Analysis Of The Relation Between Investor Sentiment And China's Stock Market Basing On Text Mining

In recent years,with the improvement of Internet infrastructure construction,Internet services continue to infiltrate,and browsing online media such as portal websites,Weibo,WeChat,etc.has become a daily activity of Chinese people.In the financial sector,the financial website provides investors with information services on stocks,funds,bonds,insurance and other financial products,and becomes an important channel for the public to access financial news.At the same time,they open up online stock forums,places for investors who demand the same information to gathered together,exchange information,share ideas,and meet their social needs.For this emerging network information publishing platform,many investors will use the information published on the network platform as their own investment decision reference,and the rationality of this behavior has not been confirmed in the theoretical circle.In addition,the posts in online stock community reflects investor sentiment,while limited by the capabilities in processing unstructured data and technical early researchers did not incorporate the posts and reviews on the websites into the construction of integrated investor sentiment Index,causing the indicators used to measure investor sentiment at this stage do not reflect the impact of Internet media.This paper attempts to fill the gap in this field by introducing text mining technology.Based on the comprehensive factors such as daily visitor visit volume,posting volume and data accessibility,this paper selects Oriental Wealth Shares as the source of online social media data,and extracts the constituents of Shanghai and Shenzhen 300 from January 1,2014 to December 31,2018.By using sentiment analysis,the investor sentiment index reflecting the stock market sensation is constructed.Drawing on the construction method of CICSI,the Comprehensive Sentiment Index,the constructed stock market sentiment index is added to CICSI's original six individual emotional indicators,including the closed-end fund discount,the number and average first-day returns on IPOs,investor confidence index and new investor account opening.Then we use Principal Component Analysis to construct a comprehensive investor sentiment index for China's stock market that reflects online public opinion.Further,we conducted a Granger causality test on the composite investor sentiment index and the CSI 300 index yield and volatility series during the same period.The study found that the impact of investor sentiment ISX and stock index return rate is up to 4 months.Investor sentiment can predict the stock index return rate.The sharp fluctuation of stock market will affect the change of investor sentiment.In the short term,investors believe that buying and selling will fall,while in the long run they are more willing to believe in the mean return model of the stock index.Finally,based on the research results,the article puts forward suggestions for investors,listed companies and national regulatory authorities.