A Study On The Interactive Relationship Between Investor Sentiment And Stock Market Based On Text Mining

The capital asset pricing theory and the efficient market theory are major theoretical foundation of the classical financial theory, formed the basic frame of traditional financial theory. According to the traditional financial theory, investor is rational, noise trading does not exist on the market. By simplifying the operation of the financial market, traditional financial theory describe the performance of the financial market with mathematical models. But in real life, the stock market tend to be more diversified and complex, the stock market is a complex system of many influence factors. the traditional financial theory have a lot of difficulties to explain some anomalies. The hypothesis of rational man ignores the impact of the irrational factors witch plays an important role in the decision-making process of investors, and in the process of the formation of the asset price. With Further learning of the financial market, the behavioral finance theory had been gradually accepted by the mainstream economics. Behavioral finance intents to re-examine the irrational factors, introduce investor sentiment to study the financial market, questioned the traditional finance. Behavioral finance theory holds that, the market is not efficient, arbitrage cannot eliminate the deviation of asset’s value.In this paper, based on behavioral finance theory, the writer analyzed the interaction between investor sentiment and stock market in china. This paper find a innovative way to measure investor sentiment by introducing methods of text analysis. By analyzing the financial blog text, this paper mining the sentiment, views and opinions of the investor from the blog text. This paper selected the blog as the data source, downloaded blog text from 2014 to 2013 by the web crawler software, using semantic analysis and semantic rules of Chinese context, and introduced new investor sentiment proxy indicators, analyzed the interactive relationship between investor sentiment and the volatility of the market return.Using the empirical method, this paper analyzes the relationship between investor sentiment and stock market variables. And than, Multivariate BEKK-GARCH model had been used to describe volatility spillover between investor sentiment and stock returns,to study the existence of the two-way influence mechanism of sentiment and the stock market. The empirical results show that, there is a significant correlation between investor sentiment and stock market variables, and, investor sentiment volatility and stock return volatility exist two-way volatility spillover, confirmed the hypothesis. The results show that investor sentiment will affect the stock market performance, supported the research on behavioral finance theory. In addition to the conclusion, this paper also shows that investor sentiment can be influenced by the fluctuations from the stock market.