Expected Idiosyncratic Volatility Measures and Expected Returns

Is expected idiosyncratic volatility related to the cross section of asset returns? The answer depends crucially on the information set used to form the expectation. We find that idiosyncratic volatility forecasts using information available to traders at the time of the forecast is not reliably related to expected returns, contrary to several recent papers. In particular, we find no evidence of a positive relation between out-of-sample idiosyncratic volatility forecasts and expected returns. Expected idiosyncratic volatility only has a positive relationship with expected returns when forward-looking information is incorporated into the volatility estimate. This relationship is driven by the realized idiosyncratic volatility component that is unforecastable by investors a priori. Our findings are robust to several choices of volatility forecasting models and systematic factor models. Our findings are important for several reasons. First, our results have important implications for modern portfolio theory – it appears that expected idiosyncratic volatility is priced positively as hypothesized in several academic papers, so long as that expectation is formed using the full data sample. Second, our results have ramifications for portfolio management. We find that it is not profitable to build trading strategies that rely upon a relationship between expected idiosyncratic volatility (as formed using information available at the time of the trade) and expected return.