Research On The Influence Of Investor Sentiment On GEM Stock Excess Return Rate

The change of investor structure in a country’s capital market can fully reflect the transformation of its investment philosophy.Compared with the mature capital market,Chinese investors show more irrational behaviors,such as overreaction to information,overconfidence in their own investment ability and herding effect of following others’decisions.The speculative effect is more obvious,and the proportion of speculation in the market is greater than that of investment.Based on this,the study of the impact of China’s incomplete rational investor sentiment on the stock market is not only conducive to our research and judgment on the trend of the stock market,but also conducive to the relevant departments of our country to make corresponding supervision and legislation,so as to promote the gradual maturity of China’s stock market.After systematically sorting out the previous relevant literature research,this paper first constructed the investor sentiment index,and used Python software to crawl the China Africa news stock evaluation data of the gem index bar in the Oriental Fortune stock bar,the largest stock evaluation website in China.The time range is from March 2020 to the end of October 2021 after the COVID-19;After that,using the data of more than 150000 stock evaluation,the existing bosonnlp thesaurus is used as the emotional dictionary to calculate the investor sentiment index;Then,this paper makes an empirical study on the impact of investor sentiment on stock market returns,and selects the gem stock data of the same time span;After data preprocessing,descriptive statistics and correlation test are carried out respectively.In order to solve the problem of multicollinearity,redundancy test and GRS test are further carried out according to the test results to pave the way for the introduction of emotional factors;Finally,in the empirical regression analysis,the GEM stocks are grouped according to different categories,which are divided into three different stock portfolios:scale valuation,scale profit and scale investment.The Fama French six factor model is used for regression;In the robust test part,this paper uses the traditional principal component analysis method to construct a new investor sentiment index,and carries out the same grouping regression as before to carry out the relevant empirical test.The relevant conclusions of this paper are as follows:in general,the stocks with low stock market value are more vulnerable to investor sentiment,and small-scale stocks are more vulnerable to their individual retail investor sentiment because of the high proportion of individual investors;Among the stocks with low stock market value,stocks with high valuation are more likely to receive emotion.The reason may be that stocks with low market value and high valuation are mostly subject concept stocks,which are vulnerable to market speculation;The stock portfolio with low market value and high profit has great growth potential,so it is more likely to be catalyzed by market sentiment and get financial attention;In addition,the reason why the stock portfolio with low market value and low investment is vulnerable to market sentiment may be that such stocks do not have much capital attention,and the circulating chips are easier to be controlled;Finally,some policy suggestions are put forward.