Predicting the money multiplier: Forecasts from component and aggregate models

This paper compares the relative abilities of two money multiplier forecasting procedures. The components method models and forecasts the multiplier's individual ratios. The second procedure is simply to model and forecast the aggregate multiplier itself. It has been suggested, though untested, that the components approach is more accurate in forecasting, because it accounts for certain ratio-specific changes that are masked in the aggregate model. The evidence presented here, based on forecasts of the MI multiplier for the period January 1980 through December 1982, indicates that the aggregate model forecasts the multiplier as well as the component procedure.