52 Pages Posted: 4 Mar 2007 Last revised: 4 May 2010
Date Written: October 25, 2007
We assess the relevance of parameter uncertainty, model uncertainty, and macroeconomic information for forecasting the term structure of interest rates. We study parameter uncertainty by comparing Bayesian inference with frequentist estimation techniques, and model uncertainty by combining forecasts from individual models. We incorporate macroeconomic information in yield curve models by extracting common factors from a large panel of macro series. Our results show that accounting for parameter uncertainty does not improve the forecast performance of individual models. The predictive accuracy of single models varies over time considerably and we demonstrate that mitigating model uncertainty by combining forecasts leads to substantial gains in predictability. Combining forecasts using a weighting method that is based on relative historical performance results in highly accurate forecasts. The gains in terms of forecast performance are substantial, especially for longer maturities, and are consistent over time. In addition, we find that adding macroeconomic factors generally is beneficial for improving out-of-sample forecasts.
Keywords: Term structure of interest rates, Nelson-Siegel model, Affine term structure model, forecast combination, Bayesian analysis
JEL Classification: C5, C11, C32, E43, E47
Suggested Citation: Suggested Citation
De Pooter, Michiel and Ravazzolo, Francesco and van Dijk, Dick J. C., Predicting the Term Structure of Interest Rates: Incorporating Parameter Uncertainty, Model Uncertainty and Macroeconomic Information (October 25, 2007). Available at SSRN: https://ssrn.com/abstract=967914 or http://dx.doi.org/10.2139/ssrn.967914