Divisia-Geldmengenindizes | Ein Vorschlag zur theoretisch korrekten Erfassung des Geldes?

Divisia Money Supply Index. A Proposal on Theoretically Correct Monetary Recording? The present contribution analyses the proposal not to measure “money supply” by unweighted aggregate sums, but by a Divisia index. The two fundamentals of this proposal are the (microeconomic) aggregation theory and the statistical index-number theory. The aggregation theory-based considerations originate in assumptions about an economic actor’s optimisation behaviour, in the course of which “money” creates benefit for being eligible as a medium of exchange. Any benefitcreating aggregate must be consistent with the maximisation behaviour. The crucial condition for being classified as money is the poor separability of the monetary components. This can be verified with the help of both parametric and nonparametric tests. The aggregation function need not be specified for the mere reason that any discretionary function can be approximated by a group of especially appropriate (first-class) index numbers. The group of such first-class index numbers includes, inter alia, the divisia index (steady over time) and its discrete approximation (Törngvist-Theil-Index). Besides the criticism voiced about the microeconomic bases (e.g. about the model of a representative benefit-maximising economic actor or about direct monetary benefit), the main problem about the divisia approach is one of empirical implementation insofar as the price of the monetary components (user costs) is concerned. In practice, it has turned out to be difficult to make the necessary modification of the user costs ascertained on the basis of the observable market interest rates so that these can hardly be expected to correctly reflect the means-of-exchange characteristic of the monetary components and will thus lead to distorted weights in the divisia index