By analyzing the advantages and disadvantages of the GARCH class models, combining copula techniques with GARCH model. We propose the Copula-GARCH mode which can avoid defects of classical risk analysis models. Without limiting the variability of the marginal distribution, this model can capture the no-linear distribution in financial market and the flexible multivariate distribution, especially the letter can be used to analyze the portfolio value-at-risk. Based on the copula techniques, we apply Monte Carlo technique in portfolio analysis, further more, according to copula models with various marginal distribution, we sum up the advantages to investigate the multivariate Copula-GARCH model. Finally, we use k-s test method in model data analysis, solve the VaR problem from five Shenzheng stock's portfolio. The result shows the feasibility and effectiveness of the model we proposed. |