RDP 2015-02: Central Counterparty Loss Allocation and Transmission of Financial Stress 6. Conclusions

Applying an adaptation of the methodology developed in Heath et al (2013) and using actual data on banks' OTC derivatives positions, this paper has examined further the implications for the stability of the financial network of introducing central clearing and collateralisation of OTC derivatives trades.

At the time of writing, CCPs' risk management arrangements have been the subject of considerable policy debate. We have therefore modelled a series of extreme ‘tail-of-tail’ scenarios to examine the circumstances in which CCPs could feasibly be a channel for contagion in the event of wider stress in the financial system. We conclude that, while CCPs are central nodes in the financial network, maintaining CCP financial resources in accordance with international standards is consistent with maintaining system stability.

The analysis demonstrates that there could nevertheless be circumstances in which, having exhausted its pre-funded resources, a CCP transmits stress back to its participants by haircutting their variation margin gains. Although in our analysis there is little evidence of contagion, the precise scope for spillover stress is dependent on a number of factors, including the loss allocation mechanism applied, the distribution and direction of positions among participants, the magnitude of price changes across product classes and their co-movement, and the financial position of participants at the time of the shock.

Our analysis underscores the importance of understanding the level of stress that CCPs' pre-funded financial resources are designed to withstand, and also the channels by which losses could be transmitted back into the system in the event of a more extreme shock that depleted these resources. In considering the level of stress implied by these scenarios, it is important that attention is paid not only to assumed extreme price changes, but also the assumed co-movement between products and the closeout periods. Transparency of CCPs' own internal stress tests and other elements of the risk modelling framework would help to improve market participants' understanding and awareness of their exposure to such extreme tail events. To the extent feasible, this could usefully be combined with regulatory stress tests of CCPs' exposures and network analysis using individual banks' position-level exposure data.

We leave to future research further refinement of analytical techniques to deepen the analysis of how CCPs could transmit stress under alternative loss allocation mechanisms once pre-funded resources have been depleted. Alternative distributional assumptions for price changes could, for instance, be considered. Other topics for future research may include further analysis of the transmission of liquidity risk, and other channels for contagion, such as links between CCPs.

More work would also be useful on the implications of alternative loss allocation mechanisms for participant incentives. This could consider, for instance, the risk that some participants ‘walk away’ from a CCP in stress to avoid future obligations in loss allocation. There could also be implications for the use of CCPs by those with more directional positions – including perhaps investment funds and other ‘end users’ – that might be more exposed to mechanisms such as VMGH. Finally, the analysis in this paper has taken banks' OTC derivative positions, liquidity holdings and capital positions as given. In practice, it is likely that these would all change endogenously in response to alternative market structures.