Local Twitter Activity and Stock Returns

Abstract

In this study, we use geographic proximity as a measure of private information and examine the informational role of social media in stock markets. Using a large sample of individual messages (tweets) collected from Twitter during the period July 2011–March 2012, we find that local Twitter users are more likely to tweet about firms with high information asymmetry, and their Twitter activity in turn increases the trading volume of local stocks. More importantly, we find that the negative tone of local tweets predicts future stock returns and subsequent earnings announcement returns, while nonlocal tweets present no such predictability. These results suggest that local social media activity reflects new information, contributing to price discovery. We also find that the negative tone of local tweets is associated with higher bid-ask spreads and lower market depths. This finding suggests that social media – in contrast to traditional news media, which reduce firms’ information asymmetry – serve to share information with an audience in their networks and therefore increase information asymmetry among investors. Overall, our findings suggest that local social media contain value-relevant information and affect firms’ information environments.

Keywords: Twitter; social media; textual analysis; geographic proximity; trading volume; stock return; information asymmetry