Modelling money demand in the FYR of Macedonia

The empirical analysis of money demand seems to be of crucial importance for checking the existence of a stable relationship between money and prices as a prerequisite of attaching a greater role for monetary aggregates in the conduct of monetary policy. This paper provides for an empirical analysis of the demand for narrow money in the FYR of Macedonia. Specifically, the paper deals with the following issues: the empirical modelling of the demand for money in the long-run, the short-run dynamics of money, and the stability of the demand for money. The empirical analysis of the demand for money in the FYR of Macedonia covers the period from the first quarter of 1994 up to the last quarter of 2012. We work with quarterly values of the following variables: narrow money as represented by the monetary aggregate M1 (comprising cash plus sight deposits), real Gross Domestic Product, and nominal interest rate on 3-month time deposits, denominated in domestic currency. We employ the Vector Error Correction Model (VECM) where we first check for the long-run relationship between the variables and then we investigate the short-run dynamics. As part of the estimation strategy, we first run a series of unit root test, which shows that narrow money, GDP and interest rates are non-stationary. Then, we proceed by testing for the presence of cointegration between these three variables by means of the Johansen maximum likelihood approach. In order to determine the number of cointegrating vectors, we employed the two alternative test statistics: the trace of the stochastic matrix and the maximum eigenvalue. Based on both tests we can conclude that there is cointegration relationship between the variables of interest. Hence, the cointegrating vector can be interpreted as a money demand function, where money holdings are positively related to real income and negatively associated with the short run interest rate. As for the economic importance of the obtained coefficients, it can be seen that income elasticity is much lower than unity, i.e. it is closer to 0.5. This result may reflect the fact that we model the demand for narrow money, which serves for transaction purposes and not as an asset, so that economic agents tend to economise with money holdings. This finding is further confirmed by the pretty high coefficient before the interest rate, which is in line with the interest semi-elasticity usually found in other countries with less developed financial systems. As for the short-run dynamics, the results suggest very slow adjustment of the demand for money towards its longrun equilibrium level. Finally, we estimate the recursive coefficients before real income and nominal interest rates and find that they remain quite stable over time. Therefore, we take this exercise as evidence in favour of the stability of the demand for money.