Battle Lines Drawn Over CCP Resolvability

Introduction

In the context of the continuing industry and regulator discussion regarding CCP resolvability, last week ISDA published a position paper entitled “CCP Loss Allocation at the End of the Waterfall”.  The paper addresses two scenarios:

  • “Default Losses” – i.e. losses that remain unallocated once the ‘default waterfall’ is exhausted following a clearing member (“CM”) default; and
  • “Non-default Losses” – i.e. losses that do not relate to a CM default but exceed the CCP’s financial resources above the minimum regulatory capital requirements.

Default Losses

ISDA recognises the importance of central clearing for standard OTC derivatives, the difficulty of achieving optimal CCP recovery and resolution and the fact that no loss allocation system can avoid allocating losses to CMs.  It takes the view that residual CCP losses should be borne not by the taxpayer, nor solely by surviving CMs who as guarantors have no control over losses.  Rather, ISDA believes that all CMs with mark-to-market gains since the onset of the CCP default should share the burden of CCP losses.  Accordingly, ISDA is an advocate of Variation Margin Gains Haircutting (“VMGH”) being applied at the end of the default waterfall.

Under a VMGH methodology, the CCP would impose a haircut on cumulative variation margin gains which have accumulated since the day of the CM default.  In doing so, ISDA believes that:

  • losses fall to those best able to control their loss allocation by flattening or changing their trade positions;
  • CMs with gains at risk are incentivised to assist in the default management process; and
  • in the event that the CCP runs out of resources, VMGH mimics the economics of insolvency.

ISDA believes that a VMGH methodology should not have an adverse impact on the ability of a CM to net exposures or gain the appropriate regulatory capital treatment for client positions held at the CCP[1].  In contrast to contractual tear-up provisions or forced allocation mechanisms, VMGH allows a CM to assume that its portfolio of cleared transactions outstanding as of any given date will be the same as of the point of a CCP’s insolvency (because there is no mechanism by which they can be extinguished prior to any netting process).  As such, because it has certainty with respect to its legal rights in the CCP’s insolvency, the CM should be able to conclude that netting sets remain enforceable.  In addition, to the extent that VMGH provides incremental resources to the CCP, ISDA believes that it effectively protects initial margin held at a CCP and therefore strengthens segregation.

In theory, VMGH should always be sufficient to cover a defaulting CM’s mark-to-market losses in the same period.  However, if in practice this was not the case (e.g. because the CCP was not able to determine a price for the defaulting CM’s portfolio) and in the absence of other CMs voluntarily assuming positions of the defaulting CM, ISDA advocates a full tear-up of all of the CCP’s contracts in the product line that has exhausted its waterfall resources and has reached 100% haircut of VM gains.  ISDA contends that there should be no forced allocation of contracts, invoicing back, partial non-voluntary tear-ups, or any other CCP actions that threaten netting.  Furthermore, prior to the point of non-viability, ISDA believes that resolution authorities should not be entitled to interfere with the CCP’s loss allocation provisions (as detailed within its rules) unless not doing so would severely increase systemic risk.

Non-default Losses

An example of Non-default Loss (“NDL”) would be operational failure.  ISDA views NDL in a different light to Default Losses believing there to be no justification for reallocating NDL amongst CMs and other CCP participants.  Accordingly, it does not believe that VMGH (or similar end-of-the-waterfall options) are appropriate for allocation of NDL.  Rather, it considers that NDL should be borne first by the holders of the CCP’s equity and debt.

Conclusion

The ISDA paper is a useful contribution to the ongoing discussion around CCP resolvability.  It suggests a sensible CCP default waterfall,[2] but is probably most noteworthy for its opposition to initial margin (“IM”) haircutting as a resolution tool.  In ISDA’s view, IM haircutting would distort segregation and “bankruptcy remoteness”.  In doing so it would have significant adverse regulatory capital implications and would create disincentives for general participation in the default management process.  In this sense, it adopts the opposite position to that detailed by the Committee on Payment and Settlement Systems (“CPSS”) and the International Organization of Securities Commission (“IOSCO”) in their recent consultative report on the Recovery of financial market infrastructures (see this blog post for more detail).  CPSS/IOSCO see IM haircutting as an effective tool which may facilitate access to a much larger pool of assets than VMGH.

There is general agreement on the principle that the taxpayer should never again have to pick up the tab following the failure of a systemically important firm.  On this basis alone, one suspects that IM haircutting will ultimately be included in the suite of resolution tools, if only to act as additional buffer between derivatives losses and the public purse.  In fairness, it’s difficult to see how a general tear-up of contracts is consistent with one of the underlying goals of CCP resolution – to ensure the continuity of critical services.  Ultimately, however, we will have to wait to see whether the contagion which may result from ISDA’s tear-ups outweighs the regulatory impact associated with CPSS/IOSCO’s IM haircutting.


[1] Pursuant to Article 306(1)(c) of the Capital Requirements Regulation, a CM will likely have to be able to pass on the impact of a CCP default to its clients in order to attract the appropriate regulatory capital treatment

[2] See page 8

FMI Recovery Continues to Take Centre Stage

On 12 August 2013, the Committee on Payment and Settlement Systems (CPSS) and the International Organization of Securities Commissions (IOSCO) published a consultative report on the Recovery of financial market infrastructures (FMI) and a cover note detailing the specific issues on which comment is sought.  The consultation period ends on 11 October 2013.

The report provides guidance to FMIs and authorities on the development of robust recovery plans. It was produced in response to requests for additional guidance following the publication in July 2012 of the CPSS-IOSCO report on Recovery and resolution of financial market infrastructures and supplements the Principles for financial market infrastructures, published in April 2012.  It should also be read in conjunction with the recent Financial Stability Board consultation regarding the Application of the Key Attributes of Effective Resolution Regimes to Non-Bank Financial Institutions (see this blog post for more detail).

The report includes an interesting high-level discussion on the design and use of recovery tools, touching on issues such as:

  • transparency versus flexibility;
  • loss allocation waterfalls;
  • mutualisation of loss versus targeting of losses;
  • the balance between pre-funded and non-prefunded resources;
  • incentivising stakeholders to:
    • manage risk vis-à-vis FMIs;
    • assist in default management processes; and
    • maintain/increase the use of an FMI rather than settle bilaterally.

Recovery tools are separated into five categories:

  • tools to allocate uncovered losses caused by member default;
  • tools to address uncovered liquidity shortfalls;
  • tools to replenish financial resources;
  • tools to allocate losses not related to participant default;
  • tools for central counterparties (CCPs) to re-establish a matched book; and
  • tools to address structural weaknesses (not addressed further in the report).

Tools to allocate uncovered losses caused by member default include:

  • cash calls on participants;
  • position based loss allocation tools;
  • variation margin haircutting; and
  • initial margin haircutting.

There is an interesting discussion as to how cash calls should be calculated – whether fixed or as a proportion of default fund contributions, volumes or positions – and the pros and cons of capped versus uncapped calls.  The potential downsides of this tool are highlighted, particularly the pro-cyclical effects they can have as well as the credit risk inherent within the application of this tool.  Position based loss allocation tools include borrowing funds owed to participants (loans, repos etc.), variation margin haircuts, reduced payouts and contractual tear-ups.  On the plus side, they are generally regarded as being a comprehensive solution to the problem of shortfalls which can be executed immediately and involve no performance risk vis-à-vis the FMI participants.  On the down side, it was noted that they can have a negative effect on participants’ confidence in the FMI.  Variation margin haircutting in particular may not allocate losses to those best able to cope with them.  Initial margin haircutting is also regarded as an effective tool and one which may facilitate access to a much larger pool of assets than variation margin haircutting.  Unfortunately, it is also one which exposes a CCP to risk during the period where initial margin is being replenished, may well have pro-cyclical effects and would likely result in a capital charge for participants (due to the initial margin not being held in a bankruptcy remote manner).

Tools to address uncovered liquidity shortfalls mainly involve obtaining liquidity – either from third-party institutions or from non-defaulting participants.  In the case of the latter, there is a discussion as to whether participants which are owed money by the FMI should be accessed first as they would not be required to pay-in money to the FMI, thus reducing performance risk associated with the use of this tool.

In the main, the discussion regarding tools to replenish financial resources focuses on cash calls on participants (see above for more detail).

Tools to allocate losses not related to participant default range from simple loss allocation, to capital raising, insurance and indemnities.  The risks inherent with respect to insurance (the time taken to process a claim, the uncertainties of a pay-out and the capped nature of most pay-outs) are noted.

Discussion with respect to the tools for CCPs to re-establish a matched book focus, in the first instance, on incentivising participants to accept unmatched contracts.  This can be achieved by making the default fund contributions associated with these trades the last to be used in a default scenario.  Whilst this is regarded as a transparent tool which mitigates the risk of a failed auction, it is accepted that it does not represent a comprehensive solution.  As such, consideration is given to the forced allocation of contracts.  Whilst this is a comprehensive solution which involves no performance risk, it is noted that it may result in contracts being allocated to participants which are unable to manage them.  The option of contract termination is also discussed.  Whilst another comprehensive and effective solution, the use of this tool exposes participants to replacement cost risk, is disruptive to the market where contracts are terminated, may actually trigger the spread of contagion and could create disincentives for firms to participate (as their hedged positions could effectively be turned into directional ones at the option of the CCP).  As such, it is questionable whether the use of this tool actually contributes to achieving the objective of continuity of key services.  The conclusion is that the use of contract termination (at least full, as opposed to partial, termination) should be avoided to the extent practicable.  Indeed, the report suggests that the use or imminent use of such a tool may itself be regarded as a trigger for resolution.

RRP for Non-Banks – Is Your Data Up to Scratch?

On 12 August 2013, the Financial Stability Board published a consultation document regarding the “Application of the Key Attributes of Effective Resolution Regimes to Non-Bank Financial Institutions”, inviting comments by 15 October 2013.

The consultation document proposes draft guidance on how the Key Attributes should be implemented with respect to systemically important non-bank financial institutions.  It deals with three main areas:

  • The resolution of financial market infrastructure (FMI) and systemically important FMI participants;
  • Resolution of insurers; and
  • Client asset protection in resolution.

The proposed rules are, to a large extent, little more than the formalisation of existing thought and best practice regarding the resolution of non-bank financial institutions.  However, this does not detract from the value of the document.  Indeed, it highlights the practical challenge that institutions which are subject to the rules will face in providing the data necessary to facilitate the implementation of resolution measures by regulators.

Both FMIs and insurers will be required to maintain information systems and controls that can promptly produce, both in normal times and during resolution, all data needed for the purposes of timely resolution planning and resolution.  In the case of FMIs, this will include:

  • Information on direct and indirect stakeholders, such as owners, settlement agents, liquidity providers, linked FMIs and custodians;
  • Exposures to each FMI participant (both gross and net);
  • Information on the current status of obligations of FMI participants (e.g. whether they have fulfilled their obligations to make default fund contributions);
  • FMI participant collateral information, such as:
    • location;
    • holding arrangements; and
    • rehypothecation rights; and
    • netting arrangements.

Insurers will also be required to generate data regarding:

  • sources of funding;
  • asset quality and concentration levels; and
  • derivatives portfolios.

In addition, any entity holding client money, must have the ability to generate a wide variety of data that would facilitate its speedy return in a resolution scenario.  That data must be in a format understandable by an external party such as a resolution authority or an administrator and includes information on:

  • the amount, nature and ownership status of client assets held by the firm (directly or indirectly);
  • the identity of clients;
  • the location of client assets;
  • the identity of all relevant depositories;
  • the terms and conditions on which client assets are held;
  • the applicable type of segregation (e.g. “omnibus” or “individual”);
  • the effects of the segregation on client ownership rights;
  • applicable client asset protections (particularly where client assets are held in a foreign jurisdictions);
  • any waiver, modification or opting out by a client of the client asset protection regime;
  • the ownership rights of clients and any potential limitations to those rights;
  • the existence and exercise of rehypothecation rights; and
  • outstanding loans of client securities arranged by the firm as agent, including details of:
    • counterparties;
    • contract terms; and
    • collateral received.

If the experience of banks is anything to go by, the capture, analysis, delivery and updating of this type of data is a significant undertaking.  The FSB is clearly laying out its intentions and the direction of travel on this issue.  As such, non-bank financial institutions would do well to start analysing their capabilities in these areas, with a view to upgrading their data architectures where necessary.

Single Point of Entry or Multiple Point of Entry: the Choice is Yours?

Here is a link to an article in today’s FT explaining that, following the FSB guidance issued on 16 July 2013 (see this blog post for more detail), banks seem likely to be given more ‘choice’ between single point of entry and multiple point of entry.  This seems to represent a subtle shift away from the previous consensus that had been developing within regulatory circles regarding the benefits of single point of entry over multiple point of entry.  However, the quid pro quo is that banks will have to implement potentially wide-ranging changes in order to make their business models more consistent with their chosen resolution mechanism.