What Predicts Stock Returns? – The Role of Expected versus Unexpected Predictors

Stock returns could be predictable due to time-varying expected returns. In this paper, we model the conditional expected return as an unobservable state variable affected directly by the expected components of predictors. This approach separates the different roles played by expected versus unexpected predictors, while avoiding both the “error-in-variables” problem due to imperfect predictors and bias due to persistent predictors in a typical predictive regression. In addition, the feature of negative correlation between an innovation in the expected return and the unexpected stock return is built into the model from imposing the return identity. At the same time, the unexpected components in predictors are allowed to co-move with the unexpected cash flow component of the stock return. This flexible correlation structure improves the efficiency in estimating the conditional expected returns. More importantly, we are able to study the role of predictors in forming the expected return as well as predicting the cash flow component of returns. ∗The address of the corresponding author is: Yexiao Xu, School Of Management, The University of Texas at Dallas, PO Box 688, Richardson, Texas 75080, USA; Email: [email protected]