Campbell (1991) decomposes the variation in stock returns into revisions in expectations of dividend growth and revisions in expectations of future returns: (2) rt+1 - E4Pt+1 = (Ez+1E) pAd4+1+j (Et+1-E) port+1+j j=0 j=1 a) What do each of the terms in this equation mean? b) In order to empirically implement this decomposition, we need stock returns to be forecastable. This will allow us to construct Etrt+1 from a forecasting model of returns. State two different forecasting variables that have been used for returns, and state the underlying economic rationale provided for why you expect that each one would predict returns. Also state the expected sign of the coefficient B in each of the predictive regressions for these two variables. c) Explain the difference between in-sample and out-of-sample forecasting of stock returns. What is the evidence for whether stock returns are forecastable out-of-sample? Discuss with reference to the academic literature