Marek J. Karwanski, Urszula Grzybowska

Operational risk models are used to examine losses that arise as a result of operational events that can be treated as stochastic variables. One of the problems encountered in modeling is the need of taking into account correlations between events. Estimation of copulas and their application in risk models is complicated because random vectors here cannot be represented as matched pairs for which one could build correlation based models. Fortunately, copulas used in operational risk (the Archimedean copulas) are based on existence of common factors. Thanks to that it is possible to build models for correlated events.

In the article the authors present models of copulas and their application to calculating operational risk. The calculations were done both real data and on simulated data that allows for estimation of correlation influence on risk measurement.