Modeling the Yield Curve of BRICS Countries: Parametric vs. Machine Learning Techniques

We compare parametric and machine learning techniques (namely: Neural Networks) for in–sample modeling of the yield curve BRICS countries (Brazil, Russia, India, China, South Africa). To such aim, we applied Dynamic De Rezende–Ferreira five–factor model with time–varying decay parameters a Feed–Forward Network to bond market data countries. enhance flexibility model, also introduce new procedure estimate time varying that significantly improve its performance. Our contribution spans towards two directions. First, offer comprehensive investigation in examined both by maturity; working on five at once ensure our results are not specific particular data–set; second make recommendations concerning modelling estimation choices curve. In this respect, although comparing highly flexible methods, highlight superior capabilities neural network all markets then suggest can be valid alternative more traditional methods presence marked turbulence.