LAUSR

In this study, we investigate the impact of social media on future stock price crash risk. A stock price crash occurs when managers hoard bad news over an extended period and disclose all the bad news at once. Using Stocktwits data, we calculate informed tweets measure, which is the number of tweets with hyperlinks to original source of information divided by the total number of tweets. Our results demonstrate that future stock price crash risk is lower when the proportion of informed tweets is higher, suggesting that informed tweets on social media disseminate information and limit managers’ ability to hoard bad news. The results continue to hold when we address potential endogeneity issues using two-stage least squares regression, change analysis, and firm-fixed effect models. The cross-sectional analyses suggest that the effect of informed tweets on social media is stronger when the information environment is lower, further supporting the hoarding aversion effect. The results also suggest that informed tweets on social media serve as external monitoring mechanism. Controlling for alternative information acquisition channels, such as Google or the SEC EDGAR database does not change the inferences.