Research On The Impact Of Investor Sentiment On China's Stock Market Returns And Volatility

With the rapid development of China's economy,the increasing openness of the financial industry and the reform of non tradable shares,China's stock market can reflect the current situation of economic operation to a certain extent and become the "barometer" of the national economy.Therefore,the yield and volatility of the stock market has always been a hot issue in academia and industry.According to behavioral finance theory,investors' decision-making is easily affected by their own emotions and other factors.Therefore,investors' emotions should be taken into account in the study of financial markets.At present,mining the information in the network platform is an effective way to obtain investor sentiment.In order to reveal the influence of investor sentiment on China's stock market,this paper uses the Chinese sentiment analysis method based on sentiment dictionary to extract investor sentiment with different sentiment tendencies from Sina-Weibo text,and obtains the sentiment time series through quantitative analysis to study the influence of investor sentiment on the stock market.This paper is divided into two parts,the first part is the empirical study of investor sentiment and stock market returns,the second part is the empirical study of investor sentiment and stock market volatility.In the first part,based on the Bayesian quantile regression model,the impact of investor sentiment on the return of Chinese stock market will be studied.The results show that investor sentiment with different emotional tendencies has a significant impact on the yield at each quantile.When the stock market returns are low or high,such as 0.25 or 0.75,it is most vulnerable to various emotions.When the stock market returns are in the middle position,for example,in the 0.5 quantile,it is less affected by investor sentiment,and the estimation coefficients of various emotions are not significant in this quantile.In the second part,three kinds of volatility models are introduced,ARFIMA-RV model,HAR-RV model and LSTM model,and expand each model to study the impact of investor sentiment with different emotional tendencies on the volatility of China's stock market.The rolling time window method is used to predict the volatility.The results show that different investor sentiment does have a significant impact on stock volatility.In addition,use a variety of loss functions to prove that adding investor sentiment can improve the prediction accuracy of stock market volatility,and the sentiment after detailed classification is more helpful to improve the prediction accuracy.Finally,MCS test is used to prove that the extended model of HAR-RV has the best prediction effect among the three kinds of volatility models.When the conclusions of the three models are different,the conclusions of the extended model of HAR-RV are more reliable.Its practical application value lies in providing new research ideas for the estimation and prediction of stock market volatility,as well as supporting evidence for the value of network information.