Composite index of leading indicators
sequential probability recursion
Springer Online Journal Archives 1860-2000
Abstract There are very few economic variables that capture as much public attention as the Composite Index of Leading Indicators (CLI). Designed as an early warning system for signaling recessions and recoveries in business cycles, CLI now has significant influence over the expectations of decision makers and market participants. Since its inception in 1938, the index has undergone many revisions, the last of which took effect in March 1989. In this study we evaluate the new index. A number of filters, including the Bayesian sequential probability recursion and the rule of three consecutive declines in CLI, are used to evaluate the information content of CLI in forecasting the turning points in the economy. We also report the result of our forecasting experiment to gauge the marginal value of the new index to improve forecasts of variables such as the unemployment rate, index of industrial production, GNP, sales, profit, consumption and housing starts.
Type of Medium: