Separating the Wheat from the Chaff: Identifying the Signal and Predictable Error Components of Target Price Forecasts

Abstract

Using the analytical framework originated in Campbell and Shiller (1988), we show that target price implied returns are a function of expected dividends, analysts’ private information about future stock returns, and errors with respect to forecasting cash flows and discount rates. Our empirical application of the model explains approximately 24 percent of the cross-sectional variation in target price implied returns, of which around eight percent is explained by analysts’ information about expected dividends and future stock returns. We also find that analyst errors in assessing the returns to risk are more important than poor fundamental forecasts or badly aligned economic incentives in explaining the cross-sectional variation in implied returns. We then use the model to decompose target price implied returns into a signal and a predictable error component. The results show that the signal component is positively associated with future returns over short time horizons (the following two months), while the predictable error component is negatively associated with future returns up to at least six months. Two trading strategies based on the two components earn abnormal returns of 74 to 113 basis points per month. These results suggest that investors fail to react efficiently to the signal and the predictable noise components of analyst target price forecasts.