The Research On The Influence And Prediction Of Investor Sentiment On Stock Market Volatility

With the gradual development and promotion of financial information platforms such as stock review forums,stock review information has become an important reference for most individual investors to make investment decisions by virtue of its advantages of timeliness,easy understanding,and low acquisition costs.In recent years,the stock market has fluctuated greatly,and the phenomenon of sharp rises and falls has occurred frequently.If the investor sentiment contained in the stock review text misjudges the stock market volatility trend,it will inevitably affect the income of individual investors,and even cause serious losses.This paper studies the impact of investor sentiment on stock market volatility,the prediction effect and the correlation between the two,which is of great significance for guiding individual investors to establish a correct investor concept and ensuring the healthy and stable development of the financial market.The ERNIE2.0 pre-training model is used to measure the sentiment tendency of stock review texts,the EGARCH model and DCC-GARCH model are used to analyze the asymmetric impact of investor sentiment on stock market volatility and the dynamic correlation between the two,combined with ARIMAX model and Bi-LSTM model to analyze The predictive effect of investor sentiment on stock market volatility.The main content consists of six parts: the first part is the introduction,expounding the research background and significance,summarizing the current situation of domestic and foreign literature research and introducing possible innovations in the article;the second part defines the concept and expounds the relevant theories for the follow-up.The third part mainly introduces the process of analyzing the sentiment tendency of stock review texts,constructs the investor sentiment index and analyzes its validity;The dynamic correlation between investor sentiment and stock market volatility and analysis of the asymmetric impact of investor sentiment on stock market volatility;the fifth part uses different models to predict the stock market volatility by investor sentiment;the sixth part is the main conclusions drawn and Related research recommendations.This paper draws the following conclusions:(1)Investor sentiment constructed by ERNIE2.0 pre-training model to analyze the sentiment of stock review texts can effectively describe the behavioral characteristics of investors described by behavioral finance.(2)The volatility of the stock market is greatly affected by its own inertia,and investor sentiment is not only affected by its own fluctuations,but also affected by external information to a certain extent.(3)The overall negative correlation between investor sentiment and stock market volatility in the sample period,the dynamic correlation changes frequently in the short term,reflecting the characteristics of time-varying.(4)Both positive and negative investor sentiment will have a significant positive impact on stock market volatility.The reason for the negative correlation between investor sentiment and stock market volatility is that the number of negative investor sentiments is much greater than that of active investors during the sample period.Emotional quantity.(5)Combining the ARIMAX model and the Bi-LSTM model can improve the accuracy of the prediction effect of investor sentiment on the stock market volatility,and its practicability is higher than that of the single ARIMAX model or the Bi-LSTM model.