Research On The Dependence Structure Of Chinese Stock Market Based On Pair - Copula

Based on the previous research on the choice of building structure, with rattan of multivariate density function are studied based on the dependence structure between a pluralities of assets. In the high dimensional distribution, there are many kinds of decomposition method, this paper adopts two kinds of the most common rules of vine:D and canonical, we choose to use as a function of mixing function. Using the method to the high dimensional problem into a two-dimensional to study, while the two elements are the more familiar, so as to we reduce the difficulty without losing accuracy. Dependence structure decomposition method to explore the manufacturing industry, pharmaceutical industry, the real estate industry and service industry using empirical part of the paper, and the mixed function fitting, finally get the joint probability density function of a four-dimensional, Through empirical analysis that the results closer to reality than the traditional method.