Labor Market Information and Predictable Returns

Abstract

This study documents simple trading strategies that predict stock returns using the sentiment information of labor market related news articles. I identify the average sentiment scores of labor market related news articles of each firm daily and use them to predict stock returns. Unlike the sentiment information from capital market and product market related news articles, those from labor market related articles appear to incorporate into stock prices slower. I find that the sentiment information of labor market related news articles appears to outperform the predictive effects on stock returns compared to those of product market related articles. Also, the predictive effects of the sentiment information in news articles outperform those in press releases. Ultimately, the findings shed light on the speed of sentiment information assimilation on stock prices.

Keywords: Return Predictability, Sentiment Analysis, Textual Analysis, Labor Markets, Market Efficiency

JEL Classification: C55, G11, G14, G41, J01