ICI Letter to OMB About Data Used in Department of Labor Re-Proposal of Fiduciary Rule (pdf)

April 7, 2015 The Honorable Howard Shelanski Administrator Office of Information and Regulatory Affairs Office of Management and Budget 725 17th Street NW Washington, DC 20503 Dear Mr. Shelanski: On behalf of the Investment Company Institute,1 we are writing to express deep concern about the data the Department of Labor (“Department”) appears to be relying on as justification for its re-proposal of rules redefining the term “fiduciary” under the Employee Retirement Security Income Act of 1974 (“ERISA”). We understand that the Department has submitted the re-proposal to the Office of Information and Regulatory Affairs (“OIRA”) for review. 2 We explain our concerns below. We also request the opportunity to meet with you to share our research and data in greater detail as OIRA reviews the proposal’s impact on American retirement savers and the retirement system as a whole.3 1 The Investment Company Institute (ICI) is a leading, global association of regulated funds, including mutual funds, exchange-traded funds (ETFs), closed-end funds, and unit investment trusts (UITs) in the United States, and similar funds offered to investors in jurisdictions worldwide. ICI seeks to encourage adherence to high ethical standards, promote public understanding, and otherwise advance the interests of funds, their shareholders, directors, and advisers. ICI’s U.S. fund members manage total assets of $18.1 trillion and serve more than 90 million U.S. shareholders. 2 As you are well aware, Executive Order 12866, 58 Fed. Reg. 51735 (Oct. 4, 1993) requires the Department and certain other federal agencies to promulgate only such regulations as are required by law, are necessary to interpret the law, or are made necessary by compelling public need. The current administration reaffirmed the principles and structure that were established in Executive Order 12866 in Executive Order 13563. See 76 Fed. Reg. 3821 (Jan. 21, 2011). As stated in a report by the General Accounting Office (“GAO”), OIRA reviews agency regulatory rules and proposals for consistency with Executive Order 12866, among other things. See GAO, “OMB’s Role in Reviews of Agencies’ Draft Rules and the Transparency of Those Reviews,” available at: http://www.gao.gov. 3 The Institute serves as a source for statistical data on the investment company industry and conducts public policy research on fund industry trends, shareholder characteristics, the industry’s role in U.S. and international financial markets, and the retirement market. For example, the Institute publishes reports focusing on the overall U.S. retirement market, fees and expenses, and the behavior of defined contribution plan participants and IRA investors. In its research on mutual fund The Honorable Howard Shelanski April 7, 2015 Page 2 of 8 As you know, there is wide concern about the impact that an expanded fiduciary definition will have on the ability of American workers to obtain the guidance, products, and services they need to adequately prepare for their retirements through retirement plans and individual retirement accounts (“IRAs”). This concern has heightened the importance of a comprehensive regulatory impact analysis of the re- proposal, focused on showing that any restriction on access to guidance, products, and services resulting from the re-proposal’s significant regulatory expansion is justified. It is that showing that is the core of the concern underlying this letter. Based on documents available on the Department’s website and on public statements made by representatives of the Department, it appears that the Department is relying as justification for the re- proposal, at least in part, on a Council of Economic Advisor’s report (the “CEA Report”) published shortly before the re-proposal was submitted to OIRA.4 Specifically, it appears that the Department is relying on the CEA Report to establish a “compelling public need” for purposes of compliance with Executive Order 12866, and as a basis for the re-proposal’s required regulatory impact analysis. Several claims made in the CEA Report call the reliability of that report into question, in particular as justification for any regulatory expansion by the Department. First, the CEA Report argues that compensation structures prevalent in the financial services industry incentivize brokers and other financial advisers to encourage workers and retirees to shift assets out of 401(k)s and into high cost IRAs.5 If brokers and other advisers were following that practice, one would expect IRA investors to be steered into high-cost mutual funds. In fact, ICI’s data show that not to be true – IRA investors tend to concentrate assets into funds that are lower-cost than average. As the chart below shows, IRA investors tend to concentrate assets in funds whose expense ratios (the bars) are far less than the average expense ratio of all funds (the line). In 2013, the average expense ratio paid by IRA investors on equity funds was 63 basis points, or 0.63 percent, lower than the average expense ratio for all equity funds.6 This pattern also holds true in hybrid and bond funds. investors, IRA owners, and 401(k) plan participants, the Institute conducts periodic household surveys that connect directly with investors. Institute data is not only referenced in the CEA Report, but also in several studies relied upon therein. 4 See Jason Furman & Betsey Stevenson, Council of Economic Advisors, “The Effects of Conflicted Investment Advice on Retirement Saving,” available at http://www.whitehouse.gov/sites/default/files/docs/cea_coi_report_final.pdf (Feb. 2015). 5 See CEA Report at 7-8. 6 Calculation: 137 basis point average expense ratio on all equity funds minus 74 basis point average paid by IRA investors on equity funds equals a 63 basis point difference. The Honorable Howard Shelanski April 7, 2015 Page 3 of 8 IRA Investors Pay Below-Average Fees Note: Data exclude mutual funds available as investment choices in variable annuities. Sources: Investment Company Institute and Lipper Second, the CEA Report’s illustration of a fee disparity between 401(k)s and IRAs vastly exaggerates the difference in fund expenses paid by 401(k) investors and IRA holders.7 The CEA Report’s hypothetical illustration assumes that typical 401(k) plan investors pay fund expenses of only 20 basis points, but pay 130 basis points when assets are rolled over to an IRA. Though the Report acknowledges that the illustration is hypothetical, this 110 basis point difference drives its multi- billion-dollar estimate of harm to investors. Actual data, however, demonstrates that the hypothetical is not accurate, resulting in the Department’s estimate of multi-billion-dollar harm to investors being severely overstated. As shown above by actual data, not a mere hypothetical, the average fee paid by IRA investors in a stock fund is only 74 basis points. And the average fee for investing in stock funds paid by 401(k) investors is actually 58 basis points. So the real difference in fees, based on what investors are actually paying, is 16 basis points—far less than the 110 basis point difference (about 85% less) claimed by the CEA Report. Our data do show that 401(k) investors pay lower fees in mutual funds than IRA holders pay. That in part reflects economies of scale, as employer plans aggregate the savings of hundreds or thousands of workers. In addition, some IRA investors pay through fund fees for services their 401(k) plans do not 7 See CEA Report at 17-18. The Honorable Howard Shelanski April 7, 2015 Page 4 of 8 provide, such as professional investment analysis. Taking away the ability to pay for those services may mean that some IRA investors get no help at all—which could be the most costly outcome. Third, the CEA Report bases its claim of investor harm on a string of academic studies.8 The studies do not support the CEA Report’s conclusions. One basic problem is that none of the academic studies relied on by the CEA Report actually compares the costs of investing with a fiduciary adviser versus those of investing via a broker or other adviser that is not a fiduciary. A simple illustrative example using low cost investments (e.g., index funds or exchange-traded funds [ETFs]), and adding the fees charged by a well-regarded fiduciary adviser9 shows how the Report’s failure to make such a relevant comparison can lead to erroneous conclusions: Illustrative example: An investor using the advice services of this firm would pay the listed annual asset-based fees (e.g., a 200 basis point fee applies to the first $150,000 in account assets) plus the fees on the investors’ chosen index funds or ETFs. An investor with a portfolio of $150,000 invested 50% in a low-cost S&P 500 index fund and 50% in a bond index fund would pay a blended ETF fee of 6 basis points plus 200 basis points in fees to the financial adviser for a total cost of 206 basis points. The same investor using a broker might invest 50% in C shares of S&P 500 index fund that pay the adviser 12b-1 fees annually and 50% in C shares of a total bond market index fund. This investor would have a total cost of 134 basis points, based on the average fees of such funds. This is an apples-to-apples comparison of “all-in” fees (including fund management fees and 12b-1 fees collected by the fund and paid to the broker and the brokerage firm). In this example, relative to a broker, using a fiduciary adviser is more costly, giving the investor a lower return. Set forth below is a brief summary of the studies relied on by the CEA and why they offer little, if any, support for the CEA Report’s conclusion that investors purchasing funds sold through brokers receive significantly lower returns than do investors using fiduciary advisers. 8 The CEA’s reliance on the studies as the basis for its conclusions is clearly stated: “This report focuses on quantifying the impact of conflicting incentives in the particular case of financial advisers providing conflicted advice to IRA account holders. To do so it turns to the empirical literature that examines conflicted payments and investment products characterized by conflicted payments.” See CEA Report at 10. 9 The illustrative example is based on published fees charged by a well-regarded fiduciary investment adviser. Such fees can be higher or lower depending on the adviser, but the premise of the illustrative example holds true – investment advisory services from a fiduciary come at a cost and a true analysis of whether or not there is an impact to investors from “conflicted” advice should take those fees into account. The Honorable Howard Shelanski April 7, 2015 Page 5 of 8 1. Bergstresser, Chalmers, and Tufano (2009) The first paper cited by the CEA Report in support of its statement that “funds characterized by conflicted payments significantly under-perform funds sold directly to savers” is the 2009 paper by Bergstresser, Chalmers, and Tufano, which attempts to measure the returns to investors using funds sold through intermediaries before accounting for fees used to compensate the intermediary. Significantly, the paper compared intermediary-sold funds with direct-sold funds and does not measure whether a fiduciary relationship produces superior returns, net of fees, over a non-fiduciary intermediary (e.g., brokerage) relationship. The Bergstresser paper does not support the CEA’s characterization of the paper’s conclusion (i.e., that investors purchasing funds through brokers get lower returns). In fact, the evidence in the Bergstresser paper, even if taken at face value, is highly inconclusive: the paper suggests that compared to investors that purchased “directly sold” mutual funds, investors who used intermediary-sold funds earned lower returns on broad equity funds and higher returns on foreign equity funds and perhaps money market funds. (The evidence on bond funds is statistically inconclusive.) Simply put, if underperformance is due to the conflicted compensation structure of the intermediary—as the CEA Report suggests—one would expect that the underperformance would occur in all types of funds, not just one. It is also important to note that the Bergstresser paper is not about retirement accounts. It does not seek to measure the returns earned on 401(k) accounts, IRAs, or any other kind of retirement account. Indeed, the term IRA is never mentioned in the paper and 401(k) is mentioned just once. The authors use the phrase “defined contribution plan” once when they state that they explicitly seek to avoid analyzing assets in defined contribution plans: “We attempt to exclude funds sold into defined contribution plans by excluding from our study share classes that are identified as being made available to these plans.” Furthermore, the kind of analysis the Bergstresser paper seeks to undertake—comparing returns of one type of investor with those of another type—is notoriously difficult. Authors who undertake this kind of analysis (including Bergstresser and co-authors) rarely if ever have data on the actual portfolio holdings, returns, characteristics, and motives of individual investors. Consequently, any such studies at best provide circumstantial evidence about investors’ experiences with intermediaries. The Honorable Howard Shelanski April 7, 2015 Page 6 of 8 2. Del Guercio and Reuter (2014) The 2014 paper by Del Guercio and Reuter cited in the Report finds that actively managed funds sold directly to investors outperform index funds and that broker-sold index funds outperform broker-sold actively managed funds. The paper speculates that this result is driven by broker incentives, but does not provide any test of this theory. Consistent with the limitations of the Bergstresser paper, the Del Guercio paper also does not measure the net returns to investors using a fiduciary adviser versus a broker. The CEA Report acknowledges that a limitation of both the Bergstresser and Del Guercio studies is that such comparisons cannot incorporate differences other than the involvement of an intermediary versus a “direct sale” of the mutual fund. The Report explains, for instance, that “investors purchasing funds through intermediaries may be more risk-averse and less experienced with investing than those buying direct-sold shares from a mutual fund sponsor” and that, “[f]ailing to account for such differences may potentially overstate or understate losses due to conflicts of interest.”10 3. Chalmers and Reuter (2012) In order to rationalize its use of the Bergstresser and Del Guercio papers despite the aforementioned limitations, the CEA Report cites a 2012 paper by Chalmers and Reuter, which attempts to measure the cost of receiving financial advice in the Oregon University System’s defined contribution plan. Based on findings associated with this single plan, the study finds that after paying the costs associated with the financial adviser, investors’ risk-adjusted returns are reduced by 92 to 142 basis points. As shown by the simple calculation above, however, a similar result would have been achieved if the financial advisers had been fiduciaries and received the same level of compensation but had been required to place investors in index funds. Hence, the paper does not test whether the risk-adjusted returns would have been lower if the investor had invested with the assistance of a fiduciary adviser. Significantly, Chalmers and Reuter also identified certain positive attributes associated with the use of an adviser, noting that those investors using an adviser had more diversified portfolios, showed less home country bias, made greater use of index funds, and had lower allocations to the plan’s default option (a money market fund). 4. Christoffersen, Evans, and Musto (2013) The CEA Report cites a paper by Christoffersen, Evans, and Musto which purports to measure the cost to investors of investing in funds sold through brokers. The paper finds that investors who were invested in funds that compensated brokers with higher-than-average loads adjusting for a set of fund features, earned lower returns. Again, the paper does not measure or test if these returns were lower 10 CEA Report at 11. The Honorable Howard Shelanski April 7, 2015 Page 7 of 8 than those that investors would have received had they used a fiduciary adviser. Also, the paper finds that paying brokers through an annual 12b-1 fee or revenue sharing did not produce lower returns, which is inconsistent with the argument that investors using a broker are more likely to be placed in underperforming funds. 5. Foerster et al. (2014), Hackethal et al. (2012a) The CEA Report also references papers examining retail investment advice in Canada and Germany. While acknowledging that those countries’ legal regimes differ from that in the United States, the CEA Report cites the studies as support for the conclusion that advised accounts underperformed by more than 150 basis points. This incorrectly represents what the authors claim to have found and, at any rate, the papers appear to be irrelevant to the discussion here. First, while the papers do purport to study the value of advice offered by financial advisers, the papers are about investment advisers in Canada and Germany respectively, not the United States. The Foerster et al. findings are based on the authors’ analysis of statistics from the Canadian Financial Monitor survey of Canadian households. The Hackethal paper relies on a review of investor accounts at “a large German bank.” Whether the papers’ results carry over to the U.S. regulatory regime, and thus to brokers or financial advisers in the United States, is unlikely or, at most, simply not clear. Significantly, the Foerster paper concludes that advisers induce their clients to take more risk—i.e., invest a greater portion of their portfolios in equities—thereby raising expected returns. The paper also observes, however, that “the amount of risk an adviser takes in his or her own portfolio strongly predicts the risk taken by his or her clients,” suggesting that the clients’ assumption of additional risk is consistent with the Canadian adviser’s own beliefs—not an inherent conflict of interest resulting from the manner in which the adviser is compensated. The Hackethal paper, based on a review of account data from a “single German bank,” does find that accounts advised by an adviser had returns that were lower than the sample mean for self-managed accounts, reflecting lower holdings in equities by the accounts.11 Given that Germans tend to invest a lower percent of their portfolios in the stock market (23 percent of German portfolios are invested in stocks, compared to 50 percent for Americans), it is not clear that these findings reflect the outcome of broker conflicts. They might, for example, be reflective of investing norms in Germany, or the fact that, as the study also observes, “[o]lder clients (over 50) have significantly greater probability than investors 11 The advised accounts in the Hackethal study were noted “to exhibit far greater diversification than those run by individuals alone. The average share of directly held stocks among self-managed accounts is just under 60 percent, while that for [advised] accounts is about 20 percent.” See Hackethal et al. at p.15. The Honorable Howard Shelanski April 7, 2015 Page 8 of 8 between 18 and 30 of using an advisor,”12 and might be at a stage of their retirement planning where they begin to reduce their holdings in equities in favor of fixed-income securities. * * * * * As demonstrated above, the deficiencies in the CEA Report call into question the Department’s reliance on that report (or the analysis informing the report) to justify its fiduciary re-proposal. Simply put, there is little in the academic papers cited in the CEA Report, or in the extensive data readily available to the Department, to substantiate the conclusion that underperformance of funds sold through brokers is “about 1%” or any other number. This is telling, especially in light of our first two points showing that IRA investors actually concentrate assets into lower-cost mutual funds. We would greatly appreciate the opportunity to share our research and discuss the issues identified above in greater detail as OIRA performs its crucial responsibilities associated with the proposal. We look forward to hearing from you soon. Sincerely, /s/ David M. Abbey /s/ Brian Reid David M. Abbey Brian Reid Deputy General Counsel, Retirement Policy Chief Economist 12 Id. at 16.