An Empirical Study On Volatility Of Stock Market As A Whole And Multi-industry Based On SV Model

In the past two years,the Chinese stock market has experienced a rise and fall cycle,in which several considerable fluctuations are extremely rare in the history of Chinese securities.Therefore,this article mainly focuses on the Chinese stock market during this particular turbulence period throughout the year.The research on the volatility of the stock market in China,as well as the sub-sectors of many major stock markets,is of great practical significance and research value in the field of risk management and control of China's stock market.On one hand,this paper uses the traditional statistical methods and modern statistical tools to describe the volatility characteristics of China's stock market.On the other hand,based on the predecessor's research theory foundation,the standard SV model and the extended thick-tail SV model,the paper deeply and completely depicts the volatility of the yield series,trying to reveal the overall characteristics of Chinese stock market.It is also worth mentioning that the second half of this article innovative to extend the SV model to the two-dimensional model,which concerns about the sub-sector volatility situation,and also studies the link between the volatility of the industry.This has opened a new inspiration door for the follow-up study of the volatility of our stock market as a whole and the lower molecular industry,which has certain practical significance.We used two stochastic volatility models based on different hypotheses-standard stochastic volatility model and thick-tailed stochastic volatility model to model the volatility of Shanghai Composite Index respectively,and compare the advantages and disadvantages of these two models.The most difficult problem is to estimate the SV model parameters.We use the Markov Monte Carlo method(MCMC method)to process into the Bayesian thought.The specific operation is to use OpenBUGS software to program the Gibbs sampling.The difference between the accuracy and the complexity of the two SV models is compared by the parameters simulation and the DIC index.The empirical results show that the thick-tailed stochastic volatility model can better describe the volatility level of Chinese stock market as a whole.In addition,this paper also selected some sub-industry indexes of Shanghai market,such as energy,materials,industry,consumption,medicine and finance,and observed the historical data.Based on the existing research results,Including the formula generalization,theoretic thinking,and the parameter estimation MCMC method,and finally completed the empirical analysis.We used models to analyze the return volatility of selected Shanghai sub-industry,and also tried to explore the correlation between volatility of Chinese stock market in different sub-sectors.We found that the volatility of the selected sector indices is generally weakly correlated with each other,which is consistent with our common sense.Among them,the correlation coefficient between the industrial sector and the material sector is strong,while the correlation coefficient between the medical sector and the consumption sector is the lowest.This conclusion shows that we can not only simulate the volatility of the two industries through the two-dimensional SV model,but also observe and study the balance and influence of the sub-sectors through the stochastic volatility.significance.