Modeling and forecasting the yield curve by an extended Nelson-Siegel class of models: A quantile autoregression approach

摘要

This paper compares the in-sample fitting and the out-of-sample forecasting performances of four distinct Nelson-Siegel class models: Nelson-Siegel, Bliss, Svensson, and a five-factor model we propose in order to enhance the fitting flexibility. The introduction of the fifth factor resulted in superior adjustment to the data. For the forecasting exercise the paper contrasts the performances of the term structure models in association with the following econometric methods: quantile autoregression evaluated at the median, VAR, AR, and a random walk. As a pattern, the quantile procedure delivered the best results for longer forecasting horizons.