Year:    1986

Author:    Geisler, Klaus-Dieter

Credit and Capital Markets – Kredit und Kapital, Vol. 19 (1986), Iss. 3 : pp. 325–339

Abstract

Exogeneity tests are empirical methods with which reciprocally assumed dependencies among two or a small number of variables are checked with respect to their empirical correctness. They proceed from the deliberation that the future cannot influence the past (“post hoc ergo propter hoc”). The use and evaluation of exogeneity tests, however, are not without problems. Despite numerous difficulties, exogeneity tests have recently been used frequently also for the Federal Republic of Germany with respect to the interrelationships between the money supply and income trends, though up to now not for the central bank money supply that is used by the German Bundesbank as a monetary, intermediate objective, and also not for closed periods in which active money supply control was practiced by the German Bundesbank. The test results presented here indicate that in the period under investigation (1973 – 1984) fluctuations in the money supply preceded corresponding changes in the trend of the national product and were suitable for forecasting national product trends; the hypothesis of a “reverse” relationship could be demonstrated as false in the majority of cases for this period. More broadly defined money supply aggregates such as the central bank money supply and M 3 showed in these tests greater suitability with regard to intermediate objectives than the more narrowly defined money supply magnitude M 1. In comparison, from the studies carried out M 2 appears to be unsuitable as an intermediate objective. The exogeneity of the money supply in the sense of a simple “monetaristic” interpretation of the monetary transmission process, where interdependent economic relationships prevail, cannot be substantiated without theoretical deduction or recourse to structural models with far-reaching a priori restrictions. Hence, the results of such tests should be regarded as preliminary studies for more far-reaching investigations, but not as methods for checking conceptions of economic theory.