Dissenting Statement of Commissioner Dan M. Berkovitz Regarding Proposed Rulemaking on Swap Execution Facilities and Trade Execution Requirement
November 5, 2018
I. Summary of Dissenting Views
I respectfully dissent from the Commodity Futures Trading Commission’s (“CFTC” or “Commission”) notice of proposed rulemaking regarding Swap Execution Facilities and Trade Execution Requirement (the “Proposal”). This Proposal would reduce competition and diminish price transparency in the swaps market, which will lead to higher costs for end users and increase systemic risks.
The Proposal would abandon the commitments the United States made at the G20 Summit in Pittsburgh in 2009 to trade standardized swaps on exchanges or electronic trading platforms and is contrary to Congressional direction in the Dodd-Frank Act and the Commodity Exchange Act (“CEA”) reflecting those commitments. It would retreat from the progress made by the Commission and the financial industry in implementing those reforms.
The Proposal would reduce competition by cementing the oligopoly of the largest bank dealers as the main source of liquidity and pricing in the swaps markets. It would diminish transparency by removing the requirement that highly liquid swaps be traded through competitive methods of trading. By reducing competition and diminishing price transparency, the Proposal would increase systemic risks and lead to higher swaps prices for commercial and financial end-users. Ultimately, the millions of Americans who indirectly participate in the swaps market through their investments in retirement accounts, pension plans, home mortgages, and mutual funds will pay that higher cost. Finally, the Proposal would provide SEFs with too much discretion to set their own rules and in so doing, weaken regulatory oversight and enforcement capabilities.
II. Major Flaws in the Proposal
The evidence is clear that the Dodd-Frank reforms, including the Commission’s swap execution regulations, have led to more competition, greater liquidity, more electronic trading, better price transparency, and lower prices for swaps that are required to be traded on regulated platforms. Numerous academic studies and reports by market consultants have documented these benefits. The Proposal ignores this evidence and analysis.
The Proposal would jettison the regulatory foundation for the way swap execution facilities (“SEFs”) currently operate. It would delete the requirement that swaps that are subject to the trade execution mandate (“Required Transactions”) be traded either on Order Book or by a request for quote from at least three market participants (“RFQ-3”). This would undermine the Congressional directive in the Dodd-Frank Act that for Required Transactions, a SEF provide multiple participants with “the ability to execute or trade swaps by accepting bids and offers made by multiple participants in the facility or system.” Consequently, the Proposal would lead to less price transparency and less competition.
The Proposal also would gut the impartial access requirement in the Dodd-Frank Act. The statute requires SEFs to establish rules that “provide market participants with impartial access to the market.” Authorizing discrimination based on the type of entity will permit the largest bank-dealers to establish and maintain exclusive pools of liquidity for themselves. By denying other market participants access to the most favorable prices in the dealer-to-dealer market, bank dealers can prevent others from cost-effectively competing with them for customers. Eliminating competition will result in higher prices for customers. Permitting large banks and dealers to discriminate in this manner is inconsistent with sound economic principles underpinning competitive markets and the CEA’s impartial access requirement.
In pursuit of the goal of “flexibility” for SEF markets, the Proposal deletes, reverses, or waters down many key trading, access, and compliance requirements for SEFs. The wide latitude that would be granted to SEFs as to how swaps may be traded, who may trade them, the oversight of the marketplace, and the conduct of the brokers looks very much like the “light-touch” approach to regulation that was discredited by the financial crisis.
Seven years ago, as the Commission was formulating the current regulations, very little data was available on swap trading and pricing. But now, after six years of experience with those regulations, we have an extensive amount of data, collected by SEFs and swap data repositories. The Commission should base its regulatory decisions on this data and the studies and literature that have analyzed this data and demonstrated the benefits of the current swap trading requirements.
Unfortunately, the Proposal does not consider the available data and market studies that demonstrate the current RFQ-3 system is working well to provide highly competitive prices and low transaction costs. For example, the Proposal ignores the following studies and conclusions:
- CFTC economists’ study (2018). This study, conducted by four CFTC economists, concluded: “Judged from our evidence, SEF-traded index CDS market seems to be working well after Dodd-Frank—dealers’ response rates are high, the vast majority of customer orders result in trades, and customers’ transaction costs are low.” With respect to the most liquid CDS index swaps, the CFTC economists found that “the average transaction cost is statistically and economically close to zero.”
- Bank of England Staff Working Paper (2018). This Bank of England paper concluded that the CFTC’s trade execution mandate, including the RFQ-3 requirement, has led to a “sharp increase in competition between swap dealers” in dealer-to-customer transactions for interest rate swaps subject to the mandate. The study concluded that this competition had led to “a substantial reduction in execution costs,” amounting “to daily savings in execution costs of as much as $3-$6 million for end-users of USD swaps.”
- Study of “Market Structure and Transaction Costs of Index CDSs” (2017). This study found that prices customers obtained in the dealer-to-customer market through the RFQ system often were better than the prices that were available on the interdealer Order Book. “[O]ur results show that the current market structure delivers very low transaction costs. . . .
The Proposal conjectures that novel “flexible methods of execution” will benefit the trading of all swaps. The Proposal, however, does not identify any trading methodology that can provide lower costs than the RFQ-3 method as applied to interest rate swaps and index CDS subject to the current trade execution mandate. In discarding the trading requirements for Required Transactions to bring more swaps onto SEFs, the Proposal throws the baby out with the bathwater.
Today, a small number of large dealers provide liquidity to the swaps market. Five very large banks were party to over 60 percent of interest rate swap transactions. Liquidity in highly standardized swaps is fragmented between a dealer-to-dealer market and a dealer-to-customer market. There are no non-dealers in the dealer-to-dealer market. This high degree of reliance on a few large bank dealers to supply liquidity to all swaps market participants presents systemic risks as well as other types of risk that arise in highly concentrated markets.
One of the fundamental purposes of the CEA is to “promote responsible innovation and fair competition among boards of trade, other markets and market participants.” It is the CFTC’s mission, and incumbent upon this agency in carrying out that mission, to ensure that there is fair competition among all market participants. This means ensuring no market participant or limited group of participants has excessive market power. Market structure and price competition should develop in the interest of all market participants, rather than in the interest of just a few of the largest banks. The Commission should strive to remove the existing barriers to broader participation and fair competition in the swaps markets. In my view, the Proposal seeks to perpetuate existing barriers.
III. Targeted Reforms to Consider
The current system is not perfect; there are flaws that should be addressed. But the evidence is clear that the current system has provided substantial benefits over the unregulated system that existed prior to the financial crisis and the Dodd-Frank reforms. The Proposal would return the swaps market to the dealer-dominated, trade-however-you-want system heavily reliant on voice brokers that existed prior to the financial crisis. At the G20 Summit in Pittsburgh in 2009, the United States made an international commitment to move away from the dealer-dominated, voice-brokered approach and Congress expressly rejected the dealer-dominated, flexible approach when it adopted the Dodd-Frank Act.
My sense from working with and talking to swap market participants is that many do not see a need for a major overhaul of the swaps regulatory framework. The benefits of the current system are due not just to the regulations, but also are the result of major efforts and investments by market participants and operators of SEFs in electronic trading technology and personnel. Many market participants do not want to deal with another round of costs and uncertainties that wholesale regulatory changes will generate. They believe the current system is working, despite its flaws. They prefer that we consider more targeted reforms to address specific issues with the current system, rather than scrap the current system entirely. They do not want to face the possibility that the Commission will continue to engage in a repetitive cycle of de-regulation and re-regulation.
Rather than completely rewrite the SEF regulatory structure, and turn our back on the progress made in transparency and competition, I favor a more limited, data-based approach to build on our progress and improve upon the current structure. This could be accomplished by removing some of the unnecessary barriers to greater participation on SEFs. Banks and other swap dealers play a critical role in providing liquidity. We need them to participate. However, a highly concentrated dealer oligopoly is not a prerequisite for sufficient liquidity. We should seek ways to bring in more sources of liquidity and competition. Robust competition leads to healthier markets and improves the overall welfare of all market participants.
I support the goal of bringing more types of swaps onto the SEF trading environment. I could support a more narrow approach to achieve this goal that does not undermine the progress that has been made to date.
I am not persuaded that we should continue to have two separate pools of liquidity in the swaps market for all types of swaps, regardless of liquidity characteristics—one in which the dealers trade amongst themselves, and another in which the dealers trade with customers. Perhaps we should look for ways to consolidate rather than separate the swaps markets.
Specifically, I support considering the following regulatory measures to improve competition in the swaps market:
- Abolish Name Give-Up. The Commission should prohibit the practice of name give-up for cleared swaps. On many platforms that provide anonymous trading, the identity of a counterparty is provided to the dealer after the completion of a trade. Name give-up is a major deterrent to non-dealers seeking to participate on dealer-only platforms as it provides the dealers with valuable information about a counterparty’s positions. Name give-up is a relic of the pre-Dodd Frank era when most swaps were not cleared and the identity of the counterparty was necessary to manage credit risks.
- Expand Floor Trader registration. The Commission should amend the floor trader provision in the swap dealer definition to remove overly restrictive conditions. This would permit a wider range of proprietary traders to provide liquidity and compete with large bank dealers on price.
- Revise capital requirements. The Commission should work with the prudential regulators to ensure that capital requirements do not unduly restrict the availability of clearing services by futures commission merchants (“FCMs”). The current capital requirements have had the unintended consequences of discouraging FCMs from providing additional clearing services to the cleared swaps market.
- Enable average pricing. The Commission should work with market participants and facilities to enable buy-side firms to obtain average pricing for buy-side swap trades. Although average pricing is available for futures, it currently is not available for swaps, which limits the direct participation of buy-side asset managers on SEFs.
We should explore these and other ways to increase competition in the swaps market rather than retreat from the progress that has been made. What follows is a more detailed explanation of how the current regulatory system has improved the swaps market and how the Proposal would undermine those improvements.
IV. Specific Concerns with the Proposal
The Proposal raises the following specific concerns:
- Less competition
- Less transparency
- Higher prices for end-users
- Diminished CFTC supervision and enforcement abilities
A. Less competition, less transparency, and higher prices
The first three concerns—higher prices, less competition, and less transparency—arise from the repeal of two critical and inter-related provisions of the current regulations.
Elimination of Order Book/RFQ-3. The Dodd-Frank Act sets forth a Rule of Construction that the goal of the SEF regulations is “to promote the trading of swaps on swap execution facilities and to promote pre-trade price transparency in the swaps market.” A key requirement facilitating the statutory goal of pre-trade price transparency is that all Required Transactions must be traded by Order Book or RFQ-3. Under RFQ-3, a customer must request quotes from at least three dealers prior to entering into a transaction. In this manner, dealers must compete on price.
The Proposal would delete the Order Book/RFQ-3 requirement, even for swaps already traded on SEFs and subject to the trade execution requirement. Instead, the Proposal states that “a SEF may utilize ‘any means of interstate commerce’ for purposes of execution and communication, including, but not limited to, the mail, internet, email and telephone.”
Authorizing discrimination; eviscerating impartial access. Next, the Proposal flips on its head the impartial access requirement. CEA Section 5h(f)(2)(B)(i) requires a SEF to “provide market participants with impartial access to the market.” Under existing Commission Regulation 37.202, which implements this statutory provision, any SEF criteria governing access must be “impartial, transparent, and applied in a fair and non-discriminatory manner.” In the 2013 SEF rulemaking, the Commission explicitly rejected a proposed interpretation that would permit SEFs to discriminate against types of market participants. “[T]he Commission believes that the impartial access requirement of Core Principle 2 does not allow a SEF to limit access to its trading systems or platforms to certain types of [eligible contract participants (“ECPs”)] or [independent software vendors (“ISVs”)] as requested by some commenters. The Commission notes that the rule states ‘impartial’ criteria and not ‘selective’ criteria as recommended by some commenters.”
The Proposal would replace this critical requirement and allow each SEF to establish exclusionary criteria determining what types of market participants are “similarly situated market participants” that are allowed to trade on the SEF (let’s call this what it is, the “Discriminatory Access Provision”). This approach flips the statutory “impartial access” requirement on its head by empowering SEFs to build limited liquidity pools for a select few market participants such as the dealers seeking to hedge with each other.
Under the Discriminatory Access Provision, it is reasonable to expect that the large bank swap dealers would encourage discriminatory SEF participation criteria such that only large bank swap dealers would be “similarly situated market participants” able to participate in dealer-to-dealer liquidity pools. Proprietary trading firms and smaller dealers provide competition to the large banks in pricing swaps, and are one major reason customers are able to obtain favorable prices through the current RFQ process. If discrimination is permitted, these other types of firms would not be able to use the dealer-to-dealer market to effectively hedge or offset trades with customers, and therefore would not be able to compete with the large bank swap dealers in the dealer-to-customer market. In this manner, the Discriminatory Access Provision would result in a significant loss of competition in the dealer-to-customer market, which ultimately would result in higher prices for end users.
If the current trade execution requirement is repealed, dealers also could establish single-dealer platforms and call them SEFs to siphon liquidity away from the RFQ platforms. The dealers wield significant market power in the swaps market. Five dealers currently account for nearly two-thirds of the interest rate swap market, which is the largest swap product category. Although SEFs that currently offer RFQ-3 functionality might continue to do so even if the requirement is repealed, once the customers are no longer required to use that functionality, the dealers could undermine the effectiveness of the RFQ process by offering incentives to trade on single-dealer platforms or voice-brokered SEFs. This outcome would reduce liquidity for the RFQ platforms. In the long run, draining liquidity from RFQ-3 platforms to single-dealer or voice-brokered systems will result in less direct competition between dealers, less transparency, and higher costs for customers.
The Proposal asserts that all-to-all markets are “inimical” to “fundamental” swaps trading features. The Proposal also states that “market participants have rarely used Order Books to trade swaps on SEFs,” and that “this low level of swaps trading on Order Books is attributable to an Order Book’s inability to support the broad and diverse range of products traded in the swaps market that trade episodically, rather than on a continuous basis.” Following a brief discussion of why the Order Book is unsuitable for some swaps, the Proposal states that the Order Book should be eliminated for all swaps: “[B]ased in part on its experience, the Commission proposes to eliminate the minimum trading functionality requirement and the regulatory Order Book definition.”
Similarly, the Proposal eliminates the RFQ requirement because it states that this method of execution may be unsuitable for some additional types of swaps that are currently traded off SEF. “[T]he Commission believes that [Order Book and RFQ-3] would not be suitable for the broad swath of the swaps market that would become newly subject to the trade execution requirement.”
This reasoning is flawed. From the proposition that an Order Book may be unsuitable for some episodically traded swaps, it does not follow that an Order Book is unsuitable for all swaps, even highly liquid ones. Nor does it follow from the proposition that the RFQ process may be unsuitable for some swaps that it should be removed for all swaps. Yet this flawed logic appears to be the rationale for the elimination of both the Order Book and RFQ-3 functionality requirements, even for highly liquid standardized swaps.
RFQ-3 has improved competition and lowered trading costs. Empirical evidence demonstrates that the Order Book/RFQ-3 and impartial access requirements for standardized, highly liquid cleared swaps have increased competition and transparency and brought low trading costs to swap markets. The Bank of England Study found that the RFQ-3 requirement significantly improved liquidity for U.S. dollar interest rate swaps, which reduced swap execution costs for end-users by an estimated $3 to $6 million per day relative to Euro swaps, which were not traded pursuant to the trade execution mandate.
The Bank of England Study also assessed the impact of the SEF trading mandate on dealer market power. The study found that, prior to the SEF trading mandate, 28 percent of customers for US and Euro interest rate swaps that became subject to the mandate dealt with only a single dealer, and over 50 percent of customers dealt with three or fewer dealers. After the SEF trading requirements went into effect, those percentages dropped to 8 percent and 20 percent, respectively. The study states that “[w]ith the improvements in pre-trade transparency, customer search costs have fallen and it has become easier for customers to trade with the dealer showing the best price.”
Other studies have found similar results. Collin-Dufresne, Junge, and Trolle compared the prices on the Order Books used in the interdealer market with the prices generated in the dealer-to-customer market through the RFQ system. The authors found that prices customers obtained in the dealer-to-customer market through the RFQ system often were better than the prices that were available on the interdealer Order Book.
Economists in the CFTC’s Office of Chief Economist examined data regarding the customer trading of index CDS on the Bloomberg and Tradeweb SEFs, which are the leading SEFs for dealer-to-customer trading. The CFTC economists found that very little customer trading occurred on the Central Limit Order Book (“Clob”) of either facility, but rather that most of the trading occurred either by RFQ or by request-for-streaming (“RFS”). Focusing on customer trading through the RFQ mechanism, the CFTC economists found that, on average, a customer requests quotes from 4.1 dealers and gets back 3.6 responses.
The CFTC economists concluded that the current regulatory structure is working well: “Judged from our evidence, SEF-traded index CDS market seems to be working well after Dodd-Frank—dealers’ response rates are high, the vast majority of customer orders result in trades and customers’ transaction costs are low.” Specifically, the CFTC economists found that transaction costs were low for index CDS contracts:
The transaction costs of on-the-run CDX.NA.IG and iTraxx Europe have a mean around 0.2 bps and a standard deviation of 1.4 bps, so the average transaction cost is statistically and economically close to zero. For on-the-run CDX.NA.HY and iTraxx Crossover, the average costs are larger, at about 0.5 and 1.1 bps, but again not significant compared to their standard deviations of about 2.6 and 3.5 bps. The first off-the-run contracts have comparable average transaction costs but a much higher standard deviation due to the relatively few number of trades in these contracts.
Market participants have expressed similar concerns about removing the Order Book/RFQ-3 and impartial access requirements. One senior executive at a trading firm recently stated that the SEF regulations have helped halve the bid-offer spread in US dollar swaps and increased price competition. “My fear is we take too big a step back from having the competitive pricing in the market,” he said. “It is still a dealer-controlled market and if the biggest dealers simply say: ‘Great, I don’t have to put a competitive price on the screen anymore, and if someone wants my most competitive price then you’ve got to pick up the phone again,’ I don’t want to take that step backwards.”
Similarly, the CEO of one SEF cautioned, “[o]ne of the risks of this concept of ‘any means of interstate commerce’ is you have benchmarks and fixings that rely on better liquidity coming in from liquid Clobs. You wouldn’t want to go backwards in that respect.”
In 2016, Greenwich Associates reported that “the buy side feels the executions they are receiving under the current paradigm are sufficient, if not excellent.” Greenwich Associates noted that, for many asset managers, sending a request for quote to three market participants and selecting the best-priced response (no matter how many respond) “has long been considered an appropriate approach to achieving best execution.”
The Proposal does not reference any of these findings or views of market participants. In contrast to these data-based empirical studies regarding the benefits of the current regulatory system, the Proposal speculates—without any evidentiary support—that the “flexibility” afforded by the elimination of the Order Book/RFQ-3 requirement may provide various benefits. For example, the Proposal asserts “SEFs would have broader latitude to innovate and develop new and different methods of execution tailored to their markets.” The Proposal further opines that these new, flexible methods “could be more efficient,” “may lead to reduced costs and increased transparency,” and “may provide opportunities for new entrants in the SEF market.”
However, the Proposal provides no factual basis for any of these hypothetical benefits. In light of the very low execution costs that have been documented for interest rate and index CDS swaps traded through RFQ-3, it is difficult to understand why RFQ-3 should be eliminated, at least for the swaps to which is currently applies.
Effect of expanded trading mandate on liquidity. The overriding rationale for the Proposal is to attract greater liquidity formation to SEFs. The Proposal seeks to accomplish this goal by expanding the SEF trading requirement to include all mandatorily cleared swaps for which SEF trading exists, with several exceptions. Although the Proposal would expand the trade execution mandate in this manner, it also would eliminate the Order Book/RFQ-3 requirements and provide effectively unlimited flexibility as to the trading methods for all swaps subject to the expanded trading mandate. The Proposal broadly asserts, without providing any evidentiary support, that the expanded trading mandate will improve liquidity and pre-trade price transparency and reduce market fragmentation.
In asserting that the expanded execution mandate will increase on-SEF liquidity, the Proposal appears to measure liquidity solely in terms of volume. But volume does not equal liquidity. It is not apparent how simply moving this volume from off SEF to being traded within a SEF will have any effect on other traditional measures of liquidity, such as cost of transaction or price dispersion. Indeed, the only difference is that the swaps would be traded on SEF, but by the same people and using the same methods that they now use to trade them off SEF. It is not apparent how this would lead to any greater price transparency or lower costs.
How many and what types of swaps would be brought onto SEFs under the expanded trading mandate? The Proposal presents little data to answer this question. One approach would be to assume that all swap transactions that are currently subject to clearing would become subject to the expanded trading mandate under the Proposal. This amount may be significantly larger than the actual result because many swaps subject to clearing may not be easily traded on SEF. But by comparing this amount to the amount of swaps currently traded on SEF, we can estimate an upper bound on the incremental increase in on-SEF trading resulting from the Proposal.
The Proposal notes that an estimated 57% of the notional amount of interest rate swaps are being traded on SEF, and that 85% are subject to the clearing requirement. Accordingly, an upper bound of about 28% of interest rate swaps could be moved on SEF under the Proposal. This estimate is consistent with a recent estimate provided by Clarus that approximately two-thirds of the fixed/float USD interest rate swap market is traded on SEF. Examining the one-third of interest rate swaps that are being traded off SEF, Clarus found that “[g]enerally speaking, everything off-SEF is bespoke.”
Again, it is not apparent how moving the trading of bespoke swaps from being traded by introducing brokers (“IBs”) outside a SEF to being traded by swap trading specialists inside a SEF will have any effect on the prices of those bespoke swaps. It is even less apparent how the trading of these bespoke swaps within a SEF will have any impact upon the trading of the highly liquid standardized swaps already being traded within a SEF under the RFQ-3 methodology. In fact, eliminating RFQ-3 for those liquid swaps could raise the prices for those swaps, and in turn may also negatively impact pricing for less liquid swaps, because most interest rate swaps—including bespoke swaps—are priced in part on a standard rate curve developed from prices for liquid swaps at various point along the curve.
Other impacts from excessive flexibility and discretion. The Proposal establishes an overly flexible approach that allows each SEF to self-determine how it will operate in almost every respect. Among other areas, a SEF would use discretion (a word used over 150 times in the Proposal) to tailor policies and procedures regarding trading procedures and rules, access, pre-execution communication, personnel oversight and ethics training, SEF compliance requirements, trading surveillance, error trade policies, record keeping, trade documentation, internal investigations and enforcement, setting fees, financial resource requirements, and supervision of third party services. Most of these changes would loosen current regulatory requirements.
Documentation of executed swaps would no longer be required at the time of execution, but as soon as technologically possible. The Proposal acknowledges that creating flexibility for execution methods and trading technology makes simultaneous documentation “impracticable.” In other words, moving away from electronic trading back to telephones will delay the time within which counterparties receive full confirmation of price and terms, preventing precision in the time of pricing, creating a higher likelihood of errors, and leading to less pre-trade price transparency.
Many of the changes in the Proposal would allow the SEF to exercise discretion in brokering trades and establishing rules to facilitate broking away from electronic platforms. The Proposal explains that one of the reasons for granting the SEF greater discretion is to allow voice-broking to occur directly within the SEF.
Traditional introducing broking, by its nature, is slower and less transparent at establishing prices as compared to electronic trading. As a broker calls around to multiple dealers for prices, the broker might make trade adjustments over time and prices from one call to the next may change. As time passes, prices may become stale, even within seconds. Dealers and other liquidity providers will add a cushion to the spread to account for this delay. This means that as the length of time increases between when a quote is first received and when the trade is executed and the price is reported, spreads become wider and pricing becomes less transparent. For certain trades, such as block trades, timing delays in price transparency might be appropriate for reasons related to the unique nature of each trade. However, we should not be adopting regulations that would degrade the current level of transparency for liquid swaps that are being efficiently traded using an Order Book or RFQ system.
Similarly, the Proposal would allow extensive pre-trade negotiation for all swaps so long as the SEF defines it into the SEF’s trading rules. Pre-trade negotiation may be appropriate for certain bespoke or large sized swaps. However, to create flexibility in SEF trading methods, the Proposal would allow SEFs to include pre-trade negotiations for any and all types of swaps including standardized swaps currently traded electronically. However, the Proposal would allow SEFs to include pre-trade negotiations for more liquid, standardized swaps for which pre-trade price transparency is better achieved through electronic trading, as explained in the studies discussed above.
In addition, the Proposal would allow SEF trading specialists, when acting as brokers, to exercise discretion in sharing different market information with different market participants. The Proposal acknowledges that this “trading discretion exercised by SEF trading specialists may affect the manner in which market participants are treated on a facility.” The Proposal suggests that this is somehow “consistent with impartial access” because it facilitates more trading. More likely, this greater degree of sanctioned discretion—the extent of which is largely left up to the SEFs to determine—would lead to unfair treatment of different market participants and less pre-trade price transparency because SEF trading specialists can decide who gets what information pre-trade.
The statements above should not be interpreted as critical of intermediary broking services. These services provide important options for trading and pricing certain types of swaps, such as bespoke swaps, package trades, and block sizes. Rather, my concern is that these important services and the professionals who provide them may become less regulated, and that they will become intermediaries for transactions that are required to be traded electronically.
B. Diminished Oversight and Enforcement
I am also concerned that this Proposal waters down the robust, and uniform, standards of conduct and supervision to which it currently holds SEFs, IBs, associated persons (“APs”) of IBs, and other market participants. This could lead to SEFs reducing their focus on compliance, require the Commission to take on an enhanced oversight role, and constrain the Commission’s ability to investigate and prosecute abusive trade practices involving SEFs.
As previously discussed, this Proposal grants extensive discretion to SEFs to create rules governing their operations and does away with some of the specific compliance and recordkeeping obligations currently required by the regulations governing SEFs, set forth in Part 37 of the Commission’s Regulations. The Proposal suggests that providing SEFs with greater flexibility to tailor their compliance and oversight programs will mitigate compliance challenges that SEFs have encountered in implementing part 37, yet fails to describe in any detail those challenges. On the other hand, we know that our current system of oversight provides market participants and regulatory authorities with uniform and descriptive standards of conduct and compliance procedures. Enumerating these standards (1) prevents a race to the bottom, in which market participants pare back their policies and procedures to the bare minimum, and (2) provides the registrant and the Commission with the tools they need to successfully enforce compliance with those standards.
As an example, the Proposal would remove the requirement set forth in Regulation 37.203(c) that a SEF establish and maintain sufficient compliance staff and resources to (i) conduct specific monitoring, including audit trail reviews, trade practice and market surveillance, and real-time market monitoring; (ii) address unusual market or trading events; and (iii) complete investigations in a timely manner. Rather, the Proposal would only require that the SEF establish and maintain sufficient compliance staff and resources to ensure that it can fulfill its self-regulatory obligations under the CEA and Commission Regulations. Without specific requirements on what compliance resources are needed, each SEF will be free to determine what level of resources is sufficient for such a broad mandate. In essence, the SEF need not map its compliance resources to specific compliance tasks. Additionally, experience has shown that conducting oversight and examinations of the sufficiency of a registrant’s compliance resources is more difficult to undertake on a standard and fair basis across registrants when each one has a different view of what resources will meet the generalized requirement.
As another example, the Proposal eliminates the specific requirements that a SEF establish an annual audit trail review and related enforcement program, and retain certain categories of documents currently required by Regulation 37.205. The Proposal assumes, however that “SEFs would continue to fulfill their information collection burdens in a manner similar to the status quo.” If the expectation is that SEFs will continue to comply with the current requirements, then why is it necessary to remove or weaken them? Many still view the compliance function as a cost center. It is unrealistic to assume that we can remove many of the specific conduct and recordkeeping obligations and expect that market participants will continue to comply, when competitive market pressures will drive the allocation of resources elsewhere. Moreover, market participants have dedicated significant resources to developing these compliance policies and systems, and changing them without sufficient justification does not make practical sense.
As a final example, the Proposal removes some of the specific requirements in Regulation 37.204 for oversight of third-party regulatory services. SEFs would no longer be required to conduct regular meetings with, and periodic reviews of, service providers or provide records of such oversight to the Commission. Instead, SEFs are given broad latitude to determine the necessary processes to supervise these providers. When registrants delegate critical functions to third-party providers, it is imperative that the registrant maintain diligent supervision over the provider’s handling of these functions.  In my view, the Proposal does not provide satisfactory reasons for removing these unambiguous requirements, considering that doing so could hamper the Commission’s ability hold SEFs accountable for supervising third-party providers.
Equally concerning is the sweeping change the Proposal makes to the way in which SEFs and their employees and agents will be registered, and in turn, the Commission’s oversight of their conduct. Under the current system, swaps broking entities that meet the definition of an IB must be registered with the Commission as such. The individuals who are involved in soliciting or accepting orders at IBs, or involved in supervising such individuals, must register as APs of IBs. As NFA members, IBs and APs are not only subject to the applicable Commission Regulations, but are also subject to uniform rules governing swaps brokering, trade practices, reporting, minimum financial requirements, proficiency testing, training standards, and supervision. In addition, NFA monitors IBs’ swaps broking activity and compliance with all applicable statutes and rules. In furtherance of that responsibility, NFA conducts periodic examinations of swap IB member firms and has the ability to discipline IBs and APs where appropriate.
Under the Proposal, which limits the activity that can be conducted off SEF, IBs will need to register with the Commission as SEFs to continue to broker swaps transactions. Given that the majority of IBs engaging in swap transactions on SEF are affiliated with SEFs, it is likely that many of these entities, or their employees, will merge into or join the affiliated SEF. We can also expect to see the formation of new SEFs, which presumably would not be required to register as IBs. SEFs and SEF employees would be free to withdraw their IB and AP registrations and memberships with NFA, leaving a regulatory vacuum with no self-regulatory organization oversight. Already strained Commission resources inevitably would need to fill that void.
Further, the Proposal creates an entirely new category of persons: the SEF trading specialist. As proposed, SEF trading specialists will perform “core functions” that facilitate swaps trading and execution, including negotiating trade terms, arranging bids and offers, and discussing market color with market participants, or directly supervising a person who engages in such functions. In fact, the Proposal notes that broadening the SEF registration and trade execution requirements would increase the level of discretion that these SEF employees and agents would exercise in connection with swaps trading. However, despite these key, customer-facing functions, SEF trading specialists would not be required to register with the Commission.
For this reason, I am also concerned that the Proposal would weaken the supervisory function within the SEF. Regulation 166.3 imposes a duty on all Commission registrants who act in a supervisory capacity, including APs, to diligently supervise the activities of employees and agents relating to their business as a Commission registrant. However, if the SEF is not registered as an IB, and its employees are thereby not registered as APs, the SEF employees themselves will have no duty to supervise under Regulation 166.3. The Proposal imposes a separate duty on SEFs to supervise the activities of its SEF trading specialists “in the facilitation of trading and execution on the swap execution facility.” Critically, however, that duty runs only to the SEF as an entity and not to its employees, including the SEF trading specialists. As a result, SEF trading specialists or other SEF employees with supervisory duties cannot be held individually liable for failure to supervise under any Commission regulation if they are not duly registered as APs of IBs. Individual accountability is an important tool in incentivizing corporate responsibility and I think it must be preserved.
Finally, in at least one instance, the flexibility afforded to SEFs to establish a code of conduct for their SEF trading specialists is in direct conflict with the supervision rules applicable to all registrants under Regulation 166.3. The Proposal states that a SEF’s Code of Conduct “may provide” that, among other things, a SEF trading specialist “not engage in fraudulent, manipulate, or disruptive conduct.” However, Regulation 166.3 requires that Commission registrants establish and maintain meaningful procedures for detecting and deterring fraud and other prohibited conduct by their employees and agents. This could create another potential gap in our supervisory structure that could weaken the Commission’s enforcement capabilities.
This Proposal is a fundamental overhaul of the SEF regulatory regime. The changes create a trading system that is so flexible that all swaps traded on SEFs—including the most liquid—could be traded the same way they were before the Dodd-Frank reforms were adopted. The Proposal would allow the largest dealers to establish separate dealer-to-dealer liquidity pools through exclusionary access criteria. Competition would be reduced and price transparency diminished. This is not what Congress intended when it passed the Dodd-Frank Act.
I am open to appropriate, targeted amendments to the regulations, several of which I have suggested above. However, empirical studies have shown that the existing SEF regulations have made great progress in achieving the statutory goals of promoting on-SEF trading and pre-trade price transparency. With respect to the swaps markets that are working and providing low costs to the buy side and end users, we should live by the adage, “if it ain’t broke, don’t fix it.”
 See infra section II.
 7 U.S.C. 1a(50).
 7 U.S.C. 7b-3(f)(2)(B)(i).
 Lynn Riggs (CFTC), Esen Onur (CFTC), David Reiffen (CFTC) & Haoxiang Zhu (MIT, NBER, and CFTC), Swap Trading after Dodd-Frank: Evidence from Index CDS (Jan. 26, 2018) (“CFTC Economist Study”).
 Id. at 50.
 Id. at 43.
 Evangelos Benos, Richard Payne & Michalis Vasios, Centralized trading, transparency and interest rate swap market liquidity: evidence from the implementation of the Dodd-Frank Act, Bank of England Staff Working Paper No. 580 (May 2018) (“Bank of England Study”).
 Id. at 31.
 Id. The authors explain that during this period these EUR-mandated swaps were not traded on SEFs due to the fragmentation of the EUR swaps market. Id. at 28.
 Pierre Collin-Dufresne, Benjamin Junge & Anders B. Trolle, Market Structure and Transaction Costs of Index CDSs (Sept. 12, 2017) (“Collin-Dufresne, Junge, and Trolle Study”).
 Id. at 38.
 Id. at 6.
 Quantifying Interest Rate Swap Order Book Liquidity, Greenwich Associates, Q1 2016 (“Greenwich Report”), at 8.
 7 U.S.C. 5(b).
 7 U.S.C. 7b-3(e).
 17 CFR 37.9. In the 2013 rulemaking adopting the current SEF regulations, the Commission explained the rationale for this requirement: “[T]he Commission believes that an RFQ System, as defined in § 37.9, operating in conjunction with a SEF’s minimum trading functionality (i.e., Order Book) is consistent with the SEF definition and promotes the goals provided in [CEA Section 5h(e), 7 U.S.C. 7b-3(e)], which are to: (1) Promote the trading of swaps on SEFs and (2) promote pre-trade price transparency in the swaps market. The Commission notes that the RFQ System definition requires SEFs to provide market participants the ability to access multiple market participants, but not necessarily the entire market, in conformance with the SEF definition.” Core Principles and Other Requirements for Swap Execution Facilities (“2013 SEF Rulemaking”), 78 FR 33476, 33496 (June 4, 2013).
 Notice of proposed rulemaking, Swap Execution Facilities and Trade Execution Requirement (“Proposal”), section IV.I.4.b.
 7 U.S.C. 7b-3(f)(2)(B)(i).
 17 CFR 37.202(a)(1).
 2013 SEF Rulemaking, 78 FR at 33508. The Commission also stated that “the purpose of the impartial access requirements is to prevent a SEF’s owners or operators from using discriminatory access requirements as a competitive tool against certain ECPs or ISVs.” Id.
 It is unclear under the Proposal what happens to market participants subject to the SEF trading requirements who are not given access to a SEF because of the Discriminatory Access Provision.
 Greenwich Report at 8. One market participant has commented on the ability of the dealers to determine market structure through the exercise of their market power:
“There is no commercial explanation for having a market that is not open to a lot more people. It just doesn’t make any sense. But the ability of people to enforce change outside the incumbent dealers is very limited,” says the expert. “The part that frustrates me more than anything is pretending that the leverage of the incumbent dealers over this market isn’t real. When I hear people talk about the natural market evolution, I would contend that progress has been 100% prevented to date.”
Robert Mackenzie Smith, US swap trading overhaul may reinforce market split, users warn, Risk.net, Mar. 21, 2018, https://www.risk.net/derivatives/5440516/us-swap-trading-overhaul-may-reinforce-market-split-users-warn.
 In the equities market, the forced transition away from a market centered around multiple dealers improved prices substantially. See, e.g., Michael J. Barclay, William G. Christie, Jeffrey H. Harris, Eugene Kandel & Paul H. Schultz, The Effects of Market Reform on the Trading Costs and Depths of Nasdaq Stocks, Journal of Finance, Vol. 54, Issue 1, at 1-2 (1999) (“Our results indicate that quoted and effective spreads fell dramatically without adversely affecting market quality.”).
 Proposal at section VII.A.1.a.
 Id. at section IV.C.2.
 Proposal at section IV.I.4.b.
 In the Cost-Benefit Considerations, the Proposal acknowledges that “the overall amount of pre-trade price transparency in swap transactions currently subject to the trade execution requirement may decline if the Order Book and RFQ-to-3 requirement[s are] eliminated. This potential reduction in pre-trade price transparency could reduce the liquidity of certain swaps trading on SEFs and increase the overall trading costs.” Proposal at section XXIII.C.
 Bank of England Study at 31. As discussed further below, the Proposal appears to consider liquidity solely in terms of total volume of trades. The Bank of England Study measures liquidity using various price dispersion measures complemented by a price impact measure and a bid-ask spread. See id. at 4. This measure of liquidity better assesses how liquidity affects efficient execution, pricing, and timing of trading.
 Id. at section 5.
 Id. at 26.
 Collin-Dufresne, Junge, and Trolle Study at 38.
 The study reports that, according to the SEF Tracker, at the time of the study, Bloomberg held a market share of 71% and Tradeweb held a market share of 13.6%. CFTC Economist Study at 2.
 Under RFS, customers ask multiple dealers to send indicative quotes in a continuous manner, and can respond to one of them by proposing to trade at the dealers’ quote.
 Id. at 17. The study also found that customers are more likely to request quotes from dealers with whom they have a clearing or pre-existing trading relationship, although customers realize small actual price benefits from requesting quotes from relationship dealers. Id. at 5.
 Id. at 50.
 Id. at 43.
 Robert Mackenzie Smith, Sef reforms could distort new, sounder benchmark rates, Risk.net, Oct. 19, 2018, https://www.risk.net/derivatives/6049931/sef-reforms-could-distort-new-sounder-benchmark-rates (remarks of Stephen Berger, Managing Director, Government and Regulatory Policy, Citadel).
 Id. (remarks of Scott Fitzpatrick, Chief Executive Officer, Tradition SEF).
 Greenwich Report at 7.
 Id. at 11.
 Proposal at section XXIII.C.4.b(1) (emphasis added).
 Using the same method, available data from ISDA indicates that only about 4-5% of index CDS that are currently subject to mandatory clearing are not currently traded on SEF. See SwapsInfo Full Year 2017 and Fourth Quarter 2017 Review, ISDA, at 13-14 (Feb. 2018).
 What is Left Off-SEF, Clarus Financial Technology (Mar. 16, 2016), https://www.clarusft.com/what-is-left-off-sef/.
 Proposal at section IV.F.2.b.
 Proposal at section VII.A.1.a(1)(iii).
 17 CFR Part 37.
 Proposal at section I.C.
 Proposal at section XXIII.B.1.f.
 See, e.g., In re AMP Global Clearing LLC, CFTC No. 18-10, 2018 WL 898755 (Feb. 12, 2018) (consent order) (charging registrant with failing to supervise diligently its information technology provider’s implementation of registrant’s information systems security program); In re Tillage Commodities, LLC, No. 17-27, 2017 WL 4386853 (Sept. 28, 2017) (consent order) (charging registrant with failing to supervise diligently its fund administrator’s operation of the registrant’s bank account containing participant funds).
 The Proposal is not clear on whether an existing IB that now must register as a SEF, but continues to primarily conduct phone broking and other IB-related activities, and continues to meet the IB definition, would need to be dually registered.
 17 CFR 166.3.
 Proposal at section VI.A.3.f. Unlike Regulation 166.3, which applies to all activities relating to a registrant’s business, the language “in facilitation of trading and execution on the swap execution facility” is susceptible to various interpretations and could considerably narrow the conduct that is required to be supervised.
 Id. at section VI.A.3.e (emphasis added).
 See, e.g., CFTC v. Sidoti, 178 F.3d 1132, 1137 (11th Cir. 1999); Sansom Refining Co. v. Drexel Burnham Lambert, Inc., CFTC No. 82-R448, 1990 WL 10830742 (Feb. 16, 1990) (registrant has “a duty to develop procedures for the ‘detection and deterrence of possible wrongdoing by its agents.’”). Moreover, various provisions of the CEA and Commission Regulations prohibit fraudulent and manipulative conduct, so adequate supervision necessarily dictates that entities and supervisors monitor for this conduct. See, e.g., 7 U.S.C. 6b, 9.