Efficient Prediction of Excess Returns

Working Paper: NBER ID: w14169

Authors: Jon Faust; Jonathan H. Wright

Abstract: It is well known that augmenting a standard linear regression model with variables that are correlated with the error term but uncorrelated with the original regressors will increase asymptotic efficiency of the original coefficients. We argue that in the context of predicting excess returns, valid augmenting variables exist and are likely to yield substantial gains in estimation efficiency and, hence, predictive accuracy. The proposed augmenting variables are ex post measures of an unforecastable component of excess returns: ex post errors from macroeconomic survey forecasts and the surprise components of asset price movements around macroeconomic news announcements. These "surprises" cannot be used directly in forecasting--they are not observed at the time that the forecast is made--but can nonetheless improve forecasting accuracy by reducing parameter estimation uncertainty. We derive formal results about the benefits and limits of this approach and apply it to standard examples of forecasting excess bond and equity returns. We find substantial improvements in out-of-sample forecast accuracy for standard excess bond return regressions; gains for forecasting excess stock returns are much smaller.

Keywords: excess returns; forecasting; regression; macroeconomic variables

JEL Codes: C22; C53; G14

Causal Claims

Cause	Effect
augmenting variables (wt) (C29)	efficiency of predicting excess returns (G17)
valid augmenting variables correlated with regression error (C29)	predictive accuracy (C52)
ex post measures of an unforecastable component of excess returns (G17)	forecasting accuracy (C53)
augmenting variables (C39)	precision of coefficient estimates (C51)
correct specification of augmenting variables (C36)	identification of causal relationships (C22)
unforecastable component of excess returns (G17)	predictive power (C52)