Velocity of Money, i.e. the ratio of nominal income to the stock of money, is the spread at which the money changes hand in a given financial year. Keynesian and Quantity theories are two competing explanations of the aggregate money demand. The monetarists think that the stability of income velocity of money (V) is important, whereas Keynesian have criticized the notion of stability of velocity of money. Monetarists believe that velocity of money is relatively stable and changes therein are highly predictable. Neo-Keynesians are less confident and argue that either contention is an exaggeration. In this paper the velocity of money function was estimated using annual data for broad money velocity (V3) for the Indian economy covering time period from 1972-2004. Various combinations of the explanatory variables were tried, each with and without the lagged dependent variables. It was found that (i) velocity of broad money [V3] in India turns out to be highly predictable, which corroborates monetarist proposition. Prior knowledge of a set of explanatory variables such as real income, interest rate, the spread of commercial bank branch-network, and past monetary expansions together can explain more than 98 percent of the variation in velocity of broad money (ii) a comparison of the estimated velocity functions, with and without the 'institutional' variables suggests that inclusion of the 'institutional' variables improves the statistical properties of the fit i.e. 2 and Durbin Watson [DW] statistic. This implies that the long run real income elasticity of money demand is significantly reduced when 'institutional' variables are added to the array of the conventional explanatory variables such as real income and interest rate. However, out of the two institutional variables considered, 'population per bank office' is found to be highly significant in all the cases, while the other institutional variable measuring the 'degree of monetization' is not found to be significant.
Read full abstract