Analysis of Australian Banks’ Operational Risk Events Using Evolutionary Techniques

This paper analyses operational risk events from the ORIC database for Australian banks using a technique borrowed from biological evolutionary analysis. The technique groups risk events by their characteristics to indicate a hierarchical structure of the characteristics, which then allows an analysis of the relative importance of the characteristics observed across multiple risk events. Analysis of the hierarchical structure then indicates the characteristics where greater management effort could result in reduced risk events, and also permits a cost-benefit analysis as to the payoff from diverting resources to the management of the dominant characteristics. The technique is also capable of identifying the emergence of new characteristic combinations in creating risk events. This may allow for the forward projection of the next combination of characters that may produce new risk events