Data Analysis Of Hushen300Index Based On T-Copula

Copulas are functions that join or couple multivariate distribution functions to their one-deimensional marginal distribution functions. In recent years, Copula is concerned and studied by lots of researchers in many fields. This paper mainly studies the regression method based on Copula and the application of Copula regression in the decomposition of VaR (value-at-risk). In the paper, t-Copula regression model is used to analyze the data of Hushen300Index.Firstly, this paper presents an overview of introduction to Copula theory, mainly about the definition, mathematical properties and the classification of Copulas, and also about the modeling process and goodness of fit of a Copula strucutre. The regression method based on Copula is studied, and a few kinds of regression models based on different Copula structures are introduced, especially, t-Copula regression model. The data distributed according Student distribution has been used to test the goodness of fit of t-Copula regression model.Then, this paper introduces the theory of VaR and the decompositon of VaR, and discusses the computing method of VaR and its decompositon based on the assumption of Normal distribution, and also the method based on the assumption of Student t distribution.Finally, the paper studies the applicatoin of Copula regression model in the theory of decomposition of VaR. The method to obtain maginal-VaR and component-VaR of a certain portfolio with the t-Copula regression model is discussed. The data of Hushen300Index is analyzed with t-Copula regression model.