ICI Letter on CFTC's Minimum Block Trade Reproposal (pdf)

May 14, 2012 Mr. David A. Stawick Secretary Commodity Futures Trading Commission Three Lafayette Centre 1155 21st Street, NW Washington, DC 20581 Re: Procedures to Establish Appropriate Minimum Block Sizes for Large Notional Off-Facility Swaps and Block Trades (RIN 3038-AD08) Dear Mr. Stawick: The Investment Company Institute (“ICI”)1 is submitting this letter in response to the re- proposal by the Commodity Futures Trading Commission (“CFTC” or “Commission”) to specify the procedures for determining block trade sizes pursuant to section 727 of the Dodd-Frank Wall Street Reform and Customer Protection Act (“Dodd-Frank Act”).2 The Reproposal would define the criteria for determining swap categories and the methodologies for setting appropriate minimum block sizes for each swap category. The Reproposal also would include additional measures to prevent public disclosure of the identities, business transactions, and market positions of swap market participants. As part of the CFTC’s proposed rules to implement real-time reporting of swap transactions, the CFTC originally proposed two different tests for setting appropriate minimum block sizes to which a time delay would apply for real-time reporting.3 In response to a significant number of comments (including those of the ICI), the CFTC determined to replace the original proposal for establishing the 1 The Investment Company Institute is the national association of U.S. investment companies, including mutual funds, closed-end funds, exchange-traded funds (ETFs), and unit investment trusts (UITs). ICI seeks to encourage adherence to high ethical standards, promote public understanding, and otherwise advance the interests of funds, their shareholders, directors, and advisers. Members of ICI manage total assets of $13.4 trillion and serve over 90 million shareholders. 2 Procedures to Establish Appropriate Minimum Block Sizes for Large Notional Off-Facility Swaps and Block Trades, 77 FR 15460 (Mar. 15, 2012), available at http://www.cftc.gov/ucm/groups/public/@lrfederalregister/documents/file/2012- 5950a.pdf (“Reproposal”). 3 Real-Time Public Reporting of Swap Transaction Data, 75 FR 76139 (Dec. 7, 2010) (“Original Real-Time Proposal”). Mr. David A. Stawick May 14, 2012 Page 2 of 9 minimum block trade thresholds with a new methodology based on different swap categories.4 The ICI appreciates that the CFTC intended the Reproposal to provide a more tailored approach, one that would increase the transparency of swap transactions without damaging liquidity in the swap markets. We remain concerned, however, that the proposed swap categories are too broad – grouping swaps with vastly different liquidity profiles together – and that the proposed 67-percent notional amount calculation would result in too high a threshold for block trades. We recommend that the CFTC analyze the swap data that it will receive from repositories to refine further the swap categories that will pool swaps with similar liquidities. Moreover, we respectfully urge the CFTC to adopt the 50-percent notional amount calculation that was suggested by the Commission as an alternative approach and to phase-in this standard over a period of time for very illiquid categories of swaps. We also recommend some modifications to the frequency of the calculation as well as the data used in the calculation. We provide our comments in more detail below. Background and Reproposal Pursuant to Section 2(a)(13) of the Commodity Exchange Act (“CEA”), in December 2010, the CFTC proposed rules to implement real-time reporting of swap transactions as mandated by the Dodd-Frank Act.5 As part of those rules, the CFTC proposed definitions for the terms “large notional off-facility swap” and “block trade,” methods for determining the appropriate minimum block sizes for block trades, and the time delay that would apply to the reporting of block trades. In January 2012, the CFTC adopted rules relating to the reporting of swap data but deferred adoption of rules relating to the determination of block trade sizes.6 The CFTC also adopted time delays for public dissemination of block trades and interim time delays for all swaps until the minimum block sizes are finalized. As noted in our letter on the original proposal (“February 2011 Letter”), block trades enable funds (on behalf of their shareholders) to engage in large transactions with minimal disruptions to the swaps market.7 Counterparties to these transactions, which are typically dealers, are willing to provide this liquidity to funds if the dealers can offset the risks of the resulting positions at a reasonable cost. Knowledge of a block trade to other market participants before a dealer has an adequate time to lay-off those risks could provide an opportunity for those seeking to profit from this knowledge to attempt to extract a higher premium from the dealer to offset those positions. As a result, those offsetting transactions may become more difficult and costly, potentially increasing the costs of market making. This risk may cause dealers to raise the costs of block trades to compensate for the difficulty in hedging 4 In this letter, we refer to “block trades” to include block trades and large notional off-facility swaps as discussed in the Reproposal. 5 Original Real-Time Proposal, supra note 3. 6 Real-Time Public Reporting of Swap Transaction Data, 77 FR 1182 (Jan. 9, 2012) (“Real-Time Reporting Release”). 7 See letter from Karrie McMillan, General Counsel, ICI, to David A. Stawick, Secretary, CFTC, dated Feb. 7, 2011. Mr. David A. Stawick May 14, 2012 Page 3 of 9 their positions in the market. This higher cost, in turn, will be passed on to funds and their shareholders. To avoid this outcome, funds may be forced to break up block trades into smaller orders, which will create market inefficiencies and potentially diminish liquidity. In addition, opportunistic market participants may piece together information about a fund’s holdings or trading strategy, leading to front running of a fund’s trades. The Commission has correctly recognized that the “benefits of enhanced market transparency are not boundless, particularly in swap markets with limited liquidity” and that “‘outsize swap transactions’ could expose counterparties to higher trading costs.”8 With the Reproposal, the Commission seeks to provide maximum transparency while taking into account the impact on market liquidity by proposing detailed criteria to establish swap categories and a phased-in approach to determining appropriate minimum block sizes. In the Reproposal, the CFTC proposes to phase-in the regulation of minimum block sizes in two periods – initial period and post-initial period. The CFTC proposes to establish initial appropriate minimum block sizes for each category of swaps within the interest rate, credit, foreign exchange (“FX”), and other commodity asset classes.9 The proposed initial period would expire following publication by the CFTC of post-initial appropriate minimum block sizes, which would be set no less than once each calendar year. Each swap category in each asset class would be subject to a common appropriate minimum block size. The CFTC proposes to determine the initial and post-initial appropriate minimum block sizes for swaps in the interest rate and credit asset classes by using a 67-percent notional amount calculation. For swaps in the FX asset class, the CFTC proposes in the initial period to set appropriate minimum block sizes based on whether such swap is economically related to a futures contract.10 In the post-initial period, the CFTC would use the 67-percent notional amount calculation to determine appropriate minimum block sizes for each swap category of transactions in the FX asset class. For other commodity swap categories, the CFTC proposes to determine the appropriate minimum block size in 8 Reproposal, supra note 2, at 15466. 9 Proposed Appendix F to part 43 prescribes the initial appropriate minimum block sizes based on these swap categories. 10 Proposed amendments to Section 43.2 would define “economically related” to mean a direct or indirect reference to the same commodity at the same delivery location or locations, or with the same or a substantially similar cash market price series. As an example, the CFTC notes that, a swap utilizing the Standard and Poor’s (‘‘S&P’’) 500 reference price is economically related to the S&P 500 Stock Index futures contract because it is based on the same cash market price series. Reproposal, supra note 2, at 15477 n. 157. The CFTC would establish the appropriate minimum block sizes for futures- related swaps based on the block trade size thresholds set by the designated contract market (“DCM”) for economically- related futures contracts. For non-futures related swaps in the FX asset class, the CFTC would treat all non-futures related swaps in the FX asset class as block trades or large notional off-facility swaps and subject to a time delay. Mr. David A. Stawick May 14, 2012 Page 4 of 9 the initial period depending on the type of other commodity swap categories.11 For the post-initial period appropriate minimum block sizes, the CFTC would use the 67-percent notional amount calculation. No trades, however, of equity swaps would be treated as block trades and, therefore, no equity swap transactions and trades would be subject to a time delay. Swap Categories (Q2., Q9., Q15., Q17., Q21., Q36., Q37., Q38.) For the initial and post-initial periods, the CFTC proposes to group swaps in the asset classes into certain swap categories. We generally support this approach. We support the CFTC’s proposal to establish categories of swaps within different asset classes that would be subject to a common appropriate minimum block size to better calibrate the block thresholds to the relative liquidity of the swap categories in each asset class. In the February 2011 Letter, we expressed concerns that the Commission’s original proposal did not adequately distinguish the risks, trading, and liquidity associated with each individual swap category within an asset class based on type, term, and underlying asset, and that the thresholds for block trades in the swap market should account for the liquidity in each unique category of swaps.12 We appreciate the Commission’s efforts in the Reproposal to address those issues by creating buckets within each swap asset class to which the calculation formula would apply. We, however, continue to have concerns with respect to the categorization of swaps within different asset classes. As an initial matter, only two asset classes – interest rate swaps and credit default swaps – were based on any data and the data set was extremely limited (covering transactions during an approximately three-month period in 2010). For the interest rate and credit default swaps categories, the proposed categories appear to continue to group together swaps of a wide range of liquidities. We understand that the amount of daily trading can vary greatly within a proposed swap category, for example, even for swaps that may have the same tenor and spread. In the Reproposal, the Commission attempted to address two policy objectives with the proposed criteria for distinguishing among swaps: (1) categorizing together swaps with similar quantitative or qualitative characteristics that warrant being subject to the same appropriate minimum block size and (2) minimizing the number of the swap categories within an asset class to avoid unnecessary complexity in the determination process.13 Although we appreciate that the Commission intended to avoid unwarranted complexity, the proposed 11 With respect to swaps that reference or are economically related to one of the futures contracts listed in appendix B to part 43 (other than national gas and electricity swaps), the CFTC would set the appropriate minimum block size based on the block sizes for related futures contracts set by DCMs. If the swaps are not subject to a DCM block trade rule, those transactions would not be treated as a block trade or large notional off-facility swap. All non-futures-related swaps in the other commodity asset class would be treated in the initial period as block trades. 12 See February 2011 Letter, supra note 7. 13 Proposing Release, supra note 2 at 15467. Mr. David A. Stawick May 14, 2012 Page 5 of 9 categories do not go far enough with respect to the first goal and the proposed categories are too broad by grouping swaps with vastly different levels of liquidity.14 In addition, the historical data on which the CFTC relies may not be reflective of the swaps market once the Dodd-Frank requirements (including public reporting of swaps data) are fully implemented. The lack of accurate information and the potentially changing swaps landscape caution against adopting procedures for determining block trade sizes that cannot accommodate the swaps market that may evolve over time. For those asset classes for which no data is available – FX and other commodity asset classes – it is not possible to determine whether the CFTC has identified the most relevant criteria. Therefore, we recommend that the CFTC analyze the data it will receive from swap data repositories to provide more granular categories within the interest rate and credit default swap asset classes to more accurately reflect liquidity and to determine whether further refinements are necessary for the FX and other commodity asset classes.15 Moreover, we disagree with the CFTC’s proposal not to treat any trades in equity swaps as block trades. The Commission is of the view that the existence of a highly liquid underlying cash market, among other factors, supports its approach. At a minimum, we believe that there are differences in liquidity in the underlying equity cash market that would call for closer study of the data on equity swap transactions. We encourage the Commission to propose an appropriate minimum block sizes for equity swaps.16 14 There may be multiple ways by which the CFTC should provide more granularity to the swap categories to reflect appropriately the differing levels of liquidity. Although a study of the swap data that the CFTC will receive should be used to develop these categories, some suggestions include separate categories for different types of CDS or interest rate swap products, narrowing of the tenor groupings, and individual currency groupings for interest rate swaps. 15 The Commission recently recognized the importance of reevaluating the thresholds given the potential future changes in the swaps market. In the context of a de minimis threshold for swap dealers, the CFTC (jointly with the Securities and Exchange Commission) adopted a phase-in period “to consider the information that will be available about the swap market, including real-time public reporting of swap data and information reported to swap data repositories. . . . [and to] afford the Commissions additional time to study the swap markets as they evolve in the new regulatory framework . . . .” Further Definition of “Swap Dealer,” “Security-Based Swap Dealer,” “Major Swap Participant,” “Major Security-Based Swap Participant,” and “Eligible Contract Participant.” RIN 3038-AD06, available at http://www.cftc.gov/ucm/groups/public/@newsroom/documents/file/federalregister041812b.pdf (Apr. 27, 2012) at 148- 149. 16 The interim time delays for all swap transactions (adopted by the CFTC in the Real-Time Reporting Release) should continue to apply to equity swaps until the CFTC adopts appropriate categories of equity swaps and methodology for calculating minimum block sizes. Mr. David A. Stawick May 14, 2012 Page 6 of 9 Notional Amount Calculation (Q29., Q32., Q33.) The CFTC proposes a 67-percent notional amount calculation for determining the minimum block trade size. According to the CFTC, the 67-percent notional amount calculation would result in approximately 94% of trades being reported in real-time. As an alternative, the CFTC requests comment on whether it should use a 50-percent notional amount calculation methodology. We strongly urge the Commission to adopt the 50-percent notional amount calculation methodology with a phase-in period for very illiquid swaps. We agree with the Commission that it should adopt an approach that would “minimize the potential impact of real-time public reporting on liquidity risk” and implement an “incremental approach to achieve real-time price transparency in swap markets.” 17 We believe, however, that the methodology that would better represent a “more tailored and incremental step. . . towards achieving the goal of ‘a vast majority’ of swap transactions becoming subject to real-time public reporting”18 would be the alternative 50-percent notional amount calculation rather than the 67-percent notional amount calculation. We believe that the 67-percent notional amount calculation would set the minimum block thresholds too high, which could cause disruptions in the swap markets. The Commission confirms that the 67-percent notional amount calculation would result in 6% of trades being considered a block trade subject to a time delay. Given the lack of depth and liquidity in the swaps market, we believe this methodology would result in too high of a threshold. As discussed above, too high a threshold will impose real costs to funds and their shareholders. If many large trades do not qualify as “block trades,” liquidity providers will not have adequate time to lay off their risks and will face more difficulty in offsetting their positions. The higher costs to liquidity providers will either result in higher costs being passed on to market participants that engage in large trades or liquidity providers becoming reluctant to act as counterparties to large transactions. Funds and their shareholders will ultimately have to bear these additional costs associated with block trades. Moreover, if funds were forced to break up the block trades into smaller orders instead of paying higher costs, the swaps market may become less efficient and funds may pay higher trading costs. The additional time it will take to complete the execution of the entire block through smaller orders also may reveal the investment strategy of the fund to other market participants who may try to profit from this knowledge to the disadvantage of the fund and its shareholders. Therefore, to minimize the potential negative effects of imposing a threshold for minimum block trades that is too high, we respectfully recommend that the CFTC adopt the 50-percent notional 17 Reproposal, supra note 2, at 15480. 18 Id. Mr. David A. Stawick May 14, 2012 Page 7 of 9 amount calculation.19 Although the 50-percent notional amount calculation may delay for some period of time greater transparency of all swap transactions, a 67-percent notional amount calculation raises the risk of damaging market liquidity severely. Given these trade-offs, we believe it would be prudent to be more cautious rather than rushing to achieve benefits that may not materialize and could harm the swaps markets and market participants.20 In fact, we urge the Commission to phase-in the 50-percent notional amount calculation for certain very illiquid instruments (e.g., swaps with less than 3 or 4 trades per week) over a period of time by first implementing a 25-percent notional amount calculation. We believe that a gradual transition may help alleviate potential harmful effects of disclosure of large block sizes on liquidity on particularly illiquid swaps markets. Frequency of Post-Initial Period Block Calculations (Q30.) The CFTC proposes to set the minimum block sizes no less than once a calendar year and seeks comment on whether this interval would be adequate. The ICI believes that once a calendar year would be too infrequent and that the thresholds must be calculated more regularly to reflect changes in the market. Trade volumes in the swaps market can fluctuate considerably within a short period of time21 and, therefore, more frequent calculations are necessary to track liquidity. We recommend a quarterly or (at a minimum) a semi-annual calculation of the thresholds to ensure that the minimum block sizes continue to be appropriate for each category within an asset class. Data Pool for Determining Block Size (Q32.) For the 67-percent notional amount calculation, the CFTC proposes a nine-step calculation methodology based on three years of data and excluding certain large transactions from the data set.22 19 Even with the 50-percent notional amount calculation, the CFTC estimates that only 14% of interest rate swaps would be considered block trades and only 15% of credit default swaps would be considered block trades. Id. at 15481 n. 198. 20 See An Analysis of OTC Interest Rate Derivative Transactions: Implications for Public Reporting, Staff Report by Federal Reserve Bank of New York (Mar. 2012) (“FRBNY Staff Report”) at 21 (“the level of trading in many instruments is too low and widely distributed to significantly enhance price discovery”). 21 According to the FRBNY Staff Report, even the most commonly traded instruments were not traded with a high degree of frequency. See id. at 3. 22 The nine steps would include: (1) selecting all of the publicly reportable swap transactions within a specific swap category using a rolling three-year window of data beginning with a minimum of one year’s worth of data and adding one year of data for each calculation until a total of three years of data is accumulated; (2) converting to the same currency or units and using a “trimmed data set;” (3) determining the sum of the notional amounts of swaps in the trimmed data set; (4) multiplying the sum of the notional amount by 67 percent; (5) ranking order the observations by notional amount from least to greatest; (6) calculating the cumulative sum of the observations until the cumulative sum is equal to or greater than the 67-percent notional amount calculated in step 4; (7) selecting the notional amount associated with that observation; (8) rounding the notional amount of that observation to two significant digits, or if the notional amount associated with that observation is Mr. David A. Stawick May 14, 2012 Page 8 of 9 The proposed nine-step calculation requires the selection of all of the publicly reportable swap transactions within a specific swap category using a three-year window of data beginning with a minimum of one year’s worth of data and adding one year of data for each calculation until a total of three years of data is accumulated. Given the fact that swaps are not traded with a high degree of frequency, we believe using a three-year window of data may not provide an appropriate data set to calculate the minimum block threshold. We encourage the Commission to look additionally (or alternatively) at a one year set of data and one quarter set of data to determine whether the calculation would produce more accurate results. “Cap Size” Determinations (Q71., Q72., Q74.) To protect the identities of counterparties and business transactions and market positions of counterparties as required by the CEA, the CFTC is proposing several measures, including the establishment of notional cap sizes and requirements to mask the geographic details for swaps in the other commodity asset class. The CFTC proposes to define the term “cap size” as the maximum limit of the principal, notional amount of a swap that would be publicly disseminated. Moreover, the CFTC proposes to establish initial cap sizes for each swap category within the five asset classes, as well as a process for establishing post-initial cap sizes for each swap category. For the initial period, the CFTC would set the cap size for each swap category as the greater of the interim cap size provided in the Real-Time Reporting Release or the appropriate minimum block size for the respective swap category. During the post-initial period, the CFTC proposes to use the 75-percent notional amount calculation (a nine-step calculation methodology). We understand that the Commission seeks to set a threshold for the cap size that is higher than the threshold for minimum block trades because the actual notional amount for a trade that is larger than the cap size would never be publicly reported rather than publicly reported with a time delay. Because we believe that the 67-percent notional amount calculation would result in too high of a threshold for minimum block trades, we similarly believe that the 75-percent notional amount calculation would be too high for determining the cap size. Given the lack of depth and liquidity in the swaps market, requiring principal amounts that do not meet the Commission’s proposed cap size to be publicly reported may reveal the identities, business transactions, and market positions of market participants engaging in large transactions. Accordingly, we recommend that the Commission adopt a 67-percent notional amount calculation for determining cap size in the post-initial period. already significant to two digits, increasing that notional amount to the next highest rounding point of two significant digits; and (9) setting the appropriate minimum block size at the amount calculated in step 8. Mr. David A. Stawick May 14, 2012 Page 9 of 9 * * * If you have any questions on our comment letter, please feel free to contact me at (202) 326- 5815, Sarah Bessin at (202) 326-5835, or Jennifer Choi at (202) 326-5876. Sincerely, /s/ Karrie McMillan Karrie McMillan General Counsel cc: The Honorable Gary Gensler The Honorable Jill E. Sommers The Honorable Bart Chilton The Honorable Scott D. O’ Malia The Honorable Mark Wetjen