From the loss data associated with the significant operational risks of Taiwanese commercial banks during the period from 1995 to 2009, we collected a set of 323 observations to use for modeling and estimating tail parameters of the banks' operational loss distribution. By means of three copula functions, we calculated correlations between event pairs, tested for independence between different classifications of risk types in the Basel II framework and sought to understand the characteristics and concentration of commercial banks' operational risks. Further, we estimated parameters for the generalized Pareto distribution and compared its fit with those of standard parametric models. A bootstrap method was used to estimate confidence intervals for the parameter values. In addition, value-at-risk and expected shortfall calculations were performed. Our results provide important information that financial supervisory authorities can use when accounting for operational losses of commercial banks. This research also contributes to the measurement of operational risk capital for the banks.
