SPEECHES & TESTIMONY

Statement of Concurrence of Commissioner Dawn D. Stump on Proposed Rule Amendments to The Commission’s Regulations Relating to Certain Swap Data Repository and Data Reporting Requirements

April 25, 2019

The Commission is publishing for public comment “Proposed Amendments to the Commission’s Regulations Relating to Certain Swap Data Repository and Swap Data Reporting Requirements” (Proposal).  Accurate swap data reporting is vital to our ability to make appropriate policy choices.  I very much look forward to receiving feedback from all parties impacted by this Proposal to assure that the Commission has robust and accurate data, which is a lynchpin of future Commission decision-making.

However, a Latin proverb reads: Qui tacet consentire videtur, ubi loqui debuit ac potuit (he who is silent, when he ought to have spoken and was able to, is taken to agree).  While I share the Commission’s desire for accurate swap data, I do not agree with all the policy and procedural choices in this Proposal. I question certain of the underlying assumptions driving these policy changes, and the promulgation of this rulemaking in isolation and without corresponding changes to other swap data reporting rules.  I am uncomfortable with the lack of details and nebulous description of certain obligations in many parts of the Proposal, which I believe will make it difficult for the public to comment in an informed fashion. And I disagree with imposing immense additional burdens on swap data repositories (SDRs) and all types of reporting counterparties (RCPs), particularly without commensurate streamlining of regulatory obligations in the rest of the Commission’s swap data reporting rule set.

Because I share the Commission’s ultimate goal of accurate swaps data, I support the Proposal going out for comment, with the caveat that the other aspects of the swaps data “Roadmap”[1] are published in quick succession.  I look forward to feedback from all interested parties as to how that goal can best be achieved in light of my concerns about the Proposal discussed below and other options that may be at the Commission’s disposal to enhance data accuracy while appropriately balancing costs and benefits.

  1. Verification: Solution in Search of a Problem?

This Proposal is predicated upon a view that new verification procedures are needed because the swap data currently being reported to SDRs is substantially wrong and inaccurate.  Yet, the Commission has recently proffered positive reviews of the role of SDR data in enhancing its understanding of swaps markets, citing the “more complete information now available regarding certain portions of the swap market, [and] the data analytical capabilities developed since the [swap dealer] regulations were adopted”[2] as supporting its policy decision making.  Specifically, the Commission cited analysis based upon a year of SDR data sourced from data reported to the registered SDRs in its recent rulemaking concerning the de minimis exception to the swap dealer definition relating to insured depository institutions (IDIs).[3]  Given that the Commission has not voiced concern about widespread discrepancies or inaccuracies in swaps data reported to SDRs in relying upon that data in our rulemakings, I am not convinced that it is necessary to add new layers of complexity to swaps data reporting and create new burdens on market participants via the steps outlined in the Proposal.

Taken in isolation, asking RCPs to verify the accuracy of data reported to SDRs is appealing.  But how does the Commission know that a substantial portion of that data is actually incorrect?  The Proposal attempts to depict a data accuracy problem by referencing that it is not uncommon for discrepancies to be found in SDR data. However, from the universe of reported swap data that contains millions of swap transactions and exponentially more messages sent to SDRs over the course of the last five years, the Proposal mentions only two examples of errors:  “In the processing of swap data to generate the CFTC’s Weekly Swaps Report, for example, there are instances when the notional amount differs between the Commission’s open swaps information and the swap data reported for the same swap.  Other common examples of discrepancies include incorrect references to an underlying currency, such as a notional value incorrectly linked to U.S. dollars instead of Japanese Yen.”[4]  I would expect a much more extensive and egregious list of systemic, recurring errors in reported swaps data to warrant the expansive new obligations contained in the Proposal.

The Proposal strains to quantify the number of inaccuracies in reported SDR data by opining that, “[b]ased on swap data available to the Commission and discussions with the SDRs, the Commission estimates that an SDR would perform an average of approximately 2,652,000 data corrections per year.”[5]  The Proposal does not explain exactly how this figure was derived, identify the interaction between SDRs and RCPs referenced in its corrections estimate, indicate whether the “correction” refers to incomplete or inaccurate data,[6] or provide critical context as to the percentage of messages that this number represents. Indeed, it is impossible to know for certain that an RCP was intending to correct erroneously reported data based on the data schema utilized by SDRs to address changes in swaps data – which include actions such as “snapshot,” “amendment,” and “modify,”[7] but may not actually include a category of “correction” messages.

While the Proposal posits the annual number of corrections across all SDRs to be about 8 million “corrections” (3 provisionally registered SDRs * 2,652,000 annual data corrections per SDR), it lacks the total number of data submissions that are received by the SDRs.  The Paperwork Reduction Act portion of the Proposal does provide one potentially related data point, as it includes an estimate of 462,981,508 total annual responses across all SDRs for the relevant information collection.[8]  Without the benefit of further clarity, the corrections could apply to the entire universe of the collections associated with the Proposal.  If the figures are roughly rounded for the sake of simplicity, and it is stipulated for the sake of argument that all the corrections cited by the Proposal reveal data inaccuracies, then does this suggest that only approximately 2% (400 million responses / 8 million corrections) of all messages might be inaccurate?  In my opinion, the burdens that this Proposal would impose on SDRs and RCPs (including commercial end users) may be difficult to justify if the problem the Commission is attempting to rectify may equate to 2% of all messages delivered to SDRs.

I share the view that has been stated by some of my colleagues recently that the Commission should strive to make data-driven policy determinations and should avoid relying on assumptions or anecdotes when engaged in rulemaking activity.[9]  Yet, the same is true when it comes to imposing costs and burdens on market participants that are already heavily encumbered by a broad swath of regulatory obligations that continue to shift and expand.  Our recent rulemakings have referenced data driven policy making, learning from experience with Dodd-Frank implementation, and demonstrating supporting evidence for regulatory change, but the verification provisions of this Proposal deviate from that approach.  The Commission should delay this rulemaking until the other aspects of the Roadmap critical to improving swaps data reporting and lessening unnecessary regulatory burdens were ready to be proposed. But, short of that, I welcome public comment and data evaluating the breadth and depth of inaccuracies in SDR data.[10]  Such information would help to determine how much reported SDR data is actually incorrect before the Commission requires SDRs and RCPs to build additional systems and undertake significant new compliance burdens and obligations to address an accuracy problem that, at this point, has not been proved. I look forward to comments and data that demonstrate the actual need for the proposed changes.

  1. Insufficient Level of Detail for Appropriate Public Comment and Cost-Benefit Consideration

The Administrative Procedure Act (APA) requires that, in issuing its rules, the Commission “examine the relevant data and articulate a satisfactory explanation for its action including a rational connection between the facts found and the choices made.”[11]  Section 15(a) of the Commodity Exchange Act (CEA) further requires that in doing so, the Commission must consider the costs and benefits of its proposed action.[12]  A notice of proposed rulemaking affords the Commission the opportunity to gather information and build a record that will provide the reasons for the conclusions that it ultimately draws when final rules are issued.  If the Commission fails to properly exercise this responsibility, we risk having our rules set aside as arbitrary and capricious agency action.[13]

While I support the purposes and intent underlying the Proposal, I am concerned that some of the proposed rules are too vague to enable the public to provide the Commission with information necessary to adopt a sound final rule set.  For RCPs, the Proposal informs them of their general obligations, but leaves a tremendous amount of the details to future action by the Commission (often delegated to staff) and the SDRs to dictate the operational work flows that RCPs will have to adhere to in order to comply with the Commission’s rules.  RCPs reading the proposed rules still would not know what changes are being proposed in what they have to report, when they must report by, and how they are to deliver that information to SDRs.  The proposed rules are often amorphous, lacking specificity as to the actual processes and procedures to be imposed, with RCPs left to comment without really knowing what much of this would actually require of them in the future.

The same is true for SDRs. For example, proposed § 49.9 covering open swaps reports to be provided to the Commission is quite opaque, and provides no detail as to any potential future instructions from the Commission that “may include, but are not limited to, the method, timing, and frequency of transmission as well as the format of the swap data to be transmitted.”  Similarly, proposed § 49.17(c)(1) would require an SDR to transmit all swap data requested by the Commission, but provides that the SDR will receive instructions that may include, but are not limited to, the method, timing, and frequency of transmission, and the format and scope of the SDR data to be transmitted, at a later time.

How can RCPs and SDRs prepare for, budget, build, test, and implement systems to comply with these requirements without ample information ahead of time as to what these requirements entail?  Indeed, it is not clear to me how RCPs and SDRs can even meaningfully comment on either the merits or the costs and benefits of the proposed rules when these critical elements of the requirements are left for future determination.

But the proposed rule that troubles me most in this regard is proposed § 49.13, which addresses an SDR’s duty to monitor, screen, and analyze data upon the request of the Commission.  The Proposal explains that in its original consideration of current Regulation 49.13,[14] the Commission received comments that the rule does not sufficiently describe the specific tasks that SDRs are expected to perform.  The Commission decided to later establish specific monitoring, screening, and analyzing duties when its knowledge was more fully developed, and that is where we find ourselves presently.  Yet, despite the Commission’s experience with swaps data over the last five plus years, this Proposal still fails to delineate specific duties that would enable an SDR to provide appropriate budget, technological development, and staff resources to assure an ability to comply with the demands that may be made upon it.

Proposed § 49.13(a)(1) requires SDRs to be prepared to comply with Commission requests for monitoring, screening, and analyzing of data. Several of the tasks alluded to in the proposal rule could impose significant, albeit wholly undefined, obligations on SDRs. For example, proposed § 49.13(a)(1)(iv) contemplates assessments of risk, which is not particularly an SDR function and which can be a very complicated exercise that is defined and calculated differently by different market participants.  Proposed § 49.13(a)(1)(viii) would appear to render SDRs an arm of the Commission’s enforcement program, as it would require them to provide information about compliance with Commission regulations without clarifying how SDRs could do so, and despite the fact that SDRs are not self-regulatory organizations.

Proposed § 49.13(b), in turn, requires SDRs to “establish and at all times maintain sufficient information technology, staff, and other resources to fulfill” these Commission requests.  Yet, proposed § 49.13(a)(2) provides that the content, scope, and frequency of all monitoring, screening, and analyzing requests shall be at the discretion of the Commission (to be exercised by staff pursuant to delegated authority); further, in addition to the 11 types of potential Commission requests identified in the proposal, SDRs also would have to be prepared to comply with other, unspecified, types of requests for monitoring, screening, and analyzing as well.  How can an SDR be expected to efficiently allocate capital and meet the standards of proposed §49.13(b) with respect to information technology, staff, and “other” (undefined) resources when it does not know what the actual requirements will be, when it will be expected to deliver, at what frequency, and the exact form and manner of the deliverable?

Finally, proposed § 49.30 would mandate that “a swap data repository shall submit SDR data reports and any other information required under this part to the Commission, within the time specified, using the format, coding structure, and electronic data transmission procedures approved in writing by the Commission.”  I cannot begin to fathom the uncomfortable (and unenviable) position of an SDR under rules whereby the Commission can ask for almost anything under proposed § 49.13, and then demand its submission whenever and however it wishes under proposed § 49.30.

The Proposal states, somewhat incredibly, that it “expects specifying these topic areas [in proposed § 49.13] would not impose substantial new fixed costs on SDRs. . .”[15] It is wishful thinking to claim that the extensive list of undefined, open-ended tasks hypothesized in proposed § 49.13(a)(1) that SDRs must prepare to build and deliver will not represent a meaningful burden.  Although it is not clear how SDRs could quantify the costs of compliance with such vague obligations, it is likely that the costs incurred by SDRs will be significant – and that their clients, including commercial end-users, ultimately will pay the price.

I appreciate that it is not possible to foresee all future circumstances when proposing a rulemaking, and I recognize the need for flexibility in aspects of the Commission’s day-to-day administration of the Dodd-Frank swap regulatory regime. Nevertheless, I am concerned that the Proposal fails to inform the public as to the full nature of the responsibilities that the Commission intends to impose upon RCPs and SDRs so that they can provide appropriate comment and feedback to drive the best final rule outcome possible.  I wonder how the Commission can produce a complete cost-benefit consideration without specifying the actual scope and technical details of the requirements it is proposing to impose, particularly with respect to requests to SDRs to be made via proposed § 49.13.  In sum, I fear that in proposing several rules where critical elements are left for future specification (often by staff), the Commission will not receive informed and meaningful public comments (including comments on costs and benefits) that are necessary to provide the foundation on which our rules ultimately must rest.

  1. Suboptimal Policy Choices

Certain elements of the Proposal rest on questionable policy choices that I wish to highlight in order to garner public input as part of the comment process.

First, the Proposal would remove a longstanding market practice of trusted sources when it comes to verification of data accuracy without demonstrating why such a change is necessary, or appropriate.  The Proposal states:  “The Commission provided an exception to the requirement that SDRs ‘confirm with both counterparties to the swap the accuracy of the data that was submitted’ in § 49.11(b)(1)(ii) for swap creation data and § 49.11(b)(2)(ii) for swap continuation data when swap data is received from a [swap execution facility, or ‘SEF’], [designated contract market, or ‘DCM’], derivatives clearing organization (‘DCO’), or from a third-party service provider acting on behalf of the swap counterparty, under certain conditions.”[16]  The Proposal’s departure from this policy means that SDRs would no longer be able to rely on an exception from the requirement to affirmatively confirm with both counterparties where (1) the SDR forms a reasonable belief that the data is accurate, (2) the reporting identifies that both counterparties agreed to the data submitted, and (3) the SDR provides both counterparties with a 48-hour correction window.

The Proposal argues, without citing any evidence, that, “based on the Commission’s experience with swap data submitted by SEFs, DCMs, DCOs, and third-party service providers since the rule was adopted, the Commission believes that such swap data has not been consistently complete and accurate in some instances, and the swap data accuracy is not sufficient to justify the exception to the requirement that SDRs confirm the reported swap data’s accuracy with swap counterparties.  The current requirements have had a negative effect on swap data accuracy and consistency, which has hampered the Commission’s ability to carry out its regulatory responsibilities.”[17]  I do not believe that trading venues, which value execution certainty and must deliver accurate trade details to clients, or clearing organizations, which must have verified trade details available for risk management purposes, would report systematically or consistently inaccurate swaps data to SDRs, given their level of technological expertise and concern for reputational risk.  At a minimum, I would not eliminate the existing exception absent evidence establishing that this is the case.

Second, the Proposal would mandate in proposed §§ 43.3(e) and 45.14(b) that corrections of errors and omissions be performed by SEFs, DCMs, and RCPs “regardless of the state of the swap that is the subject of the swap data.”  The Proposal defines an “open swap” as “an executed swap transaction that has not reached maturity or the final contractual settlement date, and has not been exercised, closed out, or terminated.”  Thus, the Proposal is requiring additional reporting for “dead” swaps without demonstrating a relevant use-case to warrant such a requirement.

It is more difficult for RCPs to correct dead/expired swaps that are no longer on their books and records.  SDRs also face additional challenges and complexity in modifying swaps that are no longer what the Proposal defines as an “open swap.”  The Proposal does not identify a Commission or public use-case that justifies the increased burden and challenge associated with correcting data on dead/expired swaps.  The financial crisis that precipitated Dodd-Frank was not caused by, nor could it have been prevented by, regulatory oversight of dead swaps, but rather was the result of active risk.  Again, absent an identified justification with evidentiary support, I do not support imposing additional regulatory burdens that force market participants to shift resources from the management of active risks to the reporting of dead swaps.

Third, I would prefer a more sensible approach to the duration of the recordkeeping requirements for SDRs. Proposed § 49.12(b)(2) would require SDR records -- including SDR data, timestamps, and messages -- to be readily accessible following final termination of the swap for five years, and then for a period of ten additional years in archival storage, which, of course, has an associated cost.  Unless the Commission can clearly articulate the use-case and regulatory purpose that would justify requiring archival storage up to 15 years after the expiration of the swap, I believe the Commission should consider reducing the recordkeeping time frame for SDRs.

  1. Process Foul to Address Only One Aspect of the Complex Swap Data Reporting Puzzle

I also am uncomfortable with the sequencing of this Proposal and the rush to publication on a stand-alone basis rather than as part of the contemplated overhaul of all the swaps data reporting rules.

I expressed a similar view about the application of a holistic approach to interrelated regulations during last November’s Open Meeting concerning SEFs when I noted that “I would prefer that the Commission be able to opine on a final SEF rule and a final rule on name give-up at the same time.  Acting on all aspects impacting SEF trading contemporaneously would benefit all entities involved.”[18]  The same principles apply to swap data reporting, as both the public and the Commission would benefit from holistically addressing the entirety of the swap data reporting universe.  Unfortunately, the Commission continues to propose regulations that are interrelated and that would govern the same activity in an inefficient, piecemeal manner.

Swap data reporting is a complex web of interrelated processes and systems that must all work in sync in order to generate complete and accurate data in a timely and cost effective manner.  Many tasks in reporting are sequential in nature, and it takes all participants in the reporting ecosystem to coordinate and cooperate with a complete understanding of all the swap data reporting regulations from the Commission.  For example, SDRs have to scope out and create policies and procedures and build systems/templates for any new requirement.  RCPs cannot adequately prepare for, much less build and test, systems on how to comply until they receive final feedback and instructions from the SDR.  For this reason, implementing reporting changes – which invariably is quite costly to both SDRs and RCPs in terms of the expenditure of time, energy, and money – must be orchestrated and timed very carefully.

SDRs and RCPs have previously expressed to the Commission the importance of being made aware of anticipated future modifications to reporting so that they can understand the expected end-game that the Commission has in mind.[19]  Market participants also have commented on the need to understand the entire policy idea and all the associated pieces before committing time and energy to provide the Commission with meaningful comments and input.[20]

I appreciate that the Proposal states that “[w]hen the Commission proposes the next two rulemakings, the Commission anticipates re-opening the comment period for this proposal to provide market participants with an opportunity to comment collectively on the three rulemakings together, because the proposals address interconnected issues.”[21]  But I do not see the benefit of proceeding in such an inefficient manner. Issuing the Proposal now does provide notice of the Commission’s intentions with respect to one piece of the swaps data Roadmap, but no notice of what else from the Roadmap might come to pass.  Such “partial notice” does not enable parties to evaluate, and comment upon, the full picture of their new compliance obligations, including their costs and burdens.[22] Under these circumstances, I would not be surprised if market participants simply waited for all of the reporting rules to be proposed before providing feedback to the Commission on the whole of what is being proposed.

In addition, if, as the Proposal suggests, there actually is a significant problem with inaccurate swap data being reported to SDRs, the piecemeal issuance of these rulemakings makes it more difficult for the Commission to evaluate whether that problem can be rectified by allowing other facets of the swaps data Roadmap to gain traction.  Query whether the Commission generating a technical specification removing uncertainty as to what must be reported and how, harmonizing with other regulators and implementing unique identifiers (Unique Transaction Identifiers and Unique Product Identifiers) and critical data elements from CPMI-IOSCO work streams, minimizing the number of fields required to be reported, and affording RCPs more time to report would organically resolve a large proportion of any inaccurate data reporting problem that may exist.  The manner in which the Commission has elected to proceed will make it challenging for SDRs and RCPs to comment appropriately on these questions, and I fear will place the Commission in a predicament as it attempts to make informed policy decisions on how best to proceed.

  1. Lack of Harmonization with the SEC

Market participants of all shapes-and-sizes – even those that are often on opposing sides of most regulatory debates – all agree on a common theme that has been repeatedly urged upon the Commission via every imaginable medium since the enactment of Dodd-Frank:  The Commission and the Securities and Exchange Commission (SEC) should coordinate and harmonize their respective derivatives regulations to the maximum extent possible, and especially concerning entities that have already incurred systems and compliance costs in connection with the corresponding requirements of the related agency.  All types of market participants have implored both the Commission and the SEC to minimize compliance burdens on potential dual registrants in connection with the derivatives rules, such as swap data reporting.  And yet, notwithstanding the current emphasis on CFTC-SEC harmonization,[23] the Commission is proposing a swap data reporting rule that appears to take an approach that is the opposite of, and in direct contrast to, the SEC’s thinking on the same issue.

The SEC published a proposed rulemaking in December 2018[24] that specifically discusses, among other things, verification of the terms of reported security-based swaps – as does the Proposal.  Yet, while the Proposal would increase regulatory burdens on all entities in its amended regulatory reporting scheme, the SEC is considering a more pragmatic approach.  The SEC, in its proposal, “believes it to be an appropriate time to revisit and request comment on an issue previously identified in connection with the rules . . . [that] require[] each registered SDR to ‘confirm with both counterparties to the security-based swap the accuracy of the data that was submitted.’”[25]

Specifically, the SEC in its proposal states that “SDRs may be able to reasonably rely on certain third parties to address the accuracy of the transaction data.  For example, the Commission previously stated that if an SDR develops reasonable policies and procedures that rely on confirmations completed by another entity, such as a third-party confirmation provider, as long as such reliance is reasonable the SDR could use such confirmation to fulfill its obligations under certain SDR rules.  Because the two relevant provisions that we are proposing today generally relate to the obligation of [Security-Based Swap, or ‘SBS’] Entities to take certain steps in the reconciliation and documentation processes related specifically to the reporting of the relevant security-based swap data to an SDR…the Commission believes that...these measures, taken together, could provide an SDR with a set of factors to assess the reasonableness of relying on an SBS Entity’s ability to independently provide the definitive report of a given security-based swap position, thereby providing a basis for the SDR to satisfy its statutory and regulatory obligations to verify the accuracy of the reported data when the SBS Entity’s counterparty is not a member of the SDR.”[26]

In other words, the SEC is considering whether the reconciliation process undertaken by security-based swap dealers of their swaps portfolios could satisfy the statutory obligation to confirm the accuracy of data reported to SDRs.  This sensible approach being considered demonstrates deference to trusted sources for swap data accuracy when a third-party service provider is employed to address the confirmation of swaps data, similar to the exceptions in Regulations 49.11(b)(1)(ii) and 49.11(b)(2)(ii) that the Proposal would eliminate.

As discussed more fully in Section VI below, based on the Commission’s reporting hierarchy in Regulation 45.8,[27] swap dealers (SDs) are the RCP and transmit required swap data elements to an SDR for the vast preponderance of swap transactions.  These same SDs are already subject to another regulatory obligation relating to verification of the terms of their swap transactions, as they must conduct a portfolio reconciliation exercise on a regularly recurring basis via Regulation 23.502.[28]  Portfolio reconciliation forces the “[e]xchange [of] the material terms of all swaps in the swap portfolio between the counterparties” and requires the parties to “[r]esolve any discrepancy in material terms and valuations”.[29]  Since SDs already must check the accuracy of their portfolios through a reconciliation exercise, and since SDs report almost all swaps, then the Commission, like the SEC, should consider leveraging this existing process and afford SDs that undertake such an exercise enough time for it to run its course and then submit that same accurate and verified data set for SDR reporting purposes.  Leveraging this existing regulatory process, rather than creating yet another process that compliance officers and operations staff must adhere to, may offer a “good government” solution, assuming the existence of a systemic problem with SDR data accuracy.  If SDs represent that the same data reconciled with counterparties per Rule 23.502 is reported to SDRs, then the Commission might not need to impose the burdensome new requirements set out in the Proposal.

It is unfortunate that the Commission did not propose – or even request comment on – the less burdensome approach to verification that the SEC is considering in light of our stated commitment to harmonizing the agencies’ derivatives rules.  And it is even more mystifying to me why we are proposing these rule amendments in the inefficient, piecemeal manner described above when delaying the issuance of this Proposal would not only enable us to issue the various proposed amendments to our swap data reporting rules as a unified package, but also to learn from comments on the SEC’s data verification discussion (the comment period closed on April 16) whether the SEC may have identified a better option for fostering accurate reported swaps data.

  1. Outsized Burden Placed Upon SDRs and RCPs, Including End-Users

Swap market participants have repeatedly emphasized to the Commission that the swap data reporting rules are overly complicated, difficult to implement, and a significant operational burden and compliance challenge for all concerned, including end-users.[30]  Yet, the Proposal would add more layers of complexity to reporting workflows, and require SDRs and RCPs to commit more time and money to submit more reports and undertake additional obligations.

In particular, the Commission has heard from many end-users about the immense nature of their reporting burdens, how regulatory capture on end-users has impacted their business models and their ability to hedge via derivatives markets, and the unintended consequences of the initial implementation of the Dodd-Frank swap reporting regime.  In response, the Commission, commendably, has made considerable progress in addressing reporting issues and limiting burdens on end-users via the various tools at our disposal when consistent with our regulatory responsibilities.  It is not clear to me why this Proposal would break from those efforts and go in the opposite direction by placing new and burdensome swap data reporting obligations on end-users.

End-user RCPs would bear several onerous obligations under this Proposal.  End-user RCPs would have to commit considerable resources to create more sophisticated and elaborate reporting systems in order to be compliant.  The Proposal estimates that 1,585 RCPs are neither SDs, major swap participants (MSPs), nor DCOs.[31] As a result of the Proposal, all of these end-user RCPs would have to acquire or build additional processes and hire more staff to comply with these new reporting regulations, regardless of the number, notional amount, asset class, or risk profile of the swaps for which they are the RCP.  To provide some perspective, staff has indicated that of new transactions in January 2019, trades with at least one SD counterparty (which would serve as the RCP) per asset class represented 99.6183% of the 22,446 CDS trades; 98.2466% of the 137,499 IRS trades; 97.0540% of the 603,696 FX trades; 99.9998% of the 471,657 Equity trades; and 85.3056% of the 60,021 Commodity trades.  In other words, the 1,585 RCPs that are not SDs, MSPs, or DCOs reported, at most, 86 CDS, 2,454 IRS, 18,325 FX, 1 Equity, and 10,339 Commodity swaps during this time period.  Given the limited number of swaps for which end-users are RCPs compared to the overall swaps market, I question whether imposing on all end-users that may serve as an RCP the additional burdens of preparing for compliance with the requirements of this Proposal reflects an appropriate consideration of costs and benefits.

The Commission has made strides post the initial roll-out of its Dodd-Frank rulemakings to fix unintended consequences of its swap data reporting rules and minimize the burdens on end-users where appropriate.  This Proposal, unfortunately, errs in the other direction.  I welcome suggestions via the public comment process on the appropriate role for end-user RCPs to play in assuring the accuracy of reported swap data short of imposing the burdens set out in the Proposal.

  1. Alternate Approaches for Further Consideration

To be clear, my concern with the Proposal is not simply that it would impose costs on market participants; all necessary regulatory requirements do so.  Rather, my concern is with the extent of the burdens that the Proposal would impose on market participants, including end-users, in light of the prospects that the Proposal will meaningfully improve the quality of reported swap data.  As discussed above, the Proposal does not establish that there actually is a systemic problem in that regard.  But assuming that to be the case, consider the following fact pattern and whether any errors would be found and rectified under the Proposal:

  • RCP submits data to an SDR from its regulatory reporting databases;
  • SDR creates Open Swaps reports based upon the data received;
  • SDR provides a mechanism for the RCP to verify the accuracy of the Open Swaps report; and
  • RCP checks the Open Swaps report against the data that it submitted to the SDR.

In other words, if the original data set utilized by the RCP contains an inaccuracy, the Proposal could simply impose a futile exercise based on circular logic.  The end result of the new burdens placed upon RCPs and SDRs would merely be a false positive in this scenario.  If the RCP’s data is inaccurate in the first place, then the Proposal would be successful only in making swap data reporting more complicated and expensive, without actually improving the accuracy of the data reported to the SDR.[32]

Accurately reported swap data is, of course, crucial to the Commission’s performance of its regulatory responsibilities and the effective operation of the Dodd-Frank swap regime.  That is why I am concurring in the issuance of the Proposal – because I support the Commission’s efforts to determine whether appropriate improvements can be made to its swap data reporting rule set.

This Proposal provides an opportunity for the public to suggest other, perhaps better, solutions to more efficiently produce the desired outcome of accurate swap data for purposes of conducting the Commission’s work, facilitating risk oversight and management, and fostering robust swaps markets.  I strongly encourage SDRs, SDs, DCOs, end-users, and the public in general to take advantage of this opportunity and provide not only feedback on the Proposal, but also their ideas on how to appropriately balance the need for accurately reported swap data with the costs and burdens associated with obtaining it.  The Commission should consider any alternate approaches that can satisfy the policy goal of improving the quality of SDR data while limiting the impact on market participants already saturated with complex and repetitive reporting obligations.

I would like to offer, and invite comment on, a few alternatives with respect to RCPs. CEA Section 21(c)(2) provides that SDRs shall “confirm with both counterparties to the swap the accuracy of the data that was submitted.”[33]  As a result, a clear obligation exists as to what SDRs must do. The statute is less clear on what RCPs must do, if anything.

Under the Commission’s current regulations, all RCPs must submit hundreds of fields per transaction to their respective SDRs.  Some RCPs have thousands of open swaps that would be captured under this Proposal and require recurring verification.  I hope that commenters will address whether a smaller number of swaps and/or a limited subset of essential fields that must be verified would enable the Commission to conduct its regulatory functions without indiscriminately requiring verification of all swap data elements.

Another option on which public comment would be helpful is requiring RCPs to verify only the accuracy of a statistically significant portion of their Open Swaps report and then decide, based on the level of accuracy, whether the entirety of Open Swaps must be analyzed.  Still another option might be to require verification of only a limited set of the most important fields required to understand the basic terms of plain-vanilla swap transactions.  Finally, commenters could address a possible de minimis level that must be exceeded before the new reporting obligations in this Proposal would apply.  For example, if an RCP has less than X swaps per year, or less than Y notional transacted per year, then it would not have to perform these verification functions.

With respect to end-user RCPs in particular, where the ability to build reporting systems and the cost of doing so on a per swap basis is much different than for SDs, MSPs, and DCOs, comment would be beneficial on whether end-user RCPs should have more time than proposed, both for replying to Open Swaps reports with a “verification” or “notification of discrepancy” message and correcting errors and omissions.  Also, commenters may wish to address the frequency of how often end-user RCPs should be required to participate in this labor-intensive process.  I recognize that the Proposal includes less stringent obligations on end-user RCPs in comparison to SDs, MSPs, and DCOs that are RCPs, but I welcome comment on whether the Commission should strive to do more in this regard.

As written, the Proposal would impose a number of new, often undefined, obligations with respect to swap data reporting.  The potential alternatives noted above, together with others that commenters may suggest, could represent a common sense approach to addressing concerns regarding swap data accuracy while appropriately calibrating the costs and burdens associated with verification of SDR data.

 


 

[1] See Roadmap to Achieve High Quality Swaps Data (DMO July 10, 2017), available at https://www.cftc.gov/sites/default/files/idc/groups/public/@newsroom/documents/file/dmo_swapdataplan071017.pdf, published with CFTC Letter 17-33, Division of Market Oversight Announces Review of Swap Reporting Rules in Parts 43, 45, and 49 of Commission Regulations (DMO July 10, 2017), available at https://www.cftc.gov/sites/default/files/idc/groups/public/@lrlettergeneral/documents/letter/17-33.pdf.

[2] De Minimis Exception to the Swap Dealer Definition -- Swaps Entered Into by Insured Depository Institutions in Connection With Loans to Customers, 81 FR 12450, 12452 (April 1, 2019) (IDI De Minimis Rulemaking).

[3] Id. at 12454 and n.59 (“The Commission believes that end-users would primarily benefit from the IDI De Minimis Provision by entering into [interest rate swaps, or ‘IRS’], [foreign exchange, or ‘FX’] swaps, and [non-financial commodity, or ‘NFC’] swaps with IDIs to hedge loan-related risks. SDR data indicates that IDIs that have between $1 billion and $50 billion in [aggregate gross notional amount, or ‘AGNA’] of swaps activity primarily enter into IRS, FX swaps, and NFC swaps, as measured by AGNA and transaction count.

[4] Proposal, text accompanying n.239.

[5] Id., at section VII.B.3.v.

[6] Incomplete data is not the same things as inaccurate data. Thus, “corrections” of incomplete data would not be relevant to the verification with respect to inaccurate data that is the subject of this Proposal.

[7] DTCC SDR templates, for instance, include the following message and action types. The modify action type allows for the valid modification or correction to an existing trade that has previously been reported by the submitting party. However, firms could reflect a correction using other methods. The snapshot message allows participants to report the current state of the swap in their portfolio as a “point-in-time” view of the position. The reported position should reflect all post-trade events and non-position forming amendments that the submitter may wish to reflect on their trade record. The amendment transaction type could be utilized as an indication of a confirmable amendment, via a negotiated agreement, to a previously confirmed and reported trade. As a result, it would be difficult to conclude with any certainty the actual number of corrections without a critical review of contrasting terms related to a particular trade on each type of action, message, or transaction type submission. 

[8] Proposal at section VII.B.3.xi.

[9] See, e.g., IDI De Minimis Rulemaking at 12467 (Statement of Chairman J. Christopher Giancarlo) (“As I have said many times before, I believe that CFTC policy is best when it is driven by data and not assumptions.”).

[10] The cost-benefit consideration in the Proposal loosely references and mischaracterizes information contained in three public studies that allude to challenges in SDR data. Unfortunately, these studies are from 2015 or earlier and are based upon data from the initial roll-out of SDR reporting. These studies address incomplete rather than inaccurate data and do not belong in this Proposal that focuses on verification of data. See fn. 6, supra.  The Roadmap explained that validations should be utilized to reject swap data reports with missing data fields, and these issues would be better served by a holistic implementation of the Roadmap and do not require the onerous verification aspects of the Proposal. Furthermore, some of these identified issues also would be resolved by the technical specification detailed in the Roadmap and, again, if proposed in unison, would provide RCPs with clear definition, form and manner, and allowable values. The reference to the third study also fails to mention that the two soybean swaps referred to were removed from a universe of 39,622 agricultural swaps.

[11] Motor Vehicle Mfrs. Ass’n of U.S., Inc. v. State Farm Mut. Auto Ins. Co., 463 U.S. 29, 43 (1983).

[12] 7 U.S.C. 19(a).

[13] See APA, 7 U.S.C. 706(2)(A).

[14] 17 CFR 49.13.

[15] Proposal at section II.I.

[16] Id., text accompanying n.70.

[17] Id., text immediately following n.73.

[18] See Opening Statement of Commissioner Dawn D. Stump before the CFTC Open Meeting, November 5, 2018, available at https://www.cftc.gov/PressRoom/SpeechesTestimony/stumpstatement110518.

[19] In late 2015, CFTC staff issued a request for comment on draft technical specifications for certain prioritized swap data elements and sought input on 80 enumerated questions addressing 120 data elements for several swap data reporting topics.  See Draft Technical Specifications for Certain Swap Data Elements (December 22, 2015), available at https://www.cftc.gov/sites/default/files/idc/groups/public/@newsroom/documents/file/specificationsswapdata122215.pdf and https://www.cftc.gov/PressRoom/PressReleases/pr7298-15. In responding to staff’s request for comment, SIFMA stated that it “view[s] the Draft Technical Specifications as one component of a broader initiative to enhance swap data reporting” and that the “interrelationships among the Draft Technical Specifications and these other workstreams, as well as their shared dependencies on the same technology and human resources, necessitate a well-planned and sequenced approach to enhancing swap data reporting requirements. Prioritizing among the various enhancements under consideration will help to avoid inadvertent inconsistencies and associated potential for erroneous data and unnecessary infrastructure costs.” Letter from Kyle Brandon, SIFMA, at 2 (March 7, 2016), available at https://comments.cftc.gov/PublicComments/ViewComment.aspx?id=60702&SearchText=.

[20] SIFMA and ISDA jointly commented on the swaps data Roadmap and suggested that the Commission align the anticipated timeframes for swaps data reporting changes: “[G]iven the interconnection between SDR functions and the counterparties’ reporting workflows, we believe that any proposed rule amendments and final rules associated with Tranche 1 and Tranche 2 should be issued at the same time.” Their letter then went on to comment: “Alternatively, should the Commission decide to publish the proposed rule amendments to the SDR rules first in Tranche 1, then we recommend that the public comment period for this release remain open for at least 90 days following publication of the proposed rule amendments to the reporting workflow rules in Tranche 2. This extended comment period would provide market participants with a comprehensive and holistic understanding of whether the two proposals achieve the desired policy outcomes and account for operational costs and possible additional builds to comply with a modified reporting regime.” Letter from Steven Kennedy, ISDA, and Kyle Brandon, SIFMA, at 3-4 (August 21, 2017) (footnote omitted), available at https://comments.cftc.gov/PublicComments/ViewComment.aspx?id=61288&SearchText=.

[21] Proposal, text immediately following n.23.

[22] The Commission’s disjointed delivery of proposed changes to its swap data reporting rules also raises questions as to its consideration of relevant costs and benefits. Cost-benefit considerations, by their very nature, must evaluate the proposed changes in comparison to the status quo – including the present state of other relevant regulations. As a result, the cost-benefit portion of the Proposal could be deemed obsolete to the extent it does not incorporate any of the modifications to other swap data reporting requirements in parts 43 and 45 of the Commission’s regulations that the Commission intends to propose and act upon. The failure to propose all the swaps data reporting rule amendments in unison would seem to necessitate a refresh of the accompanying cost-benefit portion of this Proposal, and further public comment.

[23] See, e.g., Memorandum of Understanding Between the U.S. Securities and Exchange Commission and the U.S. Commodity Futures Trading Commission Regarding Coordination in Areas of Common Regulatory Interest and Information Sharing (July 11, 2018) (specifically addressing the regulatory regime for swaps and security-based swaps), available at https://www.cftc.gov/sites/default/files/2018-07/CFTC_MOU_InformationSharing062818.pdf and https://www.cftc.gov/PressRoom/PressReleases/7745-18.

[24] Risk Mitigation Techniques for Uncleared Security-Based Swaps, 84 FR 4614 (February 15, 2019) (proposed rules).

[25] Id. at 4633-4634 (footnote omitted).

[26] Id. at 4634 (footnotes omitted).

[27] 17 CFR 45.8.

[28] 17 CFR 23.502.

[29] 17 CFR 23.500(i)(1), (3).

[30] In responding to staff’s request for comment on the Draft Technical Specifications, see fn. 19, supra, ISDA stated: “Endusers which either have reporting obligations or which would be compelled to provide data to the reporting counterparty necessitated by the proposed fields would be particularly burdened by the requirements and many will lack the technological capability to capture, transform and report or provide data as required. The small to midsized commodity producers, processors, merchants and other endusers that use swaps to mitigate commodity, interest rates, foreign exchange or other price risks will require additional technology, compliance and legal support in order to accommodate additional reporting requirements. This will impose significant, unjustified costs to endusers….ISDA, on behalf of commercial endusers, requests the CFTC to avoid imposing changes and additional reporting requirements on endusers by maintaining their obligations under the current Reporting Regulations to the greatest extent possible.” Letter from Tara Kruse, ISDA, at 7-8 (March 7, 2016), available at https://comments.cftc.gov/PublicComments/ViewComment.aspx?id=60713&SearchText=.  ISDA continued to advocate against placing additional burdens on end-users through its joint comment letter with SIFMA to the Swap Data Roadmap and suggested the Commission “should not require non-reporting counterparties, end-users, and smaller firms to perform reconciliations because these entities generally do not have the resources to effectively validate their swap transactions.” See fn. 20, supra, at 6.

[31] Proposal, text accompanying n.226.

[32] To be sure, the Proposal might identify situations in which the SDR inexplicably alters the data that it receives from an RCP. But current Regulation 49.10(c), 17 CFR 49.10(c), already prohibits such activity since an SDR “shall establish policies and procedures reasonably designed to prevent any provision in a valid swap from being invalidated or modified through the confirmation or recording process of the swap data repository. The policies and procedures must ensure that the swap data repository's user agreements are designed to prevent any such invalidation or modification.”

[33] 7 U.S.C. 24a(c)(2).