This study examines the problem of forecasting an aggregate of cointegrated
disaggregates. It first establishes conditions under which forecasts of an
aggregate variable obtained from a disaggregate VECM will be equal to thos
e from an aggregate, univariate time Series model, and develops a simple pr
ocedure for resting those conditions. The paper then uses Monte Carlo simul
ations to show, for a finite sample, that the proposed test has good size a
nd power properties and that whether a model satisfies the aggregation cond
itions is closely related to out-of-sample forecast performance. The paper
then shows that ignoring cointegration and specifying the disaggregate mode
l as a VAR in differences can significantly affect analyses of aggregation,
with the VAR-based test for aggregation possibly leading to faulty inferen
ce and the differenced VAR forecasts potentially understating the benefits
of disaggregate information. Finally, analysis of an empirical problem conf
irms the basic results. Copyright (C) 2000 John Wiley & Sons, Ltd.