VAR: risk mitigant or amplifier?

Value-at-risk is a far-from-perfect risk measure. Jon Danielsson, Ulrich Klueh and Laura Kodres take a close look at the lessons to be learnt from its use in stressful or volatile periods, such as the current financial crisis

The current financial market crisis has elicited accusations that mechanistic adherence to risk management systems, such as value-at-risk (VAR) market risk measures, may have been a contributing factor. The model in the box opposite explores how, in combination with desired capital levels, risk management techniques - including VAR-type techniques - can lead to destabilising asset price behaviour in certain circumstances.

It is true that VAR-type techniques can help risk managers judge and potentially mitigate risks, thereby protecting their individual institution from adverse events. However, the interaction of rational responses from individual institutions holding similar positions during market stress can collectively cause detrimental asset price dynamics.1 While difficult to anticipate, risk managers need to be aware that the rigorous use of some risk management techniques can have negative systemic implications.

The VAR measure is one of several measures that seek to unify traded positions across a number of different assets to calculate the potential loss on a portfolio that would exceed a given dollar level a certain percentage of the time. That is, VAR is an estimate of the expected loss that an institution is unlikely to exceed in a given period with a particular degree of confidence, often assumed to be 95% or 99% of the time. It is used broadly in the financial industry as one of a number of metrics to assess market risk.

It is easy to demonstrate that the VAR measure - or VAR-type measurements - increases for a given portfolio of assets in stressful times when, as is typically the case, the volatility of the underlying assets rises or the correlations among them increases.

VAR: risk mitigant or amplifier? (PDF, 223KB)