The World Federation of Exchanges (WFE) published new research on measuring the procyclicality of central clearing counterparty (CCP) initial margin models. It studies the standard measures of initial margin model procyclicality, which are random variables subject to uncertainty.
To date, this random element has drawn little attention but has significant consequences, as the ability to adequately measure the responsiveness of a model is fundamental to the discussion and implementation of any procyclicality mitigation tool. Therefore, to be robust, any decision or policymaking based on these measures must consider the impact that uncertainty will have on expected outcomes.
The paper aims to estimate such impact, examining the case of some typical margin models, both empirically and within a Monte Carlo simulation setting.
Initial margin models are a fundamental part of CCPs’ management of counterparty risk: margin requirements ensure that the exposure to a failing member is sufficiently collateralized and, in this way, contributes to the CCP buffer against financial contagion. On the other hand, attempts to mitigate procyclicality by altering the responsiveness of a margin model are limited by the fact that models need to be risk sensitive to ensure that the CCP remains adequately collateralized at all times, and by the need to keep central clearing economically efficient.
The results presented in the WFE Research Working Paper show that, in addition to the above limitations, there is a significant amount of uncertainty when measuring model responsiveness. This has important implications:
- To be robust, a decision or policy-making process based on the standard procyclicality measures requires quantifying the presence of uncertainty in the measurements, for example, by estimating the sensitivity to the choice of scenario.
- From a policy perspective, without quantifying the uncertainty surrounding procyclicality measurements, there is a risk of prescribing rules with an unknown but significant chance of being ineffective.
- It is also difficult to judge whether a behavior that may be deemed “too procyclical” in one particular scenario reflects a failure of the model’s anti-procyclicality credentials or is the likely consequence of the uncertainty in the measurements.
- When estimating the trade-off between costs, procyclicality, and risk sensitivity, the additional parameter uncertainty that some anti-procyclicality (APC) tools bring, plus the uncertainty in the estimation of costs, increases the probability that, in some future scenarios, these APC tools could be inefficient, or even detrimental, from a cost-benefit perspective.
- The greater the uncertainty in responsiveness measurements, or the more sensitivity to the underlying scenario, the less certainty we can place on the benefits of approaches to mitigate future responsiveness based on one-fits-all, hard-rules.
Pedro Gurrola-Perez, head of Research at the WFE and the author of the paper, said in a statement: “Without doubt, CCPs should continue adopting margin arrangements that, ‘to the extent practical and prudent, limit the need for destabilizing, procyclical changes’, as established in the PFMIs. But, as this paper shows, uncertainty in model procyclicality forecasts limits what can be achieved through model-focused, hard-rule approaches. It underlines the importance of expert judgement to address model responsiveness on a case-by-case basis.”
Nandini Sukumar, WFE’s chief exec, said in a statement : ‘’The WFE Research Working Paper shines a light on a little studied, yet critical, element of margin models. We believe that policy positions should always be based on fact, empirical evidence and data. As an industry and organization, we look forward to engaging in conversation with our stakeholders on the findings of the paper.”