Exchange rate variability, pressures and optimum currency area criteria: some empirical evidence from the 1990s

Abstract

This paper estimates the medium-term determinants of the bilateral exchange rate variability and exchange rate pressures for 20 developed countries in the 1990s. The results suggest that optimum currency area criteria explain the dynamics of bilateral exchange rate variability and pressures, to a large extent.

Acknowledgements

The author thanks Zdeněk Čech, Jarko Fidrmuc, Július Horváth and Luboš Komárek for helpful comments. The usual disclaimer applies. The views expressed in this paper are not necessarily those of the Czech National Bank. This work was supported by Czech National Bank Research Project No. A2/2003.

Notes

¹ Weimann (2003) argues that BE methodology may entail some econometric difficulties. He claims that there are censored as well as uncensored dependent variables. Nominal exchange rates are bounded by fluctuation margins in the ERM, while free elsewhere. This is right in the bilateral target zone. In fact, it does not have to be in the multilateral target zone. Indeed, Serrat (2000) shows that the exchange rate volatility in multilateral target zones may increase as compared to free float.

² We collected data on variable VOL_ij and FIN_ij from IMF's IFS, SD(Δy_i − Δy_j ), SIZE_ij and OPEN_ij were computed from the World Bank databases, TRADE_ij was calculated using the Directions of Trade – IMF and the World Bank databases and variable DISSIM_ij from the Monthly Statistics of Foreign Trade – OECD.

³ We use only Shea's measure of relevance of instruments, despite the fact that this test lacks distributional theory. All other tests rely on the assumptions not met in this paper. Hahn and Hausman (2002) provide the test only for one or two endogenous regressors and we identified four endogenous regressors. Yet, the test by Stock and Yogo (2004) assumes homoskedastic disturbances.

⁴ We downloaded these instruments from the website of Andy Rose: www.haas.berkeley.edu/~arose. The set of instruments is as follows: log (distance) and its square, regional trade agreement dummy, common language dummy, common border dummy, size of the economies, USA dummy and exogenous variables.

⁵ The results with the ERM participation dummy are largely similar to those with Europe dummy.