The drawbacks of VaR's, or risk management's Byzantine discussion

Abstract

Crises, such as the present economic downturn, are also usually times for much needed critical thinking about wrongdoings and best practices. Yet these periods are commonly full of commentators saying ‘I told you so’. Much like the attacks on value at risk (VaR) in the 1990s, today's discussions seem to be looking for a risk management tool to blame, at least in part, for the catastrophic losses experienced by banks and other financial intermediaries. But is the technique really to blame? Whatever the answer, looking for a scapegoat will help neither the financial system nor the risk management profession and, more importantly, will not prevent future generations from making the same kinds of mistake. As such, this paper does not set out to defend or propose enhancements to a particular model. Instead, it holds that tools are only as good as the framework, judgment and expertise used to implement them. Consequently, a constructive debate on risk management should not focus on the appropriateness of certain tools, but rather on the approach and assumptions used to implement them, thereby characterising uncertainty more robustly. In short, and although VaR is used as an illustrative case, this paper calls for more realistic assumptions and modelling paradigms, for now and for the years ahead.