The Impact Of Investor Sentiment On Stock Market Returns

In recent decades,numerous scholars have done a lot of research on the rise and fall of stock prices.Including the early traditional financial model,to find the traditional financial model can not fully explain the phenomenon of the market,then direct their attention to the rational behavior of investors,and find that take emotional factors into the traditional financial model can make the market interpretation ability greatly enhanced.Although the related research in China is still emerge in an endless stream,but in 2015 the market fluctuated heavily and early 2016 brought "crash" event,the investor sentiment research for Shanghai stock return and volatility is still worth of further study.Based on this,this paper chooses indexes from three aspects of direct emotion questionnaire,direct and indirect market sentiment index market sentiment index,the correlation between calculation and return,and return on the next trading day forecast.First,crawling from the web,preprocessing the text,emotion classification and emotion index generation of xueqiu users,and the extraction and pretreatment of the other emotional indexes.It is found from the study results,which shows significant to the Shanghai stock index returns includes: the Shanghai stock exchange rate over the same period,dongfang sentiment research over the same period,the snowball sentiment over the same period,difference margin trading over the same period,Shanghai stock exchange over the same period,lagged difference margin trading,lagged difference margin trading,lagged Shanghai stock exchange,lagged two phase margin trading.Through the multiple linear regressions of emotional variables and yield,it is found that the direct fitting of sentiment to the yield value is not ideal.This article will turn emotional predictions of return to predicting two classification of yields,the index is divided into two classification that greater than 0 and less than 0 points,find significant variables through logistic regression and clustering: dongfang sentiment research over the same period,the snowball sentiment over the same period,lagged Shanghai stock exchange rate,lagged two phase difference margin trading,by constructing the linear SVM model to predict the rate of return,The prediction accuracy of the emotional indexes extracted from the three categories has a certain predictive ability and the result is 64%.From the study we can found that for the individual and institutional investors,can choose to buy stocks when sentiment index like margin trading and dongfang sentiment research showed a gradual increase,and sell stocks when investors form consensus expectation;for the government,can control the market by considering from capital side,if the margin trade index is too high,the government can limit the access mechanism of funds,strengthen the control of funds and leverage from capital side;on the contrary,when the market sentiment and margin trade show low level,the government can relax the access mechanism of funds and guide the funds into the stock market,so as to bring vitality to the market.