Are shocks to real effective exchange rates permanent or transitory? Evidence from Pacific Island countries

Abstract

In this article, we consider the stability of the real effective exchange rates for four Pacific Island countries using the Lee and Strazicich (2003a, b) unit root test, which allows one to incorporate at most two structural breaks in the data series. Our main finding is that for Papua New Guinea and Samoa, exchange rates are stable, implying that shocks will have a transitory effect on real effective exchange rates, while for Fiji and the Solomon Islands we find exchange rates to be unstable, implying that shocks will have a permanent effect on real effective exchange rates.

Acknowledgements

Helpful comments and suggestions on earlier versions of this article from an anonymous referee of this journal and from Professor Mark Taylor are acknowledged. Any errors or omissions are, however, our own.

Notes

¹ For excellent surveys of the PPP literature, see Taylor and Taylor (2004), Taylor (2003, 2006).

² In a recent contribution Coakley et al. (2004) generalize the concept of long-run relative PPP to the case where the long-run elasticity of the nominal exchange rate with respect to relative national prices is unity, without restricting the innovation sequence to be stationary. They term this general relative PPP and develop panel regression techniques to test for PPP. This study may lead to a new group of studies on PPP.

³ See also Taylor and Taylor (2004) for an excellent tour on PPP including discussions on nonlinearities in real exchange rates.

⁴ Between 1995 and 2002, Fiji's economic growth averaged only 2.4% while per capita growth was 1.3%. Fiji's growth rate in 2004 was 3.8% compared to only 3.0% in 2003 (Narayan and Narayan, 2004). This growth was generally led by recovery in the agriculture, forestry and fisheries sector (ADB, 2005). The growth rate for GDP for 2005 has been revised downward to 1.3% from an initial forecast of 1.5% and for 2006 and 2007, the forecasts are 0.7 and 1% respectively.

⁵ Samoa's exchange rate and the peg is supportive of the economic growth prospects (IMF, 2003).

⁶ Arguing in favour of using a time trend in testing for a unit root in real exchange rates, Shively (2001) states that ‘rejection of the random walk hypothesis in favour of the trend stationary alternative is evidence that the real exchange rate returns to a long-run equilibrium and therefore evidence in favour of a true long-run purchasing power parity relationship’. For further importance of incorporating a time trend in the model (Marcela, 2003).