Real-Time Tests of the Leading Economic Index: Do Changes in the Index Composition Matter?

This paper reports real-time out-of-sample tests of the ability of the U.S. Index Leading Economic Indicators (LEI) to forecast the economy using "composition-changing" "as-published" versions of the LEI. It is an extension of recent work that focused on forecasts with a "composition-constant" LEI. The results demonstrate that the LEI helps forecasts and compositional change in the LEI does not account for poor real-time out-of-sample forecast performance found in earlier work. Reviews of the historical record reinforce the findings.