For most of the period since the mid-1970s, the Federal Reserve has expressed its monetary policy intentions by announcing the target growth rates of three principal monetary aggregates: the simple-sum Ml, M2 and M3. However, the sweeping changes and the deregulation in the financial industry have greatly affected the relevance of these traditional monetary aggregates. The unusual behaviour of the simple-sum monetary aggregates has forced the Federal Reserve to stop setting target range for M1. The measuring of monetary aggregates has become a controversial question. This paper constructs the new-benchmark Divisia monetary indexes which reflect 'moneyness' more accurately than the old Divisia indexes. I demonstrate that the historical trends of the Divisia monetary indexes are sensitive to the benchmark rates chosen in constructing these indexes. In addition, I compare the forecasting performance of the new-benchmark Divisia monetary indexes with the simple-sum and the old Divisia monetary indexes in the estimated money demand functions. I find that the new-benchmark Divisia monetary indexes provide the best static forecasting performance. The results indicate that the new-benchmark Divisia monetary indexes should be considered as alternative measures of money in studying the relationship between money and the economy. © 1991, Taylor & Francis Group, LLC. All rights reserved.
