The Application of EVT-Copula in Operational Risk Quantification

As the operational risk of Chinese banks becomes increasingly outstanding, we need urgently to find out a accurate measurement method. This article takes one commercial bank as an example, analyses the tails´ characteristics of three types of operational risk loss event. It measures the fat-tail and light-tail loss by the POT model of EVT and logarithm normal model. Since the operational risk sample is small, it estimates with Bayesian inference based on MCMC, meanwhile it uses t-Copula and normal Copula function to aggregate each kinds of operational risk and gets the distribution of total operational risk It turns out that, when estimate in this way, the result is more stable, smaller and reliable.