Uncertain Earnings Betas
We examine how firms’ exposures to changes in market-wide earnings shape investors’ expectations of firms’ earnings and returns. Following Heinle, Smith, and Verrecchia (2018), we model changes in firms’ expected earnings as a function of two uncertain components: changes in market-wide expected earnings and firms’ exposures to those changes. We incorporate these two sources of uncertainty into a rolling-window Bayesian estimation to approximate investors’ expectations about firms’ exposures from past firm-specific and market-wide earnings forecasts. From this estimation, we obtain posterior distributions of future earnings components, as expected by the investors in our simple model. We document that the resulting distributions exhibit non-normality and that their higher moments (e.g., standard deviation and skewness) appear to be priced. Consistent with changes in higher moments containing discount rate news, we also document an association between and changes and contemporaneous returns.