Bayesian Simulation And Application Of Stochastic Volatility Models Using Scale Mixtures

There is a common phenomenon of the volatility of Economic and financial time series, which is a core problem description of financial markets research, by the variance of the rate of return financial measures.Volatility contains the information of market changes and the risk of investment. In real financial market, return on assets volatility sequences often show "volatility clustering", volatility in the form of continus change over time, don’t spread to the infinite namely volatility in a fixed scope change, the volatility of prices have risen sharply and a sharp fall in different reaction namely "leverage effect", the characteristics of distribution of return on assets also show the characteristics of the "heavy-tail". The second part of this paper studies the bayesian theory, MCMC algorithm and gibbs sampling, WinBUGS software.In the third intorduce stochastic volatility (SV) model and the development of common SV model are given, is deduced in detail in this paper, by using the model, the basic properties of the model is studied, and a simulation analysis was carried out on the model;In the fourth chapter we carried on the empirical analysis to the Shanghai Composite Index and Compositional Index of Shenzhen Stock Market, first has carried on the pretreatment to the yield of the two cities and descriptive statistical analysis, found that the yield of the two cities are rush the nature of the thick tail.Then use the model study of the two cities respectively, this paper finds out the fluctuations in the two indexes of outliers, some of these outliers from earnings, some are from the unpredictable fluctuations.