Forecasting the Brazilian Yield Curve Using Forward-Looking Variables

Abstract

This paper proposes a forecasting model that combines a factor augmented VAR (FAVAR) methodology with the Nelson and Siegel (NS) parametrization of the yield curve in order to predict the Brazilian term structure of interest rates. Importantly, we extract the principal components for the FAVAR from a large data set containing a range of forward-looking macroeconomic and financial variables. Our forecasting model improves on the predictive accuracy of extant models in the literature significantly, particularly at short-term horizons. For instance, the mean absolute forecast errors are 15–40% lower than those of the random walk benchmark on predictions at the three-month horizon. The out-of-sample analysis shows that the inclusion of forward-looking indicators is the key to improving the predictive ability of the model.