Tests of different monetary aggregates for the monetary models of the exchange rate in five ASEAN countries

Abstract

This study examines the usefulness of divisia money, relative to simple sum money, for exchange rate modelling in a period of rapid financial deregulation. This comparison is conducted using the monetary model of the exchange rate. In the long-run modelling, the divisia money is significantly superior to simple sum money in the case of Malaysia and the Philippines while indifferent for Indonesia, Singapore and Thailand.

Notes

¹ Husted and MacDonald (1999) employed unrestricted flexible-price monetary model to examine the exchange rates’ misalignments for nine Asian countries, namely Australia, India, Indonesia, Korea, Malaysia, New Zealand, the Philippines, Singapore and Thailand; while Chinn (2000a, b) estimated a monetary model of exchange rates augmented by productivity trends for Indonesia, Korea, Singapore, Thailand and Taiwan.

² Divisia money calculations were originated by Barnett (1980).

³ The 10 developed countries are Australia, Canada, France, Germany, Ireland, Japan, Netherlands, Switzerland, UK and USA, while the 10 developing countries are Indonesia, Malaysia, Mauritius, Morocco, Nigeria, the Philippines, Singapore, South Korea, Sri Lanka and Thailand.

⁴ This is because the monetary model may not hold exactly in the short run because of the short-run influence of technical analysts or chartists who do not look at fundamentals (e.g. Allen and Taylor, 1990; Taylor and Allen, 1992). The monetary model might hold in the long run because long-run PPP holds (e.g. Taylor and McMahon, 1988; Lothian and Taylor, 1996) and the government may intervene if the exchange rate gets too far away from the fundamental equilibrium (e.g. Sarno and Taylor, 2001).

⁵ For a comprehensive discussion see MacDonald and Taylor (1992) and Taylor (1995).

⁶ Similar data from Habibullah (1999) had been used by Nourzad (2002).

⁷ To choose the number of cointegrating vectors, we followed the approach taken by McNown and Wallace (1994), where the vectors were found to be significant at 1% in both the tests considered.

⁸ The FPMM without proportionality between the coefficients of exchange rate and relative money for Thailand in both simple sum and divisia M2 specifications failed to pass the diagnostic tests.