UvA

The cointegration literature suggests that forecast errors may be reduced by incorporating the knowledge of cointegrating relationships into linear models to generate forecasts. We show that the long-term (one- to sixty-steps ahead) forecasting performance can further be enhanced by applying nonlinear equilibrium correction models. In particular, we focus on a bivariate threshold vector equilibrium correction model with the same unknown cointegrating parameter vector in both regimes (TVECM), and a bivariate cointegration model with regime-specific cointegration vectors (LTVECM). Based on simulation experiments as well as two real data sets, and using a variety of evaluation measures, we find that the forecasting performance of the LTVECM outperforms the TVECM and the usual linear specification of the equilibrium correcting mechanism. This result holds for forecasts generated by bootstrapping and Monte Carlo simulation.