Analysis Of Co-movement Between Investor's Sentiment And The Chinese Stock Market

This paper makes an empirical study on the investor 's emotional effect in China' s stock market,based on historical data of the Sina blog issued for the year 2015.The research of this thesis belongs to the scope of behavioral finance.In the course of research on behavioral finance,the subjective factors such as investor psychology are taken into account.China's stock market from 1989 when it started to be on test run to today has gone through more than 20 years of ups and downs,but compared to some developed countries' stock market it is not mature enough.Some of the traditional theory of finance,such as market news theory,is difficult to work in China's stock market and are unable to explain some of the unreasonable and abnormal phenomenon in China's stock market.Therefore,some domestic experts and scholars introduced behavioral finance theory in the study of China's stock market and through the study they found that investor sentiment has a significant impact on the performance of the stock market.With the rapid development of Internet technology,social networks(such as microblogging,blog,WeChat,forum,etc.)slowly step into people's daily life,and play an increasingly important role in people's lives.It makes people change from passive receivers into active creators in the process of information transmission,this change greatly mobilized the enthusiasm of Internet users.Internet users in the social network has a large number of users of the original data generated every day.These data embodies the public view of social issues,such as financial media,financial practitioners in the social media real-time release of the stock market information.Investors receive these information in real time and express their views on the market and stocks,and discuss with a participant in other social networks about a market issue.The original data of these users represent the estimation of the whole market trend in both current and future period and has great significance to the forecast of the stock market trend.More and more scholars are involved in the research on the original data mining of the stock market users.As an important application in social media and in recent years having been rapidly developed,Microblogging bacame a platform of information circulation with its high user coverage,high speed at information spreading and originality on the content.This paper attempts to use the LDA model to extract the theme of the massive text resource on the microblogging platform and measure the investor's emotion according to the microblogging theme,and at last analyze the interaction between investor sentiment and stock market performance in different market environment.The main research results are as followed:(1)Combine the network reptile technology and Sina microblogging site search to obtain investors published by the stock market related to Sina microblogging blog.(2)Through the investors published microblogging using text analysis technology and LDA model to measure the investor sentiment.And get the investor's sentiment value of2015.(3)Through the correlation analysis,this paper compares the relationship between investor sentiment and stock in two different market conditions of bull market and bear market.In the bull market,investors' sentiment and the current period and lagged one period of the Shanghai Composite Index turnover,the Shanghai Composite Index yield was significantly positive correlation,with the correlation coefficients of 0.27 and 0.76 respectively.In the bear market when the investor sentiment is only with the current and lagged one of the Shanghai Composite Index yield is significantly positive correlation,with the correlation coefficients of 0.75 and 0.1 respectively.(4)By comparing the standardized mean square error of the model of adding the investor's emotional value and the model which does not join the investor's emotional value,it is found that adding the investor sentiment value can reduce the degree of deviation of the predicted value from the actual value,that is,improve the goodness of fit.According to the data of this paper,adding the investor sentiment value of the Shanghai Composite Index yield forecast model through 5-fold Cross Validation,the training set has a better fit of 92%.At the same time,the investor's emotional prediction model has betterfit in the bear market than the bull market.Therefore,investor sentiment can have an impact on the future rate of return on the Shanghai Composite Index,especially in the bear market.