Estimating The Demand For Money In A Developing Country: The Case Of South Africa

Abstract

In this paper, the demand for real money, M3, is estimated for South Africa for the period 1965 to 2003. The paper employs an Autoregressive Distributed Lag (ARDL) model using a two equation technique that includes cointegration and an errorcorrection model (ECM). The cointegration model estimates the long-run relation that might exists between the dependent variable and the explanatory variables, and the ECM determines the short-run relationship between money demand and its determinants. Recent studies (Nell, 1999; Moll, 1999; Jonsson, 2001; and Nell, 2003) found inconclusive results with regard to the stability of M3, as a monetary policy tool. While some researchers (Jonsson, 2001 and Nell, 1999) found significant relationship between M3 and its determinants, others (Moll 2000; Nell, 2003) concluded that there was no stable relationship between M3 and its determinants, with the level of inflation used as one of the determinants. Nevertheless, the test for stability conducted in this paper shows that the money demand function employed in the model is stable. The paper also found a strong and stable positive relationship between money demand and inflation, which shows that M3 as a policy tool is stable and can be used to target the level of inflation in South Africa. While previous studies did not employ statistical techniques that establish a long- and a short-run relationship between M3 and its determinants, this paper’s contribution to literature is to establish a stable long- and short-run relationship between money demand and its determinants in the context of structural breaks.