Study On Prediction Of China’s Stock Market Based On Internet Search

Using Baidu search engine data, this paper tries to answer this key problem:Can internet predict China’s stock market. We argue that investors’ internet search is a composite of investor sentiment, investor attention and other factors, which has a stronger power to explain stock market compared to conventional indicators of investor sentiment and investor attention. A theoretical model is constructed to elaborate this viewpoint.In this paper, the highest market share of Chinese search engine Baidu search data around the Internet search can predict China’s stock market, a study of this issue. This paper argues that investors in Web search is a composite of investor sentiment, investors are concerned about a number of other factors as well as gray boxes, and compared to conventional indicators of investor sentiment and investor concerns, the stock market Web search explanatory power and predictive power may be stronger, this paper constructs a theoretical model in detail.Based on CSI-300 index constituent stocks and Baidu search index from 2006.6 to 2012.5, we find:(1) Internet search intensity can predict the stock’s short-run return, trading volume and cumulative return, which implies a positive relationship between internet search and stock market.(2) Internet search intensity can affect stock market further and predict stock market effectively. The problem of endogenous has little impact on the prediction effect. Moreover, a simulation analysis reveals that a stock portfolio constructed according to the internet search will yield excess abnormal return.(3) Internet search intensity index is strongly correlated with traditional investor sentiment index and investor attention index, and is more powerful to explain and predict stock market.(4) Internet search intensity index has little power to explain and predict stock market when stock index is in rising phase, but has strong power when in decline and adjustment phase.This paper expands the existing research in academic theory and empirical analysis. For individual investors and institutional investors, the investment strategy and the comparative analysis of different phase have important reference value. For the securities regulatory authorities, it is possible to improve the technical means of stock market regulation and supervision.