Research On Interactive Relationship Among Investor Sentiment,Stock Market Return And Volatility In Different Market States

With the continuous emergence of market anomalies and the development of behavioral finance theory increasingly mature,the traditional financial theory system which on the basis of the efficient market and rational investors as assumptions has been challenged from both practical and theoretical.From the empirical results of all the stock markets around the world,the price of the stock market is affected by the investor sentiment to a certain extent.In 1986,Black proposed the famous noise trader hypothesis and Delong etc.established the DSSW noise trade model in 1990.Since Shanghai stock exchange founded in 1990,China's securities market has developed rapidly but still has a lot of problems such as market information asymmetry,unbalanced structure of investors,"policy market" characteristics significantly etc.Research on the interaction relationship between investor sentiment and stock market returns has contribute to improve the system of behavioral finance theory theoretically,and provide some useful references for solving those problems in theory.In reality,it can help China's policy makers to grasp market trends so as to reduce the impact of the "policy market" as far as possible and improve investors' analyzing ability to promote the healthy development of China's stock market.Based on the existing research results,this paper refers to the research ideas and methods of He Xingqiang and Zhou Kaiguo,according to the thought of Pegan and Sossounov's non-parametric test methods,dividing the different market conditions of the Chinese stock market.This paper selects four proxy indicators of investor sentiment,construct the comprehensive investor sentiment index by using principal component analysis method.In the part of empirical analysis,this paper uses the VAR model,Granger causality test and impulse response analysis etc.to study different samples respectively in different market states;Using TARCH-M(1,1)model with three data samples to tests the influence of investor sentiment on stock returns in the stock market and its fluctuation state inspection under different environment.In the end,this paper puts forward some policy recommendations and summed up some shortcomings in order to provide useful reference for the follow-up research.