Predicting Recession Using the Yield Curve: An Artificial Intelligence and Econometric Comparison

Abstract

Normally the yield curve has an upward slope and is slightly convex. There are times when the curve becomes flat or slopes downward, which signals an approaching economic slowdown soon. Although simple, the yield curve has often been more successful at predicting recession than larger and/or more complex econometric models. Using an artificial intelligence methodology called "neural networks" and quarterly data from 1959 to 1997, the purpose of this study was to use the yield curve for forecasting recessions. The results confirm earlier findings that used econometric modelling, and suggest that the yield curve is a very useful tool for the prediction of recession. Furthermore, the results of the out-of-sample simulation imply that the forecast of the artificial intelligence method is more accurate than that of the traditional econometric model.