Abstract:
For over five decades, diffusion indexes have been widely used by statistical and economic agencies as an instrument to summarize the dynamics of a group of disaggregated time-series economic data. In this note we revise the methods for constructing diffusion indexes, propose a novel generalized diffusion index and apply it to the U.S. State Coincident Indexes published by the Federal Reserve Bank of Philadelphia. We show that the proposed index is more informative and conclusive regarding the stage of the aggregate business cycle than the traditional indexes used by some statistical agencies. Moreover, one of the unique properties of the generalized diffusion index is that it allows a consistent reading of the contributions of its constituent units.