Memo #
31821

CPMI and IOSCO Issue a Discussion Paper on CCP Default Management Auctions
| Print

[31821]

June 24, 2019

TO: Derivatives Markets Advisory Committee
ICI Global Trading & Markets Committee RE: CPMI and IOSCO Issue a Discussion Paper on CCP Default Management Auctions

 

The Committee on Payments and Market Infrastructures (CPMI) and the International Organization of Securities Commissions (IOSCO) recently published a discussion paper on central counterparty (CCP) default management auctions.[1] The discussion paper describes the current practices at various CCPs and identifies the types of factors that CCPs consider when planning and executing default management auctions.

Comments on the discussion paper are due by August 9, 2019. If you would like to share your views on the discussion paper, please contact George Gilbert at george.gilbert@ici.org and Giles Swan at giles.swan@ici.org by Monday, July 8, 2019. This memorandum provides background on CCP default management auctions, reviews the key aspects of default auctions identified by the discussion paper, summarizes the portion of the report that addresses client participation in CCP default auctions, and provides, in the annex, the questions posed by the CPMI and IOSCO.

Introduction to CCP Default Management Auctions

A central counterparty, commonly referred to as CCP or clearinghouse, is a financial market infrastructure that concentrates on “trade management, position management, collateral and risk management, and delivery management.”[2]  Its primary objective is to provide stability and efficiency to the financial markets by reducing counterparty, market, legal, and default risks for traders. To “clear” a trade, the counterparties submit their transaction to a CCP (either directly or indirectly, e.g., if one or both of the counterparties is not a clearing member of the CCP). If the CCP accepts the transaction for clearing, the original trade between the buyer and the seller is novated and replaced by two new transactions, with the CCP becoming the buyer to the seller and the seller to the buyer. Through the novation process, the counterparty credit risk between the buyer and seller is mutualized among the CCP and its clearing members.

Expeditious and competent management of a clearing member’s default is essential to a CCP’s viability. Upon a clearing member’s default, a CCP must take “timely action to contain losses and liquidity pressures and continue to meet its obligations” in regard to its other clearing members.[3] A CCP “should have effective and clearly defined rules and procedures” that guide such action and help limit the effect of one member’s failure on other participants.[4]

When a clearing member defaults, a CCP must attempt to restore a matched book, which it typically will do by transferring a defaulting participant’s positions, or subset thereof, to a non-defaulting party. A default management auction is one of the tools that a CCP may use to effect such a transfer.[5] The auctions typically require active involvement of a group of auction participants, which creates a necessity for clarity and understanding of the auction procedures, the roles and responsibilities of the participants, and other operational considerations. If an auction fails, a CCP may use other tools to restore a matched book, including partial or full tear-ups of the positions held by other clearing members or clients of clearing members. 

Key Aspects of the Default Management Auctions of a CCP

The paper focuses on the following five key aspects of the CCP default management auctions:

  • Governance (Chapter 3). The paper discusses the roles and responsibilities of key stakeholders in a CCP’s default management auction. The paper explains that an effective auction process includes specifying the roles and responsibilities of the auction participants and a CCP’s board of directors, management, and other personnel, who may be involved in the auction process.
  • Considerations for a successful default management auction (Chapter 4). The paper outlines considerations of a successful default management auction and, by contrast, identifies scenarios in which a CCP may determine an auction to be unsuccessful. It also identifies key elements a CCP considers when designing an auction process, including the hedging strategy and the auction design elements, such as the auction format, timing, and bidding requirements.
  • Operational considerations (Chapter 5). The paper describes the operational issues a CCP considers when planning and executing a default management auction, such as effective communication and testing exercises under business-as-usual conditions.
  • Client participation (Chapter 6).  For a more detailed discussion, please refer to the next section of this memorandum.
  • Default of a common participant across multiple CCPs (Chapter 7). The paper identifies potential issues inherent when two or more CCPs conduct auctions concurrently, thus creating further operational and/or financial strains on auction participants.

Client Participation in Default Management Auctions

A CCP may permit clients of its clearing members to participate in a default auction. This could increase the number of potential bidders in the auction and, potentially, increase the competitiveness of the bids as well as the likelihood of the auction’s success. In addition, allowing clients of clearing members to participate in a default auction may diversify the types of bidders and distribute the risk of the portfolio more broadly across market participants. The primary incentive for the participating clients is to improve their financial position, e.g., by obtaining gains from acquiring positions, or avoiding potential negative outcomes that could arise from a failed auction, such as a partial or full tear-up of their existing positions with the CCP.

The discussion paper explains that client participation in a default management auction generally takes one of two forms:

  • A direct participation model, in which a client communicates directly with the CCP to participate in the auction and to submit its bids with the consent or authorization of the client’s clearing member. The client in this model resembles a clearing member with respect to its direct communications with the CCP.
  • An indirect participation model, in which a CCP permits a client to participate in an auction by submitting its bids through its clearing member. This model more closely resembles typical clearing arrangements, in which a clearing member intermediates all communication between its client and the CCP.

The discussion paper notes that a CCP and its clearing members generally address five factors when deciding whether to permit client participation:

  • Liability of clearing member. When a client submits the winning bid in a default auction—the client’s clearing member becomes ultimately responsible for the obligation arising from the bid. To mitigate this risk, a CCP generally requires some form of consent or authorization from the clearing member before allowing clients to participate in the auction. The CCP also may require a clearing member to approve or deny a client’s bid during the auction before the bid is accepted by the CCP.
  • Client incentives to bid competitively. The discussion paper notes that clients might have less incentive to bid competitively than clearing members because they do not contribute to the CCP’s default fund and, consequently, face fewer direct financial consequences from an auction than clearing members. To increase the likelihood that clients will bid competitively, a CCP may require a client to contribute an established amount prior to participation in an auction, which a CCP may use to cover losses in a manner similar to the default fund contributions of a clearing member. In addition, in the event of a failed auction, clients may become subject to a CCP’s recovery tools, such as variation margin gains haircutting and position tear-ups. .
  • Legal readiness. A CCP generally requests that clients execute agreements, such as non-disclosure agreements or agreements to comply with auction rules.
  • Operational readiness. Success of a default management auction often depends on timely pricing and bidding. In the case of direct client participation, a CCP generally establishes and tests communication channels with participating clients on an ex ante basis. A CCP may further take steps to ensure that participating clients have sufficient experience in the market to evaluate the portfolio and provide competitive bids.
  • The risk of information leakage. All auction participants receive confidential information about the portfolio that will be auctioned. Leakage of this information may create undesirable consequences, such as a client trading against the defaulted portfolio. The tools that a CCP may employ to avoid this result include splitting the defaulted portfolio, requiring two-way pricing, or enforcing non-disclosure agreements.

The paper further discusses several factors that may make client participation feasible only for the “largest and most sophisticated” clients. First, clients must be able to risk-manage exposures related to the auction portfolio, which may limit participation to clients with a significant amount of assets under management and/or capital. Second, clients must be able to price and bid on the auction portfolio in a compressed time frame, which requires a certain level of operational capabilities and the investment of resources to ensure operational readiness. Third, the structure of the auction portfolio might not be suitable for all types of clients.

Annex: Questions Posed by the Discussion Paper

Roles and responsibilities (Chapter 3)

Q1. What are the considerations for a CCP’s board when determining whether and how to assign tasks related to the planning and conduct of default management auctions within the CCP’s risk management framework? How does the CCP’s board identify potential limits to the assigned responsibilities? Q2. What different considerations may apply when a CCP’s board establishes procedures for consulting external experts, such as independent consultants or clearing members, when designing or conducting a default management auction? How does a CCP’s board address such concerns?

Considerations for a successful auction (Chapter 4)

Q3. Do you agree with the description of a successful auction in the discussion paper? Do you agree with the scenarios identified that would constitute an unsuccessful auction, and are there additional such scenarios? Q4. What are the primary challenges to achieving a successful default auction? In addition to those included in the discussion paper, are there other elements in the design of an auction that a CCP could consider in order to increase the likelihood of a successful auction? Q5. What process/set of factors, including applicable governance, is used/considered to determine whether an auction is successful or unsuccessful? What governance would apply to this determination, including the decision whether to run an additional auction (as opposed to using other tools) and why? Q6. What are the considerations for CCPs in choosing to utilise auctions as a default management tool? What product categories are most suitable for auctions and what product categories are least suitable for auctions? How do you assess suitability? Q7. In addition to those outlined in the discussion paper, are there other considerations that may be useful for a CCP to take into account when designing its hedging strategy, including circumstances where a CCP may wish to delay hedging? Q8. How do you incorporate cross-margining arrangement considerations in the hedging strategy and in the broader auction design process? Q9. The discussion paper notes that, with respect to hedging, execution methods vary and depend on a CCP’s choice of hedging instruments. What methods are used for hedging, and what is the rationale for implementing (or not implementing) a particular method? Q10. What factors, other than those identified in the discussion paper, do you see as relevant when determining how to split a portfolio? Are there situations where certain factors would be more important than others? Please provide examples. Q11. The discussion paper describes two common auction formats. Are there other auction formats not included that could be beneficial for a CCP to consider employing? What factors influence the decision to implement (or not implement) a particular auction format?
  1. Besides promoting competitive bidding, are there other considerations for choosing two-way pricing? Are there circumstances where it would be beneficial or circumstances where it might not be appropriate? If so, please describe.
  2. What are the considerations for choosing to use a reserve price in an auction? Are there circumstances where it would be beneficial or circumstances where it might not be appropriate? If so, please describe.
Q12. The discussion paper highlights two factors that affect the amount of time auction participants may need to evaluate a portfolio and submit bids. Are there other factors that are important to consider? Is there a minimum time period that a CCP should consider providing to auction participants? Q13. If a clearing member contributes a “significant” part of the default fund, should that clearing member automatically be included in the auction process? What reasons are there for not including the clearing member? Q14. The discussion paper discusses the trade-off between flexibility and predictability. How do you assess these trade-offs? Can you elaborate on the ways you provide for predictability while still maintaining flexibility (e.g., establishing rules and conditions to govern the determination of auction parameters)? Q15. If a CCP uses juniorisation as an incentive to encourage competitive bidding, and in a scenario where the CCP has invited only a subset of participants to an auction, how will the CCP apply the juniorisation to the clearing participants who were not invited?

Operational considerations (Chapter 5)

Q16. CCPs may distribute information that would help auction participants estimate the potential impact of a successful auction bid on their margin requirements. Besides those that members and clients would having during BAU, what information (and at what level of detail) or tools would be most useful for calculating these estimations and why? Q17. The sharing of confidential information (i.e., the defaulted participant’s portfolio) carries potential risks, as discussed in the paper. What are the potential risks associated with information leakages, how does the CCP balance such risks with other objectives (e.g., sharing sufficient information for a successful auction), and what are the measures the CCP uses to address such risks? Q18. CCPs use various modes of information transmission during a default management auction (e.g., email, web-based portal). Can you elaborate on which modes are the most effective in which circumstances and whether it varies depending on the type of information, and why? Would you consider web-based portals a best practice? If so, why? Q19. What are the challenges and trade-offs of creating a realistic default management testing exercise? What processes are used to create the scenarios used in such exercises? Q20. There may be benefits in pursuing greater standardisation and harmonisation across CCPs in relation to certain operational elements which support execution of an auction.
  1. For example, should auction portfolio files be in a standard (or partially standardised) format (for different product types)? If so, which aspects of the portfolio file would benefit the most from cross-CCP standardisation (e.g., file type, layout, order of information, or content)?
  2. Besides CCP portfolio files, which other operational elements would benefit (the most) from greater standardisation and harmonisation across CCPs?
  3. Are there specific operational elements or areas where standardisation and harmonisation may not be helpful?

Client participation(Chapter 6)

Q21. For which markets, asset or product classes and client types would client participation be most feasible and/or desirable? What would be the incentives for clients to participate in auctions? Does this differ for direct vs indirect client participation? Please elaborate in your response. Q22. The discussion paper describes some ways to address the risks borne by a clearing member arising from its clients bidding in an auction. Are there additional ways to address the risks? Are there incentives that a CCP could employ to encourage client participation in an auction (e.g., ways to encourage clearing members to facilitate their clients’ participation)?
  1. One option for addressing a disparity in incentives between clearing members and clients is to require clients to contribute an established amount to the default fund prior to participating in an auction. What are the implications of this requirement (such as regulatory, economic or contractual implications) and how can a CCP address these implications?

Default of a common participant across multiple CCPs (Chapter 7)

Q23. The discussion paper suggests that the conduct of multi-CCP default management exercises may provide useful insights into the hedging and auction procedures, should these be conducted by multiple CCPs concurrently. Can you elaborate on what specific insights could be obtained in relation to hedging and auctions via these multi-CCP default management exercises, if possible with concrete examples? Q24. Feedback from the industry suggests that introducing a cap on the number of traders that can be seconded to multiple CCPs from a particular common clearing participant at any one time may mitigate the potential burden on clearing participants’ participation in DMGs. How could such caps be instituted and implemented in practice? What could be the challenges of introducing such caps? Apart from caps, are there other options a CCP could consider to mitigate this potential burden? Q25. Are there efficiencies or benefits to be gained from CCPs coordinating their respective default management auctions or hedging arrangements? If so, how?
  1. Are there any arrangements that could be coordinated ex ante (e.g., cross CCP netting arrangements)? How could these arrangements be established? What would be the challenges with these arrangements? How could these challenges be mitigated?

General

Q26. Are there any additional points of consideration that would contribute to a successful auction that are not mentioned in this discussion paper? If so, what are they? Q27. What are the potential areas in the context of default management auctions where guidance might be most welcome?

 

Sofya Bakradze
Legal Intern

Giles Swan
Director of Global Funds Policy, ICI Global

George M. Gilbert
Assistant General Counsel

 

endnotes

[1] Committee on Payments and Market Infrastructures and Board of the International Organization of Securities Commissions, A discussion paper on central counterparty default management auctions (June 2019), available at https://www.iosco.org/library/pubdocs/pdf/IOSCOPD631.pdf.

[2] For further discussion, see Functional Definition of a Central Counterparty Clearing House (CCP), published by European Securities and Markets Authority, available at https://www.esma.europa.eu/file/5752/download?token=1Icjakov.

[3] See Principles for financial market infrastructures (April 2012), Principle 13, published by IOSCO, available at https://www.iosco.org/library/pubdocs/pdf/IOSCOPD377-PFMI.pdf.

[4] Id.

[5] Other ways include a sale on an exchange or a negotiated bilateral sale.