Long-run validity of purchasing power parity and rank tests for cointegration for Central Asian countries

Abstract

This study finds that Purchasing Power Parity (PPP) holds in the long-run for Azerbaijan, Kazakhstan and Kyrgyzstan, based on Breitung's (2001) rank tests for cointegration. Results from further analysis indicates that nominal exchange rates and relative prices are nonlinearly interrelated. Trade barriers, transportation costs and government intervention in the pricing system in these countries may have resulted in the establishment of the above-mentioned nonlinear relationship.

Acknowledgements

The authors are thankful to two anonymous referees and an editor for their constructive comments and encouragement on the earlier draft of this article. Full responsibility for any error that may remain rests on the authors.

Notes

¹Taylor (1988), Taylor and McMahon (1988) and Mark (1990) are among the first studies to test for long-run PPP using the most recently available unit root (residual-based test for cointegration) and multivariate cointegration tests in the late 1980s.

²There is a general understanding that the power of these tests may be low in the context of short-span of data; interested readers may read Taylor and Taylor (2004) for a briefing on the related power problem. In this respect, Diebold et al. (1991) and Lothian and Taylor (1996, 1997, 2000) are examples of attempts to improve the power of unit root tests for long-run PPP by using long-span of data.

³Besides, nonlinear adjustment in real exchange rate (which implies nonlinear adjustment of exchange rate towards PPP equilibrium) could also arise due to the costs of arbitrage in international goods (e.g. Dumas, 1992; Zussman, 2002; Juvenal and Taylor, 2008), or perhaps the effects of the use of technical analysis in the foreign exchange market (e.g. Allen and Taylor, 1990; Kilian and Taylor, 2003; Sager and Taylor, 2006; Menkhoff and Taylor, 2007).

⁴Recently, it has been argued that these linear testing procedures may be defective if the PPP holds with nonlinear adjustment (see, for instance, Taylor and Peel, 2000; Taylor and Taylor, 2004; Taylor, 2006). Evidences of nonlinear adjustment of real exchange rates in developed and developing economies are provided by Taylor and Peel (2000), Taylor et al. (2001), Kilian and Taylor (2003), Liew et al. (2003, 2004, 2009), Anoruo et al. (2006) and Lothian and Taylor (2008), to name some. The current study for the first time considers nonlinearity in the PPP study for Central Asian countries.

⁵Briefly, Breitung (2001) proposes two kinds of rank tests, one for the detection of cointegration, and the other one to distinguish linear from nonlinear relationship if cointegration exists.

⁶These tests had been for the first time adopted by Liew et al. (2009) to provide evidence supporting the of nonlinear PPP for the East Asian economies.

⁷To deal with plausible serially correlated errors and endogenous regressors, residuals in Equations 3 and 4 are estimated using the dynamic OLS procedures of Stock and Watson (1993), see Breitung (2001) for details.