Predicting the Cross-Section of Project Values

Abstract

To compute the value of a project, the baseline textbook recommendation is to use the Present Value (PV) formula of expected cash-flows, with a discount rate based on the CAPM. In this paper, we ask what is, empirically, the best discounting method. To do this, we study listed firms, whose expected cash-flows are measured through EPS forecasts, and whose actual prices are observed. We compare different discounting approaches of computing the PV. Our metric is the ability of each discounting approach to predict actual market prices. We find that discounting with an imputed Internal rate of return (IRR), computed on comparable firms, obtains very good results. Discounting based on expected returns (such as the CAPM or multi-factor), or a simple multiple approach, performs less well. Significant, but small, improvements can be obtained by allowing, in a simple and actionable way, for a more flexible term structure of expected returns. We benchmark all of our results with flexible, purely statistical models of prices based on Random Forest algorithms. These models do barely better than our preferred, NPV-based methods.