Investor Behavior And Its Impact On The Stock Market Based On Big Financial Data

When Internet Technologies has been widely applied at a snowballing pace in a growth development,numerous financial scholars try to distinguish useful information from Internet.There are also many scholars believe that Internet forums contain effective information that can reveal the fundamentals of the company.This paper employs data from the stock BBS of Oriental Wealth,the most powerful and mostvisited online stock BBS in China's Stock Markets from January 2011,to June 2014.Using text classification algorithm based on Na?ve Bayes,we classify the messages into three types: positive,neutral,negative,and extract messages to several firm-specific variables including Bullishness,Agreement,Quality.This paper mainly studies two problems:On one hand,this study explores the relationship between Online Message Board and stock market,we analyze roughly 8.8 million stock?related messages on a daily basis.We find a bidirectional prediction association between stock BBS sentiment and stock returns,message volume and trading volume,as well as agreement and volatility by Fama-Macbeth cross-section regressions.On the other hand,by another perspective: the relationship between information quality and prediction ability,we indirectly study the possibility of effective information contained in stock BBS.we find that the investor's sentiment in China stock BBS significantly predict the stock return in next day with negligible R-square and there is no distinction at predicting outcomes between the sentiment prediction of high quality stock BBS and the ones of low quality BBS.Moreover,we find that the mean quality of posts is below 0.5,which illustrates the predicting capacity of an individual post cannot exceed random prediction.Due to the autocorrelation of stock return,we consider that stock BBS predict the stock return because of its reflection on the stock return in the same period rather than the revelation on valuable information.