The Application Of Copula In Finance-Dependent Analysis And Estimation Of VaR

In modern financial analysis, Copula study has extremely attractive since it can be used in dependence analysis of multi-asset portfolios , because the Pearson's linear correlation coefficient is no valid for the most situations in practice .In this paper, the concept and character of Copula inference methods for parameters of Copula , the choice of Copula and Copula-EGARCH model are introduced in detail.We study the degree and patterns of dependence for two different portfolios composed of the ShangHai A and B stock index, the ShangHai A and the ShenZhen stock index,respectively. The empirical results show that dynamic Copula model is better than constant Copula in the ability of description and prediction of dependence between financial series.The tail dependence of portfolios is obvious whenever market is up or down ,it is different from the foreign empirical results .We apply Monte Carlo simulation technique to estimated the 0.95, 0.99 VaR for two different portfolios .In addition,we use the initial part of the sample to estimate the models, and the the remaining part to compare the out-of-sample performances of the different approaches, using various back-testing techniques.The empirical analysis showed that correct marginals specification is absolutely crucial for VaR forecasting, while Copula specification plays a minor role.