UC Berkeley

How does social media change the information landscape and what are the effects on various stakeholders? This dissertation contains two essays that study this question empirically by investigating how social media affects investors, firms, influential individuals, and the stock market.

In the first chapter of the dissertation, I study the implications of social media comment visibility on retail investor consumption and production behavior. I exploit a shock in the comment display sorting algorithm in 2018 on Daily Discussion posts on Reddit’s WallStreetBets. By comparing the market reactions to comments before and after the change, I find that going from a more curated but less timely display (Best Regime) to a less filtered but more timely display (New Regime) increases absolute abnormal returns and abnormal retail trading volumes in the five minutes after comment publication. Following the initial five minutes, Best Regime comments see a stronger price drift while New Regime comments see a slight return reversal. These results are driven primarily by firms with small market capitalization and high Robinhood user ownership. In addition, I find that changes in comment display also affect comment production timing and volume.

In the second chapter of the dissertation, I investigate the social media posting behavior of influential individuals, namely CEOs at S&P 1500 firms, in the context of quarterly earnings announcements and provide evidence of CEO strategic behavior. Quarterly earnings announcements followed by an earnings tweet from the CEO correspond with a 1.5-2.6% higher 3-day industry-adjusted announcement return conditional on the same level of earnings surprise and are not followed by return reversals. An intraday event study around earnings tweets shows that CEOs time their tweets to ‘take credit’ for quarters with positive stock price reactions above and beyond the earnings news. I attempt to tie this behavior pattern to CEO career management concerns and find evidence that suggests that CEOs can reduce their likelihood of being fired conditional on performance measures by leveraging social media.