| Limited Attention Theory in behavioral finance states that investors'attention to stocks leads to Attention-Driven Buying and Over Transaction,which will bring the short-time pressure on stock price.However,the rising price is not supported by favorable fundamentals.Therefore,the price will go down later.This phenomenon is called the Over Attention Underperformance.A large number of scholars have done their studies on this topic,but they all face the same difficulty:How the attention should be measured.Conventional studies often use the number of news or trading volume as the measurement of attention,while the drawbacks are obvious.In this paper,I use the search index of search engine to measure the attention which is more accurate.In addition,this article has its advantage on data sampling,model construction and regression method.I use Chinese Growth Enterprise Market?GEM?data,spaning from January 1st,2014 to December 31st,2016,and get the following conclusions:Firstly,compared to other types of search engine index,search index based on stock code can reflect the change of investor attention more precisely.Compared to conventional measurements of attention,search index is more in time and more accurate.Secondly,search index based on stock code has a significant positive effect on stock return in the same day,which means that a higher search index leads to a higher stock return during the same period.This effect is still significant even if we control for conventional proxy variables such as trading volume.As for estimation method,we use Fama-MacBeth regression and Clustered standard error estimation to estimate the coefficients and construct for more robust statistics.Thirdly,cumulative attention in non-trading days has significant influence on the absolute price jump the next trading day.At last,Over-Attention Underperformance exists in the Chinese GEM with its time lag.In addition,we take multi-collinearity into consideration to make the conclusion more reliable. |
PDF Full Text Request