December 4, 2014
Thank you for that kind introduction, and my thanks as well to the Futures Industry Association for having me here to speak at this year’s Asia Derivatives Conference. I am honored to be with you in Singapore. I want to give a special thanks to my good friend, Walt Lukken, who has shown tremendous leadership in his role at the FIA.
While traveling in Asia this week and meeting with members of the derivatives community, I’ve been struck by both the vastness and diversity of the region. Notwithstanding this, all of us involved in these complex marketplaces face similar issues, which is why conferences like this one are so important.
Today I want to talk about clearing, but first I’d like to highlight a relatively recent market event. In December 2013, a small, Korean securities firm experienced a trading error that caused it to suffer substantial losses in some of its positions on the Korea Exchange’s listed-derivatives platform. The securities firm defaulted, creating a loss at the affiliated clearinghouse.
The clearinghouse used a portion of its guaranty fund to cover the defaulter’s losses, which resulted in clearing members losing some portion of their default-fund contributions; the clearinghouse itself did not suffer a loss because its skin in the game came after the non-defaulting members’ contributions.
I mention this incident for two reasons. First, the clearing members were institutions based all over the world, including the United States. While the losses were small relative to the size of the members, that losses occurred at all served as a wakeup call for firms who are members of multiple Central Counterparties (CCPs) around the globe. Over the last year, discussions about CCP risk management, recovery and resolution have intensified.
Second, the Korea Exchange incident raised a number of issues for policymakers concerning appropriate incentives related to risk mutualization within the default waterfalls administered by clearinghouses.
Today, I will focus my remarks on these issues in particular. I should add that it is only coincidental that I am here with you today in Asia, and the incident I just described happened to take place on the Asian continent nearly a year ago. It could have happened anywhere, and the same questions about clearinghouse recovery likely would have been raised.
Indeed, one could say that the fundamental framework for loss-mutualization was tested and proved sufficiently resilient at the Korea clearinghouse given the fact that losses of the defaulting securities firm were absorbed, and trading on the exchange resumed relatively quickly.
Nonetheless, we should always ask whether the regulatory framework can be improved.
The CFTC Updated its Rule Set Implementing the PFMIs
In 2009 the G20 determined that standardized swaps should be cleared on CCPs. Much progress has been made to implement this goal. Swap clearing mandates are in effect in the United States and Japan, and other jurisdictions are making significant progress towards their own clearing mandates. (Earlier this week, Japan phased in another portion of its clearing mandate.)
Central clearing promotes financial stability by mitigating counterparty credit risk, enhancing transparency and facilitating more efficient use of capital through clearing’s netting effects. With this in mind, it’s encouraging to note that market and regulatory efforts to increase centralized clearing for OTC swaps have largely been successful, with the volume of cleared swaps continuing to grow throughout the past few years. As clearing volumes increase, however, we need to be cognizant of, and effectively address, the resulting increased concentration of risk in the cleared space.
In 2013, the Commodity Futures Trading Commission (“CFTC” or “Commission”) completed a set of rulemakings to establish regulations for clearinghouses that are consistent with the Principles for Financial Market Infrastructures, or PFMIs. The CFTC regulations and the PFMIs are intended to enhance the safety and soundness of systemically important clearinghouses, thereby supporting the resilience of the global financial system.
This does not mean, however, that our work is done.
Focus of Future Regulatory Efforts by the CFTC
To address the concentration of risk in the cleared space, CCPs must be subject to sound risk-management requirements that are periodically assessed and refined. There are several risk-management areas that could benefit from regulatory and market scrutiny. I would like to discuss three of these areas today:
1. Improving transparency, in particular with respect to standardizing stress tests;
2. Assessing loss mutualization by considering a requirement for CCP capital contributions to the guarantee fund, as well as the appropriate allocation of losses in the default waterfall; and
3. Ensuring that recovery and wind down plans are effective and realistic, including whether to prohibit CCPs from allocating losses to customers in their recovery plans.
Clearinghouses mitigate counterparty credit risk by becoming the buyer to every seller, and the seller to every buyer. Because clearing members trade with a single counterparty (the CCP), due diligence efforts and the risk-management burden that counterparties normally face in the bilateral OTC markets are reduced.
However, this means that clearing members have an even greater need to understand a CCP’s risk profile. Thus, it is crucial for clearing members and other market participants to be able to conduct effective due diligence to understand the risks they face as clearing participants. This understanding not only helps market participants choose among CCPs for clearing, it also allows them to take appropriate steps to manage the risks that they identify.
In order to facilitate that due diligence, clearing members and other market participants must have ready access to relevant information from and about the CCP. The CFTC already has adopted measures to improve transparency for clearing members. Under CFTC rules, a Derivatives Clearing Organization (DCO) is required to provide market participants with information concerning its rules, and the operating and default procedures governing its clearing and settlement systems.
DCOs also are required to publicly disclose the terms and conditions of each contract, agreement, and transaction cleared and settled by the DCO. They also must disclose each clearing fee and other fees charged to members, the DCO’s margin-setting methodology, daily settlement prices, and other matters relevant to participation in the DCO’s clearing and settlement activities.
Similarly, the PFMIs state that a financial market infrastructure, or as relevant here a systemically important clearinghouse, should disclose its rules and procedures to participants, so that participants can have an accurate understanding of the risks, fees, and other material costs they incur by participating in the clearinghouse. Under the PFMIs, a systemically important clearinghouse should also publicly disclose fees, basic operational information, and other relevant information.
It is principally through these disclosures that members and customers will be able to understand a CCP’s risk profile. Generally speaking, the careful review of a CCP’s rule book should allow market participants to understand the operating and default procedures governing a CCP’s clearing and settlement systems.
Nonetheless, some have argued that current CCP disclosure requirements are not enough. They argue that they cannot fully assess their risk exposure to a CCP without additional disclosures, including stress test results.
Others argue that the disclosure of these results in and of themselves will result in further confusion because of differences in stress test methodologies. If two CCPs clear similar products, but use different methodologies for their stress tests, comparability of the stress test results could be meaningless or, worse, misleading without a clear understanding of the methodologies.
With this in mind, some have suggested that stress test scenarios be standardized to achieve a degree of uniformity for comparability purposes, including policy makers at the Federal Reserve. This focus on appropriate transparency will help enhance CCP risk management, but we need to carefully balance a number of potentially competing interests.
Good stress test procedures make it more likely that a DCO will be able to understand the risks posed by its members, and enable it to meet its obligations promptly. In 2011, when the Commission completed an initial set of DCO regulations, the Commission thought it appropriate to allow the DCO discretion in designing stress tests because their design inherently entails the exercise of judgment at various stages. The Commission also made clear, however, that it has the authority to evaluate the testing and require changes as appropriate.
While standardization and uniformity are appealing, they could inadvertently impede innovation and thoroughness. Would we start to teach to the test instead of evaluating and refining the stress test methodologies as appropriate?
Additionally, while in general greater transparency would seem to benefit market participants in their evaluation of CCPs, I am mindful that disclosure of stress test results without further information about the scenarios and assumptions underlying the methodologies, may cause greater confusion and hamper, rather than enhance, the due-diligence process. Regulators and markets should exercise caution and thoughtfully consider the potential consequences of providing additional transparency.
For example, if CCPs are required to disclose the assumptions and stress scenarios underlying their stress tests, could this disclosure (together with the other publicly available information) permit someone to reverse engineer the position information of targeted market participants? Could these disclosures be used as a tool for market manipulation? Does disclosing results on an anonymous basis sufficiently mitigate these concerns? Could these disclosures inadvertently damage market confidence? If global regulations are not harmonized, what could be the impact on competition?
While each of these considerations should not be ignored, I believe regulators still should pursue greater standardization for stress tests. While it might not be possible in the end for every CCP around the globe to use precisely the same assumptions and methodologies, we should be able to agree on more granular principles or guidelines than what we have today.
More standardized stress tests would enable greater coordination among global regulators in assessing the risk-management practices of CCPs, which in turn would enhance transparency for clearing members who belong to multiple CCPs. Along these lines, the CFTC should begin a public dialogue to consider these issues, and should do so through a concept release seeking comment or through one of its advisory committees as soon as practicable.
Loss mutualization is another important component of clearing. A CCP mitigates the credit risk that it takes on behalf of the original counterparties to a trade through the use of multilateral netting and by marking-to-market positions through the day with an appropriate amount of initial margin. Where a clearing member fails to meet its obligations, a CCP can allay its market risks by liquidating the defaulter’s positions in an orderly manner either on the market, or through a special auction. In ordinary circumstances, the defaulting clearing member’s initial margin on deposit equals or exceeds the losses incurred on liquidation. Where initial margin is not enough, a CCP will turn to the collateral pledged to it by the defaulter.
In the extraordinary event that the defaulter’s collateral is insufficient to cure the default, loss mutualization may be implemented. A CCP has multiple options: it could apply its own capital or “skin in the game” if it includes skin in the game in its waterfall; it could begin applying the guaranty fund contributions from its non-defaulting clearing members and place its skin in the game at the end of the waterfall; it could apply its own capital and clearing member guaranty fund contributions on a ratable basis; or it could do some variation thereof.
CFTC rules do not require skin in the game, although in fact all of the systemically important clearinghouses registered as DCOs do include their own capital in the default waterfall. For example, CME has a corporate contribution equal to $100,000,000 in its Base (futures) Guaranty Fund, and ICE Clear Credit has a priority contribution to the General Guaranty Fund of $25,000,000 and an additional pro rata contribution to the General Guaranty Fund of $25,000,000.
In the event of a clearing member default, a DCO will generally apply this capital to cover the defaulter's losses before, or on a pro rata basis with, the guaranty fund contributions of non-defaulting clearing members. That said, the amount of capital that a DCO puts at risk, and where this capital rests in the waterfall, varies.
Concerns have been raised that insufficient skin in the game results in inappropriately aligned incentives at the CCP. Many believe that a CCP’s capital contribution to the guaranty fund should always be used before the contributions of non-defaulting clearing members. Further, some argue that the size of the CCP’s skin in the game should be a fixed percentage of the guaranty fund, or equal to the largest single clearing member contribution.
A CCP’s default rules should clearly state how, when and in what order the CCP will use its default resources. In the extraordinary event that default losses are so great that a CCP exhausts the defaulter’s collateral, non-defaulting clearing members should never be surprised by how and when their guaranty fund contributions are used. The CFTC currently is actively discussing with DCOs their recovery plans, which are required to be submitted pursuant to our Part 39 rules.
I believe it would be helpful to ensure clarity for the CFTC to consider a rule addressing appropriate CCP capital contributions to the default waterfall. If we were to consider a rule in this area, I believe global harmonization on this point may be appropriate. Again, further considering this policy would be appropriately pursued through a CFTC release seeking comment or one of its advisory committees.
Recovery and Wind-down
As risk concentration increases in the cleared space, CCPs need to have a plan to either recover from severe shocks, or for the CCP’s operations to be shutdown in an orderly manner.
Here again, transparency is critical. Explicit recovery and wind-down plans help avoid an ad hoc response to a systemic event, such as a dramatic and fast market crash. Explicit plans also mean clearing members get a better picture of exactly how much risk they might face. Clearing members must take care to understand these plans, and clearinghouses should maximize details in order to enhance predictability for clearing members.
This is not to say that clearing is an inherently risky venture. As we’ve discussed, CCPs are, and will continue to be, subject to comprehensive risk-management requirements. Rather, we must recognize that no system is perfect; that despite the best efforts of regulators, the CCP, and the industry, safety measures may be overwhelmed and the CCP may face financial distress. In this event, recovery and wind-down plans can make sure that critical services can continue to be provided, or orderly wind-down can be conducted.
In addition, this loss-allocation planning may help identify critical interdependencies that would otherwise be addressed on an ad hoc basis with potentially negative impacts on market confidence. Take, for example, a bank holding company with multiple material subsidiaries that are all active at a single CCP. One subsidiary could act as the CCP’s primary custodian, another could be a clearing member with a large number of positions at the CCP, and a third could be part of a lender consortium that has agreed to provide the CCP with short-term funding in the event of a liquidity shortfall.
If an idiosyncratic event threatens the stability of the bank holding company and its material subsidiaries, the CCP’s operations and ability to meet its obligations could be severely impacted. A comprehensive recovery plan would help the CCP identify a path forward in the event of such a scenario.
Clear and detailed recovery and wind-down plans should also be designed to minimize the contagion risk to the broader market. To this end, it is imperative that regulators work closely with CCPs to help assess the effectiveness of these plans. Where a CCP operates in multiple jurisdictions, cross-border regulatory coordination and collaboration is crucial.
Clearinghouse recovery and wind-down plans are still in development. CFTC regulation 39.39 requires a systemically important DCO (as well as any DCO that elects to opt-in to the heightened SIDCO risk management standards) to maintain viable plans for recovery and orderly wind-down. Moreover, just last month, CPMI-IOSCO released their final report on the recovery of financial market infrastructures.
While market and regulatory focus typically rests with systemically important CCPs, I believe that all CCPs should be required to articulate how they would deal with unfunded credit losses and liquidity shortfalls, how they would then replenish their prefunded resources, and how they would cover losses that are not related to a participant default.
I also believe that we should carefully consider whether customer collateral provided by non-defaulting members to a clearinghouse should ever be part of a CCP's recovery plan through the use of tools such as variation-margin haircutting. I recognize that the CFTC’s rules currently do not prohibit reliance on these contributions by a DCO, but this is another issue that deserves further consideration by the CFTC and other policymakers around the globe.
It’s appropriate to ask whether pensioners and endowment beneficiaries should ever bear this risk of loss when risk management is controlled by the CCP, its clearing members, and other market participants. Removing this risk to such customers would not be inconsequential, so this would be another appropriate topic for public comment through a CFTC release or advisory committee discussion.
To conclude, clearing provides market benefits. It mitigates counterparty risk and promotes transparency. Through the use of waterfalls, recovery plans and wind-down plans, CCPs are able to mutualize losses in a predictable way, making it easier for market participants to manage large losses and liquidity shortfalls. This transparency can help bolster market confidence.
But as clearing volumes increase, the concentration of risk in CCPs will continue to grow. I believe this risk can be mitigated through both market discipline and global regulatory coordination. Regarding the latter, the CFTC should begin exploring soon whether to (1) further require standardized stress tests by DCOs to improve transparency and better enable global coordination; (2) specifically require DCOs to contribute capital to the guarantee fund and specify when it should be tapped vis-a-vis non-defaulting member contributions; and (3) prohibit a DCO from allocating losses to customers in its recovery plan.
I intend to call a meeting of the CFTC’s Global Markets Advisory Committee this winter where market participants and regulators can meet to further examine, discuss and produce a recommendation regarding these important CCP risk-management issues.
There are other issues that deserve attention from regulators as well. For example, I understand that there are potential issues surrounding the treatment of segregated margin in the leverage ratio calculation under the Basel III leverage ratio framework and disclosure requirements. I look forward to learning more about this issue over the coming months to see how we can help make sure that the tools implemented to make the financial system stronger do not inadvertently result in increased costs to customers, reduced access to the cleared markets, and a reduction in the number of FCMs willing to do business in this space.
Last Updated: December 4, 2014