Based On Extended Kalman Filter, China's Securities Investor Sentiment Index Research

Some assumptions of traditional finance,such as investors' rationality,homogeneity,and market efficiency,are increasingly questioned by people.The study of behavioral finance has attracted widespread attention in the academic and practical world.As a typical immature emerging market,China's stock market is very different from the mature foreign stock market,and the “financial vision” it generates is even more prominent.Investor sentiment is an important factor reflecting the investor's psychology and will inevitably have a major impact on investors' decisions and behaviors.Investor sentiment as a qualitative indicator,due to the incomprehensib ility of measurement,has yet to reach consensus on investor sentiment measurement methods.And because C hina's stock market started relatively late and was controlled by government policies,even if Western countries' research on investor sentiment matures,their methods are not entirely feasible in China's stock market.Therefore,based on the unique characteristics of China's securities market,it is very necessary to find an investor sentiment measurement method that suits the characteristics of China's securities market.This article first combines C hina's national conditions,select the price-earnings ratio(PER),turnover rate(TURN),the number of new A shares opened(N AAC),IPO circulation,volume(TUR),consumer confidence index(CCI)these six single The emotional metrics serve as the basis for constructing comprehensive measures of investor sentiment.After removing the influence of economic macro factors,the extended Kalman filter method and principal component analysis method were used to construct a comprehensive sentiment index.A comparative analysis of the correlation between the sentiment index constructed by these two methods and the market stock index,and the effectiveness of the two sentiment indexes were analyzed using the AH stock premium index based on the “twin stock” phenomenon.Analyze and compare the consistency and superiority of the emotional index constructed by these two methods.The analysis of consistency and superiority shows that the investor sentiment index constructed by these two methods has a positive correlation with the market broad index CSI 300 index and the investor sentiment index based on the extended Kalman filter method.It is better than the sentiment index constructed based on the principal component analysis method;the Granger causality test shows that the investor sentiment index constructed based on the extended Kalman filter has a two-way Granger causality relationship with the Shanghai-Shenzhen 300 stock index return rate,which shows that in China There is an interaction between investor sentiment and the stock market.In a bull market,there may be a price bubble,and in a bear market,it may cause a downward trend.