Original sin and the exchange rate regime debate: lessons from Latin American and transition countries.

Jonas, Jiri

INTRODUCTION

The financial crises of the 1990s have ignited an intense debate about the international financial architecture. The question of what exchange rate regime is optimal for emerging market countries (EMCs) becoming integrated with the international capital market has attracted special attention. The fact that all of the EMCs suffering a financial crisis in the 1990s had some kind of fixed exchange rate arrangement has discredited the so-called soft peg as a preferable or even a viable exchange rate arrangement. Even though views continued to differ about what exchange rate arrangements are appropriate for EMCs, (2) a consensus seemed to be emerging favouring the so-called corner solutions, either a flexible regime or a hard peg such as a currency board arrangement (CBA), dollarisation, or creation of a monetary union.

The collapse of Argentina's CBA, accompanied by a serious currency, banking, and debt crisis, will probably have important effects on the debate about what exchange rate regimes are appropriate for EMCs. For most of the 1990s, Argentina was considered a successful economy. The view seemed widely be held that the CBA had contributed much to Argentina's successful stabilisation and resumption of growth. However, after its collapse in late 2001, history will probably judge its CBA less kindly.

In this paper, I will highlight the link between the inherent vulnerability of the EMCs (the so-called original sin hypothesis) and the operation of different exchange rate regimes. I will begin with a brief review of the latest discussion about exchange rate arrangements for EMCs following the recent financial crises. Then I will discuss the structural weaknesses that typically make EMCs vulnerable to negative external shocks. I will illustrate how these weaknesses affect different currency regimes, using examples such as Argentina with its CBA, Brazil with its floating currency, and Panama with dollarisation.

In the last part, I will discuss why the accession countries in Central and Eastern Europe are much less sensitive to external shocks no matter what their exchange rate regime. I will argue that the main reasons for this difference are smaller role of debt-creating flows in the financing of current account deficits, higher domestic savings, and--most important of all--the prospect of accession to the European Union (EU) and Economic and Monetary Union (EMU), which provides an important anchor for market expectations and sends a credible signal of countries' commitment to disciplined economic policies.

ETERNAL CURRENCY REGIME DEBATE

The debate of optimal currency arrangements has a long history (eg, see Cooper, 1999). The frequent recurrence of this reflects the apparent fact that no currency arrangement operates satisfactorily at all times and in all places.

The spectacular collapse of fixed exchange rates in the 1990s, in connection with financial crises in several EMCs, has illuminated in a very destructive way the risks of maintaining fixed exchange rate regimes in EMCs that are becoming integrated into the world capital market and are therefore exposed to large swings in capital flows. The conclusions drawn from the experience of the 1990s is that it is significantly more difficult to maintain a successful fixed exchange rate regime in a world of relatively free capital movements. The prerequisites for maintaining a successful fixed exchange rate regime have increased so much that a soft peg is no longer a viable option for most, if not all, emerging market countries.

After the collapse of the fixed currency regimes in Asia and elsewhere, a consensus began to emerge in favour of the so-called 'corner solutions.' This view holds that EMCs open to capital flows should choose either a freely floating exchange rate or a hard currency peg, that is, either dollarisation or a CBA. Some economists expressed no preference; others supported either one or the other. (3) It appears that most EMCs are moving toward more flexible currency regimes (see Fischer, 2001), but some authors have documented that the movement toward more flexibility is not so clear when one examines the actual practices of exchange rate management as opposed to the declared currency regimes. (4)

What is noteworthy in this debate is that all corner solutions, that is, the currency board, dollarisation, and the float appear to have serious problems that can complicate exchange rate management and economic policy in EMCs. It is possible to raise serious objections against any of these currency regimes, and to find evidence of problems in practice.

* CBA: Even before the collapse of Argentina's CBA, there were serious reservations about whether this currency regime could be applied to more than a handful of EMCs. Roubini (2001) argues that the reason for the success of countries with CBAs was not the currency regime but the fact that they were pursuing correct policies otherwise. Countries pursuing correct policies do not need to adopt a CBA. Goldstein (2002) lists the potential drawbacks of a CBA, and concludes that this regime is appropriate only for countries with poor policy track records but solid fiscal and financial sector fundamentals. The collapse of Argentina's CBA gives the views of these skeptics a further boost.

* Dollarisation: Dollarisation was sometimes seen as a solution in cases where neither a CBA nor a floating currency was an attractive alternative. However, as Edwards (2001) argues, there is little empirical evidence to justify the view of dollarisation's proponents that it is a workable and attractive currency arrangement for more than a small number of countries.

* Floating currencies: Calvo and Reinhart (2000) have argued that floating is less advantageous for EMCs than its proponents suggest. They note that EMCs that float their currencies nominally exhibit relatively more variable interest rates and relatively less exchange rate variability than developed countries, and that they pursue procyclical interest rate policies. In their words, EMCs exhibit a 'fear of floating'--they are unable or unwilling to enjoy the most important benefit of a floating currency, which is the ability to implement their own monetary policy. Policymakers in EMCs are concerned about the effects on their economies of large exchange rate fluctuations, and try to limit these fluctuations through more aggressive interest rate policies and interventions in the foreign exchange market. Under such conditions, the exchange rate does not serve as an absorber of external shocks, but is rather seen as a potential channel of instability. Furthermore, for EMCs depreciations can be contractionary, because the adverse balance sheet effect of a weaker currency outweighs its positive effect on net exports. Calvo and Reinhart argue that the middle ground is not actually disappearing, and that in practice, 'corner' solutions are still rare.

The debate about the optimal currency arrangement for EMCs suggests one conclusion: the inherent vulnerabilities of these countries makes the choice of the currency regime especially difficult, and increases the probability that whatever regime is chosen will not operate smoothly.

EMERGING MARKET COUNTRIES ARE VULNERABLE BY DEFINITION

By definition, EMCs are very vulnerable. 'Emerging' means that these economies are abandoning the protective shield of administrative controls by liberalising economic activity and opening themselves to trade and financial flows. In the process they become more dependent on external economic and financial conditions. They become more exposed to external competition, more sensitive to capital flows, and more dependent on economic activity in the rest of the world. While this integration should ultimately bring many benefits, these economies must travel a long road before they can reach the degree of stability and resistance to external shocks which the developed economies enjoy. This road is full of danger. (5)

There are several reasons for EMCs' financial fragility (see Eichengreen and Hausmann, 1999). Their inherent vulnerability has even given rise to an 'original sin' hypothesis. For various reasons, including their histories of price and financial instabilities and a large degree of currency and interest rate volatility, the financial markets of EMCs are insufficiently developed. They do not provide their borrowers with many opportunities to borrow long-term in the domestic currency, particularly from nonresidents. The borrowers must therefore borrow either short-term, or in foreign currency. Such a composition of borrowing makes debtors sensitive and vulnerable to external shocks, complicates the conduct of monetary policy, and makes any currency regime potentially problematic. If exchange rate fluctuates a lot in a floating regime, the domestic currency value of external debt in foreign currency will also fluctuate, threatening financial distress and even bankruptcy to borrowers with external debt. At the same time, efforts to limit exchange rate fluctuations with a more active use of interest rate policy would affect the costs of borrowing in domestic currency. It would immediately increase the borrowing costs for floating rate debt, or increase the costs of the new debt that has to replace the maturing debt. All this can cause financial problems for borrowers whose debt is denominated in domestic currency. (6)

In principle, there are two solutions to the "original sin' problem. (7) An immediate and apparently simple solution would be to get rid of the domestic currency and dollarise. As argued by Eichengreen and Hausmann (1999), this would do away with currency mismatches and reduce maturity mismatches, because of the possibility to borrow more long-term in dollars. However, as we have noted, dollarisation is not a free lunch: it can have its own problems, and may be appropriate for only a small number of countries.

A second solution would be to seek redemption from the original sin by correcting the financial market deficiencies that produced the vulnerable structure of borrowing. It is not clear how easily this can be accomplished. All depends on the cause of the original sin. Eichengreen and Hausmann argue that foreign creditors are unlikely to lend to a sovereign in its own currency, if the sovereign is capable of manipulating the exchange rate and thus the real value of its external debt. The authors speculate that until there is a strong political constituency that would provide a sufficient guarantee that the sovereign will not pursue such a strategy, external borrowing in domestic currency will not develop. Cooper (1999) provides some additional reasons for scepticism about the possibility that local markets will develop. He argues that a broad and well-functioning financial market is unlikely to develop under a floating exchange rate regime. He postulates the existence of a vicious circle, in which imperfectly developed financial markets contribute to large exchange rate fluctuations, and in their turn these large exchange rate fluctuations prevent the development of deeper and broader domestic currency markets. The reason is that large exchange rate fluctuations cause equally large fluctuations in real value of assets denominated in domestic currency, thus giving residents an incentive to invest in a more stable foreign currency. (8) In contrast, Goldstein (2002) argues that the original sin could be cured, which in turn should permit the successful operation of a currency regime called 'managed float plus.' Goldstein's 'plus' includes inflation targeting to provide a credible nominal anchor for the economy, and measures to reduce currency mismatches in the economy.

The problems that EMCs are facing in connection with external borrowing are not limited to the terms of external borrowing arising from the original sin problem. There is a related problem involving the quantity of external borrowing. External borrowing can become a problem when domestic savings are already very high, in which case it could finance excessive investment and ultimately inefficient projects. It can also be a problem when domestic savings are too low, in which case the country would become too dependent on external borrowing, and its investment activity could fluctuate excessively in response to fluctuations in the availability of external savings.

The first problem was typical for some Southeast Asian countries that had high domestic savings before the crisis, while the second problem is typical of many Latin American countries with low domestic savings.

In the EMCs of Southeast Asia, domestic savings were traditionally high, well over 30 percent of GDP, so that the inflows of foreign savings augmented the already abundant savings available to finance domestic investments. The result was very large investment-to-output ratios in these economies, sometimes approaching 40 percent. This eventually led to overinvestment, excess capacity, and even to bubbles. The subsequent correction of these excessive investments caused enormous swings in current account balances, which moved from deficits of 6-8 percent of GDP to like-sized surpluses 1 or 2 years later as these countries stopped using external savings and began to export some of their domestic savings and to accumulate foreign reserves.

For most EMCs in Latin America the situation was different. Large inflows of external savings were a welcome supplement to very low domestic savings. Here, the problem was that a relatively large share of domestic investment was dependent on the availability of external savings.9 This made domestic investment and economic activity much more sensitive to the availability of external savings. At the same time, large fluctuations in economic activity could reduce the availability of external savings, further worsening the external vulnerability resulting from low savings. Over time, heavy reliance on external savings results in high external debt and external vulnerability. As concern grows about the debtor country's ability to continue servicing its external debt, the rise in external debt will gradually begin to inhibit access to external savings.

There is good reason why some of the savings should flow from the highly developed industrial economies to the less developed EMCs. EMCs are correctly seen as a having potential for higher-return investment opportunities. The stock of capital and the capital-labour ratio are typically lower in EMCs than in developed economies, and the marginal productivity of capital is higher. (10) The differences between the capital-labour ratios and the marginal products of capital in developing countries and developed countries are so large that in a world conforming to the neoclassical growth model, we should see much higher capital flows to developing countries (see Lucas, 1990; Lipschitz et al., 2002). However the original sin-related weaknesses in the structure of EMCs' external borrowing, and the problems of excessive investment and low domestic savings, are among the market imperfections that cause the reality to differ from the predictions of the neoclassical growth model. The EMCs face a dilemma. Given their relatively low level of economic development, they would normally be recipients of foreign savings coming from the more developed countries, but the structural vulnerability (see the original sin) of EMCs implies that foreign savings often come in forms that make the borrower country more sensitive to exchange rate movements and other shocks, thus increasing their external vulnerability.

THREE CURRENCY REGIMES--THREE PROBLEMS

It should be noted that the concern about the operation of individual currency arrangements has been largely based on the examination of countries in Latin America. This reflects probably the fact that among EMCs, those located in Latin America are especially vulnerable to adverse changes in the external environment, and their vulnerability to external shocks complicates policy management under any currency regime.

We will now look at the experience of three Latin American countries with different currency regimes that illustrate this point. We start with the collapse of CBA in Argentina, which we examine in more detail.

Argentina and the CBA

Argentina has a long history of financial instability, high inflation, and defaults on its debt. After a particularly devastating episode of hyperinflation at the end of the 1980s, Argentina introduced the CBA. The stabilisation programme brought positive results. Between 1989 and 1990, the public budget deficit fell from 7.6 to 2.3 percent, and inflation fell from 1344 in 1990 to 3.9 percent in 1994. Real GDP grew rapidly and foreign capital began to flow into the country. The CBA seem to have delivered the long-awaited stability.

The CBA was seriously tested during the Mexican crisis in 1994-1995, but it survived--arguably even strengthened. When the Asian financial crisis broke out in 1997, Argentina was initially unaffected and continued to enjoy good access to international capital market. However, the Russian crisis in August 1998 and Brazil's crisis and floating of its currency in 1999 made themselves felt in Argentina as well. Argentina's need to borrow did not diminish, but the terms of borrowing began to deteriorate as investors became more risk-averse and less willing to invest in emerging markets generally. While some other countries were able to adjust to the scarcity of external borrowing by adjusting their exchange rates and current accounts, Argentina's ability to adjust was limited by the policy constraints of the CBA and the relatively closed nature of its economy. In 2001, Argentina's situation quickly deteriorated and ended in a currency, debt, and banking crisis.

The collapse of the CBA resulted from a vicious circle created by the interaction of domestic vulnerabilities and a series of adverse external shocks, combined with the economy's limited ability to respond flexibly to these shocks.

The first and, to many analysts, the most prominent cause was the role of fiscal policy (see Mussa, 2002). Maintaining fiscal discipline is an important condition for the successful operation of a CBA. The CBA-based stabilisation programme launched in March 1991 brought reductions of public spending and the public deficit. The CBA hardened the government budget constraint by ending the monetary financing of government deficits, but this tightening began to leak over time. As Argentina's macroeconomic fundamentals and credibility improved, its access to international borrowing improved as well. Two new sources of public deficit financing appeared: privatisation revenues and borrowing in the international capital market. There was no dramatic weakening of fiscal discipline during the 1990s, but budget deficits began to creep up and public debt increased (Table 1).

Fiscal discipline at the provincial level remained weak as well. An important source of provincial government revenues were transfers from the federal government, maintained at a minimum level regardless of the volume of the revenues collected by the central government. This arrangement provided insufficient incentive for fiscal discipline on the part of the provincial governments. Table 1 shows that the provincial budget balance remained in deficit throughout the 1990s.

The second factor contributing to the crisis and the collapse of the CBA was Argentina's chronic dependence on external savings. Argentina's domestic savings was relatively low compared with other countries at a similar stage of economic development, and declined continuously during the 1990s, forcing the country to rely on foreign savings (see Table 2).

Argentina's heavy dependence on external savings has had two negative consequences. First, the availability of external savings was an important determinant of both domestic investment and economic growth in Argentina. When a worsening external situation made investors less forthcoming and reduced capital inflows, Argentina's economic activity weakened rapidly. Argentina's heavy reliance on external savings imposed uncertainty, subjected economic growth to sudden large swings and increased vulnerability to shifts in investor sentiment. Second, Argentina's prolonged reliance on external savings has resulted in rapidly and persistently increasing external debt (see Table 3).

As we have noted, it is not unusual for EMCs to rely on external savings and thus run current account deficits. In addition to the problem of the quality of emerging market borrowing discussed in the previous section (ie, the maturity and currency composition), there was another factor that increased Argentina's vulnerability and contributed to the crisis. Although Argentina became very open to capital flows in the 1990s, it remained relatively closed to trade flows. (11) Argentina's exports in the 1990s were very low, not even reaching 10 percent of GDP. Argentina's long-term reliance on external financing and the rapid growth of its external debt were not accompanied by matching rapid increases in competitiveness and export growth that would permit servicing the growing external debt and thereby maintain Argentina's access to international capital markets (Table 4).

This lack of openness became a major weakness. Measured in terms of GDP, Argentina's external debt never reached excessive levels (before the currency depreciation). In 1999, it was over 50 percent of GDP, high but not excessive. However, by the end of the 1990s, Argentina's external debt reached more than 400 percent of its annual exports. Argentina's external debt continued to increase, but its ability to increase foreign currency revenues from exports of goods and services was growing more slowly.

In short, its 10-year sojourn under the CBA did not change Argentina's status very much as an emerging market country, with all the attendant weaknesses and vulnerabilities. The CBA allowed Argentina to import financial stability, and thereby enabled Argentina to access the international capital market and increase its external borrowing. Meanwhile the stock of external debt was increasing, and the economy remained vulnerable to external shocks. It remained mostly closed to foreign trade, which made it hard to offset changes in the capital account balance by adjusting the current account balance. Such adjustments were made still more difficult by the currency regime, which precluded nominal devaluation of the currency.

Brazil and the float

The recent developments in Brazil illustrate that even a successful exit from the pegged currency regime may not make the country immune to original sin-related fragilities. Brazil's exit from its exchange rate peg in early 1999 was relatively painless. In spite of replacing the peg with a currency regime described as a managed float, Brazil exhibits the standard 'fear of float' syndrome. It tries to prevent large exchange rate fluctuations, seeks to limit depreciation when the currency is under pressure, and is unable to pursue an independent monetary policy. Brazil's monetary policy is strongly procyclical. Why cannot a country with a floating exchange rate benefit from the pursuit of an independent monetary policy?

The main reason is Brazil's exposure to external vulnerabilities similar to Argentina's. Brazil's external debt is relatively large, and its economy is relatively closed. In 2001, Brazilian exports of goods and services were only about 13 percent of GDP, and although its external debt was about 42 percent of GDP, it was over 300 percent of its exports. Brazil's public sector is also financially vulnerable. A large part of Brazil's public debt (about 80 percent by the end of 2002) is indexed either to SELIC interest rate (central bank's policy rate) and thus sensitive to central bank's interest rate changes, or to the currency's exchange rate, and thus sensitive to exchange rate fluctuations. Currency depreciation and high interest rates explain why the net public debt-to-GDP ratio has risen from 35 percent in 1997 to about 55 percent in 2002. (12)

The political uncertainty that preceded the presidential elections in October 2002 illustrates the policy dilemmas created by this vulnerable structure of public debt. The polls that suggested that the presidential elections might be won by a candidate who is less favoured by the markets have administered a significant shock to a market confidence, which has contributed both to a weakening of the Brazilian currency and to a significant widening of the spread on Brazilian debt. (13) As a result, the monetary authorities had to choose between two unpalatable alternatives. The first alternative was to respond to the increased market uncertainty by raising interest rates to prevent a sharp currency depreciation but increasing the costs of servicing that part of the public debt that is indexed to the SELIC interest rate. The second alternative was to keep policy interest rate low and let the currency to depreciate, which would result in an increase in the cost of servicing that part of the external debt that is linked to the exchange rate. Letting currency depreciate would also carry the risks of increasing inflation (and inflationary expectations), which would cause higher market interest rates to reflect the higher inflation premium. Moreover, higher inflation would require higher policy interest rates in the future to bring inflation back down.

Despite weakening economic growth, the central bank has decided to tighten monetary policy and raised interest rates sharply, from 17 percent in October 2002 to 26.5 percent in February 2003, but this did not prevent a currency from weakening. Combination of higher SELIC rate and a more depreciated currency contributed to an increase in net public debt to GDP ratio by about 5 percentage points in 2002. In turn, continued rise in the debt burden weighed further on market confidence. Only the clear statement and policy steps of the new administration indicating the continuity of stability-oriented policy and the determination to continue servicing Brazil's external debt have alleviated somewhat the dilemma as they helped to improve market confidence in early 2003. However, even though Brazil's spreads fell to below 800 points in May, Brazil still remains vulnerable to any future adverse shocks and the floating currency regime will not insulate it from these.

Panama and dollarisation

What about the opposite end of the spectrum of currency regimes? Could dollarisation improve economic performance and bring the desired stability and policy credibility? So far, there is little long-term empirical evidence about the operation of dollarised economies. Panama is an exception and has therefore been thoroughly analysed. The supporters of dollarisation point to Panama's history of low inflation and relatively strong growth. Furthermore, Panama is the only country in Latin America that has been able to issue 30-year mortgages.

Not everything is perfect in Panama, however. A more sceptical analysis of Panama's performance is worth quoting: 'In spite of not having a central bank, or a currency of its own, for years Panama failed to maintain fiscal discipline. Initially, these large deficits were financed through borrowing from abroad. And when foreign debt became too high, the IMF stepped in with fresh resources. And when it was not enough, Panama restructured its debt.' (Edwards, 2001). If this experience looks familiar, it is not surprising. It could as well be applied to Argentina's experience under the CBA in the 1990s. The similarities underscore the point that regardless of the currency regime, a government that does not have the political courage or will to adjust public spending to available resources will always find some ways to finance spending in excess of its resources, and in the process will make its economy vulnerable to adverse shocks and undermine its long-term growth performance.

Another country in the region that has introduced dollarisation, Ecuador, has not yet accumulated enough experience with this currency regime to permit drawing definitive conclusions. However, Ecuador's problems in the first years of dollarisation only confirmed that a change in the currency regime per se, even introduction of dollarisation, does not guarantee a pursuit of disciplined policies.

These observations may leave one with a rather pessimistic conclusions about the prospects of EMCs. But even though original sin can have long-lasting and serious consequences for EMCs, it is not something they are doomed to live with forever. To obtain a better understanding of what policies can help them achieve a rapid redemption from original sin, it may be useful to look at a different group of EMCs.

ACCESSION COUNTRIES IN EUROPE: A DIFFERENT STORY?

The problems of Latin American countries with various exchange rate arrangements, discussed in the previous part, are not typical of all EMCs. The accession countries of Central and Eastern Europe seem much less vulnerable to adverse shocks or changes in financial markets' sentiment--regardless of their currency arrangements.

Table 5 shows that in terms of GDP, the level of the external debts and current account deficits of the accession countries is relatively high, generally similar to what we see in the countries of Latin America.

Accordingly, one might assume that these countries suffer from problems similar to those of Argentina, Brazil, and other countries in Latin America, that is, that they would be vulnerable to an adverse external environment, that the price of their external debt would increase with investors' increasing aversion to risk, that they would find international borrowing more difficult to obtain during times of adverse external shocks, that their currencies would be weakening, and that they would be unable to pursue independent monetary policies and countercyclical fiscal policies.

However, the reality is somewhat different. Even though these economies are very open to capital flows, there is little evidence of external vulnerability we saw in some Latin American countries. Although they have suffered temporary contagion from financial crises elsewhere, particularly the Russian crisis of 1998, this contagion was limited in scope and duration. Their asset prices suffered only a relatively small and temporary setback (International Monetary Fund, 1999, Chapter III). The relative absence among these countries of vulnerability to turbulences in the world economy and in other EMCs is illustrated by the behaviour of their external borrowing spreads, shown in Table 6.

The relative resilience to adverse external conditions of the financial indicators of the countries of Central and Eastern Europe means that when the accession countries need to borrow abroad at all, they can do so on much better terms than countries in Latin America. Table 7 shows the coupon rates of latest bond issues in domestic currency and in the euro for selected countries in the region.

First, the countries that issued euro-denominated public debt were able to do so at relatively long maturities--mostly 10 years--and at reasonably low yields. Second, for countries with a CBA--Bulgaria and Lithuania--there is not much difference in yields between domestic and euro-denominated debt, which reflects low currency risk. For countries with floating currencies--the Slovak Republic and Poland--their relatively higher yields on domestic currency debt compared to the yields on euro-denominated debt reflect a higher currency risk. (Note that the euro-denominated Slovak bond shown in Table 8 was issued in 2000. If it were issued in 2002, market analysts estimate that the yield would be in the range 5.5-6.0 percent.)

Third, governments of these countries have largely been able to finance their borrowing needs from local markets, by issuing debt denominated in domestic currency. This has an important advantage: financing these governments' borrowing needs is less dependent on access to the international capital markets. (14) Market capitalisation of the euro-denominated and dollar-denominated bonds is relatively small (see Table 8). Market capitalisation of domestic currency debt is relatively high in Poland and Hungary (around 30 percent of GDP) and somewhat less in the Czech Republic (about 15 percent of GDP). However, these countries have basically no restrictions on nonresident purchases of domestic-currency government debt, and it is estimated that foreign investors hold 30-35 percent of Hungary's outstanding local currency bonds, 15-20 percent of Poland's, and 10-15 percent of the Czech Republic's.

The relatively small external vulnerability of the accession countries is also evidenced by their conduct of monetary policy. Accession countries with flexible currency regimes are able to pursue rather independent monetary policies. (15) To the extent that the fear of floating is present, it is a different 'fear,' namely fear of too strong a currency that would damage competitiveness and economic activity. The relative independence of monetary policy in the accession countries can be seen by comparing changes in policy interest rates in these countries with those in the euro area. Figure 1 shows that during a period of stable or even increasing ECB interest rates in 1999 and 2000, the central banks of the Czech Republic and Hungary were able to reduce interest rates significantly. In fact, in July 2002, Czech interest rates fell below euro area interest rates. Higher interest rates in Hungary and especially in Poland were partly explained by efforts to bring down higher inflation, but even these countries have been able to cut domestic interest rates regardless of the ECB's interest rate policy. (16)

[FIGURE 1 OMITTED]

For comparison, Figure 2 shows policy rates in the United States (the targeted federal funds rate) and Brazil (SELIC rate) during the period of Brazil's currency float. Two things stand out. First, adjusted for the inflation differential, the difference between the two rates is significantly greater than in the accession countries (perhaps with some exception made for Poland). Second, when the Fed began to cut interest rates aggressively in January 2001, the Banco Central do Brazil not only did not follow, but instead had to increase its interest rate significantly, so that the difference between the US and Brazilian policy rates widened from about 10 to 25 points.

[FIGURE 2 OMITTED]

Why are the accession countries less sensitive to external developments and to financial contagion? Why are they able to borrow at much better terms? Why are they able to pursue relatively independent policies (in countries with a floating currency) or why do their CBAs seem to enjoy relatively high credibility? There are several reasons for this relative resilience to contagion.

First, while the size of the accession countries' external debt is not significantly different from that of countries in Latin America when measured in terms of GDP, it is much smaller when measured in terms of exports (see Table 9). The accession countries are much more open, and thus their debt-to-exports ratios are generally much lower than Latin America's (except for Poland, exports usually represent some 50 percent or more of GDP). Therefore, for a given level of the external debt to GDP ratio, there is less concern about the availability of foreign currency earnings to service it. A higher degree of trade openness also makes changes in the real exchange rate more effective as an instrument for adjusting the current account balance. Moreover, only a small fraction of their external debt is in the form of market borrowing and liable to be adversely affected by changing sentiment in the international capital market. This may reflect the fact that the accession countries do not have long histories of involvement with the international capital markets, and so have simply not had enough time to build up high levels of marketable external debt. They have also implemented the extensive privatisation programmes that have been a welcome source of financing.

Second, while the accession countries are generally no less dependent on external savings than the countries of Latin America, as shown by their relatively large current account deficits, they are able to cover a large part of these deficits with non-debt-creating foreign direct investment. Reliance on foreign direct investment is not without its risks and problems, but it is clearly a less risky form of capital inflows than debt-creating instruments. (17) Table 10 compares the extent to which net inflow of foreign direct investment covers the current account deficits of Argentina, Brazil and accession countries. While Argentina and Brazil were also able to finance their current account deficits from net inflow of foreign direct investment, they still had to devote a significant part of their export revenues to servicing previously accumulated external debt and had to resort to more external borrowing to refinance maturing debt.

Third, and closely related, domestic savings in accession countries are generally much higher than in Argentina and Brazil (and other Latin American countries as well). To the extent that higher growth requires higher investment, low domestic savings force countries to rely more on external savings as a source of financing investment spending. This has two consequences: first, when for some reason foreign savings become less available, investment has to be cut unless domestic savings increase. This could result in increased output volatility. Second, prolonged extensive reliance on foreign savings in the form of debt-creating capital inflows can result in increased debt burden and external vulnerability. This is the case particularly when economies remain relatively closed to trade flows and do not generate enough foreign currency earnings to service, without problems, external debt. It is exactly this interaction of low domestic savings and reliance on external savings (that is, opening to capital flows) and low share of exports in national output that makes any currency regime vulnerable, as in case of Argentina and Brazil. This contrasts with the combination of relatively high domestic savings and large openness to trade flows that makes most accession countries less vulnerable to adverse market sentiment.

Fourth, and perhaps most important, the accession countries benefit from the confidence-building effects of their future EU/EMU accession. The EU/ EMU membership requirements ensure that they will pursue macroeconomic policies and structural reforms that will produce financial stability and solid economic performance. The prospect of EU/EMU membership serves as an important policy anchor and disciplining influence, and thus helps increase the credibility of policy in the accession countries. This holds regardless of which exchange rate regime is in place. For countries with CBAs (Bulgaria, Estonia, Lithuania), EU/EMU membership provides guarantees that their exit from the CBA will be from a position of strength (since they must meet the strict qualification criteria for EMU membership), rather than from a position of weakness and crisis. For countries with flexible currency regimes, it helps anchor inflation expectations, and thus limits exchange rate fluctuations.

In short, the accession countries have entered a virtuous circle: future EU/EMU membership provides an anchor for economic policies, improves policy credibility, and contributes to positive expectations about future economic performance. This promotes the development of local financial markets and reduces dependence on external financing. Less need of external financing, together with large inflows of FDI, is rewarded with better access to the international capital markets. The favourable terms and diminished need for external financing make it possible to avoid the build-up of external weaknesses that increase economies' vulnerability to adverse external shocks.

This is not to deny that the prospect of EMU membership does not bring temporary policy problems of its own, such as the unstable capital flows and exchange rate fluctuations that precede EMU entry and enroisation, but as long as the underlying policies are sound, these problems should be manageable.

CONCLUSION

After the collapse during the past decade of pegged currency regimes in many EMCs, a consensus has emerged that the so-called soft peg regimes are not suitable for EMCs that have and become integrated into the international capital market. There seems to be a further consensus that these unsuitable regimes should be replaced by one or another of the so-called 'corner solutions'--a floating currency, or the hard peg in the form of a CBA, or dollarisation, or membership in a monetary union--as the appropriate regime for an EMC.

The collapse of the CBA in Argentina has raised new questions about the optimal currency arrangements for EMCs. Argentina's experience has shown that even a CBA may not deliver the degree of policy discipline and stability that its proponents had hoped for. We have also briefly discussed the experience of Brazil with a managed float and Panama with dollarisation--to emphasise the point that a floating currency does not necessarily provide room for an independent monetary policy, and that using someone else's currency does not establish necessary policy discipline or prevent a build-up of financial vulnerability.

The conclusion is that no particular currency arrangement can per se explain why a country finds itself vulnerable to shifts in market sentiment that cause large fluctuations in economic activity, and why it is unable to pursue independent policies to smooth out these fluctuations. These vulnerabilities result not from a particular currency arrangement, but rather from low domestic savings, heavy reliance on external savings, and borrowing needs that often cause a build-up of external debt. Economies that are not sufficiently open to trade and rely on external borrowing will attain unsustainability sooner rather than later.

We have discussed why EMCs suffer from vulnerabilities that make the operation of any exchange rate regime potentially difficult, but different EMCs suffer from these vulnerabilities in different degree. We have compared the rather negative experience of the countries in Latin America with the experience of accession countries in Central and Eastern Europe. The accession countries were relatively untouched by recent financial crises and turbulence in the international capital markets. They continued to enjoy good access to external borrowing, and the prices of their financial assets, including the exchange rate, suggests no loss of credibility.

The relatively strong credibility and resilience of the accession countries to external shocks does not depend on their choice of a currency regime. The currency regimes of the accession countries vary widely. For countries with CBAs or currency band, it is important to have a clearly defined exit strategy. For the accession countries this means their entry into the EMU. The EMU entry provides an important signal of the commitment to pursue disciplined policies. To countries with a flexible currency regime, this commitment serves as a potent nominal anchor that helps them to achieve and maintain price stability, and promotes the development of local financial markets that make them less dependent on external financing, and thus less vulnerable to exchange rate movements. This allows them to pursue relatively independent monetary policy tailored to the needs of their domestic economies.

The success of the process of EU/EMU accession for the candidate countries could provide an important lesson for the EMCs that have trouble establishing the credibility of their policies and therefore remain unduly sensitive to conditions in the international capital markets. Replacing the domestic currencies with the dollar or some other currency is not the same as joining EMU. The advantage of EMU membership is not just that it brings to its new members a stable currency enjoying high credibility. It also means membership in a club with strict rules for maintaining stability-oriented policies, which is something that dollarisation alone cannot accomplish.

Table 1: Argentina: public sector balance
and public debt (in percent of GDP) (a)

 1991 1992 1993 1994 1995

Consolidated -3.2 -0.5 -0.8 -2.3 -2.3
Federal govt -2.5 -0.2 0.9 -0.5 -0.9
Provincial govt -0.7 -0.2 -0.8 -0.9 -1.4
Public debt $bn 64.7 68.8 77.6 90.3 101.5

In percent of GDP 35.8 30.3 30.1 32.1 36.3

 1996 1997 1998 1999 2000

Consolidated -3.2 -2.1 -2.1 -4.2 -3.6
Federal govt -2.5 -1.6 -1.3 -2.5 -2.4
Provincial govt -0.7 -0.5 -0.8 -1.6 -1.1
Public debt $bn 114.4 111.6 123.5 133.9 144.8

In percent of GDP 38.5 38.1 41.3 47.3 50.8

Source: IMF

(a) Balance of federal and provincial governments does not always
add up to the consolidated balance due to the operation of trust
funds and some other items that affect consolidated balance.

Table 2: Argentina: savings, investment
(in percent of GDP) and GDP growth rate

 1993 1994 1995 1996 1997 1998 1999 2000

Domestic savings 17.1 18.6 19.7 16.2 15.9 15.5 13.9 13.5
External savings 4.1 4.8 0.9 2.7 4.7 5.6 5.1 4.1
Domestic investment 21.2 23.4 20.6 18.9 20.6 21.1 19.0 17.6
Real GDP 6.0 7.1 -4.4 5.5 8.1 3.9 3.4 -0.5

Source: The Economist Intelligence Unit Country Report

Table 3: Argentina: external debt and debt service (a)

 1993 1994 1995 1996 1997 1998 1999 2000

Debt/GDP 27.7 27.8 39.2 42.0 45.7 48.6 53.7 52.7
Debt/exports 395.4 368 336.2 338.6 358.7 379.0 435.7 487.9
Debt service/
 exports 36.8 31.8 30.4 39.5 50.2 57.5 75.8 93.5

Source: The Economist Intelligence Unit Country Report

(a) Data here for Argentina are not directly comparable with those
in Table 4, because Table 4 uses a present value of external debt,
while this table uses a nominal value.

Table 4: External debt and debt service, 1998 (a)

 Debt service/
 Debt/GDP Debt/exports (b) exports (b)

Argentina 52 406 58.2
Brazil 29 340 74.1
Mexico 41 111 20.8
Indonesia 169 252 33.0
Thailand 76 123 19.2
Korea 43 84 12.9
Czech Republic 45 71 15.2
Hungary 63 107 27.3
Russia 62 186 12.1

Source: The World Bank (2000), World Development Indicators

(a) Present value of debt.

(b) Exports of goods and services.

Table 5: Accession countries: indicators of external vulnerability
(in percent of GDP)

 Total external debt Current account balance

 1998 1999 2000 2001 1998 1999 2000 2001

Czech Republic 40.4 44.8 42.6 36.6 -2.4 -3.0 -4.5 -4.8
Estonia 53.3 58.7 59.4 61.0 -9.2 -5.8 -6.8 -6.8
Hungary 55.8 64.5 66.8 65.1 -4.8 -4.4 -2.9 -2.1
Latvia 50.9 57.5 65.7 73.5 -10.7 -9.8 -6.9 -9.7
Lithuania 34.8 42.4 42.9 44.8 -12.1 -11.2 -6.0 -6.7
Poland 37.3 42.2 44.0 39.5 -4.3 -7.5 -6.3 -4.0
Slovak Republic 55.9 53.4 56.7 53.9 -10.0 -5.0 -3.6 -5.4
Slovenia 24.1 26.9 34.3 36.0 -0.8 -3.9 -3.3 -0.8

Source: IMF

Table 6: Selected countries: sovereign debt spreads

 1998 1999 2000 2001 Oct 2002 Dec 2002

Poland 198 199 197 185 250 210
Hungary 155 107 116 78 x x
Argentina 705 548 773 1566 7000 6150
Brazil 825 920 749 863 2200 1460

Source: Bloomberg

Table 7: Selected transition economies: yield on government debt
(in percent)

 Local currency Euro

Bulgaria 10Y BGL 7.75% (April 02) 10Y [euro] 7.5% (March 02)
Croatia 7Y HRK 6.25% (Jan 02) 7Y [euro] 6.25% (Feb 02)
Hungary 10Y HUF 7.1% (May 02) 10Y [euro] 5.625% (June 01)
Lithuania 10Y LIT 6.15% (March 02) 10Y [euro] 5.875% (May 02)
Poland l0Y PLZ 8.1% (May 02) 10Y [euro] 5.5% (March 02)
Slovak Republic 10Y SK 7.3% (May 02) 10Y [euro] 7.375% (2000)

Source: Bloomberg

Table 8: Selected transition economies: market capitalisation in
billions (May 2002)

 In local currency Of euro bonds Of dollar bonds
 ($ equivalent) (in [euro]) (in $)

Bulgaria 0.8 1.1 3.8 (a)
Czech Republic 8.0 0.0 0.0
Croatia 2.0 2.5 1.1
Hungary 17.0 3.0 1.5
Poland 40.0 2.1 3.7 (a)
Slovak Republic 6.0 1.5 0.5

Source: Bloomberg

(a) Old Brady bonds.

Table 9: Accession countries: external debt in percent of exports

 1996 1997 1998 1999 2000 2001

Czech Republic 66.6 69 68.3 64.3 56.4 50.9
Estonia 52.5 70.9 70.1 71.9 62.6 65.5
Hungary 138.6 96.2 102.3 106.7 94.9 94.6
Latvia 80 96 99.3 131.1 144.1 161.6
Lithuania 57 63.2 74.8 107.1 95.6 87
Poland 170.8 160.4 175 218.8 214.6 104.2
Slovakia 70.4 83.8 81.8 86.3 76.5 74.5
Slovenia 38 39.4 44.2 51.3 58.1 59.5

Source: European Bank for Reconstruction and Developments (2002)

Table 10: Net FDI inflow in percent of current account deficit

 1998 1999 2000 2001

Argentina 32.1 65 121.3 70.2
Brazil 86.5 112.5 133.3 97
Czech Republic 276.9 442.8 185.2 185
Estonia 120 89.8 110.2 101
Hungary 65.2 81 84.6 191
Latvia 45 48.5 81.1 23.1
Lithuania 71 40 54.4 78.6
Poland 73.5 54.3 81 93
Slovakia 17.6 64 300 83.3
Slovenia 166.7 17.5 18.3 340

Source: European Bank for Reconstruction and Developments (2002),
International Monetary Fund (2002)

Table 11: Gross domestic savings in percent of GDP

 1998 1999 2000 2001 2002

Argentina 15.2 13.8 13.1 12.5 18.8
Brazil 15.3 15.4 17.6 16.4 16.0
Czech Republic 27.3 15.4 24.4 25.3 25.4
Estonia 20.1 19.8 18.3 20.2 20.7
Hungary 24.8 14.4 27.2 24.5 24.0
Latvia 18 17.1 19.8 18.9 19.3
Lithuania 12.3 11.5 14.6 17 16.6
Poland 21.9 18.9 19.9 18.5 18.9
Slovakia 25 23.4 22.8 22.9 22.5
Slovenia 25.4 26.9 24.2 24.7 24.7

Source: IMF

(1) The views expressed here are author's own and do not reflect the official position of the IMF.

(2) As Larrain and Velasco (2001) observe, 'a good deal of enthusiasm over currency boards owes to the experience of one country, Argentina, over a fairly brief period of time. All other experiences, except for Hong Kong's, have been too short-lived to be informative.'

(3) For examples of the agnostic approach to currency regimes, see Cooper (1999) and Frankel (1999). Steve Hanke is one of the strongest proponents of CBA or dollarisation, but many economists seem to prefer the float. For arguments against CBAs, see Roubini (2001). For a comprehensive discussion of various views on optimal exchange rate arrangements, see Goldstein (2002).

(4) For example, Masson (2001) argues that there were frequent crossings between the different exchange rate regimes in the 1990s, rather than a shift from middle to corner regimes. See also Calvo and Reinhart (2000).

(5) Being an emerging market is like being a teenager. The innocence of childhood is gradually being lost, but the mature judgement of adulthood is not yet well enough established to guide the teenager through this sensitive period when he/she becomes exposed to the temptations and pitfalls of life.

(6) Moreover, faced with an adverse external shock, EMCs with a large share of debt denominated in foreign currency and/or of short-term debt will not be able to pursue the same shock-mitigating policies used by advanced economies with well-established policy credibility and better developed financial markets. Often, EMCs must pursue procyclical policies to avoid a loss of confidence and financial instability, in spite of the cost to output and employment. See Mishkin (1996).

(7) Eichengreen et al. (2002) recently proposed another solution to the original sin problem: the World Bank and other developments banks would sponsor a mechanism that would allow the EMCs to issue more debt in their own currencies. However, the main problem with this mechanism is that it does not provide strong incentives to address the fundamental weaknesses and problems of EMCs leading to the original sin problem.

(8) This conclusion holds for very open economies, where changes in the exchange rate have a strong effect on domestic prices. The obvious question is whether a fixed (and therefore stable) exchange rate is more helpful in overcoming the original sin problem. Again the prospects are not promising, as shown by the experience of Argentina. The CBA helped speed the establishment of price stability and growth, and allowed Argentina to borrow abroad, but even a decade of exchange rate stability could not improve Argentina's credibility enough to redeem it from its original sin. The former central bank governor Pedro Pou (1999) complained in 1999, before the crisis escalated, that despite having played by the currency board's rules for 9 years, Argentina could not borrow in domestic currency from international investors except at short-term maturities.

(9) To illustrate, if domestic savings are being used to finance investments equal to 30 percent of GDP, and the country is also using external savings (running a current account deficit) equal to 5 percent of GDP, then external savings are financing one-seventh of total domestic investments (which equal 35 percent of GDP). If domestic savings can finance investments worth only 15 percent of GDP, then the same amount of external savings would be financing one-fourth of total domestic investment.

(10) However, Baer (2001) correctly observes that it is not only the return, but also the variance (risk) that determines investors' willingness to invest in EMCs. High risk can partly or fully cancel the attractiveness of a high return.

(11) It can be assumed that the low level of domestic savings was an important factor pushing Argentina towards the liberalisation of capital flows.

(12) Moreover, sonic authors argue that net debt may underestimate the extent of vulnerability of Brazil's public sector because its assets may not be as liquid as its liabilities. A more realistic appraisal of liquidity of government assets could imply a higher level of net debt. For a discussion of this issue, see Goldstein (2003).

(13) The Brazilian currency, the real, has lost 35 percent of its value in 2002, while the benchmark C-bond fell from about $83 in early 2002 to the low of $49 in mid-October 2002, with spread on Brazilian debt rising to about 2500 points.

(14) At present, the Czech Republic has no outstanding foreign currency-denominated sovereign bond, even though there are indications that demand for such an instrument would be high. In late 2001, Hungary changed its borrowing strategy from borrowing in foreign currency to borrowing in local currency, and in November 2001, it issued a heavily oversubscribed 15-year Hungarian forint bond.

(15) Of course, it is another question whether the independent conduct of monetary policy brings the desired results. Butter and Grafe (2002) warn that in most accession countries, underdeveloped debt markets will prevent monetary policy from being very effective for stabilising output. However, local debt markets are developing rapidly, at least in the more advanced transition countries.

(16) However, at the end of 2002 and in early 2003, Hungary faced a different problem, increased tension between its commitment to keep exchange rate within the [+ or -] 15 percent fluctuation band and meeting its inflation target. For a more detailed discussion, see Jonas and Mishkin (2003).

(17) Baer (2001) discusses some of the potential problems connected with FDI inflows.

REFERENCES

Baer, GD. 2001: Risk and capital flows to the emerging markets. Paper presented at the Conference organized by the Croatian National Bank, Dubrovnik, 28-30 June 2001.

Buiter, WH, and Grafe, C. 2002: Anchor, float or abandon ship: Exchange rate regimes for the accession economies. Paper prepared for the 10th Anniversary Conference of the European Bank for Reconstruction and Development, London, 13-14 December 2001.

Calvo, G and Reinhart, C. 2000: Fear of floating. NBER Working paper no. 7993, Cambridge, Massachusetts.

Catao, L and Terroned, M. 2000: Determinants of dollarization: The banking side. IMF Working Paper no. 146, IMF: Washington, DC.

Cooper, R. 1999: Exchange rate choices. In: Little, JS and Olivei, GP (eds). Rethinking the International Monetary System. Federal Reserve Bank of Boston Conference Series No. 43, Boston, Massachusetts.

Edwards, S. 2001: Dollarization and economic performance: An empirical investigation. NBER Working paper no. 8274, Cambridge, Massachusetts.

Eichengreen, B. 1998: Toward a new international financial architecture. Institute of international Economics: Washington, DC.

Eichengreen, B and Hausmann, R. 1999: Exchange rates and financial fragility. NBER Working paper no. 7418, Cambridge, Massachusetts.

Eichengreen, B, Hausmann, R and Panizza, U. 2002: Original Sin: The Pain, the Mystery and the Road to Redemption, Mimeo University of Berkeley, California.

European Bank for Reconstruction and Developments. 2002: Transition Report 2002.

Fischer S. 2001: Exchange rate regimes: Is the bipolar view correct? Journal of Economic Perspectives 15(2): 3-24.

Frankel, JA. 1999: No single currency regime is right for all countries or at all times. NBER Working paper no. 7338, Cambridge, Massachusetts.

Goldstein, M. 2002: Managed floating plus and the great currency regime debate. Paper presented at the IMF Institute Economics Training Seminar, January 2002, Washington, DC.

Goldstein, M. 2005: Debt sustainability, Brazil and the IMF. Institute of International Economics, Working Paper 03-1, Washington, DC.

International Monetary Fund, 1999: International capital markets: Developments, prospects and policy issues. IMF: Washington, DC, Chapter III.

International Monetary Fund. 2002: World Economic Outlook, April Chapter 2.

Jonas, J and Mishkin, FS. 2003: Inflation targeting in transition economies: Experience and prospects. Paper presented at the NBER Inflation Targeting Conference, Bal Harbor, FL, 21-23 January 2003.

Larrain, F and Velasco, A. 2001: Exchange rate policy in emerging markets: The case for floating. Paper presented at the Conference "When is a National Currency a Luxury?" London Business School, 16-17 March 2001, London.

Lipschitz, L, Lane, T and Mourmouras, A. 2002: Capital flows to transition economies: Master or servant? IMF Working paper no. 11, IMF: Washington, DC.

Lucas, R. 1990: Why doesn't capital flow from rich to poor countries? American Economic Review 80(2): 92-96.

Masson, P. 2001: Exchange rate regime transitions. Journal of Development Economics 64:571-86

Mishkin, F. 1996: Understanding financial crises: A developing country perspective. NBER Working paper no. 5600, Cambridge, Massachusetts.

Mussa, M. 2002: Argentina and the Fund: From triumph to tragedy. Institute for International Economics: Washington, DC.

Pou, P. 1999: Is globalization really to blame? In: Little, JS and Olivei, GP (eds). Rethinking the International Monetary System. Federal Reserve Bank of Boston, Conference Series No. 43, Boston, Massachusetts.

Roubini, N. 2001: The case against currency boards: Debunking 10 myths about the benefits of currency boards. See http://www.stern.nyu.edu/globalmacro/CurrencyBoardsRoubini.html.

JIRI JONAS (1)

International Monetary Fund, Office 11-302, 700 19th Street NW, Washington, DC 20431, USA. E-mail: [email protected]