Re-examining purchasing power parity for East-Asian currencies: 1976–2002

Abstract

We investigate the behaviour of real exchange rates of six East-Asian countries in relation to their two major trading partners–the United States and Japan. These countries, except, Singapore were affected by the financial crisis of the fall 1997. Using monthly frequency data from 1976 to 2002 and the autoregressive distributed lag (ARDL) cointegration procedure we test for the long-run purchasing power parity (PPP) hypothesis. We find no evidence for the weak form of PPP in the pre-crisis period, but strong evidence in the post-crisis period. For the post-crisis period, we also find very small persistence of PPP deviations as indicated by very small half-lives (<7 months) and narrow confidence intervals with an upper bound of 1 year or less in most countries. Our findings reveal that the East Asian countries are returning to some form of PPP-oriented rule as a basis for their exchange rate policies.

Acknowledgements

The first author would like to thank MOSTE for funding this project [Project no: 05-020-0532-EA001]. Earlier versions of this article were presented at a seminar at the Faculty of Economics & Management of University Putra Malaysia and the Asia Pacific Economics and Business Conference (2004). We are grateful to the participants for their comments and suggestions on the earlier drafts of this article. The usual disclaimer applies.

Notes

¹ Mollick (1999) utilized the standard unit root test to examine the behaviour of the real exchange rate in Brazil over the period 1855 through 1990 (136 years). Contrary to the mean reversion behaviour of the real exchange rates reported in Lothian and Taylor (1996) for the industrialized countries, the evidence based on the Brazilian data is mixed.

² The argument here is that real exchange rates tend to be more volatile under floating than under fixed rates, and hence the econometric implications of mixing data from the two exchange rate regimes are unclear (Mussa, 1986).

³ Several recent papers have provided evidence of nonlinear mean reversion of real exchange rates during the post-Bretton Woods period (Dumas, 1992; Taylor and Peel, 2000). These studies find that the speed of convergence to PPP increases with the distance of real exchange rates from their means.

⁴ Asian-6 is increasingly becoming a key player in the global market, with high export and import levels recorded. All six countries are member of the Asian Pacific Economic Corporation (APEC). The same six economies, except South Korea, also belong to Association of South East Asian Nations (ASEAN) trading bloc.

⁵ The beginning of general floating is in 1973 (quarter 2) but most researchers have found that the initial period of floating 1973–1976 is problematic as it was characterized by excessive exchange rate volatility. To avoid this problem we begin the sample period in January 1976.

⁶ The real effective exchange rate is stationary for South Korea based on both the ADF and Phillips–Perron test while the real effective exchange rates in Singapore and the Philippines are stationary based only on the Phillips–Perron test.

⁷Patel (1990) argues that since different countries use different weights in constructing price indices, the unity constraint on the price indices may not be satisfied even if PPP holds.

⁸Pesaran and Shin (1995) demonstrate that valid asymptotic inferences on short- and long-run parameters can be made under least squares estimates of an ARDL model, provided the order of the ARDL model is appropriately augmented to allow for contemporaneous correlation between the stochastic components of the data generating processes included in estimation. Hence, ARDL estimation is applicable even where the explanatory variables are I(0) or I(1).

⁹ The result for Malaysia is not surprising given the fixity of its currency to the US dollar in the post-crisis period, meaning that the only channels for real exchange rate adjustment are changes in the domestic and foreign price level.