An application of PCC model: risk measurement of extreme event in Chinese stock market

摘要

We focus on analyzing extreme event on stock market with the pair-copula constructions (PCC) based multivariate models with GARCH (p, q) margins. We utilize the PCC model to get the estimation of joint PDF parameters. Then, we use six indices construct the decomposition of the PCC copula. As for different tail dependence of these log-returns series, we build the estimating model with bivariate t-copulas. Finally, we apply Monte Carlo method to simulate the extreme loss with parameters estimated from decomposing steps.