On forecasting cointegrated seasonal time series

We analyze periodic and seasonal cointegration models for bivariate quarter ly observed time series in an empirical forecasting study. We include both single equation and multiple equation methods for those two classes of mode ls. A VAR model in first differences, with and without cointegration restri ctions, and a VAR model in annual differences are also included in the anal ysis, where they serve as benchmark models. Our empirical results indicate that the VAR model in first differences without cointegration is best if on e-step ahead forecasts are considered. For longer forecast horizons however , the VAR model in annual differences is better. When comparing periodic ve rsus seasonal cointegration models, we find that the seasonal cointegration models tend to yield better forecasts. Finally, there is no clear indicati on that multiple equations methods improve on single equation methods. (C) 2001 Elsevier Science B.V. All rights reserved.