貨幣需求函數與結構改變：ASEAN-4的實證研究

Abstract

The study on the money demand stability has occupied the field of monetary economics for decades. We re-investigate it in order to understand money demand function under influence of structural changes. Previous studies which have neglected structural changes may have distorted results; that is the reason which motivates this research. To do what previous studies have failed, in this paper, we re-examine the money demand stability of four ASEAN countries (hereafter ASEAN-4) for the period from 1970Q1 to 2004Q4; those are Indonesia, Thailand, Malaysia, and Singapore, respectively. Conclusions drawn from the empirical findings suggest that in implementing monetary policy, monetary authorities may have to also consider structural change on money demand function. Therefore, this study may serve as a guideline for monetary authority of ASEAN-4 countries to conduct monetary policy more effectively. Our money determinants inclusion which differs with Chaisrisawatsuk et. al. (2004) and Dekle and Pradhan (1999) as they did not consider structural change in unit root tests, therefore their empirical results might be biased. Evidences show that breakpoints of most series are happened on the aggression of Asian financial crisis. Next, the sign and magnitude of each money determinants can be considered by the government of ASEAN-4 countries in choosing correct monetary instrument and determining the proper time and dosage of intervention, but it depends on the parameter stability of the cointegrating relationships. Furthermore, most of money demand functions are found to be not stable, and then there is harder for each government to set up the monetary policies towards money equilibrium, except for Singapore. Our instability results differ with Dekle and Pradhan (1999) which treated structural changes as known information by using dummy variable, because we regard the structural break events as unknown information to avoid subjectivity. Most of graphical results show that Asian financial crisis is the source of instability, and several cases can be referred to the changing of national leadership. At last, several outcomes of Gregory and Hansen (1996, hereafter GH) test contradicts with Johansen (1988) test results, due to the traditional cointegration test did not consider structural break into calculation. But without any confirmation from weak exogeneity test, we cannot assure whether cointegrating relationships are money demand functions or not. Most of breakpoints are outsourced from Asian financial crisis and located within the instable regime, consistent with the Hansen (1992) oscillatory results.