Usefulness of the newly revised composite index of leading indicators as a quantitative predictor

This paper compares the performance of the current and earlier versions of the Commerce Department's Composite Index of Leading Indicators as predictors of quantitative (percentage) changes in real GNP. Almon-type distributed lag functions are estimated over a sample period covering the years 1948–1970. Tests of predictive accuracy focus on the post-sample period 1971–1975. The results, in terms of three commonly accepted measures of forecast accuracy (root mean square error, mean absolute error and Theil's U), indicate that the new index performs marginally better than the best of the older versions.