The fault in the stars: exploring employee sentiment and stock returns through textual analysis

Prior research has used Glassdoor.com’s five-star rating system to measure employee satisfaction and explore its relationship to stock returns. Due to a 2012 internal Glassdoor.com policy change and subsequent ratings drift, we demonstrate how this five-star system suffers from abnormalities that limit its usefulness. To account for these issues, we employ textual analysis via multinominal inverse regression to create monthly firm-specific satisfaction measures for public US firms from 2009-2019. We find evidence that high minus low portfolios sorted by our sentiment measure produce significant risk-adjusted returns in contrast to the raw star rating sort portfolios. While using star ratings was an important first step in investigating the relationship between employee sentiment and corporate performance, future research should rely on methodologies free from collection biases and ratings drift.