Term Structure Forecasting of Government Bond Yields with Latent and Macroeconomic Factors: Does Macroeconomic Factors Imply Better Out-of-Sample Forecasts?

This study examines the role of macroeconomic and stock market variables in the dynamic Nelson-Siegel framework with the purpose of fitting and forecasting the term structure of interest rate. We find that incorporating the macroeconomic indicators in yield curve model leads to a better in-sample fit of the term structure. The out-of-sample predictability also improves significantly for all maturities for the short horizon forecasts, however regarding the longer horizons forecasts, the forecast performance of yields-macro and yields-only models is same for maturities beyond 5 years. The one-step state-space estimation approach employed to the yields-macro model produces accurate forecasts and outperforms the results of earlier related studies. Especially, the autocorrelation of the forecasts errors and in-sample residuals persistency across maturities, which is a common phenomenon in the statistical class of term structure models, can be reduced to a greater extent by inclusion of macroeconomic factors in the yield model.