On The Application Of Copula-MCMC Method In The Securities Portfolio

There are two significant topics of the portfolio research topics: the optimal allocation of portfolio proportion and the measure for portfolio risk’s value. This paper will do the research from the above two perspectives.The financial markets have the features of high kurtosis, fat tails, fluctuation, skewed and so on. The emerging Copula function can describe these related characteristics in financial markets well, which the traditional portfolio analysis method can’t describe for it fails to consider the dependence of financial market exactly. In addition, the traditional method for allocating the asset portfolio generally adopts the experts’method and the Markowitz model, both of which have their disadvantages respectively. The experts’method has strong subjectivity, while the Markowitz model has too many unrealistic assumptions. Therefore, this article creatively combines the Copula theory and the MCMC method together to analyze portfolio by choosing four stocks from Shanghai securities market as samples and doing the empirical study of the asset allocation in portfolio and risk measurement. Compared to the traditional portfolio analysis method and simply using one of the two methods to analyze the financial markets, we can conclude that the Copula-MCMC method has better effects in improving asset allocation and risk measurement. The main contents of this thesis can be summarized as follows:Firstly, this thesis introduces the contributions made by scholars in the existing portfolio research, and studies the possibility and significance for further relevant research on the basis of previous studies.Secondly, according to the limitations and deficiency of traditional methods in previous studies, this paper puts forward the Copula-MCMC method which can better improve the portfolio. This paper mainly introduces the related theories of Copula, the related methods of measurement, the sampling method and algorithm of MCMC method.Then, on the theoretical basis of the above two methods, this paper constructs the Copula-MCMC model by using the relevant correlation coefficient of copula to improve the computing method of VaR according to the financial market’s feature and the GARCH model.At last, combined with securities market’s features of China, this paper selects four stocks from Shanghai securities market to make the empirical analysis. And the Copula-MCMC method is used to calculate the asset allocation in portfolio and the VaR though the actual samples. To sum up, the above analysis leads us to the conclusion that the Copula-MCMC method can better improve the securities portfolio than the traditional methods.