RDP 2015-02: Central Counterparty Loss Allocation and Transmission of Financial Stress 2. Background and Relevant Literature

2.1 The Systemic Importance of Central Counterparties

A CCP assists institutions in the management of counterparty credit risk by interposing itself between the counterparties to trades in securities and derivatives markets – becoming the buyer to every seller, and the seller to every buyer. These arrangements support anonymous trading, deepen market liquidity, and generally maximise the netting of exposures across participants. They also deliver operational efficiencies and help to coordinate actions in the event of a market participant's default.

At the same time, however, they result in significant concentration of risk in the CCP. This risk can crystallise if a participant defaults on its obligations to the CCP, since the CCP must continue to meet its obligations to all of the non-defaulting participants. The CCP does this by replacing the trades of the defaulted participant, but may incur losses should the replacement trade be executed at an unfavourable price. This is known as replacement cost risk.

Importantly, a participant default on obligations arising in a CCP will typically be the result of financial difficulties that it experiences, or constraints that it faces, outside of the CCP. The CCP is then a potential channel for the transmission of stress to other participants and the financial system more widely, rather than an initial trigger for stress. Policymakers acknowledge that confidence in underlying markets could be severely tested if a financial (or indeed an operational) shock disrupted a CCP's activities. These markets might then cease to function, leaving market participants unable to establish new positions or manage existing exposures.^[3]

How widely stress could ultimately be transmitted will depend crucially on the CCP's design and its risk management arrangements: whether the CCP is sufficiently collateralised; how quickly it can liquidate non-cash collateral assets or re-invested cash collateral; how effectively it manages the default and closes out the risk associated with the defaulted participant's trades (e.g. by entering into a replacement contract with a new counterparty to rebalance its position); etc. Any shortcomings in the design or risk management framework of the CCP could, in the event of a shock, have spillover effects throughout the system (RBA 2014).

The need for sound risk management arrangements to address concentration risks is well recognised both by policymakers (Tucker 2011, 2014; Bailey 2014; Cœuré 2014; Powell 2014) and by industry participants (JP Morgan Chase 2014; ISDA 2015). Accordingly, international policymakers and standard-setters have focused increasingly on CCP resilience in recent years. The G20-led international initiative to expand the scope of CCP clearing to OTC derivative markets added impetus to these efforts (G20 2009).

In particular, new international standards have been developed for the design, operation and risk management arrangements of CCPs and other financial market infrastructures (FMIs). These Principles for Financial Market Infrastructures (PFMIs), developed by the Committee on Payments and Market Infrastructures (CPMI; formerly the Committee on Payment and Settlement Systems (CPSS)) and the International Organization of Securities Commissions (IOSCO), were published in 2012. Among other things, they establish minimum requirements in all areas of risk management: e.g. credit, liquidity, investment, business, legal and operational risks (CPSS-IOSCO 2012).

2.2 Collateral and Netting

CCPs typically manage replacement cost risk through the use of variation and initial margin. Variation margin is typically exchanged at least daily – usually in cash – to reflect mark-to-market price changes on participants' outstanding positions. Initial margin is collected to cover, with a high probability, potential future exposure arising between the last variation margin payment and the closeout or replacement of a defaulted counterparty's trades. Initial margin requirements may be met either in cash or using high-quality non-cash assets that carry low credit, market and liquidity risk. Consistent with the PFMIs, initial margin is typically calibrated to at least a 99 per cent confidence interval. Only the defaulted participant's initial margin can be used in the event of a default.

A CCP's initial margin resources are supplemented with a pool of resources, typically pre-funded by contributions from all participants (along with a layer of CCP equity). This default fund is managed as a mutualised resource and sized to ensure that, in combination with the defaulted participants' margin, the CCP could withstand the default of its largest participant (Cover 1) or, in the case of CCPs that are systemically important in multiple jurisdictions, the largest two participants (Cover 2) in ‘extreme but plausible’ market conditions. There is not yet a consistent interpretation of ‘extreme but plausible’, but some CCPs target market stress equivalent to a ‘once-in-30-years’ price change.

In recent years, the use of margin has also become more commonplace in non-centrally cleared markets, although bilateral collateral agreements have to date typically covered only variation margin and not initial margin (ISDA 2014). This is set to change in light of new regulatory standards under which the exchange of both variation and initial margin will become mandatory between bilateral derivative market counterparties. These standards are due to be phased in from December 2015 (BCBS-IOSCO 2013). The BCBS-IOSCO standards will also establish a minimum level of initial margin coverage of 99 per cent.

The expansion of CCP clearing to OTC derivative markets, and margining of non-centrally cleared derivative transactions, will increase market participants' demand for high-quality assets and change how collateral markets operate (Singh 2013). There have been a number of attempts to estimate the magnitude of this increase in demand (Heller and Vause 2012; ISDA, IIF and AFME 2012; Levels and Capel 2012; Sidanius and Zikes 2012; CGFS 2013; Duffie, Scheicher and Vuillemey 2014). These studies have delivered a wide range of estimates, which largely reflect assumptions about the underlying volatility of OTC contracts, the share of the market that is ultimately centrally cleared, and the netting efficiency of alternative clearing arrangements (Cheung, Manning and Moore 2014).

Netting efficiency depends on the product and counterparty scope of a given clearing arrangement, the profile of positions, and the margining methodology applied:

• Variation margin is calculated as a net payment/receipt, based on observed price changes across all products covered by the clearing arrangement. In the case of non-centrally cleared trades, separate variation margin payments/receipts are calculated vis-à-vis each bilateral counterparty. In the case of central clearing, variation margin payments/receipts are multilaterally netted across all counterparties.
• Initial margin is similarly calculated separately vis-à-vis each bilateral counterparty in non-centrally cleared arrangements, and multilaterally across all counterparties where positions are centrally cleared. There is, however, typically less scope for netting across products in calculating initial margin requirements. For CCPs, the PFMIs require that so-called ‘margin offsets’ are limited to combinations of products where prices are significantly and reliably correlated.

In general, netting efficiencies are likely to be greater if trades are centrally rather than non-centrally cleared. However, the netting advantage of central clearing will be smaller the more concentrated is activity across counterparties, the more fragmented is central clearing, and the more directional are participants' positions (Duffie and Zhu 2011; Heath et al 2013).

2.3 CCP Recovery and Loss Allocation

Reflecting the central and systemically important role that CCPs play, policymakers have also made progress on initiatives to enhance arrangements for the recovery and resolution of CCPs and other FMIs, including providing guidance on the requirement in the PFMIs that CCPs develop recovery plans.

In the event that the market conditions prevailing at the time of bank default were more extreme than the market scenarios the CCP considered when calibrating its additional resources, or that multiple banks were simultaneously in stress, the defaulted banks' initial margin and the CCP's default fund resources could be fully depleted. To deal with such scenarios, the PFMIs require that CCPs' recovery plans include arrangements to fully address any uncovered losses and liquidity shortfalls (FSB 2013; CPMI-IOSCO 2014).

One mechanism for uncovered loss allocation that has been widely debated, and in some cases adopted, is ‘variation margin gains haircutting’ (VMGH) – see Elliott (2013), Gibson (2013), ISDA (2013), CPMI-IOSCO (2014), and Duffie (2014). This involves writing down a CCP's variation margin outflows in proportion to the amount owed to each ‘winning’ participant, so as to fully allocate the loss.

One benefit of this approach, as Gibson (2013) demonstrates, is that VMGH mimics the allocation of losses to creditors that would otherwise arise in insolvency. This reflects that a CCP does not typically issue debt; rather, its obligations arise solely from clearing on behalf of its participants. At any point in time, therefore, the participants that are owed variation margin are the CCP's creditors. It is also comprehensive; to the extent that a CCP's only obligations are variation margin payments to winning participants, these can be fully met by uncapped VMGH (Singh 2014).^[4] Finally, VMGH is reliable; since VMGH operates via a write-down of outgoing payments from the CCP, participants do not need to raise liquidity to meet their obligations in loss allocation.

However, as Duffie (2014) notes, VMGH may lead to ‘unequal and unpredictable’ loss allocation, since those who bear the loss are those that just ‘happen to be’ on the winning side of a trade on the day the CCP enters stress. Further, to the extent that some participants rely on the amounts written down in order to fund other obligations – e.g. hedges – such loss allocation could stress the solvency of participants. Those with highly directional positions vis-à-vis the CCP – including those hedging exposures outside of the CCP – are more likely to have net gains or losses than those with more balanced positions. Accordingly, directional participants will be more exposed to loss allocation under VMGH.

An alternative loss allocation mechanism is ex post calls on participants – otherwise referred to as ‘default fund top ups’ or ‘emergency assessments’. Such calls would typically be allocated proportionally with each participant's contribution to the pre-funded default fund, or its share of initial margin. Accordingly, applying this mechanism, losses may be allocated more widely, more equally and more predictably. If ex post calls were uncapped, this mechanism would also be comprehensive. However, particularly relative to VMGH, participants could face liquidity challenges in meeting their obligations.

Footnotes

Wendt (2015) describes a range of contagion channels in the event of a shock – either to a participant of a CCP, or to the CCP itself – that arise from the connections that a CCP has with its ‘ecosystem’ (i.e. connections with the financial markets that a CCP serves, its participants, and linked CCPs and other financial market infrastructures). [3]

One qualification to this is that a CCP may incur a loss in closing out its exposures beyond the mark-to-market revaluation reflected in the variation margin obligation. [4]