Capital Market Effects of Fake Social Media Accounts
Abstract
Fake social media accounts, or bots, are often used to spread or amplify content on social media platforms. I examine the effect of this artificial amplification on capital markets. I obtain a large dataset of posts from the social media platform Twitter (now X) and use several methods to identify accounts that are likely bots. The descriptive analysis suggests bot accounts produce about a third of all tweets mentioning company tickers. Bot tweets are associated with higher market activity, particularly among retail traders. Bot tweets are generally not associated with return reversals, but those with stock price-related content are. Bot tweets are generally associated with increased liquidity, but those with earnings-related content are associated with reduced liquidity, consistent with higher informed trading. Using an exogenous shock where Twitter suspended a large number of bots, I find evidence consistent with bots impacting capital markets. Overall, the results suggest that fake social media accounts may amplify both helpful and harmful signals to investors.