Research On VaR Of Portfolio Based On Copula Theory

Right investment decision requires reliable predictions of return and risk,and reliable predictions can only be obtained if the underlying statistical model tests on realistic assumptions. The traditional portfolio theory is based on the nomal distribution for the real distribution of stock market in China. But very little has the true obedient multi-variate normal distribution data in the financial domain, moreover between various assets income has non-linear relations, Therefore, asset portfolio risk analysis under the normal distribution hypothesis is bigger difference from the actual situation. We need to know asset portfolio earning ratio distribution function uniting , but the distribution function uniting so are sometimes unable now casually express out. The equal ratio investment reasoning with the assets income marginal distributions and their relevance structure separating the exponent handling , verifying fitting 180 exponents and Shenzhen component share by the aid of Copula function constitutes the main body of a book having carried out risk magnanimity.In this paper, first, the basic principle and main methods of VaR computation are researched. The analytical method, history simulation method and Monte-Carlo Simulation are contrasted .Once more, the basic theories of Copula function, specially the nature and characteristic of two-dimensional Copula function and single parameter two-dimensional Archimedean Copula function have been introduced. Emphatically we discussed the Copula function parameter estimation, Use the nonparametric density estimation technique when estimate the Copula function, the condition on the distribution of assets can be relaxed. Compared with the former methods ,such as maximum likelihood law, the canonical maximum likelihood law, the marginal distribution function inference law, it is a kind of improvement. Finally, the varied ratio portfolio formation verifying 180 exponents and Shenzhen component share exponents owing to that Copula function theory fits has carried out the risk magnanimity. Wield the Copula function , the assets income marginal distribution and their relevance structure is separated studying. The book adopt t-3 to scatter as two stock exponent avails has scattered , Copula has described the relevance structure between two stock exponent avails with three kinds single parametric Archimedean coming the function . The Quasi Monte carlo simulation methods and the traditional Monte-carlo simulation methods have been used to calculate the VaR of the portfolio under each kind of Copula function separately, and carried on the appraisal to the obtained result.