Abstract:

Recent work by Clements and Hendry has shown why forecasting systems that are in terms of differences, dVARs, can be more accurate than econometric models that include levels variables, ECMs. For example, dVAR forecasts are insulated from parameter non-constancies in the long run mean of the cointegration relationships. In this paper, the practical relevance of these issues is investigated for RIMINI, the quarterly macroeconometric model used in Norges Bank, which we take as an example of an ECM forecasting model. We develop two dVAR forecasts for the period 1992.1-1994.4. In addition, we compare forecasts from the full scale models with those of univariate dVAR type models. The results seem to confirm the relevance of several important theoretical insights. dVAR forecasts appear to provide some immunity against parameter non-constancies that could seriously bias the ECM forecasts. On the other hand, for open systems like the RIMINI model, the misspecification resulting from omitting levels information seems to generate substantial biases in the dVAR forecasts. Therefore, the incumbent ECM performs comparatively well over the forecast period investigated in this paper.