Mean VaRiance…

It is a truism that past is not prologue; and that one should be careful not simply to project what happened (or did not happen) before as the basis of a decision or expectation about the future.

This fact was admirably demonstrated yet again recently when banks’ Value at Risk (VaR) models were shown to be inadequate, misleading and, therefore, dangerous in the wake of the Swiss National Bank’s decision to end the Swiss Franc’s peg to the Euro. Criticism of the VaR approach to risk management is not new. It was amply aired before and during the post-2007 financial crisis, when events that would supposedly not happen more than once in billions of years were a little too common! Therefore, it is a little disconcerting that banks (and their regulators) seem to continue to rely upon them to the extent that they do.

Those of us that have long experience and retain our memory of past events know that in the real world, extreme and unexpected events have a tendency to happen rather more frequently than quantitative and VaR-like models predict, because those models are usually rooted in observing the past over a relatively short period of time; and, thus, often do not record events or volatility that a sufficiently-long institutional memory would record and factor in.

Perhaps, risk managers, who by temperament are probably often Stoics (or should be!) might practice on a daily basis, one of the key (and most difficult) tenets of Stoicism- the premeditatio malorum, which involves imagining all the worst things that could occur in the immediate future and how one would deal with them? The idea is to develop resilience. Not surprisingly, this bears a remarkable resemblance to asking an individual what keeps them up at night in terms of events that could damage or destroy their business, franchise or ability to regulate or govern. There are developments of VaR, such as Tail VaR and Extreme Value Theory (EVT) that are intended to address standard VaR’s limitations, but the prevalence of the “once in 100 years” (i.e., 99.5% confidence level) approach as the basis for stress-testing and setting capital levels used in many financial regulatory regimes remains somewhat surprising; perhaps partly explicable because using more robust measures would require more capital and so reduce returns.

Naturally, we at Awbury are not saying that models such as VaR have no value or purpose, or that stress-testing models are invalid, but rather that their constraints and limitations need to be recognized and understood. Given our focus on E-CAT, providing our clients with ways to manage exposures to significant and complex financial and economic risks, we are careful both to identify and assess the impact of all the risks that could have a material adverse impact, and to try to think in ways that are not constrained by any particular worldview or intellectual framework, while avoiding indulging in scenarios that are simply not realistic. Of course, we also try to retain a healthy paranoia that we may not have foreseen a key risk; and so structure our transactions in ways that provide a balanced ratio of reward to risk, while addressing our clients’ key needs and concerns.

– The Awbury Team


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.