JCMS 2007 Volume 45. Number 2. pp. 459–485 Exchange Rate Regimes and Exchange Market Pressure in the New EU Member States* ANDRÉ VAN POECK University of Antwerp JACQUES VANNESTE University of Antwerp MARET VEINER University of Antwerp Abstract Economic theory has stressed the vulnerability to currency crises of intermediate exchange regimes. ERM II constitutes a fixed but adjustable pegged exchange rate arrangement and can therefore be categorized as an intermediate regime, in contrast to polar regimes such as currency boards and freely floating exchange rates. Our regression results for eight new EU Member States reveal the role of economic fundamentals in explaining exchange market pressure in these countries and confirm the bipolar view on exchange rate regimes. We conclude that the new EU members should not enter ERM II before their fundamentals are strong enough to compensate for the vulnerability of the exchange rate regime. Otherwise the condition for entering EMU, i.e. preceding participation in ERM II without devaluation or serious tensions on the exchange market, could be jeopardized. Introduction Eight CEECs – the Czech Republic, Estonia, Hungary, Latvia, Lithuania, Poland, the Slovak Republic and Slovenia – changed to a market economy and joined the European Union in the spring of 2004. It is to be expected that they will also join the Economic and Monetary Union (EMU) in the near future. * This research was partly supported by the National Bank of Belgium and the Fortis Bank Chair. We greatly benefited from the comments of the anonymous referees of this journal. We also benefited from discussions with the participants of the 25th SUERF Colloquium (Madrid, 14–16 October 2004). © 2007 The Author(s) Journal compilation © 2007 Blackwell Publishing Ltd, 9600 Garsington Road, Oxford OX4 2DQ, UK and 350 Main Street, Malden, MA 02148, USA 460 ANDRÉ VAN POECK, JACQUES VANNESTE AND MARET VEINER One of the selection criteria to judge whether a country is allowed to join EMU is related to the stability of its exchange rate in the preceding period. According to the Maastricht criteria, in order to be eligible the currency of the candidate country has to participate fully in the EMS. In the present context this means that the country has to join ERM II. Moreover, the currency should not be subject to serious pressures within that system in the two years that precede entry into EMU. In this respect the European Commission has already made it clear that some of the current exchange rate regimes in the CEECs are not acceptable (see Commission, 2001). As a matter of fact the following regimes are not compatible with ERM II: crawling pegs, independent floats or managed floats without a mutually agreed central rate and pegs to anchors other than the euro. This means that all CEECs except Estonia and Lithuania (who have currency boards) will have to modify their exchange rate arrangements when joining ERM II. Slovenia has already abandoned its managed float. The Czech Republic, Hungary, Poland and the Slovak Republic currently use exchange rate regimes that can be classified as floating. In our opinion, the change towards ERM II constitutes a potential danger for these countries of increased susceptibility to currency crises. In the terminology of this article they have to change their exchange rate regime from one classified as ‘extreme’ (floating exchange rate) to one classified as ‘intermediate’ (fixed but adjustable exchange rate). Economic theory has recently stressed the vulnerability to currency crises of intermediate exchange rate arrangements (see Fischer, 2001). Intermediate exchange rate arrangements, such as conventional fixed pegs, fall between the two extreme exchange rate systems: monetary union or currency board which can be classified as credible hard pegs (irrevocably fixed exchange rate systems) on the one hand and floating exchange rates on the other. The reason for the high vulnerability of intermediate exchange rate systems is that they can never be made fully credible (as long as the central bank cares about domestic objectives) and that the central bank has an incentive to devalue once economic agents expect it to happen (see De Grauwe and Grimaldi, 2002). Increased capital mobility and capital import dependency further increase this vulnerability. In a world of high capital mobility the required increase in the domestic interest rate to counter a currency crisis becomes prohibitively high. In this respect it should be noted that the CEECs have removed all impediments to capital mobility and have increasingly benefited from capital inflows. Following the transition to a free market, capital inflows increased from $3.2 billion to $22.3 billion in 1990–95. During the second half of the 1990s they averaged $23.2 billion per year, equivalent to 6.3 per cent of the © 2007 The Author(s) Journal compilation © 2007 Blackwell Publishing Ltd EXCHANGE RATE REGIMES AND EXCHANGE MARKET PRESSURE 461 CEECs’ gross domestic product. These capital inflows make the candidate countries more vulnerable to sudden capital withdrawals. This concern has been voiced by a number of economists (see Buiter and Grafe, 2002). Several countries, such the Czech Republic in 1997, have already experienced rapid capital outflows on a smaller or bigger scale in the past (see Vanneste et al., 2003). An important research question therefore is to what extent does ERM II membership for the CEECs, which constitutes a compulsory transition to an intermediate exchange rate regime for a number of them, increase their vulnerability to currency crises and reduce their chances of fulfilling the Maastricht exchange rate criterion? If ERM II does increase their vulnerability to currency crises, then ERM II membership could be counterproductive. In this article we test the bipolar view of currency crises with data for the CEECs and investigate whether in the past currency crises have been more frequent in CEECs with intermediate systems, compared to those with credible fixed and flexible systems. Although there already exists an extensive empirical literature on explaining or predicting the occurrence of foreign exchange crises (see Flood and Marion, 1998), few studies have tested the bipolar view. Notable exceptions are Bubula and Otker-Robe (2003), using data for developed and emerging market countries, and Effenberger (2004), based on data for the CEECs. These studies provide some support for the proponents of the bipolar hypothesis. Darvas and Szapáry (2000), however, found no empirical evidence that the spillover effects of the global financial crises of 1997–99 on five small open economies (including three CEECs) were primarily influenced by the exchange rate regime in place. To answer the research question we focus on exchange market pressure (emp) as a general measure of tensions on the foreign exchange market and countries’ susceptibility to crisis. We compute a quarterly measure of emp for the various CEECs over the period 1990–2002 and relate it to the exchange rate system under which they operated. The advantage of this measure is that it enables a comparison of exchange market pressure in the different exchange rate regimes applied by these countries or over time within the same country. We also compute the proportion of currency crises to which the CEECs have been subjected. In a subsequent regression analysis, exchange market pressure is explained by a number of fundamental economic variables and the exchange rate regime. The rest of the article is organized as follows. In section I we give an overview of the current exchange rate regimes in the CEECs and consider the compatibility of the current choice with ERM II. We also stress the difference © 2007 The Author(s) Journal compilation © 2007 Blackwell Publishing Ltd 462 ANDRÉ VAN POECK, JACQUES VANNESTE AND MARET VEINER between being a member of ERM II and compliance with the Maastricht exchange rate criterion. In section II we present a brief overview of currency crises models and explain the bipolar view. Section III computes a measure of exchange market pressure (emp) for the countries under investigation on the basis of which crises quarters are subsequently defined. We compare the average value of emp and the occurrence of crises quarters for different exchange rate regimes. In section IV we show the results of a regression analysis explaining exchange market pressure by a set of fundamental economic variables. We test whether the exchange rate regime offers an additional explanation. Finally, we make some concluding remarks giving more perspective to our findings. I. ERM II Membership and the Maastricht Exchange Rate Criterion The current exchange rate regimes of the eight new Member States are shown in Table 1, together with their ERM II-status. On 27 June 2004 the currencies of Estonia and Lithuania, who keep their currency boards, joined ERM II. This transition is not biting since the characteristics of a currency board are more stringent than ERM II. Slovenia on the other hand had to abandon its managed floating regime and shift towards a less flexible regime, that is a horizontal band. Table 1: Exchange Rate Regimes in the New EU Member States and ERM II-Status (2006) Country Exchange rate regime Czech Republic ° Managed float ° Inflation target ° Pegged to euro ° Currency board Estonia Hungary Latvia ° Crawling peg ° Fixed exchange rate Lithuania ° Pegged to euro ° Currency board Poland Slovak Republic Slovenia ° Managed float ° Managed float ° Horizontal band Source: Authors’ own data. © 2007 The Author(s) Journal compilation © 2007 Blackwell Publishing Ltd ERM II-status ° Not compatible ° Member since June 2004 with 15% fluctuation band ° Hot compatible ° Compatible ° Not member ° Member since June 2004 with 15% fluctuation band ° Not compatible ° Not compatible ° Member since June 2004 with 15% fluctuation band EXCHANGE RATE REGIMES AND EXCHANGE MARKET PRESSURE 463 Of the five remaining countries, four currently operate under an exchange rate arrangement that is not compatible with ERM II and therefore has to be adjusted in the future. It should be clarified that ERM II membership and the fulfilment of the Maastricht exchange rate stability criterion are not the same.1 In ERM II the central rates and fluctuation bands of participating countries’ currencies against the euro are set by common procedure. The standard fluctuation band is ⫾15 per cent, while not excluding the possibility of setting a narrower band (Estonia, Lithuania and Slovenia have a 15 per cent fluctuation band, while Denmark – the only other member – has 2.25 per cent fluctuation margin). Intervention support of the ECB for the national central banks (NCB) is automatic at the margins of the band (marginal interventions); any intervention within the band (intra-marginal interventions) need not to be – but may be – supported by the ECB. Further, the ECB and the NCBs have a formal right to suspend intervention should the price stability objective be jeopardized. In ERM II realignments of central parity are made by common procedure, which both the ECB and the Member States have the right to initiate. The Maastricht exchange rate criterion states that a Member State has to respect the normal fluctuation margins provided for the exchange rate mechanism of the European Monetary System without severe tensions for at least the last two years before the examination. In particular, the Member State shall not have devalued its currency on its own initiative in the same period. Participation in ERM II for at least two years at the time of the assessment is therefore compulsory. But more is required: there should be no downward realignment of the central parity within the two-year examination period (upward realignment of the central parity is implicitly possible). Moreover, the exchange rate has to be maintained within a fluctuation margin of 2.25 per cent (in other words, narrower than the standard band of 15 per cent) around the central parity in ERM II ‘without severe tensions’. In other words, maintaining the exchange rate within the narrow margin of 2.25 per cent ‘at any cost’ by means of excessive interventions or non-market measures will not necessarily be assessed as successful fulfilment of the exchange rate stability criterion. If the exchange rate moves outside this band, a distinction is to be made between a breach of the upper margin and breach of the lower margin (a breach of the upper margin being implicitly more admissible). In such a case it is necessary to 1 See Czech National Bank (2003). The other Maastricht convergence criteria with respect to the inflation rate, the interest rate and the government deficit and debt position are not considered in this article. © 2007 The Author(s) Journal compilation © 2007 Blackwell Publishing Ltd 464 ANDRÉ VAN POECK, JACQUES VANNESTE AND MARET VEINER examine the duration of the deviation, the reasons for it, as well as interest rates and intervention policy at the time of the deviation. II. Currency Crises and the Bipolar View Since the collapse of the Bretton Woods system in the 1970s and especially since the major currency crises of the last decade, currency crises have been a widely studied subject. The theories of currency crises can be broadly divided into three generations of models. Weak macroeconomic fundamentals are the main cause of currency crises according to the first generation models pioneered by Krugman (1979). Krugman’s model considers the situation where the government pursues an overly expansionary monetary policy in a fixed exchange rate system. The excessive expansion of domestic credit results from monetization of fiscal deficits or from supporting a weak banking system. It leads to a fall in the domestic interest rate, capital outflow and a loss of international reserves. Rational speculators regard this situation as unsustainable and find the maintenance of a fixed peg impossible in the long run. They anticipate the devaluation and launch an attack long before the international reserves are completely exhausted. In conclusion the crisis stems from monetary or fiscal policies that are inconsistent with the fixed exchange rate regime. Since such crises result from macroeconomic imbalances, they are in principle predictable. However, the crisis of the EMS at the beginning of the 1990s has demonstrated that countries with relatively sound macroeconomic fundamentals can be victims of speculative attacks too. According to Eichengreen, Rose and Wyplosz (1994) there was no considerable difference between the behaviour of macroeconomic variables during the EMS pre-crisis and the crisis periods, and the crisis could not be explained by overly expansionary monetary policies in the countries concerned. So the EMS crisis gave support to the second generation models which incorporate the phenomenon of self-fulfilling expectations and multiple equilibria. In this model the central bank has an incentive to devalue when speculators expect devaluation. For example, in Obstfeld (1994) a loss in confidence sharply increases the interest rate and raises government’s cost of debt servicing. Finally, the central bank abandons the peg as the cost of maintaining it exceeds the cost of abandoning it. The core of the previous models is combined in the so-called third generation models. Third generation models were developed after the South East Asian crisis of 1997. Besides macroeconomic fundamentals (the © 2007 The Author(s) Journal compilation © 2007 Blackwell Publishing Ltd EXCHANGE RATE REGIMES AND EXCHANGE MARKET PRESSURE 465 inconsistency of economic policy with the fixed exchange rate regime) and the role of expectations, these models stress problems in the banking sector and weak institutions in general as the main driving forces behind banking crises which spill over into currency crises. Weak supervision of the banking sector and implicit government guarantees, leading to moral hazard problems and over-indebtedness abroad, are crucial elements in these models. Although different in structure, the three generations of models suggest that pegged exchange rate systems are intrinsically vulnerable to currency crisis. In an adjustable pegged exchange rate regime, a central bank always has the temptation to renege on its promise not to devalue and hence to abandon the peg. This temptation increases with the importance that the central bank attaches to other objectives than defending the peg. This can be any economic objective that is conceivably part of the central bank’s (or the government’s) social welfare function and whose attainment involves a tradeoff with the fixed peg (e.g. output stabilization or employment growth). The temptation to renege further increases with the size of the shock to the economy (e.g. the depth of the recession) and the cost of defending the peg (e.g. output loss as a result of the interest rate increase to defend the peg). The temptation to abandon the peg decreases with the cost of the devaluation. An important cost of the devaluation is the loss of credibility of the monetary authorities after the devaluation. Hence countries should prefer extreme exchange rate arrangements (such as a currency board or a freely floating exchange rate) to intermediate exchange rate arrangements (such as an adjustable fixed peg system). This is the bipolar view on exchange rate arrangements and currency crises. According to this view countries should opt for corner solutions to avoid currency crises. Flexible exchange rate regime advocates moreover stress that this system offers greater financial stability and allows transition countries to reform their economies more smoothly. Masson (1999) adds that structural differences between the CEECs and the EU countries make adjustable pegs in the CEECs especially vulnerable to speculative attacks. III. Exchange Market Pressure and Currency Crises in the CEECs In this section we investigate in a more formal way to what extent exchange markets in the CEECs have been subject to tensions between 1990 and 2003 and whether the choice of the exchange rate regime has mattered in explaining different experiences. More specifically, we test the relevance of the bipolar view for these countries. To this end we compute a quarterly measure © 2007 The Author(s) Journal compilation © 2007 Blackwell Publishing Ltd ANDRÉ VAN POECK, JACQUES VANNESTE AND MARET VEINER 466 of exchange market pressure (emp) for each of the CEECs.2 We also compute the proportion of currency crises quarters to which the CEECs have been subject. A crisis quarter is defined as one in which the value of emp takes an exceptionally high value. The notion of exchange market pressure was introduced by Girton and Roper (1977). They started from the insight that excess demand or supply on the foreign exchange market can result in a change in the price of foreign exchange as well as in a change in the level of foreign reserves. The interesting feature of the concept is that it is applicable to all exchange rate systems and to different degrees of exchange rate management. We use the extended version which includes the change in the interest rate differential, in addition to reserve and nominal exchange rate changes (see Eichengreen et al., 1995). As a matter of fact, interest rates have been frequently used in CEECs to alleviate exchange market pressure.3 Exchange market pressure therefore also captures those attacks that have been successfully resisted by the monetary authorities (since it includes reserve and interest changes). It is therefore a good measure of crisis proneness. We also take into account the different volatility of the components by using variance smoothing weights. The weights on the intervention and interest rate terms are the ratio of the standard error of the percentage change of the exchange rate over the standard error of the percentage change of reserves and the interest rate differential respectively. Exchange market pressure is thus defined as: emp = eɺ − σ eɺ σ eɺ ⋅ rɺ + ⋅ is − is* σrɺ σ(is −is* ) ( ) (1) where: ė rate of depreciation of domestic currency; ṙ increase in domestic international reserves; is − is* change in short term interest rate differential; sė, sṙ, σ(is −is* ) standard error of the variables respectively. 2 For a measurement and explanation of exchange market pressure in the original EU Member States see Pentecost et al. (2001). 3 In the spirit of the emp-measure one might also include (changes in) capital controls, since capital controls have also been used to alleviate exchange market pressure. The empirical research testing the effectiveness of capital controls often relies on data from the IMF’s Annual Report on Exchange Arrangements and Exchange Restrictions. However, this source does not measure the intensity of controls and only reports on capital outflows. In the case of CEECs it is especially hard to connect increased capital controls with increased speculative pressure on their currencies. The eight EU joiners have achieved almost full alignment with the aquis’ requirements on free capital movement. Hence during the last decade there has been almost a constant trend of liberalization of capital flows. For these reasons we do not include capital controls in our emp-measure calculations. © 2007 The Author(s) Journal compilation © 2007 Blackwell Publishing Ltd EXCHANGE RATE REGIMES AND EXCHANGE MARKET PRESSURE 467 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 Czech Republic Estonia Hungary Latvia Lithuania Poland Slovak Republic Slovenia 1990 Table 2: Exchange Rate Regimes in the New EU Member States 1990–2002 3 na 3 na na 3 3 na 3 na 3 na na 5 3 (7) 3 2 3 (8) (8) 5 3 7 3 2 3 (8) (8) 5 3 7 3 2 3 3 2 5 3 7 3 2 6 3 2 6 3 7 4 2 6 3 2 6 4 7 7 2 6 3 2 6 4 7 7 2 6 3 2 6 7 7 7 2 6 3 2 6 7 7 7 2 6 3 2 7 7 7 7 2 6 3 2 7 7 7 7 2 6 3 2 7 7 7 Sources: von Hagen and Zhou (2002); IMF Annual Report on Exchange Rate Arrangements and Exchange Restrictions and authors observations for 2000 onwards. For a description of the major characteristics of the different exchange rate regimes, see Ghosh et al. (2003, table 1.1). Notes: End-year observations; Codes in parentheses refer to the periods when the newly-introduced national currencies have not yet assumed the status as the sole legal tender. The meanings of the codes are: na = not available, 1 = currency union (no separate legal tender), 2 = currency board arrangements, 3 = conventionally fixed pegs (adjustable pegs, de facto pegs), 4 = horizontal bands, 5 = crawling pegs, 6 = crawling bands, 7 = managed floating without preannounced path for the exchange rate, 8 = independent floating. The data used to compute emp are derived from IMF International Financial Statistics (see Appendix I). Changes in the exchange rate are computed relative to the German mark using the fixed euro conversion rate after 1999. We also use Germany as anchor to compute the changes in the short-term interest rate differential. We drop the possibility of intervention by foreign authorities, which is quite realistic for the CEECs. So only the unilateral intervention measure is used. Appendix II shows the evolution of the different components used to compute the emp-measures and Appendix III shows the resulting empmeasure4 for each CEEC. The choice of the exchange rate regimes by the CEECs over the period 1990–2002 is reflected in Table 2 by an indicator ranging from 1 (currency board) to 8 (independent floating). These are end of year observations according to the official IMF classification (for the computations and the regression analysis we use end of quarter observations5). For our purpose, we prefer the IMF classification over other classifications. The IMF classification is a de jure classification based on the stated policy intentions of 4 We tried several emp-measures starting with the classical simple Girton-Roper and ending with the more sophisticated Eichengreen et al. (1995) measure. All measures proved to be highly correlated. 5 When a currency crisis occurs and the authorities opt for another exchange rate regime after the crisis but within the same quarter, we characterize that quarter with the regime that existed at the outbreak of the crisis. © 2007 The Author(s) Journal compilation © 2007 Blackwell Publishing Ltd 468 ANDRÉ VAN POECK, JACQUES VANNESTE AND MARET VEINER the monetary authorities (with some corrections for overly clear deviations between stated intentions and practice). The drawback of this approach is that policy practices may indeed diverge from promises. As an alternative, in Appendix IV a de facto classification by Reinhart and Rogoff (2004) is used. De facto classifications are based on actual movements of the exchange rate. They are essentially backward-looking and may reflect exchange rate policy intentions very poorly. For a comparison of both types of classification and a defence of the de jure approach, both because of the central role of policy intentions for expectations formation and significant conceptual and practical problems associated with de facto classifications, see Ghosh et al. (2003). Table 3 reports the average emp-measure computed for each of the exchange rate regimes that occurred in the CEECs. We also compute the number of crisis quarters. A crisis quarter is defined as one in which the emp-measure exceeds the mean value by 1.5 standard deviation. However, the general picture and conclusions are not significantly altered by using weaker (1 standard deviation) or stronger (2 standard deviations) definitions of a crisis. We also show the averages over all countries and quarters for the total fixed regimes, total intermediate exchange rate regimes and total floating regimes. Currency unions and currency board arrangements are thereby classified as credible fixed exchange rate regimes; conventional fixed pegs, horizontal bands and crawling pegs as intermediate regimes; and crawling bands, managed floating and independent floating as flexible regimes. A rough indication in favour of the bipolar view that exchange market pressure is higher in intermediate regimes than in the extreme cases is given by the observation that on average exchange market pressure was higher for CEECs operating under intermediate regimes as compared to credible fixed or flexible regimes (-0.22, -2.12 and -1.86, respectively). This is also illustrated in Figure 1, which is suggestive of a hump-shaped relationship between emp and the exchange rate system. The analysis of variance confirms that the exchange rate regime does influence the average emp-measure as the F-statistic is high and significant.6 The last column of Table 3 shows that the CEECs with intermediate regimes also experienced more currency crises than those that operated under a credible fixed or a flexible regime, when judged by the proportion of crisis quarters. The incidence of crisis quarters is much higher in intermediate regimes compared with the corner regimes. 6 This conclusion also holds for the Reinhart-Rogoff classification, but the results are statistically much weaker (see Appendix IV). © 2007 The Author(s) Journal compilation © 2007 Blackwell Publishing Ltd Exchange rate regimeb Observations Average emp-measure Standard deviation Currency unions (1) – – – Currency board arrangements (2) 79 -2.12 5.04 Total fixed regimes 79 -2.12 5.04 Conventionally fixed pegs (3) 73 -0.60 7.80 Horizontal bands (4) 14 1.26 7.94 Crawling pegs (5) 17 0.22 11.74 Total intermediate regimes 104 -0.22 8.50 Crawling bands (6) 51 -3.22 6.44 Managed floating (7) 94 -1.37 5.14 Independent floating (8) – – – Total flexible regimes 145 -1.86 5.99 Results of one-way ANOVA to test the means between fixed, intermediate and flexible systems F-statistic 2.781 Significance of the F Number of ‘crisis’ quartersa Proportion of ‘crisis’ quarters (%) – 1 1 8 2 4 14 1 3 – 4 – 1.3% 1.3% 11.0% 14.3% 23.5% 13.5% 2.0% 3.2% – 2.8% EXCHANGE RATE REGIMES AND EXCHANGE MARKET PRESSURE © 2007 The Author(s) Journal compilation © 2007 Blackwell Publishing Ltd Table 3: Exchange Rate Regime and Emp-Measure in New EU Member States 1990–2002 0.063 Source: Authors’ calculations. Notes: a The ‘crisis’ periods occur when the time series exceeds the sample mean by 1.5 standard deviation; b The figures between brackets refer to the IMF classification in Table 2. 469 ANDRÉ VAN POECK, JACQUES VANNESTE AND MARET VEINER 470 Figure 1: Exchange Rate Arrangement (IMF Classification) and Average Emp 1.5 1.26 0.5 Exchange rate arrangement 0.22 1 2 3 4 5 6 7 8 -0.60 emp -0.5 -1.37 -1.5 -2.12 -2.5 -3.22 -3.5 Source: Table 3. III. Regression Analysis In order to test for the validity of the bipolar view applied to the CEECs’ crises experience we additionally ran a number of regression equations explaining exchange market pressure by a set of fundamental variables and adding an exchange rate regime dummy. This exchange rate regime dummy takes the value of 1 for the extreme exchange rate regimes and 0 for the intermediate regime. Hence the expected coefficient of this dummy is negative. As for the choice of the fundamentals, we based ourselves on the findings of the theoretical and empirical literature (see Flood and Marion, 1998, for an overview). The following explanatory variables were used in a regression based on panel data for the eight CEECs over the period 1990 (Q1) to 2003 (Q1): current account balance (as % of GDP), domestic credit growth rate, real depreciation rate, inflation differential (with Germany) and growth rate of government borrowing. We did not include international reserves or the money supply-reserves ratio since reserves also enter in the computation of the emp-measure. We also did not include structural variables (although the use of country dummies could potentially improve the regression results). Our regression results are presented in Table 4. They confirm to a large extent the existing empirical literature on the determinants of currency crises: the current account and domestic credit growth significantly contribute in explaining exchange market pressure. The results can be marginally improved by adding the real depreciation rate, the inflation differential or the growth © 2007 The Author(s) Journal compilation © 2007 Blackwell Publishing Ltd Regression no. Constant Current account 1 -1.126 (-1.576) -2.581a (-3.501) -3.009a (-3.681) -2.592a (-3.412) -1.580a (-1.963) -1.150 (-1.605) -2.622a (-3.547) -3.029a (-3.701) -2.599a (-3.418) -0.201a (-2.820) -0.246a (-3.569) -0.267a (-3.760) -0.243a (-3.489) -0.223a (-3.039) -0.206a (-2.867) -0.252a (-3.642) -0.272a (-3.815) -0.247a (-3.532) 2 3 4 5 6 7 8 9 Domestic credit growth 0.0637a (5.374) 0.0636a (5.363) 0.0652a (5.442) Explanatory variables Real Inflation depreciation differential Government borrowing 0.0996 (1.209) -0.0498 (-0.779) 0.105 (1.221) 0.0641a (5.401) 0.0639a (5.387) 0.0652a (5.415) 0.0955 (1.156) -0.0388 (-0.587) 0.0002 (0.602) 0.0003 (0.845) 0.0002 (0.769) 0.0002 (0.655) Statistics R2adj Exchange rate arrangement dummy R2 -1.841a (-2.387) -1.606a (-2.166) -1.523a (-2.047) -1.704a (-2.288) -1.753a (-2.265) -1.859a (-2.405) -1.628a (-2.193) -1.547a (-2.076) -1.715a (-2.300) 0.040 0.034 6.820 0.119 0.111 14.562 0.123 0.112 11.303 0.120 0.109 10.996 0.045 0.036 5.050 0.041 0.032 4.658 0.121 0.110 11.091 0.124 0.111 9.149 0.121 0.108 8.867 F EXCHANGE RATE REGIMES AND EXCHANGE MARKET PRESSURE © 2007 The Author(s) Journal compilation © 2007 Blackwell Publishing Ltd Table 4: Regression Results: Dependent Variable – Exchange Market Pressure Source: Authors’ calculations. Notes: a Statistically significant at 5% level; Dataset consists of 328 quarterly observations for eight countries over the first quarter of 1990 to the first quarter of 2003. For the sources and description of data see Appendix I; t-statistics are given in parentheses. 471 ANDRÉ VAN POECK, JACQUES VANNESTE AND MARET VEINER 472 rate of government borrowing. This being said, we should keep in mind Flood and Marion’s (1998) remark that it is hard to generalize such results, since the empirical literature has shown that the relative importance of various fundamentals can vary over time for a single country and across countries during a single period. In Table 4, the exchange rate regime dummy is entered in an additive way. It is readily seen that the regime dummy significantly contributes together with the expected sign to explaining exchange market pressure: countries with intermediate regimes, ceteris paribus, experience higher exchange market pressure than countries with extreme exchange rate regimes, either credibly fixed or flexible rates. Alternatively, we have made use of interaction variables, testing the hypothesis that the effect of the economic fundamentals on exchange market pressure differs according to the exchange rate regime in operation. Therefore, we re-estimated the regression equations which revealed the two most important economic fundamentals for exchange market pressure in our data set, that is current account balance and domestic credit growth where the dummy was entered in a multiplicative way. The hypothesis of a different effect depending on the exchange rate regime is not contradicted by the estimation results, as shown by the regression below: emp = −3.62 − (0.39 − 0.24 D)CA + (0.083 − 0.028 D) DC (6.91) (3.87) ( 2.06 ) (3.81) (1.17) R 2 = 0.128; R 2 adjusted = 0.118; F = 11.893, t-values beetween brackets According to the above regression equations, a 1 percentage point increase in the current account deficit as a percentage of GDP increases exchange rate pressure by 0.15 in credibly fixed and floating regimes and by 0.39 in intermediate ones. An increase in the domestic credit growth rate of 1 per cent leads to an increase in exchange rate pressure of 0.08 when the economy operates under an intermediate regime and only 0.05 in the extreme regime cases. All this confirms the bipolar view on exchange rate arrangements applied to the CEECs experience. Conclusion In this article, we argued that the transition to EMU of the Central and Eastern European countries that became members of the EU in 2004 could be turbulent because they are required to enter ERM II and to fulfil the Maastricht convergence criterion, more specifically the exchange rate © 2007 The Author(s) Journal compilation © 2007 Blackwell Publishing Ltd EXCHANGE RATE REGIMES AND EXCHANGE MARKET PRESSURE 473 stability criterion, at the same time. Membership of ERM II constitutes for some of them a shift from a flexible to an adjustable peg system. Several economists have argued that adjustable peg systems are more vulnerable to currency crises than credible fixed pegs or freely floating regimes. This is the so-called bipolar view of exchange rate arrangements and currency crises. The empirical part of the article confirmed this view for the CEECs in the period 1990–2003. The results of the regression estimations can be interpreted in two ways. One is that the entry of the CEECs in ERM II increases their vulnerability to speculative attacks and therefore should be avoided. Buiter and Grafe (2002), for example, believe that there is a good economic case for the leading CEECs to become members of EMU without having to go through a two-year period of formal ERM membership. But this is not a real option anymore since ERM II membership has been made compulsory for those CEECs that want to become a member of the EMU. We therefore prefer another interpretation, that is that the CEECs should not enter ERM II before their fundamentals are in the ‘safe’ region so that they are less candidate for speculation. It is therefore of the utmost importance that their public finances are under control, inflation is subdued and that the CEECs enter ERM II with sustainable central parities. The real exchange rates of CEECs, especially those with pegged currencies, have been appreciating during the last decade. This might eventually require a correction for some of them (Czech Republic, Hungary, Slovakia) at the moment entering into ERM II (see Pictet, 2004; and Coudert and Couharde, 2002, for a different view). After entry into ERM II with a credible central parity, participation in the system may yield additional credibility and hence stability. As De Grauwe and Grimaldi (2002) have argued, the probability of a currency crisis in a fixed exchange rate regime such as ERM II may be lower when participation in that regime without devaluation is a condition for entry into the EMU. This is because in that case the cost of devaluation for the government sharply increases. Since this is known to the market it is likely to affect the expectations of speculators in a positive way. In this respect the experience of the former South European EMS members can be illustrative. As is well known, these countries went through a smooth transition course into the EMU, once an entry date and a credible conversion rate for their currencies was established.7 7 In the case of Greece, however, entry into ERM II (March 1998) was accompanied by a devaluation of its currency by 13.5 per cent. Subsequently, the drachma never came under pressure in the two years of participation in ERM II. © 2007 The Author(s) Journal compilation © 2007 Blackwell Publishing Ltd 474 ANDRÉ VAN POECK, JACQUES VANNESTE AND MARET VEINER Correspondence: André Van Poeck Department of Economics Universiteit Antwerpen Prinsstraat 13, 2000 Antwerpen, Belgium Tel +32-3-220 45 38 Fax +32-3-220 45 85 email: andre.vanpoeck@ua.ac.be Appendix I – Data Sources ė – Rate of depreciation of domestic currency. Defined as the percentage change of exchange rate vis-à-vis Deutsche Mark. Calculated on the basis of IMF International Financial Statistics line ae. ṙ – Proportional change in domestic international reserves. Defined as the change in the level of reserves divided by money base of previous period – IMF International Statistics financing of the balance of payments (line 79dad), for Poland the change in net foreign assets (line 11–line 16c) was used; the whole was deflated by inherited money base (IFS line 14). is − is* – Change in the short-term interest rate differential with Germany – IMF International Statistics money market rate (line 60b) for Czech Republic, Estonia, Latvia, Lithuania, Poland, Slovak Republic and Slovenia. Treasury bill rate (line 60c) for Hungary. CA – Current account (as a % of GDP) – IMF International Financial statistics line 78ALD and line 99b. DC – Domestic credit growth rate – percentage change compared to previous period. Calculations based IMF International Financial statistics line 32 (domestic credit). Real depreciation rate – defined as q̇ = ė - ṗ + ṗ*, where ṗ and ṗ* are domestic and German inflation rates respectively. The time series are lagged for 1 period. IMF International Financial statistics line 64 consumer price index and line ae bilateral dollar rate Inflation differential (with Germany) – IMF International Financial statistics line 64 consumer price index. Growth rate of government borrowing – percentage change compared to previous period. IMF International Financial statistics lines 12a and 22a (claims on central government). D – Exchange rate dummy – dummy variable takes the value of 1 for the exchange rate arrangements 1, 2, 6, 7 and 8 (see Table 1) and the value of 0 for the exchange rate arrangement 3, 4 and 5. © 2007 The Author(s) Journal compilation © 2007 Blackwell Publishing Ltd EXCHANGE RATE REGIMES AND EXCHANGE MARKET PRESSURE Appendix II – Components of Exchange Market Pressure 2Q 1991 2Q 1991 2Q 1991 4Q 1992 4Q 1992 4Q 1992 4Q 1998 2Q 2000 2Q 2000 2Q 2000 2Q 2000 4Q 2001 4Q 2001 4Q 2001 4Q 2001 80 100 Estonian kroon 4Q 2001 60 4Q 2001 40 4Q 2001 Latvian lat Polish zloty 2Q 2000 0 3Q 1993 Slovenian tolar 2Q 2000 2Q 2002 2Q 2000 1Q 2001 4Q 1998 4Q 1999 4Q 1998 3Q 1998 4Q 1998 4Q 1998 Czech korona 4Q 1998 4Q 1998 475 Figure 2: Index of Nominal Exchange Rate (National Currency per Euro before 1999 ECU) 1993, Q1 = 100. 2Q 1991 120 100 80 4Q 1989 20 120 100 80 60 40 20 0 250 4Q 1992 Hungarian forint Lithuanian lit 200 150 100 0 50 4Q 1992 60 40 20 0 300 250 200 150 100 50 0 120 100 80 60 0 4Q 1992 2Q 1997 2Q 1997 2Q 1997 2Q 1997 40 20 2Q 1991 1Q 1996 4Q 1995 4Q 1995 4Q 1995 Slovak korona 200 180 160 140 120 100 80 60 2Q 1991 4Q 1994 2Q 1994 2Q 1994 2Q 1994 140 120 100 80 60 40 20 4Q 1989 Source: Authors’ calculations. 4Q 1989 © 2007 The Author(s) Journal compilation © 2007 Blackwell Publishing Ltd 2Q 1991 2Q 1992 4Q 1989 2Q 1997 2Q 1997 2Q 1997 2Q 1997 1Q 1991 4Q 1989 4Q 1995 4Q 1995 4Q 1995 4Q 1995 4Q 1989 4Q 1989 2Q 1994 2Q 1994 2Q 1994 2Q 1994 4Q 1992 4Q 1989 4Q 1989 1Q 1991 1Q 1991 1Q 1991 3Q 1993 3Q 1993 4Q 1994 4Q 1994 4Q 1994 1Q 1996 2Q 1997 3Q 1998 1Q 1996 1Q 1996 4Q 1994 1Q 1996 2Q 1997 2Q 1997 2Q 1997 3Q 1998 3Q 1998 3Q 1998 4Q 1999 4Q 1999 1Q 2001 1Q 2001 1Q 2001 1Q 2001 2Q 2002 2Q 2002 2Q 2002 4Q 1999 4Q 1998 2Q 2000 1Q 2001 2Q 2002 4Q 2001 2Q 1992 3Q 1993 3Q 1993 4Q 1994 1Q 1996 4Q 1994 1Q 1996 2Q 1997 2Q 1997 3Q 1998 3Q 1998 4Q 1999 4Q 1999 1Q 2001 1Q 2001 2Q 2002 2Q 2002 Estonia 3Q 1998 2Q 1997 2Q 1992 Latvia 2Q 1997 4Q 1995 Poland 1Q 1996 Slovenia 4Q 1994 2Q 1994 1Q 1991 20 3Q 1993 1Q 1991 15 4Q 1992 4Q 1989 10 2Q 1991 2Q 1992 4Q 1989 5 1Q 1991 0 4Q 1989 60 50 40 30 20 10 0 70 60 50 40 30 20 10 0 80 70 60 50 40 30 20 10 0 4Q 1989 ANDRÉ VAN POECK, JACQUES VANNESTE AND MARET VEINER 4Q 1999 4Q 1999 2Q 2002 Czech Republic 3Q 1993 Hungary 2Q 1992 3Q 1993 Lithuania 2Q 1992 2Q 1992 Slovak Republic 2Q 1992 476 4Q 1989 4Q 1989 Figure 3: Short-Term Interest Rates 30 25 20 15 10 5 0 40 35 30 25 20 15 10 5 0 80 100 60 40 0 1Q 1991 20 35 30 25 20 15 10 5 0 Source: IMF International Financial Statistics. © 2007 The Author(s) Journal compilation © 2007 Blackwell Publishing Ltd 4Q 1989 4Q 1993 1Q 1995 1Q 1995 2Q 1996 3Q 1997 4Q 1998 4Q 1998 1Q 2000 1Q 2000 2Q 2001 2Q 2001 3Q 2002 3Q 2002 0.2 0.15 0.1 0.05 0 −0.05 −0.1 −0.15 −0.2 0.3 0.2 0.1 0 −0.1 1Q 1990 1Q 1990 2Q 1991 3Q 1991 3Q 1992 1Q 1993 3Q 2000 3Q 1997 4Q 1998 1Q 2000 3Q 1991 1Q 1993 1Q 1993 3Q 1994 3Q 1994 1Q 1996 3Q 1997 1Q 1999 1Q 1999 3Q 2000 3Q 2000 3Q 2002 1Q 2002 1Q 2002 477 3Q 1997 2Q 2001 1Q 2002 1Q 1996 Estonia 1Q 1999 2Q 1996 3Q 1991 Latvia 3Q 1997 1Q 1995 Poland 1Q 1996 4Q 1993 Slovenia 3Q 1994 2Q 1996 3Q 1997 EXCHANGE RATE REGIMES AND EXCHANGE MARKET PRESSURE 4Q 1993 Czech Republic 2Q 1991 3Q 1992 Hungary 2Q 1991 Figure 4: Change in the International Reserves (Billions of US Dollars) 6 5 4 3 2 1 0 −1 −2 1Q 1990 3Q 1992 −0.2 6 5 4 3 2 1 0 −1 −2 1 0.8 0.6 0.4 0.2 0 −0.2 −0.4 −0.6 1Q 1990 1Q 1990 5 1Q 2002 4 1Q 2002 3 1Q 1999 3Q 2000 2 3Q 1997 3Q 2000 1 1Q 1999 1Q 1996 0 3Q 1997 3Q 1994 1Q 1990 Lithuania 1Q 1996 Slovak Republic 3Q 1994 −1 1Q 1993 −2 1Q 1993 0.6 3Q 1991 0.4 3Q 1991 0.2 1Q 1990 0 −0.2 −0.4 1 0.5 0 −0.5 −1 Source: IMF International Financial Statistics. © 2007 The Author(s) Journal compilation © 2007 Blackwell Publishing Ltd 1Q 1990 ANDRÉ VAN POECK, JACQUES VANNESTE AND MARET VEINER 478 Appendix III Figure 5: Exchange Market Pressure in the Czech Republic Currency crisis in May 1997 30.00 20.00 Lowering of country rating in IV quarter 1997 10.00 4Q 2002 1Q 2002 2Q 2001 3Q 2000 4Q 1999 1Q 1999 2Q 1998 3Q 1997 4Q 1996 1Q 1996 2Q 1995 3Q 1994 4Q 1993 1Q 1993 2Q 1992 3Q 1991 4Q 1990 −10.00 1Q 1990 0.00 −20.00 fixed peg managed float −30.00 Source: Authors’ calculations. Figure 6: Exchange Market Pressure in Hungary Mini-crisis in the beginning of 1995 30.00 The impact of Russian crisis August 1998 20.00 Speculation on appreciating forint 2003 10.00 fixed peg −20.00 −30.00 Source: Authors’ calculations. © 2007 The Author(s) Journal compilation © 2007 Blackwell Publishing Ltd 4Q 2002 1Q 2002 Increases in official reserves and appreciating pressure on forint 2001 crawling band 2Q 2001 3Q 2000 4Q 1999 1Q 1999 2Q 1998 3Q 1997 4Q 1996 1Q 1996 2Q 1995 3Q 1994 4Q 1993 1Q 1993 2Q 1992 3Q 1991 4Q 1990 −10.00 1Q 1990 0.00 EXCHANGE RATE REGIMES AND EXCHANGE MARKET PRESSURE 479 Figure 7: Exchange Market Pressure in Poland 30.00 20.00 10.00 4Q 2002 1Q 2002 2Q 2001 3Q 2000 4Q 1999 1Q 1999 2Q 1998 3Q 1997 4Q 1996 1Q 1996 2Q 1995 3Q 1994 4Q 1993 1Q 1993 2Q 1992 3Q 1991 4Q 1990 −10.00 1Q 1990 0.00 −20.00 crawling peg managed float crawling band −30.00 Source: Authors’ calculations. Figure 8: Exchange Market Pressure in Slovak Republic 30.00 The impact of Russian crisis August 1998 20.00 10.00 −20.00 conventional peg −30.00 Source: Authors’ calculations. © 2007 The Author(s) Journal compilation © 2007 Blackwell Publishing Ltd horizontal bands managed float 4Q 2002 1Q 2002 2Q 2001 3Q 2000 4Q 1999 1Q 1999 2Q 1998 3Q 1997 4Q 1996 1Q 1996 2Q 1995 3Q 1994 4Q 1993 1Q 1993 2Q 1992 3Q 1991 4Q 1990 −10.00 1Q 1990 0.00 ANDRÉ VAN POECK, JACQUES VANNESTE AND MARET VEINER 480 Figure 9: Exchange Market Pressure in Slovenia 30.00 20.00 10.00 1Q 1999 4Q 1999 3Q 2000 2Q 2001 1Q 2002 4Q 2002 4Q 1999 3Q 2000 2Q 2001 1Q 2002 4Q 2002 2Q 1998 3Q 1997 1Q 1999 −20.00 4Q 1996 1Q 1996 2Q 1995 3Q 1994 4Q 1993 1Q 1993 2Q 1992 3Q 1991 4Q 1990 −10.00 1Q 1990 0.00 managed float −30.00 Source: Authors’ calculations. Figure 10: Exchange Market Pressure in Estonia 30.00 20.00 10.00 −20.00 currency board −30.00 Source: Authors’ calculations. © 2007 The Author(s) Journal compilation © 2007 Blackwell Publishing Ltd 2Q 1998 3Q 1997 4Q 1996 1Q 1996 2Q 1995 3Q 1994 4Q 1993 1Q 1993 2Q 1992 3Q 1991 4Q 1990 −10.00 1Q 1990 0.00 EXCHANGE RATE REGIMES AND EXCHANGE MARKET PRESSURE 481 Figure 11: Exchange Market Pressure in Latvia 30.00 The impact of Russian crisis 1998 20.00 Pressure on lat in 1999 Banking crisis in 1995 10.00 3Q 2000 2Q 2001 1Q 2002 4Q 2002 3Q 2000 2Q 2001 1Q 2002 4Q 2002 4Q 1999 1Q 1999 2Q 1998 3Q 1997 4Q 1996 1Q 1996 2Q 1995 3Q 1994 4Q 1993 1Q 1993 2Q 1992 3Q 1991 4Q 1990 −10.00 1Q 1990 0.00 −20.00 conventional peg −30.00 Source: Authors’ calculations. Figure 12: Exchange Market Pressure in Lithuania 30.00 20.00 10.00 −20.00 currency board −30.00 Source: Authors’ calculations. © 2007 The Author(s) Journal compilation © 2007 Blackwell Publishing Ltd 4Q 1999 1Q 1999 2Q 1998 3Q 1997 4Q 1996 1Q 1996 2Q 1995 3Q 1994 4Q 1993 1Q 1993 2Q 1992 3Q 1991 4Q 1990 −10.00 1Q 1990 0.00 482 ANDRÉ VAN POECK, JACQUES VANNESTE AND MARET VEINER Appendix IV – Emp-Measure and the Reinhart-Rogoff Classification of Exchange Rate Regimes Table 3b: Exchange Rate Regime (Reinhart-Rogoff Classification) and Emp-Measure in New EU Member States Observations Average Standard Number Proportion emp-measure deviation of ‘crisis’ of ‘crisis’ quartersa quartersa (%) No separate legal tender (1) – – – – – Pre-announced peg or currency board arrangement (2) Total fixed regimes Pre-announced horizontal band that is narrower than or equal to +/- 2 per cent (3) De facto peg (4) Pre-announced crawling peg (5) 75 -1.92 3.95 0 0% Pre-announced crawling band that is narrower than or equal to +/- 2% (6) De facto crawling peg (7) De facto crawling band that is narrower than or equal to +/- 2% (8) Pre-announced crawling band that is wide than or equal to +/- 2% (9) De facto crawling band that is narrower than or equal to +/- 5% (10) Total intermediate regimes Moving band that is narrower than or equal to +/- 2% (i.e. allows for both appreciation and depreciation over time) (11) 75 -1.92 3.95 0 0% 36 -2.44 6.06 2 5.6% – – – – – – – – – – – – – – – – – – – – 78 -0.15 6.18 4 5.1% 17 -2.36 5.30 0 0% 48 -1.13 8.91 7 14.6% 179 -1.08 6.94 13 7.3% – – – – – © 2007 The Author(s) Journal compilation © 2007 Blackwell Publishing Ltd EXCHANGE RATE REGIMES AND EXCHANGE MARKET PRESSURE 483 Table 3b: (Continued) Observations Average Standard Number Proportion emp-measure deviation of ‘crisis’ of ‘crisis’ quartersa quartersa (%) Managed floating (12) 52 -2.56 5.66 1 1.9% Freely floating (13) – – – – – Freely falling (14) 15 -2.90 12.78 3 20.0% Total floating regimes 67 -2.64 6.66 4 6.0% Results of one-way ANOVA to test the means between fixed, intermediate and flexible systems F-statistic 1.488 Significance of the F 0.227 Source: Authors’ calculations. Note: a The ‘crisis’ periods occur when the time series exceeds the sample mean by 1.5 standard deviation. Figure 1b: Exchange Rate Arrangement (Reinhart-Rogoff Classification) and Average Emp exchange rate arrangement 0 1 2 3 4 −0.5 emp −1 −1.5 −2 −2.5 −3 Source: Table 3b. © 2007 The Author(s) Journal compilation © 2007 Blackwell Publishing Ltd 5 6 7 8 9 10 11 12 13 14 484 ANDRÉ VAN POECK, JACQUES VANNESTE AND MARET VEINER References Bubula, A. and Otker-Robe, I. (2003) ‘Are Pegged and Intermediate Exchange Rate Regimes More Crisis Prone?’ IMF Working Paper No. 233, November. Buiter, W.H. and Grafe, C. (2002) ‘Anchor, Float or Abandon Ship: Exchange Rate Regimes for the Accession Countries’, CEPR Discussion Paper No. 3184. Commission of the European Communities (2001) ‘Exchange Rate Aspects of Enlargement’, Directorate-General for Economic and Financial Affairs, European Economy, Feb, Supplement C. Coudert, V. and Couharde, C. (2002) ‘Exchange Rate Regimes and Sustainable Parities for CEECs in the Run-up to EMU Membership’, CEPII, No. 15, December. Czech National Bank (2003) ‘ERM II and the Exchange Rate Convergence Criterion’. Internal Document (Information material for the Czech Government), 15 July. Darvas, Z. and Szapáry, G. (2000) ‘Financial Contagion in Five Small Open Economies: Does the Exchange Rate Regime Really Matter?’ International Finance, Vol. 3, No. 1, pp. 25–51. De Grauwe, P. and Grimaldi, M. (2002) ‘Exchange Rate Regimes and Financial Vulnerability’. European Investment Bank Papers, Vol. 7, No. 2, pp. 33– 48. Effenberger, D. (2004) ‘Institutional Vulnerability indicators for currency crisis in Central and Eastern European Countries’. In Balling, M., Lierman, F. and Mullineux, A. (eds) Financial Markets in Central and Eastern Europe Stability and Efficiency (London: Routledge). Eichengreen, B., Rose, A.K. and Wyplosz, C. (1994) ‘Speculative Attacks on Pegged Exchange Rates: an Empirical Exploration with Special Reference to the European Monetary System’. NBER Working Paper No. 4898. Eichengreen, B., Rose, A.K. and Wyplosz, C. (1995) ‘Exchange Market Mayhem: the Antecedents and Aftermath of Speculative Attacks’. Economic Policy, No. 21, October, pp. 249–312. Fischer, S. (2001) ‘Exchange Rate Regimes: is the Bipolar View Correct?’ Journal of Economic Perspectives, Vol. 15, No. 2, pp. 3–24. Flood, R. and Marion, N. (1998) ‘Perspectives on the Recent Currency Crisis Literature’, NBER Working Paper No. 6380. Girton, L. and Roper, D. (1977) ‘A Monetary Model of Exchange Market Pressure Applied to the Postwar Canadian Experience’. American Economic Review, Vol. 67, No. 4, pp. 537–48. Ghosh, A., Gulde, A.-M. and Wolf, H. (2003) ‘Exchange Rate Regimes: Classifications and Conequences’. In Ghosh, A., Gulde, A.-M. and Wolf, H. Exchange Rate Regimes, Choices and Consequences (Cambridge: MIT Press). Krugman, P. (1979) ‘A Model of Balance-of-Payments Crises’. Journal of Money, Credit and Banking, Vol. 11, pp. 311–25. © 2007 The Author(s) Journal compilation © 2007 Blackwell Publishing Ltd EXCHANGE RATE REGIMES AND EXCHANGE MARKET PRESSURE 485 Masson, P. (1999) ‘Monetary and Exchange Rate Policy of Transition Economies of Central and Eastern Europe after the Launch of EMU’, IMF Policy Discussion Paper No. 5. Obstfeld, M. (1994) ‘The Logic of Currency Crises’. NBER Working Paper No. 4640. Pentecost, E.J., Van Hooydonk, C. and Van Poeck, A. (2001) ‘Measuring and Estimating Exchange Market Pressure in the EU’. Journal of International Money and Finance, No. 20, pp. 401–18. Pictet (2004) ‘Ten More Stars – Enlargement of the European Union’. Pictet & Cie, Economic Research Department, May. Reinhart, C.M. and Rogoff, K. (2004) ‘The Modern History of Exchange Rate Arrangements: A Reinterpretation’. Quarterly Journal of Economics (forthcoming), available at: «http://www.puaf.umd.edu/faculty/papers/reinhart/ papers.htm». Vanneste, J., Van Poeck, A. and Veiner, M. (2003) ‘Capital Reversals and the CEEC’s Transition to the EMU’. Bank- en Financiewezen, Vol. 7, October, pp. 390–401. Von Hagen, J. and Zhou, J. (2002) ‘The Choice of Exchange Rate Regimes: An Empirical Analysis for Transition Economies’. Centre for European Integration Studies, ZEI Working Paper No. B02-03. © 2007 The Author(s) Journal compilation © 2007 Blackwell Publishing Ltd