Forecasting an aggregate of cointegrated disaggregates

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 those from an aggregate, univariate time series model, and develops a simple procedure for testing those conditions. The paper then uses Monte Carlo simulations and an empirical example to examine how analysis of forecasting an aggregate might be affected by a failure to correct for cointegration. The Monte Carlo and empirical analyses indicate the effects of ignoring cointegration vary sharply with model parameterization. When the aggregate of the error correction coefficients is small, ignoring cointegration will not have large effects.