11. Econometric and Fuzzy Modelling of Indonesian Money Demand

Noer Azam Achsani, Oliver Holtemöller, and Hizir Sofyan
March 3, 2005

Money demand is an important element of monetary policy analysis. Inflation is supposed to be a monetary phenomenon in the long run, and the empirical relation between money and prices is usually discussed in a money demand framework. The main purpose of money demand studies is to analyze if a stable money demand function exists in a specific country, especially when a major structural change has taken place. Examples for such structural changes are the monetary union of West Germany and the former German Democratic Republic in 1990 and the introduction of the Euro in 1999. There is broad evidence that money demand has been quite stable both in Germany and in the Euro area, see for example Wolters, Teräsvirta and Lütkepohl (1998) and Holtemöller (2004a).

In this chapter, we explore the M2 money demand function for Indonesia in the period 1990:1-2002:3. This period is dominated by the Asian crises, which started in $ 1997$. In the aftermath of the crisis, a number of immense financial and economic problems have emerged in Indonesia. The price level increased by about $ 16$ percent in $ 1997$ compared to the previous year. In the same period, the call money rate increased temporarily from $ 12.85$ percent to $ 57.10$ percent and the money stock increased by about $ 54$ percent. Additionally, Indonesia has faced a sharp decrease in real economic activity: GNP decreased by about 11 percent. Given these extraordinary economic developments, it may not be expected that a stable money demand function existed during that period.

The main contribution of this chapter is twofold. Firstly, we provide a careful analysis of Indonesian money demand, an emerging market economy for which only very few money demand studies exist. Secondly, we do not only apply the standard econometric methods but also the fuzzy Takagi-Sugeno model which allows for locally different functional relationships, for example during the Asian crisis. This is interesting and important because the assessment of monetary policy stance as well as monetary policy decisions depend on the relationship between money and other macroeconomic variables. Hence, a stable money demand function should be supported by various empirical methodologies.

In Section 11.1 we discuss money demand specification generally and in Section 11.2 we estimate a money demand function and the corresponding error-correction model for Indonesia using standard regression techniques. In Section 11.3, we exploit the fuzzy approach and its application to money demand. Section 11.4 presents conclusions and a comparison of the two approaches.