The Dynamic Relationships Among Medium of Exchange, Monetary Index Numbers, and the Macroeconomy

In this paper, we study the interrelationships among a newly constructed set of U.S. Divisia monetary aggregates and economic activity, including both income and the price level. Our methods permit nonstationarity, nonlinearity, asymmetry, and time variation in the relationships. Via unobserved components models, we estimate permanent and transitory components for each series, with potential structural changes in the dynamics. We then estimate a nine-equation Markov-switching structural vector autoregressive model in which each series is permitted two states. Our goal is to characterize the time-varying lead/lag and “causality” relationships among the series and their components, particularly during business cycles.