Forecasting with the Index of Leading Indicators
NBER Working Paper No. 244
The composite index of leading indicators is found to be a valuable tool for predicting not only the direction but also the size of near- term changes in aggregate economic activity. This conclusion is based on assessments of the leading index as a predictor of (1) business cycle turning points as dated by the National Bureau of Economic Research and (2) quantitative changes in real GNP and the composite index of coincident indicators. Specific smoothing rules are identified which reduce the frequency of false signals but still provide adequate early warning of cyclical turning points. Simple regression models based on first differences in the logarithms produce a comparatively good record of forecasts one and two quarters ahead. The best results are obtained by using predictive chains whereby, e.g., quarterly changes in the lagging index (inverted) for Q[sub t] are used to forecast changes in the leading index in quarter Q which in turn are used to forecast changes in real GNP (or the coincident index) in Q[sub t+2].
Document Object Identifier (DOI): 10.3386/w0244