A tale of idiosyncratic volatility and illiquidity shocks: Their correlation and effects on stock returns

Abstract

This paper documents that the idiosyncratic volatility shock is not a stand-alone anomaly compared to the illiquidity shock. Our empirical evidence shows that the illiquidity shock Granger-causes idiosyncratic volatility shock and is statistically and economically significant in predicting stock returns. More importantly, it subsumes most of the predictive power of idiosyncratic volatility shock. In addition, we find that it is the illiquidity shock, not idiosyncratic shock, that is significantly affected by earnings surprise and is more consistently related to other information-signaling variables such as changes in analyst dispersion, number of sell-side analyst coverage, and institutional ownership.