Beyond the headlines: Sentiment divergence and financial distress

In the age of social media, corporate financial health is increasingly influenced by diverse and often conflicting information signals from social media and traditional news sources. This study introduces sentiment divergence, the difference in sentiment expressed on social media (X, formerly Twitter) versus traditional news media, as a novel predictor of financial distress. Analyzing 1823 U.S. firms from the first quarter of 2015 to the first quarter of 2021, the results reveal that a one standard deviation increase in sentiment divergence decreases the one-year probability of default by 7-basis points, supporting theories that diverse information enhances market efficiency. Conversely, a one standard deviation increase in the volatility of this divergence increases the default probability by 46-basis points, highlighting the destabilizing influence of fluctuating sentiment, consistent with noise trading theories. Furthermore, heightened institutional investor attention dramatically amplifies financial distress, as a one standard deviation increase in institutional investor attention increases the likelihood of default by 869-basis points, underscoring how herding behavior and informational cascades can amplify financial distress. This study contributes to behavioral finance by demonstrating the complex interplay between information diversity, sentiment volatility, and investor behavior in shaping corporate financial outcomes, offering crucial implications for investors, managers, and regulators.