Comment Letters2018 | 2017 | 2016 | 2015 | 2014 | Show more years
Ms. Jean A. Webb
Commodity Futures Trading Commission
Three Lafayette Centre
1155 21st Street, N.W.
Washington, D.C. 20581
Re: Performance Data and Disclosure for Commodity Trading Advisors; 64 Fed. Reg. 41843 (August 2, 1999)
Dear Ms. Webb:
National Futures Association (NFA) welcomes this opportunity to comment on the Commission's proposed revisions to its rules regarding performance calculations and disclosure by CTAs. We strongly support the Commission's proposed change to the rate-of-return (ROR) calculation, which recognizes that a CTA's performance does not vary with its clients' cash management strategies. We also support many of the Commission's proposed disclosure requirements, all of which are designed to provide clients with more information so that they can make more informed decisions when deciding whether to hire particular CTAs to manage their futures accounts. We are, however, concerned that some of those requirements will not achieve their objectives. Our comments follow.
A. The Basis for Computing ROR
NFA commends the Commission for recognizing the most basic tenet of NFA's notional funding proposal, that how well a CTA performs is tied to the amount the client asks it to manage rather than to the actual amount of cash the client deposits with its FCM. In theory and by definition, ROR is a measure of the CTA's performance. ROR should accurately reflect the results of a CTA's trading decisions over time, not the clients' differing cash management strategies1. The actual funds method currently used to calculate ROR does not measure the CTA's performance for those CTAs with clients who, for cash management reasons, choose not to deposit the full amount allocated to the CTA with the FCM. In contrast, the proposed method provides an accurate measure of performance for all CTAs and should be adopted.
As NFA has stated on numerous occasions in the past, all similarly traded accounts should have the same ROR, regardless of the amount of funds on deposit with the FCM. Assume that two clients open accounts under a CTA's $250,000 trading program, both clients begin trading at the same time, both clients have the same commission and incentive fee structure, and both clients receive identical trades. In this situation, the CTA generates the same absolute profits and losses for both accounts regardless of funding level. The CTA does not do a better job for one client than for the other. If ROR is calculated using the current actual funds method, however, the ROR for a client with $50,000 on deposit will be five times higher than that for a client with $250,000 on deposit if the CTA makes money and five times lower if the CTA loses money. The Commission's proposed method for calculating ROR, on the other hand, recognizes that the CTA's performance is the same in both instances and provides an accurate measure of the CTA's performance.
We are aware of concerns raised in the press that the Commission's proposal would understate volatility and lead clients and potential clients to believe that futures trading is less risky than it actually is2. These concerns show a lack of understanding of the proposal and are groundless.
If using notional funds to calculate ROR downplayed the risk in futures trading, NFA would be the first to object to using that method. As the Commission is aware, NFA has always been proactive in trying to ensure that customers and potential customers know that futures trading is risky. NFA Compliance Rule 2-29 specifically prohibits Members and Associates from stating that futures trading is appropriate for everyone and from using promotional material that discusses the possibility of profit without an equally prominent statement regarding the risk of loss. Other provisions of Compliance Rule 2-29 that prohibit fraudulent or potentially fraudulent conduct have also been used as the basis for numerous disciplinary actions against Members or Associates for failing to disclose or downplaying risk. NFA Compliance Rule 2-30 specifically requires FCM and IB Members to provide customers with at least the risk disclosure statements required by CFTC Regulations 1.55 and 33.7. However, if the customer has no business trading in the futures markets at all, Rule 2-30 requires the firm to tell him that. Ensuring that customers understand the risk in futures trading is one of NFA's number one priorities.
The truth is, however, that the proposed method of calculating ROR does not understate either volatility or risk. If an account is being traded as a $250,000 account, it has the volatility of a $250,000 account, regardless of the amount of funds on deposit3. In contrast, the current method overstates volatility by exaggerating profits and losses. It also creates widely divergent RORs for similarly traded accounts based solely on the client's different cash management policies and ignores the practical reality that both the CTA and the client consider the account size to be equivalent to the amount committed to trading, not the amount deposited for margin. The proposed method of calculating ROR presents a more accurate picture of volatility and has none of the other limitations inherent in the current method.
In our experience, unwary customers are more likely to be lured into the futures markets by allusions to large profits than by information implying that futures trading is a conservative investment4. In regard to CTAs in particular, we have seen far more abuses involving inflated RORs than artificially depressed volatility. The current method of calculating ROR feeds these abuses by inflating ROR for marginally profitable, as well as highly successful, CTAs. The proposed method will not.
We urge the Commission to give little credence to concerns that the Commission's proposal would understate volatility and lead clients and potential clients to believe that futures trading is less risky than it actually is. As explained above, these concerns are groundless. Even if they were not, however, they would not justify perpetuating an inaccurate measure of performance.
B. Other Issues
Although the overall effect of the proposed amendments is to provide customers with a more accurate picture of the CTA's performance, several of the Commission's proposals may actually provide clients with confusing or distorted information. As explained below, NFA believes that these proposals are neither necessary nor effective for preserving essential customer protections.
1. Interest Earned on Actual Funds (§ 4.35(a)(6)(i)(B)(1)(iii))
NFA supports the Commission's proposed rule defining net performance to include interest income on funds deposited with the client's FCM. However, the Commission's discussion of the proposal indicates that the Commission is not convinced that it has taken the correct approach. NFA urges the Commission to adopt the provision as proposed. Our reasons follow.
The Commission's release gives two reasons why it may not be appropriate to include interest income in the CTA's ROR. The first reason is that "[s]ince the objective of performance reporting is to convey the results from the trading which a CTA performed on behalf of a client, it may be misleading to include interest earned on investments managed by the FCM (as opposed to the CTA). . . ." This reasoning might be persuasive if the CTA had no impact on the amount of interest earned in the account. While that might be true for interest on funds not deposited with the FCM, it is not generally true for those funds held by the FCM.
Both the CTA's trading strategy and its performance can affect the amount of interest earned on the funds held by the FCM. Assume, for example, that one CTA's trading strategy requires 70 percent of a $250,000 account to be deposited as margin while a second CTA's trading strategy requires only 5 percent of a $250,000 account to be deposited as margin. If clients for both CTAs partially fund their accounts at the margin level using interest-bearing instruments, the first CTA's clients will earn more interest than the second CTA's clients will. Similarly, if a CTA makes money and, as a result, the FCM holds more interest-bearing instruments in the client's account, the client will make more interest in the account.
The second reason given in the release for not including interest in ROR is that the interest on the funds in one client's account will cause the performance results for that client to vary from the performance in a similarly traded account for a different client that either does not earn interest on the funds in its account or funds the account at a different level. However, this is an equally strong argument against requiring CTAs to subtract commissions from their net performance. Commissions vary from client to client based on which FCM the client chooses and the client's bargaining power. In fact, interest and commissions may be part of the same negotiated deal - e.g., the FCM may agree to eliminate or charge lower commissions in exchange for collecting the interest on the funds in the client's account. Requiring CTA's to subtract the expenses the client incurs with the FCM without allowing the CTA to add the income the client earns as part of that same relationship simply ignores reality.
Including interest earned on funds deposited with the FCM when calculating net performance is a longstanding industry practice and one that has had the Commission's blessing to date. It is also consistent with the Association for Investment Management and Research's (AIMR) performance presentation standards, which require cash returns from cash and cash equivalents held in portfolios to be included in return calculations. Furthermore, the cash equivalents held in clients' accounts bear low rates of interest, and the amount of interest earned on those funds does not usually have a material effect on overall performance when compared to other advisors. For all of these reasons, NFA respectfully asks the Commission to adopt this portion of the rule as proposed5.
2. Funds Under Management (§ 4.35(a)(1)(iv))
NFA is confused by the proposed requirement that a CTA use the amount of funds deposited with the FCM as the amount of funds under management. On the one hand, the Commission accepts and endorses the concept that the true measure of a CTA's performance is the amount committed to the CTA's trading program since that is the amount the CTA bases its trading decisions on. On the other hand, the Commission chooses to ignore this amount when determining how much money the CTA has under management. The two positions are philosophically and pragmatically inconsistent.
NFA believes that it is misleading to tie funds under management to funds on deposit with an FCM. A CTA manages all of the funds a client commits to its trading programs, even if they are located elsewhere for the time being. Decisions to leverage accounts at a certain level and not to actively trade all available funds are themselves a form of management. In addition, the amount a CTA actively trades may vary widely depending upon market conditions and the CTA's views on what the market is going to do. Since the amount of funds on deposit in a partially funded account is usually based on margin requirements, tying the amount under management to those funds ignores the differences in management style that are the very reason why a client chooses one CTA over another.
Using the amount committed to trading as the amount of funds under management is also consistent with AIMR's performance standards. For example, where a client agrees with an investment adviser on trading size and decides to trade on margin, the adviser's performance is reported on the basis of the full account size even if the client margins the account at 50 percent.
Nor do the clients who use partial funding view it any differently. These clients understand that the CTA is managing the full amount they have agreed for the CTA to trade for them. Although these clients take the funds that are not needed for margin and employ them elsewhere, the clients understand that the funds are committed to the CTA's trading program, and they transfer those funds to their FCM accounts as they are needed for margin.
NFA does not see any reason to require CTAs to disclose the amount of funds that are on deposit with FCMs. The amount of funds on deposit changes frequently, so this requirement would place a burden on CTAs without providing any useful information to current or potential clients. The same amount of funds on deposit could represent a number of small clients with fully funded accounts or a number of large, institutional clients with partially funded accounts. Nonetheless, if the Commission believes that this is useful information, we do not object to having it mentioned briefly. However, the Commission should use terminology that more accurately reflects the nature of the funds and, therefore, is less confusing. The term "actual funds" implies that any other funds are fabricated or hypothetical, when they are in fact very real but are simply held somewhere other than at the FCM. If the Commission requires disclosure of these amounts, it should call them "funds on deposit" rather than "actual funds."
On the other hand, the amount of funds under management, measured by the total amount of funds committed to the CTA, can be useful information. Although it is not infallible, large amounts of funds under management usually indicate that the advisor has gained the confidence of either a large number of clients or of several large institutional clients, both of which indicate a level of trust in the advisor. Conversely, some people believe that performance begins to drop once a program gets too large, and those people would also benefit from knowing how much money has been committed to the advisor's trading program.
The amount of funds a CTA has under management is equal to the amount of funds that govern its trading decisions. Therefore, the Commission should define funds under management as the amount the client and the CTA have agreed to commit to the CTA for trading (the aggregate nominal account size in the Commission's release) and eliminate the requirement to disclose the amount of funds on deposit with FCMs.
3. Drawdowns (§ 4.35(a)(1)(v), (vi), & (ix))
Section 4.34(p) (3) of the proposed rules require CTAs to disclose in general terms the effect that partial funding has on leverage and the likelihood and frequency of margin calls. NFA fully supports this requirement.
The proposed rules also require CTAs to provide additional information regarding drawdowns as part of the performance capsule. NFA supports adding any information that provides truly meaningful information to clients. However, the drawdown figures being proposed contribute to information overload and could confuse clients without providing a corresponding benefit.
First, NFA objects to the proposed requirement that CTAs provide the worst monthly and peak-to-valley drawdowns for the life of the program as well as the last five years. Many CTAs refine and modify their trading programs over time. Although any single change may not be material, the cumulative effect results in declining materiality as the information gets older. It is also a simple fact of life that the longer a trading program has been in existence the more drawdowns it will have and the higher the probability that some of those drawdowns will be large. Requiring firms to provide drawdowns for the life of the program can mislead potential clients who are comparison shopping by making shorter track records look more attractive than more established programs that, given the same period of time, may actually be less volatile.
Second, NFA questions the usefulness of requiring CTAs to provide drawdown figures based on the lowest ratio of funds deposited with the FCM to amount committed to trading or, in some cases, at an imputed rate of 20 percent. If a client is interested in this information, the client already has all of the information it needs to do its own calculation - a calculation that can be done by any 7th Grade pre-algebra student.
As NFA has stated on many occasions in the past, providing too much information makes clients' eyes glaze over and deters them from reading the disclosure document or the performance capsule at all, which also keeps them from reading the information that is truly useful in making an investment decision. To add additional requirements with no real merit will simply serve to over-emphasize the importance of performance disclosures in relation to the other disclosures required by Commission rules. Therefore, NFA asks the Commission not to require CTAs to provide additional drawdown information.
4. Range of Rates of Return (§ 4.35(a)(1)(viii))
The Commission's proposal would require CTAs to disclose the range of lifetime RORs for closed accounts in the offered program. This issue was not raised in the concept release and is not a notional funding issue. Therefore, NFA asks the Commission to sever it from the proposals regarding calculating ROR (and making related disclosures) and consider it at a later date.
Contrary to the Commission's statement in the release, this portion of the Commission's proposal would create a huge burden for CTAs. Counting the number of accounts that closed with profits and the number that closed with losses is quite different from calculating the ROR for each separate account in order to determine which ones have the largest and smallest positive ROR and the largest and smallest negative ROR. And even though the proposal only requires information from those accounts that closed within the last five years, calculating the information could require the CTA to use information that goes well beyond that - as, for instance, if an account that closed four years ago had been open for the previous 10. In fact, some of the information necessary to make those calculations may no longer be available.
More importantly, the information the Commission is considering requiring is of extremely limited use, if any, to a client or potential client. RORs in individual accounts depend on a number of factors (e.g., how long the account was open and how volatile the markets were at that particular time) that may not apply to the particular client reading the information. On the contrary, range of ROR in individual accounts is confusing and contributes to information overload.
Requiring CTAs to calculate individual RORs has two major pitfalls. First, the proposed requirement is inconsistent with the very concept of providing performance information on a composite basis - a concept endorsed by both the Commission and AIMR. Second, requiring CTAs to calculate individual RORs assumes that the performance in closed accounts is more relevant than the performance in open accounts - an assumption that we disagree with. However, if the CTA added back the performance in open accounts during that same period, the CTA would have the same ROR that is reflected in the current composite. In other words, the current composite contains all of the information necessary to show the CTA's ROR6.
NFA respectfully requests the Commission to sever this portion of the proposal from the notional funding issues. In any event, NFA urges the Commission not to require CTAs to disclose a range of RORs in individual accounts that have closed during the last five years.
5. Monthly Performance (§ 4.35(a)(2)(ii))
The Commission's proposal would also require CTAs to present their monthly RORs both numerically and in a graph. This is another proposal that is not related to notional funding issues. Therefore, NFA asks the Commission to sever this proposal as well.
NFA has consistently maintained that requiring 60 plus monthly ROR figures rather than five annual ROR figures plus year-to-date makes a client or potential client's eyes glaze over and, therefore, is counterproductive. Nonetheless, since clients must live with that requirement, they should not be bombarded with even more information by requiring the CTA to present the information twice - once in a numerical table and again in a bar graph7.
NFA supports the current regulations that allow CTAs to provide ROR information in either a numerical table or a bar graph. We do, however, prefer the numerical table since it is more precise and less subject to distortion.
To illustrate our concerns, we have attached two bar graphs that both use the same numerical information. Appendix A shows monthly rates of return for January 1994 through April 1999 in what might be a typical presentation. The performance depicted in the graph is relatively volatile. If the CTA wanted to downplay this volatility, it could do so by increasing the range of RORs covered by the vertical axis, as shown in Appendix B, and even by increasing the size of the graph. Furthermore, it is hard to determine the actual ROR numbers on Appendix B, where the range between grids is 20 percent.
The Commission's proposed Regulation 4.35(a)(2)(iii)(B) does require that the bar graph "be scaled in such a way as to clearly show month-to-month differences in rate-of-return." However, given the vast differences in ROR figures from trading program to trading program, it would be nearly impossible to come up with guidelines for presenting this information uniformly, and those guidelines would of necessity be very complex. This is not a problem with the numerical presentation, since actual numbers are much harder to distort. Therefore, NFA recommends that the Commission leave well enough alone.
6. Commodity Pool Disclosure (§ 4.25(a)(1)(ii)(H))
NFA also has concerns about that portion of the Commission's proposal that would require CPOs to include additional performance information in pool disclosure documents where "the CPO allocates, to any of the pool's CTAs, an amount of actual funds which is less than the nominal account size stated in the pool's written agreement with the CTA." 64 Fed. Reg. at 41846. First, we have a conceptual problem with equating the amount allocated to the CTA with the amount of funds on deposit with an FCM. In our view, the amount allocated to the CTA is the amount committed to the CTA's trading program in the agreement between the pool and the CTA.
Second, we do not believe this information is helpful to pool participants. Pool participants are not concerned about whether or not the pool puts all of its assets in trading accounts or whether it temporarily leaves them somewhere else in order to maximize return. In fact, put that way, pool participants may be more concerned about leaving excess funds in a trading account where they receive a lower return.
Many, if not most, pools, acting as clients, partially fund their accounts. Therefore, many, if not most, pools would have to disclose the information required by the Commission's proposal. We believe that this is simply irrelevant information that will add to the length and complexity of disclosure documents without any corresponding benefit8.
On the other hand, NFA's proposed Compliance Rule 2-34 (c) required a CPO to disclose certain information if the pool allocates its assets in such a way that the aggregate amount allocated to CTAs exceeds the pool's assets. Participants may want to know when they are investing their money in a pool that uses leverage not only in the instruments it trades but also in how it allocates its funds to trading. This is better and more concise disclosure than what the Commission is proposing. However, this issue may be better addressed as part of the upcoming proposals in connection with the LTCM study.
7. Documentation of Nominal Account Size (§ 4.33(c))
NFA supports the proposed changes to CFTC Regulation 4.33 to add a new section (c). However, as the Commission is aware, it is not always easy to get existing customers to respond to requests to sign and return documents in order to comply with new regulatory requirements. Therefore, NFA recommends that the Commission allow CTAs to use a negative consent procedure with existing fully funded accounts. Under this procedure, the CTA would send the client a copy of the information required by Regulation 4.33(c) and advise the client that it will be considered part of the agreement unless the client objects within 10 days (or whatever period the Commission believes is appropriate).
NFA applauds the Commission for revising the basis for calculating ROR to provide an accurate picture of a CTA's performance. We also appreciate the opportunity to express our concerns about other aspects of the proposal and hope that our comments will assist the Commission in ensuring that clients receive understandable and accurate information concerning both the CTA's performance and the clients' own accounts. Finally, we look forward to working with the Commission to implement its rule amendments.
Daniel J. Roth
1 CTA performance should not be confused with the amounts shown on monthly account statements, profit and loss statements, and other reports furnished to a client by its FCM. Those reports merely show the actual dollars made on each trade, the amount charged as fees, and the amount held for the client in an account at its FCM.
2 These concerns were also raised in several comment letters filed with the Commission in response to its concept release and request for comments on NFA's rule proposal. 63 Fed. Reg. 33297 (June 18, 1998).
3 In a real-life situation, there is nothing hypothetical about this amount. The trading account size chosen by the client and agreed to with the CTA is the amount the CTA bases its trading decisions on.
4 Although we do occasionally see situations where Members or Associates state or imply that futures trading is a conservative investment, those representations are usually coupled with glowing predictions of large profits.
5 Although we do occasionally see situations where Members or Associates state or imply that futures trading is a conservative investment, those representations are usually coupled with glowing predictions of large profits.
6 If the proposed requirement is intended to show volatility, the drawdown figures already provide that information.
7 Our review of disclosure documents indicates that the overwhelming majority of CTAs choose to present ROR figures in a numerical table.
8 The Commission's position on this issue also appears to be inconsistent with Interpretive Letter 88-1, [1987-1990 Transfer Binder] Comm. Fut. L. Rep. (CCH) ¶ 24,058 (December 16, 1997).