Revisiting purchasing power parity in 34 OECD countries: sequential panel selection method

Abstract

Since introduction of unit roots, researchers have tried to solve the purchasing power parity (PPP) puzzle or its failure by testing for stationarity of the real exchange rates. Failure to support the PPP is mostly said to be due to low power of these tests. Panel unit root testing is more prevalent due to at least increased size of the sample and relatively more power of the tests. Within the panel, since some rates could be stationary and some nonstationary, Sequential Panel Selection Method (SPSM) should be used to distinguish the series. Furthermore, since some series could suffer from unknown structural break, the SPSM should be supplemented with Fourier function. We apply these proposed procedures to real effective exchange rate data from 34 OECD countries from January 1994 to June 2012. We find that the PPP is supported in most members of the panel.

Keywords:

• purchasing power parity
• OECD countries
• Sequential Panel Selection Method
• Fourier function

JEL Classification:

• F31

Notes

¹ See Bahmani-Oskooee et al. (2007, 2008).

² For a review article, see Bahmani-Oskooee and Hegerty (2009).

³ For the list of countries, see Table 1. Monthly real effective exchange rates over the period from January 1994 to June 2012 are used and they are constructed by the Bank for International Settlements. Each of the real effective exchange rate series was transformed into natural logarithms before the econometric analysis.

⁴ This section closely follows Bahmani-Oskooee et al. (2013).

⁵ The results are available from the authors upon request.

⁶ For some other purchasing power parity-related articles, see Beach et al. (1993), Apergis (1998), Arize et al. (2003), Baffoe-Bonnie (2004), Bleaney (1992), Hojman (1989), Holmes (2002), Horne (2004), Jung (1995) and Sjölander (2007).