همگرایی در سیاست های پولی از پتانسیل کشورهای الحاق EMU : یک تجزیه و تحلیل هم انباشتگی با رژیم های متغییر

This paper investigates monetary policy convergence between the reference country (Germany) and the new Central and Eastern European EU member countries as well as Malta and Cyprus during the process of joining the European Monetary Union (EMU) and the four candidate countries, Bulgaria, Romania, Croatia and Turkey. Monetary policy convergence is examined through testing the uncovered interest parity (UIP) hypothesis. The long-run relationship between interest rates, a necessary condition for testing the UIP hypothesis, is examined using a cointegration test that considers the presence of structural breaks. The empirical findings of this paper provide significant evidence to support that German interest rates and interest rates in six sample countries, Croatia, Estonia, Hungary, Romania, Slovak Republic, and Turkey are stochastically converging. The UIP hypothesis, however, is not rejected only for Estonia, Croatia, and Turkey.

In May, 2004, ten new member countries (Cyprus, the Czech Republic, Estonia, Hungary, Latvia, Lithuania, Malta, Poland, the Slovak Republic and Slovenia) have joined the European Union (EU). Among the acceding countries eight are Central and Eastern European countries (CEECs) which have undergone major changes in their economic and political system during the transition to market economies in the 90s. Four other countries have applied for membership: Bulgaria and Romania are expected to join the union by 2007; Croatia is expected to join the EU by 2010; Turkey which is not a transition economy, however, is hoping to be a member in near future. Although there has been some catch-up of the CEE countries in comparison to the EU member countries, large economic differences still exist. For example, despite relatively high rates of growth in the second half of 1990s, the average per capita income is still significantly less than the average member country in EU. However, economic integration of the accessing countries is an integral part for the functioning of the EU. Moreover, many of the CEE countries have expressed their strong intention to join the European Monetary Union (EMU). Unlike Denmark and the United Kingdom, the new member countries do not have a special status with respect to the European Monetary Union (EMU). They have joined the EMU with the status ‘countries with derogation’ and are supposed to adopt the Euro as soon as economic convergence is achieved (see Holtemoller, 2005). For instance, sufficiently similar exchange regimes, real economic structure and monetary policies are necessary for successfully enlarging the Euro currency area. Hence, empirical evidence regarding the state of monetary policy convergence will be helpful for political decision makers. The aim of this paper is to examine monetary policy convergence between incumbents and the new Central and Eastern European EU member countries as well as Malta and Cyprus during the process of joining the European Monetary Union (EMU) and four candidate countries, Bulgaria, Romania, Croatia and Turkey. Monetary convergence represents the narrowing and finally closing the gaps in macroeconomic stability currently existing between new members, candidates and incumbents. The issue of monetary policy coordination is of major importance for the European Monetary System (EMS). Policy coordination and the resulting monetary policy convergence would be necessary for successfully enlarging the Euro currency area. In this paper, the monetary policy convergence is examined through testing the uncovered interest parity (UIP) for the new EU members and candidates. The uncovered interest parity hypothesis has been receiving a great attention for both international finance scholars and practitioners in recent years and states that the difference between domestic and foreign interest rates should correspond to the expected exchange rate change plus a risk premium. When reaching monetary integration, this risk premium should disappear such that the development of the risk premium can be interpreted as a measure of monetary integration. Of course, this is only one aspect of monetary convergence; fiscal policy indicators and inflation rate are other important factors which are not considered in this paper. The examination of monetary convergence among the EU countries has received considerable attention in recent years and has been studied from a variety of approaches. The first approach tries to determine the degree of monetary integration by calculating the correlation coefficients between the interest rates of member countries (see for example, Lemmen and Eijffinger, 1993). The second approach, similar to the first one, tries to study the dispersion of interest rates across countries (see for example, Pigott, 1994). Finally, the third approach uses time series techniques, which relate interest rate convergence to the stationarity or non-stationarity of the variables. Karfakis and Moschos (1990), Katsimbris and Miller (1993) and Edison and Kole (1994) used bivariate cointegration framework to test for interest rate linkages of Germany with other member countries in the union. The results of these papers are more discouraging than expected due to the absence of cointegration for the whole group of countries. Hence, evidence of no cointegration is compatible with the process of convergence itself being still occurring. Fountas and Wu (1998) also used bivariate cointegration framework to test relationship between the short-term interest rates of Germany and six other European countries. They found convergence for four countries when they allow for a one-time discrete break in the cointegrating relationship. Camarero et al. (2002) analyzed the convergence process followed by the European countries in order to fulfill the interest rate criterion set in the Maastricht Treaty. They considered continuously time-varying cointegrating relationships and found the evidence of interest rate convergence for all EU countries except for Italy. Holtemoller (2005) examined deviations from uncovered interest rate parity for new member countries over the period 1994–2004. Using recursive statistical tests and error correction models, he examined co-movement of interest rates between new member countries and incumbents. Their results indicate that Estonia and Lithuania seem to exhibit the highest degree of monetary integration with the Euro area. Brada et al. (2005) examine real and monetary convergence between the EU's core and recent member countries using a rolling cointegration approach. To test monetary convergence they use base money, M2 and the CPI between Germany and the recent EU members. Their results suggest that countries that joined the EU previously exhibit time-varying cointegration with the core countries over the 1980–2000. Cointegration for the transition economies was comparable for M2 and prices, but not for monetary policy and industrial output. Kocenda et al. (2006) examine the nominal and real convergence of all recent 10 EU members to the EU standards. Their results indicate strong inflation and interest rate convergence. Finally, Sander and Kleimeier (2006) examine interest rate pass-through convergence for the eight CEECs that joined the EU recently. Their results show the evidence for convergence across the CEECs with market concentration, bank health, foreign bank participation and monetary policy regime as conditioning factors. This paper makes the following contribution: We provide evidence about monetary policy coordination and convergence between new transition-economy members, transition-economy and a market economy candidates and the EU, using Germany as the reference country (see for example Camarero et al., 2002, Fountas and Wu, 1998, Brada and Kutan, 2001 and Brada et al., 2005). The aim of this paper is achieved in three steps. In the first step, detailed unit root tests are performed to establish the order of integration of the data series. The main idea of this exercise is to check whether, in the presence of structural breaks in the data, the series are integrated of order one. Zivot and Andrews (1992) one break test for unit roots is used. In the second step, we test whether there is a long-run relationship between interest rates, a necessary condition for testing the UIP hypothesis. To accomplish this, the Gregory and Hansen (1996) cointegration test that takes account of the possibility of instability in long-run relations is used. This methodology allows both a more general specification of long-run relations than conventional cointegration tests (Johansen, 1988 and Engle and Granger, 1987) do and the estimation of the date of the structural change. In addition to Gregory and Hansen (1996) test, bounds test of Pesaran et al. (2001) is also employed when the time series have different order of integration. In the third step, the long-run elasticities are estimated using the dynamic ordinary least squares (DOLS) estimator of Stock and Watson (1993) and autoregressive distributed lag (ARDL) model of Pesaran and Pesaran (1997). Then, we test whether UIP holds in sample countries. The paper is structured as follows: Section 2 presents the theory on the concept of the UIP. Section 3 discusses econometric methodology. The data set and empirical results are presented in Section 4. Finally Section 5 contains some concluding remarks.

The monetary policy convergence between Germany, which have been considered as a proxy for the EU economy, and the new Central and Eastern European EU member countries as well as Malta and Cyprus during the process of joining the European Monetary Union (EMU) and the four candidate countries, Bulgaria, Romania, Croatia and Turkey has been examined. Monetary convergence represents the narrowing and finally closing the gaps in macroeconomic stability currently existing between new members, candidates and old members. The monetary policy convergence has been examined through testing uncovered interest parity (UIP) hypothesis that has been receiving a great attention from both international finance scholars and practitioners in recent years. The UIP states that the difference between domestic and foreign interest rates should correspond to the expected exchange rate change plus a risk premium. When reaching monetary integration, this risk premium should disappear. Hence, cointegration of interest rates might provide evidence on the monetary convergence. To test whether there is a long-run relationship between German interest rates and the interest rates in sample countries, we implemented the Gregory and Hansen (1996) residual-based cointegration test, which explicitly allows for a structural break in cointegration relations. The Gregory and Hansen (1996) test detects several equilibrium relations that are omitted by the conventional Engle and Granger (1987) test. Hence, the findings of this paper suggest that a more general view of cointegration provides stronger evidence in favor of significant long-run relation between the German rates and the rates in six sample countries, Croatia, Estonia, Hungary, Romania, Slovak Republic, and Turkey. This indicates that German interest rates and interest rates in these countries are stochastically converging. The long-run elasticities were also estimated, using the dynamic ordinary least squares (DOLS) estimator of Stock and Watson (1993) and autoregressive distributed lag (ARDL) model of Pesaran and Pesaran (1997). Then, we tested whether the UIP holds in sample countries. The results suggest that the UIP hypothesis is not rejected only for Croatia, Estonia and Turkey. Overall, the results show that the interest rates in Croatia, Estonia, Hungary, Romania, Slovak Republic, and Turkey adjust towards the German interest rates. The UIP holds only for Estonia, Croatia and interestingly Turkey. The results suggest that these countries have shown significant success of achieving monetary convergence. In other words, Estonia, Croatia, and Turkey exhibit stable relationships between domestic interest rate, corresponding German interest rate (or Euro interest rate) and exchange rate. The risk premium is small and its variability is decreasing. Although this study provides preliminary evidence on the monetary convergence and neglects other important fields of convergence, one might conclude that Estonia is ready for monetary union. This result also supports the findings of Holtemoller (2005). Surprisingly, the performance of Croatia and Turkey, which are not yet members of the EU, in the coordination of monetary polices seems much better than most of the recent member countries. As for the other member countries, rejecting the UIP implies that their interest rates exhibit more instability and in the long-run, interest rate differentials remain sizable and international capital mobility remains incomplete. Therefore, still more efforts are needed by the regulators of these countries in the coordination of monetary policies.