Employee Sentiment and Stock Return Predictability

Abstract

We propose a new measure of investor sentiment based on predictions of firms' near-term prospects, disclosed in online platforms by their employees. By aggregating this forward-looking information, we construct an Employee Sentiment Index (ESI) and find that it is a strong predictor of stock market returns with lower future returns following high employee sentiment. Its predictive ability is superior compared to existing measures of investor sentiment and commonly-studied macroeconomic variables. Further, it predicts cross-sectional returns from difficult to value and costly to arbitrage stocks. The predictive power of the ESI is explained by investors' biased beliefs about expected cash flows.

Keywords: Employee sentiment, online reviews, voluntary information disclosure, business outlook predictions, return predictability, asset pricing

JEL Classification: G12, G17, G40