Do Stock Markets Really Care About Skewness?
72 Pages Posted: 11 Sep 2014 Last revised: 31 Jan 2019
Date Written: January 30, 2019
Recent empirical studies show that statistical forecasts of a stock's return skewness negatively price stocks, apparently consistent with recent theoretical studies. While the theoretical studies, however, focus on skewness over long return horizons, the empirical studies focus on skewness over much shorter horizons. Using new estimators of realized skewness, we document that this inconsistency has important consequences for the inferences that we are able to draw from the empirical studies. Specifically, we first report that established skewness forecasts are much better in predicting short- than long-horizon skewness. More importantly, we next show that, while up to 80% of the pricing power of the forecasts comes from them predicting short-horizon skewness, only about 20% comes from them predicting long-horizon skewness. Taken together, our results suggest that the extant evidence should be interpreted as implying that short-horizon, but not necessarily long-horizon, skewness is priced in stocks.
Keywords: Asset pricing; physical skewness; realized skewness; quantile regression models
JEL Classification: G11, G12, G15
Suggested Citation: Suggested Citation