New and Old Models of Business Investment: A Comparison of Forecasting Performance.

Previous empirical evaluations of investment models have focused on the relative in-sample fit of various nonstructural models. The authors' evaluation extends this work along two dimensions. First, they augment the usual set of models with two Euler equations derived explicitly from dynamic optimization under rational expectations; one Euler equation is typical of those found in the investment literature, while the other embeds 'time-to-build' lags that yield a richer dynamic structure. Second, the authors focus on out-of-sample forecast performance as the primary criterion of model evaluation. Their results indicate that the traditional models dominate both Euler equations in forecasting investment. Copyright 1995 by Ohio State University Press.