| This paper uses crawlers to obtain relevant text data on microblogs before and after the 2015 stock market crash and the COVID-19,quantify investors’ panic,build emotional indicators,and prepare for subsequent correlation research and modeling of stock prices.The selected time span is from January 2013 to April 2022,in which the period from January 2013 to May 2015 is selected as the period before the stock market crash in 2015,the period from June 2015 to December 2018 is selected as the period after the stock market crash and before the COVID-19 epidemic,and the period from January 2019 to April 2022 is selected as the period after the COVID-19.Use the SVAR model to compare the relationship between investor sentiment,macroeconomic factors,and stock returns before and after the 2015 stock disaster,as well as the relationship between investor sentiment,macroeconomic factors,and stock returns before and after the 2019 pandemic.This article finds that investor sentiment has a positive stimulating effect on stock market returns around 2015,while the effect of stock market returns on investor sentiment is opposite before and after the stock disaster.This may be related to investors’ high-risk preferences before the stock disaster and their confidence collapse after the stock disaster.In addition,the relationship between investor sentiment and stock market returns has weakened after the epidemic,which may be related to investors shifting their focus outside the stock market. |