retirement
Portfolio Choices and Outcomes? University of Oregon and Boston College, 2014.
Chang, Tom, David H. Solomon, and Mark M. Westerfield. "Looking for Someone to Blame:
Delegation, Cognitive Dissonance, and the Disposition Effect." Journal of Finance, 2015.
Charles River Associates. "Polarisation: Research into the Effect of Commission Based
Remuneration on Advice." Financial Services Authority. London, 2002.
Chater, Nick, Steffen Huck, and Roman Inderst. Consumer Decision-Making in Retail
Investment Services: A Behavioral Economics Perspective. Decision Technology Ltd,
2010.
Chen, Xuanjuan, Tong Yao, and Tong Yu. "Prudent Man or Agency Problem? On the
Performance of Insurance Mutual Funds." Journal of Financial Intermediation 16, no. 2
(2007): 175-203.
Choi, James, David Laibson, and Brigitte Madrian. "Why Does the Law of One Price Fail? An
Experiment on Index Mutual Funds." Review of Financial Studies 23, no. 4 (2010): 14051432.
Christoffersen, Susan, Richard Evans, and David Musto. "What Do Consumers' Fund Flows
Maximize? Evidence from Their Broker's Incentives." Journal of Finance 68 (2013):
201-235.
Cici, Gjergi, and Michael Boldin. "The Index Fund Rationality Paradox." Journal of Banking
and Finance 34 (2010): 33-43.
Cooper, Robert, and Garry Frank. "The Highly Trouble Ethical Environmnet of the Life
Insurance Industry: Has it Changed Significantly from the Last Decade and if so, Why?"
Journal of Business Ethics, 2005: 149-157.
Del Guercio, Diane, and Jonathan Reuter. "Mutual Fund Performance and the Incentive to
Generate Alpha." Journal of Finance 69, no. 4 (2014): 1673-1704.
Del Guercio, Diane, Jonathan Reuter, and Paula Tkac. Broker Incentives and Mutual Fund
Market Segmentation. NBER Working Paper 16312, 2010.
Della Seta, Marco, and Sebastian. Gryglewicz. Asset Sales in Good and Bad Times. Social
Science Research Network, 2015.
Edelen, Roger, Richard Evans, and Gregory Kadlec. "Disclosure and Agency Conflict: Evidence
from Mutual Fund Commission Bundling." Journal of Financial Economics 103 (2012):
308-326.
Egan, Mark. Brokers vs. Retail Investors: Conflicting Interests and Dominated Products.
University of Chicago, 2014.

377

Elton, Edwin, Martin Gruber, and Christopher Blake. "How do Employer's 401(K) Mutual Fund
Selections Affect Performance?" Center for Retirement Research at Boston College,
2013.
Elton, Edwin, Martin Gruber, and Christopher Blake. "The Adequacy of Investment Choices
Offered by 401(K) Plans." Journal of Public Economics 90, no. 6 (2006): 1299-1314.
Evans, Richard, and Rudiger Fahlenbrach. "Institutional Investors and Mutual Fund Governance:
Evidence from Retail - Institutional Fund Twins." The Review of Financial Studies 25,
no. 12 (2012): 3530-3571.
Fama, Eugene, and Kenneth French. "Luck versus Skill in the Cross-section of Mutual Fund
Returns." Journal of Finance 65, no. 5 (2010): 1915-1947.
Ferrell, Allen. The Law and Finance of Broker-Dealer Mark-Ups. The Harvard John M. Olin
Discussion Paper Series, 2011.
Finke, Michael, John Howe, and Sandra Huston. Old Age and Decline in Financial Literacy.
Social Science Research Network Working Paper, 2011.
Finke, Michael, and Thomas Langdon, "The Impact of the Broker-Dealer Fiduciary Standard
on Financial Advice." Social Science Research Network Working Paper, 2012.
Fischel, Daniel, and Todd Kendall. Comment to the Department of Labor on a Proposed Rule
Regarding Fiduciary Status Under ERISA. Compass Lexecon, 2011.
Fitzpatrick, Sean. "The Small Laws: Eliot Spitzer and the Way to Insurance Market Reform."
Fordham Law Review 74, no. 6 (2006): 3041-3071.
Foerster, Stephen, Juhani Linnainmaa, Brian Melzer, and Alessandro Previtero. The Costs and
Benefits of Financial Advice. Western University, University of Chicago and
Northwestern University, 2014.
French, Kenneth. "The Cost of Active Investing." Journal of Finance 63, no. 4 (2008): 15371573.
Friesen, Geoffrey, and Travis Sapp. "Mutual Fund Flows and Investor Returns: An Empirical
Examination of Fund Investor Timing Ability." Journal of Banking and Finance 31, no. 9
(2007): 2796-2816.
Gallaher, Steven, Ron Kaniel, and Laura Starks. "Madison Avenue Meets Wall Street: Mutual
Fund Families, Competition and Advertising." University of Texas and Duke University,
2006.
Gamble, Keith, Patricia Boyle, Lei Yu, and David Bennett. "Aging and Financial Decision
Making." Management Science 61, no. 11 (2015): 2603-2610.
Garber, Steven, Jeremy Burke, Angela Hung, and Eric Talley. Potential Economic Effects on
Individual Retirement Account Markets and Investors of DOL's Proposed Rule
Concerning the Definition of a 'Fiduciary". RAND Corporation, 2015.
Gil-Bazo, Javier, and Pablo Ruiz-Verdu. "The Relation between Price and Performance in the
Mutual Fund Industry." Journal of Finance 64, no. 5 (2009): 2153-2183.
Gray, Welsey. Are You Trying Too Hard? Alpha Architect, 2014.
Gruber, Martin. "Another Puzzle: The Growth in Actively Managed Mutual Funds." Journal of
Finance 51, no. 3 (1996): 783-810.
378

Hackethal, Andreas, Michael Haliassos, and Tullio Jappelli. "Financial Advisors: A Case of
Babysitters?" Journal of Banking & Finance 36, no. 2 (2012): 509-524.
Harrington, Winston, Richard Morgenstern, and Peter Nelson. "On the Accuracy of Regulatory
Cost Estimates." Journal of Policy Analysis and Management 19, no. 2 (2000): 297-322.
Haziza, Mor, and Anver Kalay. "Broker Rebates and Investor Sophistication." Social Science
Research Network Working Paper, 2014.
Houge, Todd, and Jay Wellman. "The Use and Abuse of Mutual Fund Expenses." Journal of
Business Ethics 70, no. 1 (2007): 23-32.
Howat, John, and Linda Reid. "Compensation Practices for Retail Sale of Mutual Funds: the
Need for Transparency and Disclosure." Fordham Journal of Corporate & Financial
Law 12, no. 4 (2007): 685-716.
Hung, Angela, Noreen Clancy, Jeff Dominitz, Eric Talley, Claude Berrebi, and Farrukh.
Suvankulov. Investor and Industry Perspectives on Investment Advisers and BrokerDealers. RAND Corporation, 2008.
Inderst, Roman, and Marco Ottaviani. "Financial Advice." Journal of Economic Literature,
2012: 494-512.
Inderst, Roman, and Marco Ottaviani. "Misselling through Agents." American Economic Review
99, no. 3 (2009): 883-908.
Jain, Prem, and Joanna Wu. "Truth in Mutual Fund Advertising: Evidence on Future
Performance and Fund Flows." Journal of Finance 55, no. 2 (2000): 937-958.
Klein, Benjamin, and Keith Leffler. "The Role of Market Forces in Assuring Contractual
Performance." Journal of Political Economy 89, no. 4 (1981): 615-641.
Kozora, Matthew L. The Effect of Regulatory Regimes on the Provision of Retail Investment
Advice. U.S. Securities and Exchange Commission, 2013.
Krausz, Miriam, and Jacob Paroush. "Financial Advising in the Presence of Conflict of
Interests." Journal of Economics and Business 54 (2002): 55-71.
Laby, Arthur. "Selling Advice and Creating Expectations: Why Brokers Should be Fiduciaries."
Washington Law Review, 2012: 707-776.
Lazaro, Christine, and Benjamin P. Edwards. "The Fragmented Regulation of Investment
Advice: A Call for Harmonization." Michigan Business and Entrepreneurial Law Review
4, no. 1 (2015): 101-148.
Loewenstein, George, Daylian Cain, and Sunita Sah. "The Limits of Transparency: Pitfalls and
Potential of Disclosing Conflicts of Interest." American Economic Review: Papers and
Proceedings 101, no. 3 (2011): 423-428.
Lusardi, Annamaria, and Olivia Mitchell. "Financial Literacy and Retirement Planning in the
United States." Journal of Pension Economics and Finance 10, no. 4 (2011): 509-525.
Lusardi, Annamaria, and Olivia S. Mitchell. Financial Literacy: Evidence and Implications for
Financial Education. Dartmouth College and University of Pennsylvania, 2009.
Lusardi, Annamaria, Olivia Mitchell, and Vilsa Curto. Financial Literacy and Financial
Sophistication among Older Americans. NBER Working Paper 15469, 2009.

379

MacLeod, Bentley. "Reputations, Relationships, and Contract Enforcement." Journal of
Economic Literature 45, no. 3 (2007): 595-628.
Malmendier, Ulrike, and Devin Shanthikumar. "Are Small Investors Naive about Incentives?"
Journal of Financial Economics 85, no. 2 (2007): 457-489.
Marsden, Mitchell, Catherine D. Zick, and Robert N. Mayer. "The Value of Seeking Financial
Advice." Journal of Family and Economic Issues 32, no. 4 (2011): 625-43.
Morey, Matthew. "Should You Carry the Load? A Comprehensive Analysis of Load and Noload Mutual Fund Out-of-Sample Performance." Journal of Banking and Finance 27, no.
7 (2003): 1245-1271.
Mullainathan, Sendhil, Markus Noeth, and Antoinette Schoar. The Market for Financial Advice:
an Audit Study. NBER Working Paper 17929, 2012.
Munnell, Alicia H., Anthony Webb, and Francis M. Vitagliano. "The Economic Implications of
the Department of Labor's 2010 Proposals for Broker-Dealers." Journal of Retirement 1
(2013): 38-54.
Nelson, Wiliam Alan III. "Take It or Leave It: Unconscionability of Mandatory Pre-Dispute
Arbitration Agreements on the Securities Industry." University of Pennsylvania Journal
of Business Law 17, no. 2 (2015): 573-618.
Odean, Terrance. "Are Investors Reluctant to Realize Their Losses." Journal of Finance 53, no.
5 (1998): 1775-1798.
Padmanabhan, Karthik, Constantijn Panis, and Timothy J. Tardiff. Review of Comments on the
Department of Labor's Conflict of Interest Proposed Rule. Advanced Analytical
Consulting Group, 2016.
Padmanabhan, Karthik, Constantijn Panis, and Timothy J. Tardiff. Rates of Return of BrokerSold and Direct-Sold Mutual Funds. Advanced Analytical Consulting Group, 2016b.
Palaveev, Phillip. "Be Careful What You Wish For." Financial Planning, 2008: 78-81.
Palmiter, Alan, and Ahmed Taha. "Mutual Fund Investors: Sharp Enough?" The Capco Institute
Journal of Financial Transformation, 2008: 113-121.
Panis, Constantijn. Review of Study by Quantria Strategies, LLC. Advanced Analytical
Consulting Group, 2014.
Panis, Constantijn, and Michael Brien. Financial Asset Holdings of Households in the United
States in 2013. Report Prepared for the U.S. Department of Labor, Advanced Analytical
Consulting Group, 2016.
Pool, Veronika, Clemens Sialm, and Irina Stefanescu. "It Pays to Set the Menu: Mutual Fund
Investment Options in 401(K) Plans." Journal of Finance, forthcoming.
Prentice, Robert A. "Moral Equilibrium: Stock Brokers and the Limits of Disclosure." Wisconsin
Law Review, 2011: 1059-1107.
Robertson, Christopher Tarver. "Biased Advice." Emory Law Journal, Vol. 60, 2011: 653-703.
Robinson, John H. "Who's The Fairest of Them All? A Comparative Analysis of Financial
Advisor Compensation Models." Journal of Financial Planning, 2007: 56-65.
Rogerson, William. "Reputation and Product Quality." The Bell Journal of Economics 14, no. 2
(1983): 508-516.
380

Schwarcz, Daniel. "Beyond Disclosure: The Case for Banning Contingent Commissions." Yale
Law & Policy Review 25 (2007): 289-336.
Schwarcz, Daniel. "Differential Compensation and the "Race to the Bottom" in Consumer
Insurance Markets." Connecticut Insurance Law Journal 15, no. 2 (2009): 723-754.
Schwarcz, Daniel. "Insurance Demand Anomolies and Regulation." Journal of Consumer
Affairs, 44, no. 3, 2010: 557-77.
Schwarcz, Daniel, and Peter Siegelman. "Insurance Agents in the 21st Century: The Problem of
Biased Advice." In Research Handbook in the Law and Economics of Insurance, by
Daniel and Siegelman, Peter, editors Schwarcz. Edward Elgar Pub, 2015.
Schwartz, Janet, Mary Luce, and Dan Ariely. "Are Consumers Too Trusting? The Effects of
Relationships with Expert Advisors." Journal of Marketing Research 48 (2011): S163S174.
Schwartz, Janet. A, Francesca Gino, and Dan Ariely. "Experts Know Best, But for Whom?
Understanding Conflicts of Interest in the Marketplace." Society of Judgement and
Decision Marking 32nd Annual Conference, 2011.
Sharpe, William. "Capital Asset Prices: A Theory of Market Equilibrium under Conditions of
Risk." Journal of Finance 19, no. 3 (1964): 425-442.
Sharpe, William F. "Mutual Fund Performance." The Journal of Business 39, no. 1 (1966): 119138.
Sharpe, William F. "The Arithmetic of Active Management." Financial Analysts Journal 47, no.
1 (1991): 7-9.
Sharpe, William F. "The Arithmetic of Investment Expenses." Financial Analysts Journal 69,
no. 2 (2013): 34-41.
Shefrin, Hersh, and Meir Statment. "The Disposition to Sell Winners Too Early and Ride Losers
Too Long: Theory and Evidence." Journal of Finance 40, no. 3 (1985): 777-790.
Sirri, Erik, and Peter Tufano. "Costly Search and Mutual Fund Flows." Journal of Finance 53,
no. 5 (1998): 1589-1622.
Stoughton, Neal, Youchang Wu, and Josef Zechner. "Intermediated Investment Management."
Journal of Finance 66, no. 3 (2011): 947-980.
Tang, Ning, Olivia Mitchell, Gary Mottola, and Stephen Utkus. "The Efficiency of Sponsor and
Participant Portfolio Choices in 401(K) Plans." Journal of Public Economics 94, no. 11
(2010): 1073-1085.
Turner, John, and Bruce Klein. "Retirement Savings Flows and Financial Advice: Should You
Roll Over Your 401(k) Plan?" Benefits Quarterly, 2014.
Turner, John, and Dana Muir. "The Market for Financial Advisers." In The Market for
Retirement Financial Advice, by Olivia Mitchell and Kent (eds) Smetters. Oxford
University Press, 2013.
Weber, Martin, and Colin F. Camerer. "The Disposition Effect in Securities Trading: An
Experimental Analysis." Journal of Economic Behavior and Organization, 8 (1998): 167184.

381

Weil, David, Archon Fung, Mary Graham and Elena Fagatto. "The Effectiveness of Regulatory
Disclosure Policies." Journal of Policy Analysis and Management 25, no. 1 (2006): 155181.
Willis, Lauren E. "Against Financial-Literacy Education." Iowa Law Review 94 (2008); 197-286.
Zhao, Xinge. Conflict of Interests between Load Fund Investors and Brokers and Financial
Advisors. China Europe International Business School Working Paper, 2008.

382