Growth effects of currency mismatches: evidence from emerging economies

Abstract

This article investigates empirically the effects of currency mismatches on economic growth and volatility of growth in emerging market economies. Using the Arellano–Bond dynamic panel data approach, the analyses provide evidence on the presence of adverse growth effects of currency mismatches. This illustrates the crucial role played by foreign currency denominated debts in the propagation mechanism of contractionary depreciations in emerging economies in the last decade. Further evidence from estimations implies that the adverse effects of mismatches on growth and growth volatility are more pronounced with higher exchange rate fluctuations.

Acknowledgement

The author is grateful to The Scientific and Technical Research Council of Turkey (TUBITAK) for providing the research grant to carry out this study.

Notes

¹ See Datta and Agarwal (2004) and Kottaridi (2005) for a similar use of dynamic growth models in the analysis of cross-country macroeconomic issues, where a standard growth model is augmented with variables of interest, depending on the subject matter.

² Note that the draft version of the study by Eichengreen et al. (2005) was published originally in 2003 before the study by Eichengreen et al. (2003).

³ See Goldstein and Turner (2004, pp. 42–44) for the detailed definition of the AECM measure.

⁴ Note that the AECM measure of Goldstein and Turner (2004) is multiplied by ‘minus one’ in this study, in order to use currency mismatch as a counterpart of net foreign currency liability positions.

⁵ Since a broad empirical discussion on the determinants of aggregate currency mismatches is beyond the scope of this study, to be brief and also not to disturb the text flow, our relevant regression estimates are not reported here but available upon request.

⁶ As an indicator of the institutional quality, the average of the ICRG ratings on law and order, investment profile, corruption and bureaucracy quality is used in the regressions.

⁷ The choice of the explanatory variables is based on growth models in Levine and Renelt (1992), Judson and Orphaniades (1999), Levine et al. (2000), Greenaway et al. (2002) and Levy-Yeyati and Sturzenegger (2003).

⁸ XY appears with one-period lag because only the delayed effect of exports on economic growth is found statistically significant. Moreover, OPEN(−1) is excluded from this regression not to cause a multi-colinearity problem due to XY 's being a component of OPEN by definition.

⁹ When both AECM and ΔRER are included in the regressions, the negative coefficient estimates on AECM appear to be statistically insignificant (not reported here), which seems to be a result of the fact that ΔRER is the driving factor behind the predictive power of AECM in explaining the variations in economic growth.

¹⁰ REG denotes an exchange rate regime variable scaled from 1 to 8 over time for each country in the sample. It is based on IMF's classification changing from pegged to floating regimes. Data are gathered from various issues of ‘Annual Report on Exchange Arrangements and Exchange Restrictions’ published by IMF.

¹¹ These explanatory variables can be found in the growth volatility models of Beck et al. (2001), Eichengreen et al. (2003) and Levy-Yeyati and Sturzenegger (2003).