The value premium, aggregate risk innovations, and average stock returns

We test whether innovations in aggregate risk, interpolated from a vector autoregressive system that contains the Chen et al. (1986) five factors as in Petkova (2006), are common factors in cross-sectional stock returns. We provide direct evidence that innovation in industrial production growth, a classical business-cycle variable that summarizes the state of the economy, is associated with the cross-sectional return predictability of individual stocks. We conclude that the role of innovation in aggregate risk is not random, and furthermore that it provides guidance concerning an important source of nonfinancial market-based risk in asset returns.