Stock Return Comovement Based On Social Media

Stock return comovement attracts a lot of scholars to study it.With the development of the Internet,the emergence of social media provides a more convenient channel for people to get information about listed companies.This paper explores whether the social behavior of listed companies on social media is affected by their offline economic activities,which makes the stock return of listed companies with close online relationship have stronger comovement.By crawling the attention list of 2962 listed companies on twitter from December 2015 to November 2017,this paper uses the monthly attention data to build a social network of listed companies and uses complex network analysis to divide the social network of each month into 64-69 communities.Through the research,we find that the listed companies in the same community have high similarity in asset size,industry,priceto-book ratio and other factors.In addition,this paper downloads the stock data of relevant listed companies through the database of the center for security price research（CRSP）of Wharton Business School,and analysis the comovement on the stock price of Listed Companies in the same community.The results show that there is obvious comovement between the Listed Companies in every time points.In order to further prove that the online behavior of Listed Companies is affected by their offline economic activities,this paper studies the entry and exit of listed companies to the community over time.The results show that when the listed companies leave the community,they show weaker comovement with other listed companies in the original community.On the contrary,when the listed companies enter a community,they show stronger comovement with the new community.Therefore,the online interaction of listed companies is affected by their offline economic activities.The online social network can be regarded as the projection of the offline economic activities of listed companies.By extracting social networks among the listed companies in social platforms,we can provide guidance for investors and regulators.At the same time,from the perspective of the listed companies themselves,using complex network analysis and social media data to group stocks also provides a new idea for comovement research.