Statement on Proposed Rules on Funds' Use of Derivatives
Nov. 26, 2019
Yesterday the Commission proposed rules on funds’ use of derivatives to obtain leverage. Appropriate use of derivatives can produce benefits for investors, like better risk-adjusted returns or more efficient exposure to certain asset classes. But that same leverage also presents serious risks, magnifying losses for investors in times of turbulence. And the Commission’s historically piecemeal approach to these issues is insufficient given the growth in funds’ use of derivatives over the past several decades.
The Commission is long overdue in establishing a systematic approach that more meaningfully limits fund risktaking, so we support this proposal. We’re also encouraged that the proposal includes investor protection measures when leveraged and inverse funds—products that are rarely appropriate for retail use—are sold to ordinary investors. But we urge commenters to help the Commission strengthen the proposal in three areas that are critical for protecting investors from the risky use of derivatives by funds.
* * * *
First, the proposal provides some basic protections for ordinary investors who are sold leveraged and inverse exchange-traded funds. Those products are sold on the promise that they deliver a multiple, for example three times, of the returns of a particular index. These complex financial products may serve a legitimate purpose for a subset of sophisticated traders, but in a volatile market an ordinary investor can lose money in these products even when the index increases in value. The disconnect between an ordinary American family’s reasonable expectations and the reality of these products creates real risk that too many investors will find out, too late, how dangerous these products can be.
That’s especially true because the data show that too many investors buy and hold these products on the expectation that they are sound investments over the long term—when they’re usually not. Yesterday, the Commission proposed to address this problem by requiring brokers to follow certain practices when selling these products. We hope commenters will come forward with evidence about whether these steps are enough to protect ordinary investors from the risks presented by leveraged and inverse funds.
Second, the proposal requires fund boards to appoint a manager to oversee its use of derivatives. The proposal would not, however, require the board to approve the fund’s derivatives risk management program, and we are unpersuaded that hiring the risk manager is enough board-level engagement on the risks presented by derivatives use. We understand that fund boards have limited time and expertise, but insufficient board involvement in this area can prevent development of the director-level proficiency needed to provide meaningful oversight and ensure sound risk management for investors. Board approval has served investors well in a broad range of areas—from audit to executive pay to investment company risk management—and can create a structural incentive for directors to engage more actively in oversight. We hope commenters will come forward to help the Commission better understand the board’s role when it comes to derivatives.
Third, the proposal largely limits use of derivatives through constraints on funds’ portfolio risk. Those constraints generally require that a fund’s portfolio risk, including all derivatives investments, not exceed 150% of that of an index. And, to ensure that the use of an index provides meaningful limitations on funds’ leverage, the proposal prohibits the use of bespoke indices for that purpose. To measure these matters, the proposal relies heavily on value at risk (VaR), a widely-used industry metric designed to measure the level of financial risk in a fund over a specified period of time. The reliability of VaR as a risk metric is the subject of significant debate, and is in any case largely dependent on the specifics of the model used. We believe that the Commission should consider further measures to ensure that funds’ VaR models are reliable and not subject to opportunistic gaming. The proposal includes certain minimum standards for VaR calculation, and we hope commenters will help us and the Staff better understand how to make sure that funds’ VaR models reflect the actual risks posed by the use of derivatives.
* * * *
We are grateful to the dedicated Staff in the Divisions of Investment Management and Economic and Risk Analysis for their extensive work on this proposal. The questions addressed in yesterday’s proposal are exceptionally challenging ones, and the Commission, the marketplace, and investors have benefited a great deal from the Staff’s deep expertise. We now look forward to public input—especially regarding how the Commission can improve this proposal to ensure the protection of ordinary investors.
 Yesterday’s proposal follows up on the Commission’s 2015 proposed rule on funds’ use of derivatives. See Investment Company Act Rel. No. 31933 (Dec. 11, 2015). The 2015 proposal, like yesterday’s, addressed the concern that our existing guidance regarding funds’ use of derivatives—at both the Commission and Staff level—reflects a patchwork approach responding to funds’ specific questions about particular types of derivatives. That approach has led to complex and sometimes inconsistent industry practices regarding how—and whether—funds comply with the limitations on leverage drawn from Section 18 of the Investment Company Act of 1940. See 15 U.S.C. § 80a-18.
 See, e.g., FINRA, Investor Alert: Leveraged and Inverse ETFs: Specialized Products with Extra Risks for Buy-and-Hold Investors (Aug. 18, 2009) (providing real-life examples of counterintuitive investment outcomes for leveraged and inverse ETFs, including one case where an ETF seeking to provide three times the daily return of an index fell more than 50% despite the index increasing by approximately 8%).
 See, e.g., Minder Cheng & Ananth Madhavan, The Dynamics of Leveraged and Inverse Exchange-Traded Funds (working paper 2009) (noting that, despite industry claims otherwise, “[a]verage holding periods . . . are still significant for double-leveraged ETFs,” at over two weeks).
 The proposal would require broker-dealers and investment advisers to collect specific information for each account and document that the broker-dealer has a reasonable basis to believe that the account owner has sufficient knowledge and experience in financial matters to evaluate the risks of buying and selling leveraged or inverse funds. See Investment Company Act. Rel. No. 33704 (Nov. 25, 2019), at 34-35. This proposal is substantially similar to FINRA’s rules relating to the approval of accounts to engage in options trading. Compare id. with FINRA Rule 2360(b)(16).
 See Investment Company Act Rel. No. 33704, supra note 4, at 79 (citing proposed rule 18f-4(c)(5)). We appreciate that the rule makes clear that “[b]oard oversight should not be a passive activity,” and that directors “should ask questions and seek relevant information regarding the adequacy of the [derivatives] program,” see id. at 79-81, and we hope commenters will weigh in regarding whether requiring a board vote will best protect investors.
 The Commission’s 2015 Proposal would have required fund boards to vote to approve these programs, and we would be concerned if any final rule in this area signaled a retreat from the importance of boards’ responsibilities in the area of risk management. Investment Company Act. Rel. No. 31933, supra note 1, at 225-226 (“We believe that boards should understand the derivatives risk management program and the risks it is designed to manage. Accordingly, proposed rule 18f-4 would require each fund to obtain initial approval of its written derivatives management program from the fund’s board of directors, including a majority of independent directors.”).
 See, e.g., Sarbanes Oxley Act of 2002, Pub. L. No. 107-204, §301, 116 Stat. 745, 775-76 (amending 15 U.S.C. §78j-1 to require audit committee approval of certain matters).
 See NYSE Listed Company Manual §303A.05 (requiring director approval of CEO compensation); see also Jeffrey N. Gordon, Executive Compensation: If There’s a Problem, What’s the Remedy?: The Case for “Compensation Discussion and Analysis,” 30 J. Corp. L. 102, 116 (2006) (providing the theoretical case for board approval of specific executive-pay agreements requiring board-level expertise).
 See 17 C.F.R. § 270.22e-4(b)(2) (requiring a fund’s board to approve the liquidity risk management program). We note that Rule 22e-4—including the requirement for board approval of a fund’s liquidity risk management program—was adopted unanimously by the Commission only three years ago. See Investment Company Act Rel. No. 32315 (Oct. 13, 2016).
 For empirical evidence on the emergence of these benchmarks—and the problems they can present for investors—see Adriana Robertson, Passive in Name Only: Delegated Management and “Index” Investing, 36 Yale J. Reg. 795 (2019) (concluding, based on a hand-collected sample of benchmarks for U.S. mutual funds, that “the overwhelming majority of the indices in [the] sample are used as a primary benchmark by only a single fund”).
 To say, for example, that a particular position has a one-day 5% VaR of $1 million means there is a five percent chance that the position will lose up to $1 million in one day. For the seminal economic description of this measure, see Philippe Jorion, Value at Risk: The New Benchmark for Managing Financial Risk (2000).
 One issue with VaR models is that they assume that returns follow a Gaussian distribution—an assumption in some tension with the empirical evidence. This was discussed most compellingly and recently in Nassim N. Taleb, The Black Swan: The Impact of the Highly Improbable (2007), but that work built on decades-old scholarship raising questions about the validity of the Gaussian assumption in finance, Benoit Mandelbrot & Richard L. Hudson, The (Mis)Behavior of Markets: A Fractal View of Risk, Ruin and Reward (1996); Benoit Mandelbrot, The Variation of Certain Speculative Prices, 36 J. Bus. 394 (1963); Eugene F. Fama, The Behavior of Stock-Market Prices, 64 J. Bus. 34 (1965). Some evidence suggests that a simple projection can, in some cases, be as accurate as VaR. Jeremy Berkowitz & James O’Brien, How Accurate are Value-At-Risk Models at Commercial Banks?, 57 J. Fin. 1093 (2002). That’s why, among other reasons, the Basel Committee on Banking Supervision has recommended moving away from VaR models to more robust expected shortfall models. See Basel Committee, Minimum Capital Requirements for Market Risk (January 2016). We further note that notional exposure—as proposed in the 2015 release—also has the potential to inaccurately capture risk.
 Specifically, from the Division of Investment Management, we would like to thank Dalia Blass, Sarah ten Siethoff, Brian Johnson, Amanda Wagner, Thoreau Bartmann, Dennis Sullivan, John Lee, Sirimal Mukerjee, Joel Cavanaugh, Asaf Barouk, Penelope Saltzman, Tim Dulaney, Michael Spratt, Ed Bartz, Asen Parachkevov, and Jackie Rivas. From the Division of Economic and Risk Analysis, we appreciate the work of S.P. Kothari, Narahari Phatak, Alex Schiller, Wu Mi, Christian Jauregui, Rooholah Hadadi, Adam Large, James McLoughlin, and Tristan Chiappetti.