No abstract available. / Thesis (M.Comm.)-University of Natal, Pietermaritzburg, 1998.

Introduction: In recent years there has been a marked upsurge in the output of literature dealing with the demand for money, but with the exception of the North American Continent, empirical research has lagged distressingly far behind the voluminous output of theory. This dearth of empirical results has had a restrictive influence. Since any of the controversial points which are being disputed by theoreticians can only be resolved by recourse to empirical methods. The restriction of empirical research to the North American Continent has further meant that the various points under dispute have had only a limited qualification and consequently monetary theorists have had no indication as to the universalizability of their conclusions. There is thus a great need for empirical studies in other countries so that the validity of the rival theories can be tested under different conditions. It was with these thoughts in mind that the present study was undertaken. Its objectives are strictly national and no pretense of strict international comparability is made. Further, great care has been taken to avoid the pitfall so beloved of econometricians, the fallacy of reduction wherein strictly limited results are uncritically universalized. Thus no attempt has been made to draw conclusions which will have universal validity. The theoretically vital points which are going to be examined in the light of South African experience are: 1. The feasibility of distinguishing idle from active balances, and if this proves possible, the determination of the wealth and interest elasticities of these balances; 2. Dropping the explicit distinction between idle and active balances to (a) determine the role of interest rates, (b) determine the appropriate constraint on the demand function, (c) determine what effect different definitions of money have on (a) and (b); 3. To examine the stability of the demand function over time. Truth is, of course, many-sided and any uniform presentation can only aspire to present a one-sided picture, just like a photograph cannot hope to do justice to the full grandeur of nature, merely presenting a one-dimensional representation of a many dimensioned object. In spite of this restriction which is inherent in all econometrical studies, this one-sided picture seems to be justified in view of the lack of any unified and coherent treatment of the demand for money in South Africa.

A stable money demand function plays a vital role in the analysis of macroeconomics, especially in the planning and implementation of monetary policy. With the use of cointegration and error correction model estimates, this study examines the existence of a stable long-run relationship between real money demand (RM2) and its explanatory variables, in South Africa, for the period 1990-2007. The explanatory variables this study uses are selected on the basis of different monetary theories, including the Keynesian, Classical and Friedman‟s modern quantity theory of money. Based on these theories, the explanatory variables this thesis uses are real income, an interest rate, the inflation rate and the exchange rate. All variables have the correct signs, as expected from economic theory, except the inflation rate. Thus real income and inflation have positive coefficients, while the interest rate and exchange rate coefficients are negative. The results from unit root tests suggest that real income, interest rate and the inflation rate are found to be stationary, while RM2 and the exchange rate are non-stationary. Results from the Engle-Granger test suggest that RM2 and its all explanatory variables are cointegrated. Hence, we find a long-run equilibrium relationship between the real quantity of money demanded and four broadly defined macroeconomic components: real income, an interest rate, the inflation rate and the exchange rate in South Africa. Overall, the study finds that the coefficient of the equilibrium error term is negative, as expected, and significantly different from zero, implying that 0.20 of the discrepancy between money demand and its explanatory variables is eliminated in the following quarter. This evidence suggests that the speed of adjustment for money demand implies the money market in South Africa needs about four quarters to re-adjust to equilibrium. This observation agrees with the public statements of the South African Reserve Bank. Whether this will hold after November 2009 is the obvious subject of future research. / Thesis (M.Comm.) - University of KwaZulu-Natal, Pietermaritzburg, 2009.