Empirical Money Demand

Abstract

An econometric assessment of the ‘stability of money demand’ faces a number of problems which, essentially, provided the content for the previous chapters. Specifically, the notion of money demand being first of all a theoretical idea and not necessarily a hidden empirical truth is expressed in the concept of an unknown DGP introduced in chapter 2. As a methodological consequence, the proper task of a probability model is not to cast some economic theory into estimable form, but to summarize statistical properties of the analyzed data series. A probability model particularly suited to capture the dynamic interdependencies between macroeconomic data series is the VAR model introduced in chapter 3. However, even if the empirical truth of this model could somehow be established, the isolation of money demand therein remained to some extent arbitrary since other equilibrium relations could likewise be effective in the data. The reformulation of a (conditional) VAR in the light of some (arbitrary) a-priori information leads to the just-identified structural ECM introduced in chapter 4.