Spatial Autoregressive SV Model And Its Application In The Study Of Stock Market Volatility

In order to study and forecast the volatility of the stock market completely,Spatial autoregressive SV model would be established Which is a new model method based on the existing stochastic volatility model and space mode.Set the Spatial autoregressive SV model to describe the correlation between different cross section data level between stock market returns sequence.The random error term is a product of normal distribution and conditional volatility,and conditional volatility should meet the requirements of SV model in the space model,thus construct the wave equation.For the Spatial autoregressive SV model,the condition posterior distribution is derived from the Spatial autoregressive SV model based on the Bayesian Posterior Theory,construct the hierarchical model,and statistical inference by Monte Carlo Method.There has been used by the data of Shanghai Index and Shenzhen index over the past decade,the closed cross section data of Electric power index,Growth Enterprise Index and another stock over the past decade in this paper.There has been used to simulate the revenue volatility correlation of the data by the program with WINBUGS,to analyzed the case,and demonstrated the characteristics of cross section data fluctuation correlation;Then constructed the dynamic model of cross section data with these time-series data of sequences,analyzed the dynamic change of the volatility correlation between these stock.