Studies have identified employee satisfaction as an important predictor of firm outcomes. However, the information we can extract from employees goes beyond simple satisfaction. In this paper, we construct multidimensional measures of employee-generated content on Glassdoor, ranging from the platform’s numeric ratings to the tone (sentiment) and content (risk-related buzz) in written reviews. We embed our measures in a Credit Default Swap (CDS) valuation framework. We find that even after orthogonalization relative to firm fundamentals and the S pillar score in ESG, our Glassdoor measures boost the weekly cross-sectional R2 of our CDS model by 2.44 percentage points. Both sentiment and risk-related buzz channels add significant contributions, with the following distinction. The influence of sentiment exhibits a cyclical pattern, whereas that of risk buzz is steadier but becomes more important during high credit risk periods. Our additional mechanism checks and robustness tests provide converging evidence. Overall, our findings reveal the novel and robust value of employee-generated content for credit risk assessment.
