ASYMPTOTIC INFERENCE ABOUT PREDICTIVE ABILITY

This paper develops procedures for inference about the moments of smoo th functions of out-of-sample predictions and prediction errors, when there is a long time series of predictions and realizations. The aim i s to provide tools for analysis of predictive accuracy and efficiency, and, more generally, of predictive ability. The paper allows for nonn ested and nonlinear models, as well as for possible dependence of pred ictions and prediction errors on estimated regression parameters. Simu lations indicate that the procedures can work well in samples of size typically available.