The earnings announcement premium and Google searches

Abstract

This paper examines the link between the earnings announcement premium and Google searches for company tickers prior to the announcements. We use Google searches as a proxy for information demand. The analysis are performed on a sample of 733 stocks from the beginning of 2009 to end of the second quarter of 2015. We find a significant negative relationship between abnormal Google searches (ASVI) prior to the announcement week and the premium during the week of the announcement. Moreover, the relationship between ASVI and subsequent abnormal returns is significantly stronger during earnings announcements weeks compared to non-announcement weeks. Furthermore, we find a small negative relationship between ASVI and subsequent earnings announcement idiosyncratic volatility. Taken together, our findings indicate that when investors demand more information about a firm, the premium is mitigated and the uncertainty around the earnings announcement is reduced. Finally, we construct a trading strategy based on our findings and find that Google searches is important for predicting the earnings premium and has little predictive power when companies are not announcing earnings.