In May 2004, ten additional countries have joined the European Union. The new members are Cyprus, Malta, Czech Republic, Poland, Hungary, Slovenia, Slovakia, Latvia, Lithuania and Estonia. The second step of economic integration for these countries is the membership to EMU and the adoption of the single currency.1 These developments have created a long debate about the way EU will be affected by this enlargement. Among others, the stability of the single European currency is an important policy question.
In this study, we attempt to examine the likelihood of emergence of significant exchange rate fluctuations in the future for the candidate EMU countries. In doing so, we estimate the equilibrium rate of the nominal effective exchange rate for Poland, Hungary, Slovakia and Malta. If significant misalignments persist, the behavior of nominal exchange rate is expected to be unstable in its attempt to find its equilibrium rate. If the actual rate is undervalued, the domestic economy is expected to face inflationary pressures. On the other hand, if the actual rate is overvalued, the domestic economy is expected to loose competitiveness. Each of the above scenarios will cause significant problems to the process of joining EMU. In contrast, an observed exchange rate close to its equilibrium implies that we do not expect large fluctuations in the future, excluding unanticipated shocks.
This paper’s contribution to the EMU enlargement empirical literature is the way of examining exchange rate stability. In other words, our approach accepts the exchange rate convergence criterion as a necessary but not sufficient condition for successful entry into EMU.2 The intuition is that if the exchange rate is currently stable but, significantly away from its equilibrium rate, the exchange rate is likely to be unstable in the future. Moreover, a high misalignment rate can cause macroeconomic instability as well, because the unstable exchange rate will affect negatively the macroeconomic indicators. Therefore, the stability of Euro will not be weakened if the examined exchange rates are not significantly misaligned. The estimation of the equilibrium effective exchange rate is undertaken by the Behavioral Equilibrium Exchange Rate (BEER) and the Permanent Equilibrium Exchange Rate (PEER) approaches, presented by Clark & MacDonald (1998) and MacDonald (2000). The BEER approach involves the direct econometric analysis of the behavior of the exchange rate. It estimates exchange rate misalignments in accordance with the deviations of the actual exchange rate from the estimated value, derived from the relationship between the exchange rate and the macroeconomic fundamentals. The PEER approach is similar to the BEER one. The PEER differs from BEER in the way that the exchange rate is a function of variables, which have only persistent effect on it.
There is a plenty of empirical work in the literature regarding the estimation of equilibrium exchange rates. A starting point is convergence to Purchasing Power Parity in the long run as a baseline of equilibrium exchange rate. Recently, there is an increasing interest in PPP hypothesis for developing countries. Some studies apply univariate unit root tests on real exchange rates (i.e. Bahmani-Oskooee, 2000), while others apply more powerful panel unit root tests (for example Alba & Park, 2003). Some researchers apply univariate, multivariate and panel cointegration techniques on the relationship between nominal exchange rates and relative prices. Mahdavi & Zhou (1994) find evidence of weak-form PPP in 8 out of the 13 examined developing countries and state that stronger evidence exists in relatively high inflation countries. Others show that convergence to PPP equilibrium may be a non-linear instead of a linear mean reverting process (i.e. Sarno, 2000; Liew, 2003).
Besides to PPP equilibrium, numerous empirical studies estimate equilibrium exchange rates by employing alternative - to the traditional - approaches. Coudert & Couharde (2002) estimate the equilibrium of seven CEEC exchange rates. The applied methodology is the FEER3 while the estimation is undertaken by the NIGEM macroeconometric model. They find, among others, that during 2000-2001 the effective exchange rates of the Hungarian forint, the Polish zloty and the Slovak crown or their bilateral rates against Euro do not deviate significantly from their equilibrium. Similarly, Egert (2002) combining the BEER and PEER approaches estimates the equilibrium exchange rate of five Central European Countries. The estimated period is from 1992 to 2001 for the cases of Hungary and Poland, while for Slovakia the estimated period begins from 1993. He finds that the Polish zloty and the Slovak crown were overvalued, but the Hungarian forint was undervalued before its convergence.
Egert & Lommatzsch (2003) find that the bilateral exchange rates of the Hungarian forint and the Polish zloty against Euro were overvalued in the fourth quarter of 2002 by 7%-12% and 12%-15%, respectively. Bulir & Smidkova (2005) examine the real effective exchange rates of the Polish zloty and the Hungarian forint, from 1995:q1 to 2003:q1, applying the NIGEM macroeconometric model. Their findings imply that the forint was close to equilibrium until 2000, but overvalued from 2001 and now. Similarly, the zloty was not away from its equilibrium rate before 2000. Thereafter, the zloty was overvalued and in 2003 the misalignment rate decreased to less than 20%.
Egert & Lahreche-Revil (2003) combined the FEER and BEER approaches estimate a VAR-based three equation cointegration system. The estimation sample is from 1992:q1 to 2001:q2 for Hungary and Poland, and from 1993:q1 to 2001:q2 for Slovakia. The under examination exchange rates are real effective exchange rates, computed as a weighted average of US dollar and Euro4. They find that the Hungarian forint was undervalued until 1998 and close to equilibrium thereafter. The polish zloty was overvalued during the estimated period while the highest deviation is observed after 1995. The Slovak crown was very close to equilibrium until 1997. From then, it is overvalued. The highest misalignment rate was about 10% and at the end of the estimated period was observed at 8%. Other relevant studies, which focus on developing as well as developed countries, are: MacDonald (2002); Fernandez et. al. (2001); Melecky & Komarek (2005) and Frait & Komarek (2001).