A Bridge Too Var

Over recent years Value at Risk (VaR) has been accepted in the financial world, and particularly in the banking world, as the yardstick by which to measure portfolio risk. The use of VaR to assess the amount of risk capital an institution should set aside to cover its risks represents a significant increase in the importance of the role of VaR in the banking world. The science which surrounds VaR modeling can lead management to accept the results as being more accurate and reliable than they in fact are, and can also reduce awareness of the underlying factors which cause the risk in the first place. This paper demonstrates the importance of stress testing the model results themselves. Such analyses are useful in order to highlight the underlying model assumptions, the reliability of the assumptions which have been made, and (most importantly) the effect on the end results if these assumptions were to change. Stress tests are carried out on a fictitious bank to show the sort of useful information and insight they can give.