Modeling and measuring operational risk

Events such as those which occurred in Barings, Daiwa, and Sumitomo helped focus attention on an important type of risk which a bank runs, the operational risk (OR), which encompasses de®ciencies in information systems and internal controls, and includes legal and personnel events that could result in unexpected losses. This type of risk is closely associated with human error, system failure, fraud, and inadequate procedures and controls. Several banks have developed state-of-the-art market and credit risk measurement, management, and pricing tools, but they will not protect a ®nancial institution if, for example, a derivative salesman decides to input a bogus volatility ®gure in a sophisticated option pricing system to make the product more attractive for an investor. This happened recently in a large British investment bank which reported over £90 million of losses due to the poor checking of traders' inputs in their derivatives pricing system. Operational risk related losses can be massive. We show here that it is possible to develop a quantitative model, very similar to those of credit and market risk, for measuring operational risk.