Abstract:
I find empirical evidence that idiosyncratic skewness and kurtosis in addition to idiosyncratic volatility can help predict future stock returns after systematic risk and firm characteristics are controlled for. Other things being equal, stocks with high idiosyncratic volatility and kurtosis tend to have lower future returns than stocks with low idiosyncratic volatility and kurtosis. In contrast, idiosyncratic skewness affects future returns positively. I empirically test whether explanations based on equity as a call option, short-sale constraints, and coskewness/cokurtosis can help explain the significant predictive power of the idiosyncratic moments but do not find strong support for either of them. Thus, the relation between idiosyncratic moments and expected returns remains a puzzle.
