#Fail: Social Media, Firm Distress, and Going Concern Opinions

Abstract

Audit firms and regulators have both commented extensively on the potential for new sources of data to transform the audit process. We examine whether information on social media could aid auditors’ going concern deliberations and find that more negative social media sentiment, measured using message bearishness and probability of failure derived from a machine-learning algorithm, corresponds to a higher likelihood of firm failure. However, we find minimal evidence that audit opinions incorporate this public information. Additional tests suggest social media sentiment relates positively to the likelihood of several future firm events that often precipitate failure. We also provide evidence that naïve consideration of social media sentiment reduces Type II errors at a faster rate than the commensurate increase in Type I errors. Our evidence should be informative to regulators and audit firms, both of whom are currently evaluating how the proliferation of data can be useful to auditors.

Keywords: StockTwits, Social Media, Sentiment, Auditing, Going-Concern Opinions, Firm Failure

JEL Classification: M40, M41, M42, M49, G33