Footnote:
Nach Informationen von SSRN wurde die ursprüngliche Fassung des Dokuments October 1997 erstellt
Description:
We consider the forecasting of cointegrated variables, and we show that at long horizonsquot; nothing is lost by ignoring cointegration when forecasts are evaluated using standard multivariatequot; forecast accuracy measures. In fact, simple univariate Box-Jenkins forecasts are just as accurate. quot; Our results highlight a potentially important deficiency of standard forecast accuracyquot; measures they fail to value the maintenance of cointegrating relationships amongquot; variables and we suggest alternatives that explicitly do so