An Empirical Study of the Demand for Money in Uganda

Abstract

In this study, the demand for real money M1 and M2 is estimated for Uganda covering the period 1980-2004 on a quarterly basis. The modelling takes place within the framework of the ordinary least squares (OLS) single equation estimation method. The choice of this single equation estimation technique was because it is simple to use and it has been widely used with good results. To estimate the demand for money, two-equation error-correction models are constructed which contain the short-run dynamics and long-run economic equilibrium. It was found that, money demand was cointegrated with its determinants implying that M1 and M2 monetary aggregates are useful tools for long-run intermediate targeting of monetary policy. Also it was established that a stable money demand exists for both M1 and M2 monetary aggregates. In the long-run equilibrium of real M1, M2, the estimated income elasticity of money is close to unity while in the short-run equilibrium the income elasticity is less than unity. The study established that, real GDP, real exchange rate, currency-money ratio, return on physical capital, 91-day treasury bill rate and inflation rate are important in explaining the demand for money in Uganda. All the hypotheses that were set out could not be rejected except in five cases. First, the inflation rate variable was found to be stationary and it was therefore concluded that it is not a long-run determinant of real money demand. Second, the real exchange rate was found to be only a long-run determinant of money demand. Third, interest rate was found to be positively related to real money demand. Fourth, the investment ratio assumed unexpected positive sign both in the long-run and short-run implying that, the money demand model followed the MacKinnon hypothesis of positive relationship between money balances and investment suggesting that interest rates have not been adequate enough to stimulate investment following liberalisation of the financial sector. Finally, the results of the study also suggest that financial liberalisation and changes in political regimes did not have a significant impact on demand for real money balances in Uganda. The implications from our findings are first, the need for continuity of reform programs in the liberalisation of interest rates, it also calls for an appropriate monetary action such as keeping positive interest rates, strict money growth and exchange rate stability. Second, the monetary authority will need to couch their policy objectives and directions, that is, they will need to couch their policy objectives increasingly in terms of monetary aggregates.