The Impact Of Media Sentiment And Investor Attention On Stock Return

The arrival of Internet+ and the explosive growth of mobile terminals,the result is the increase of the circulation speed and the expansion of the group oriented,making the news media more and more influential on people's lives,and nowadays,with the expansion of media influence on investors,the role of investors in the relationship between the media and the stock market is becoming more and more important.At present,domestic research mostly focuses on the relationship between media sentiment and stock returns,and the relationship between investor attention and stock returns.at present,There are few studies on the impact of interaction between media sentiment and investor attention on stock returns.Moreover,the way to measure media sentiment is mostly based on the data of media reports without considering the emotions contained in the reports.this paper synthesizes media emotional indicators through three data form Youkuang net: the emotions of reporting content(positive or negative emotions towards related companies),the degree of relevance(relevance between reporting content and related companies)and the number of reports.For investor attention,most domestic research uses the data provided by Baidu search engine,but Baidu is not a professional financial search engine,so it may contain more interference information.This paper mainly use the number of posts in the stock bar community of YouKuang net as a measure,and the number of posts of snowball community as a measure in the robustness test.On the basis of predecessors,this paper uses more accurate data to measure the indicators,in order to study the impact of media sentiment and investor attention on stock returns.This paper takes CSI300 constituent stocks as samples.The time interval is from March 2016 to March 2018.The relevant weekly data is obtained for two years.The data is processed using the programming language python,and the regression analysis is carried out with stata.Find the following conclusions:(1)Sort and group the media sentiment of each week,count the weekly returns and trading volume of each group of stocks,and find that with the decline of media sentiment,the average weekly returns of each group of stocks will also decline,and the trading volume will first decline and then rise with the decline of media sentiment.Then through regression analysis of the relationship between current and lagging media sentiment data and current stock returns,we find that current media sentiment is positively correlated with current stock returns,verify the conclusion of grouping statistics,and find that lagging media sentiment is negatively correlated with current stock returns.(2)Sorting and grouping the investor attention of each week,statistics of each group of stocks weekly return and trading volume,we find that the average weekly returns and trading volumes of each group of stocks decrease with the decrease of investor attention.Then through regression analysis of the relationship between current and lagging investor attention and current stock returns,i t is found that current investor attention are positively correlated with current stock returns,while lagging investor attention is negatively correlated with current stock returns.(3)The influence of investor attention on the stock returns is far greater than the impact of media sentiment on the stock returns,and by adding the interactive variables of media sentiment and investor attention,we find that they will have an impact on earnings and can magnify each other's impact on stock returns,when the stocks have a high media sentiment,the influence of investor attention on stock returns will increase,and vice versa,when the stocks have a high investor attention,the influence of media sentiment on stock returns will also increase.