Working Paper. Sovereign Bond Defaults and Restructurings in Emerging Markets: A Preliminary Review. Luisa Palacios, Ph.D. NO.6.

Transcription

1 JBICI Working Paper Sovereign Bond Defaults and Restructurings in Emerging Markets: A Preliminary Review Luisa Palacios, Ph.D. NO.6 July 2002 JBIC Institute JAPAN BANK FOR INTERNATIONAL COOPERATION

2 Sovereign Bond Defaults and Restructurings in Emerging Markets: A Preliminary Review Luisa Palacios, Ph.D. 1 The history of sovereign lending is the history of sovereign default Vincent Truglia, Moody s, October 23, 2000 SUMMARY: 1. Sovereign bond defaults are likely to be the trend in future debt crises in emerging markets given the important share of bonds in emerging markets external debt: the IMF estimates that bonds represented 60% of public external debt in 2000 vs. 13% in 1980 (Krueger, 2002a). This reality could alter the way sovereign debt restructurings have been carried out so far in the Paris and London Clubs. 2. The important share of bonds in emerging markets external debt also poses important challenges for the financial world because bond restructurings are believed to be more problematic than official or commercial restructurings due to: a) collective action problems (the potential huge number of creditors and the rapidity in which the debt changes hands) that pose organizational issues for debt restructurings; and, b) the intrinsic rules of bond contracts which do not permit changes to the payment structure or amortization schedule which pose potential legal risks for the sovereign that restructures. 3. Bondholders have started to organize themselves. After Ecuador s unilateral debt exchange, an attempt has been made by bondholders to organize into what might be a more stable committee: the emerging markets creditors association (EMCA). The rationale behind the creation of a bondholders committee is the refusal of take-or-leave-it propositions that have been the characteristic of bond restructurings until now. So far, this bondholders association has not formally engaged in debt restructuring negotiations with a sovereign: Argentina could be the first case. 4. An Argentine Bondholders Committee (ABC) linked informally to EMCA has been formed (November 2001) and seeks to become the negotiating body representing Argentina s external bondholders in an eventual debt restructuring. ABC claims to currently represent approximately 60% of the outstanding foreign-held bonds. 5. Bondholders are pressuring for comparability of treatment not only with respect to bilateral creditors but also in relation to multilaterals: bondholders are questioning 1 Luisa Palacios is the Senior Economist of JBIC s Representative Office in New York. The author would like to acknowledge the valuable comments from Fumihiko Wada, Keiji Fukumoto, Moriyuki Aida, and Ayumu Yoshie. i

3 the seniority status of multilateral debt and the absence of an IMF debt write down if bondholders are obliged to do so. 6. The belief that bond restructurings are problematic has led to propositions on how to bring more certainty to this process: IMF s Anne Krueger has proposed a bankruptcy court type of mechanism that has received strong opposition from the private financial sector. Other less radical proposals consider the inclusion of collective action clauses into bond contracts (embraced by the US Undersecretary of Treasury John Taylor) or the use of exit amendments.* 7. Yet, despite these challenges, four countries have been able to restructure sovereign bonds or notes since 1998: Russia, Ukraine, Pakistan and Ecuador (see Table 5, page 22-23), with every sovereign restructuring bringing new issues into the debate. The use of collective action clauses in the case of Ukraine and the use of exit amendments in the case of Ecuador have set important precedents for ways in which countries can deal with debt restructurings even in the absence of formal mechanism for bond debt workouts. 8. Bonds restructurings have changed the dynamics of inter-creditor relations: With Russia the almost sacrosanct position of Eurobonds was upheld but later put in question with Pakistan s debt restructuring. In the case of Pakistan, the comparability of treatment principle was imposed by the Paris Club, forcing the country to restructure its Eurobond debt. Ecuador decision to default on its Brady bonds but not on its Eurobonds backfired and the country was legally obliged by its creditors to default on its Eurobonds as well. *Collective action clauses allow for a qualified majority of bondholders to alter the payment terms of bonds making the changes binding to all bondholders. Exit amendment or exit consents are provisions that although do not allow for changes in the payment terms of bond contracts can nonetheless allow a majority of bondholders participating in a debt exchange to alter non-payment terms of bonds as they are exiting them. This will render the old bonds less attractive. For example, the bonds could be delisted, the legal jurisdiction could be change, acceleration or default clause could be eliminated. ii

4 Contents I. Introduction... 1 II. A look at emerging markets bond debt... 1 III. New and old sources of credits and challenges of sovereign debt workouts... 7 VI. Why are bond restructurings an issue?...11 V. A proposal for an international proxy for a domestic bankruptcy court: Does Anne Krueger have a point? VI. Understanding recent bond restructurings: What have we learned? How have collective action clauses and exit amendments performed so far? VII. Argentina s gradual approach to default...25 VIII. Concluding Remarks: Some preliminary thoughts about the experience with sovereign bond defaults and restructurings References I. Introduction Since 1995, the world has seen an array of emerging market crises (Mexico, Asia, Russia, Brazil, Turkey and Argentina) that have led to a growing discussion about the need to reform the international financial architecture. Propositions for reform date back to the mid-1990s, but on the wake of the Argentina debt crisis a concrete and ambitious proposal has been put forward by Anne Krueger, first deputy managing director of the International Monetary Fund (Krueger, 2001a). Specifically, Ms. Krueger s proposition centers on the creation of a proxy to an international bankruptcy system that could help countries deal with bond debt restructurings in an orderly fashion. This proposition has encountered great criticism from the private sector but has given a new momentum to the debate on sovereign restructurings. This paper will try to venture in the complicated world of bond debt restructurings, through providing a first description of the experiences with bond restructurings at the sovereign level and the response of bondholders to what seems to be a potential trend in emerging markets crisis. It will also provide a glance into the recurrent debates and issues surrounding bond restructurings and the propositions to address the alleged complications surrounding bond workouts. II. A look at emerging markets bond debt The renewed discussion about the need to reform the international financial architecture is the result of the recent experience of international financial markets 1

5 with the new wave of sovereign debt restructurings. This new wave of debt crises that started with Russia in 1998 has recently shown its most critical and dangerous face so far with Argentina s sovereign default in December The most important characteristic of this new wave of debt restructurings is the inclusion of bond debt and the particular challenges it poses for orderly restructurings. The recent experience with bond debt crises is not accidental. It is the by-product of the successful entrance of developing countries into the international capital markets, which coincided with the wave of liberalization and privatization experienced in the early 1990 s. Emerging market countries have been able to issue approximately US$73 billion annually since Bond debt flows reached their peak in 1997: a record US$120 billions were issued representing almost a 200% increase from 1995 levels. However, the disappointing experience with Asia and Russia in and later Brazil in 1999, led to a fall in debt flows after 1997 (see Figure 1). Yet, despite the scaling back of the market, bond flows showed an important resilience in the period, with bonds issued averaging US$70 billion annually. Figure 1. Total emerging market bond issues (annually) (In million of US$) In millions of US$ Source: IIF It must be said that, despite the opening of the international bond market as a financing source for emerging countries, market access has not been universal nor has it been equally distributed. In regional terms, Latin America has had the most important share of bond issuance during the period, accounting for half of all bonds issued during this time (see Figure 2). This has also proven to be a highly concentrated market in a few individual countries as shown by the share of the 5 most important issuers in total emerging markets bond flows: Brazil, Mexico, Argentina, South Korea and Turkey together accounted for half of all the bonds issued in the 2

6 period (See Figure 3). Figure 2. Regional share of total bond issues, Africa 5% Asia 29% Latin 52% Europe 14% Figure 3: Ranking of emerging market bond issuers, (cumulative flows ) (In million of US$) Brazil Mexico Argentina S. Korea Turkey Hong Kong Russia China Philippines Hungary Venezuela S. Africa Colombia Malaysia Thailand Indonesia Chile Taiwan Lebanon India Poland Israel Singapore Panama Uruguay In millions of US$ Source: IIF This surge in debt flows is the result of both the liberalization of capital markets in the 1990 s and the rise of new borrowers that surfaced from the pool of countries that were coming out of the 1980 s debt crisis with Brady bond deals. The passing from state-led to market-oriented economies gave a new sense of creditworthiness to these countries. This is particularly so given that the debt crisis experience of the 1980 s was equated with the type of state-led development model that was to be blamed for bloated bureaucracies, hyperinflation, huge fiscal deficits, low export diversification, 3

7 corrupt state-owned companies and inefficient and protected private sectors. Thus, market reforms and macroeconomic stabilization, as embedded in the Washington Consensus, gave a certain insurance of better solvency prospects to these economies Figure 4. JP Morgan s EMBI index, daily spreads In basis points Mexican crisis Russian crisis Asian crisis Brazilian crisis Argentine crisis /2/1991 5/2/1991 9/2/1991 1/2/1992 5/2/1992 9/2/1992 1/2/1993 5/2/1993 9/2/1993 1/2/1994 5/2/1994 9/2/1994 1/2/1995 5/2/1995 9/2/1995 1/2/1996 5/2/1996 9/2/1996 1/2/1997 5/2/1997 9/2/1997 1/2/1998 5/2/1998 9/2/1998 1/2/1999 5/2/1999 9/2/1999 1/2/2000 5/2/2000 9/2/2000 1/2/2001 5/2/2001 9/2/2001 Source: JP Morgan-Chase The improved creditworthiness was reflected by spread levels of the EMBI index before the 1995 Mexican crisis and specially, before the emerging market crises of , where they reached their lowest level in the decade (see Table 1). An important aspect of the later crisis is that spreads have not return to pre-crisis levels. This could signal an interesting change in perception from bond investors of the creditworthiness of emerging markets. 4

8 Table 1. Average annual spreads, EMBI index Avg. annual spreads, in bp Source: JP Morgan-Chase An important reality check for bondholders came with the recent wave of emerging market crises that led to the discovery that despite market-reform efforts almost everywhere in the emerging world, sovereign risks continue to exist in these countries. In fact, the success of countries in accessing capital markets might have played against them, as with the surge in bond debt came as well a reversal in the improvement of debt burdens seen in the first half of the 1990 s (see Figure 5). This has led to a revision of sovereign risks. The worsening of sovereign credit risk in emerging markets is reflected in their overall rating revisions by the credit rating agencies: among a list of 22 emerging countries rated by Standard & Poor s since 1993/1994, we see that their average sovereign rating had been downgraded by almost two notches between 1993/94 and Figure 5. External debt relative to exports and GDP in emerging markets ( ) (%) External debt as a % of exports External debt as a % of GDP Source: IIF Since Russia in 1998 (who defaulted on restructured notes but not on its Eurobonds), four other countries have defaulted on their bonds, with Argentina becoming the most 5

9 important sovereign bond default in world history. Among the 5 sovereigns that have defaulted, we have two major bond issuers (Argentina and Russia) and three smaller issuers (Ecuador, Pakistan and Ukraine). The growing number of sovereign defaults after 1998 has led to a realization that bond defaults might not be isolated events. On the contrary, bond defaults are more likely to be the trend in future debt crises in emerging markets. This does not mean that we foresee sovereign debt crises emerging anytime soon. It means that in the future (as it has been the case in the recent past) credit events in emerging markets will be accompanied by bond defaults given their important share of emerging markets external debt. As Table 2 shows the average share of international debt securities (international bonds and notes plus Bradys) increased by almost 50% between 1996 and 2001 in selected emerging market countries. Table 2. Share of debt securities and Brady bonds in total external debt in selected emerging market economies ( )* Lebanon 20% 37% 47% 53% 62% 73% Argentina 48% 51% 55% 57% 58% 58% Hungary 60% 56% 57% 60% 55% 56% Venezuela 56% 57% 55% 51% 52% 54% Mexico 38% 45% 46% 51% 53% 48% Malaysia 25% 28% 32% 38% 42% 46% Brazil 41% 38% 42% 42% 46% 44% Colombia 15% 19% 23% 27% 34% 43% South Korea 29% 31% 35% 36% 38% 41% Uruguay 22% 24% 23% 28% 30% 39% Philippines 18% 22% 24% 26% 30% 37% Chile 15% 17% 20% 27% 28% 32% South Africa 19% 18% 21% 25% 28% 31% Taiwan 13% 19% 22% 25% 30% 30% Thailand 8% 11% 14% 17% 19% 25% Turkey 24% 23% 22% 25% 25% / Poland 26% 23% 20% 19% 19% 21% Russia 1% 6% 15% 16% 18% 18% Average 28% 29% 32% 35% 37% 41% * This indicator measures the share of debt securities and Brady bonds in total external debt stock (includes both private and public borrowers in the country). 3 Source: Joint BIS-IMF-OECD-World Bank statistics on external debt ( This proportion is even more important when looking only at the share of 3 The data are derived from quarterly BIS statistics on issues of money market instruments, bonds and notes in international markets and are based on information provided by various market sources (such as Euroclear, Dealogic, Thomson Financial Securities Data and ISMA). 6

10 international bonds in total outstanding public external debt of emerging markets. The IMF has estimated that the share of international bonds in emerging markets public external debt rose from 13% in 1980 to 60% in 2000 (Krueger, 2002a). 4 Table 3 shows a selection of countries where in some cases the share of bonds in public external debt has reached 70% to 80% of total public external debt. Table 3. Share of bonds in public external debt, selected countries Bonds, % of Country public external debt Source Date Notes Venezuela 79% Ministry of International bonds and Bradys as a % of direct Dec-01 Finance public sector external debt (excludes PDVSA) Brazil 76% Ministry of International bonds and Bradys as a % of direct Apr-02 Finance public sector external debt Argentina 72% International bonds and Bradys as a % of gross Ministry of Sep-01 public external debt (before the November 2001 Economy exchange of bonds into guaranteed loans) Mexico 62% Colombia 51% Ecuador 36% Poland 24% Russia 24% Peru 21% Ministry of Finance and Public Credit Central Bank of Colombia Central Bank of Ecuador Ministry of Finance Central Bank of Russia Ministry of Economy and Finance Apr-02 International bonds and Bradys as a % of gross public external debt International bonds as a % of medium and long term Dec-01 public external debt Bonds (restructured) as a % of total public external Dec-01 debt International bonds and Bradys as a % of foreign Dec-01 state budget debt Eurobonds as % of the general government external Dec-01 debt Mar-02 International bonds as a % of medium and long term public external debt III. New and old sources of credits and challenges of sovereign debt workouts Recurring financial crises has been a feature of Latin American financial markets for nearly 200 years and establishing a mechanism to process default in a smooth and predictable, if not painless, way makes sense LatinFinance, February 2002 Developing countries, and particularly Latin American governments, have had some experience with debt default. Latin America s first debt default came just few years after their long fought independence war with Spain in the mid-nineteenth century. Since then, Latin America s experience with international financial markets has been one of credit boom and busts cycles. For over a century, though, an extreme default situation could give creditors grounds for invasion or trade blockages from creditors 4 The relative importance of international bond debt for governments in emerging markets is the result not only of the growing importance of bond markets for emerging markets in the 1990 s, but also of the existing stock of Brady bonds issued in exchange for the defaulted commercial loans of the 1980 s. 7

11 wanting to be repaid. It is good to know that with time international financial markets have found more orderly ways to deal with sovereign defaults. 5 The Paris Club and London Club are an example of institutions that have emerged to better organize debt workouts. Their history and functioning are beyond the scope of this paper, 6 yet they exemplified that debt events in emerging markets have led to the creation of creditor-based institutional mechanisms for more orderly debt workouts (the London Club represents commercial banks and Paris Club official lending). More importantly, these institutions have taken a life of their own as they have outlived the credit events that led to their creation and have since then provided rules and procedures for ways to deal with debt restructurings. 7 Therefore, an argument could very well exist that bond debt workouts will be more efficient and less traumatic if bondholders were organized. Are bondholders likely to organize into a creditor s club for negotiating purposes? Bond financing had been in the pre-1930 s period the most important source of credit for emerging markets. However, after the debt crisis of the interwar period (1930 s), this source of financing dried up to emerging markets. It was not until the 1970 s that debt flows returned to emerging markets in a massive way. This credit boom was, nonetheless, bolstered by a surge in syndicated loans that resulted from the recycling of petrodollars: excess financial liquidity coming out of the oil shocks of the 1970 s that was redirected by banks into emerging markets. As other lending booms before in history, the 1970 s boom ended with the 1980 s debt crisis. The 1980 s debt defaults, however, mostly affected commercial banks and official lenders. Because the restructurings of the 1980 s did not include bondholders this gave bonds an almost de facto seniority status. This situation had until recently lessened the incentives for a bondholder s organization to deal with sovereign restructurings in the post-war era. As we have argued debt crises have been an integral part of lending which means that bondholders have been confronted before with sovereign defaults and with the 5 A credit event or default has followed each wave of foreign lending to Latin America. British financing to the newly independent Latin American countries ended up on the 1825 default crisis after which the region lost access to international capital markets for some time. The second wave of foreign lending to Latin America came during the 1850 s and 1860 s and ended with the 1873 international financial crisis, which led countries in the region to default once again on their debts. The third wave in the 1880 s involved flows from Britain and Europe to finance mainly Argentina and Uruguay and again it ended with the crash of 1890, and Argentina going into default. A fourth wave of flows to emerging markets in the 1920 s ended at the end of the decade with the collapse of commodity prices and the Great Depression. During this period, virtually all countries, with the exception of Argentina, defaulted on their debts and the last major debt crisis came during the 1980 s after a surge in credit during the 1970 s. For a more detailed history of emerging market debt crises, see Eichengreen et al. (1989). 6 It is interesting to note that the Paris Club has its origin in a debt crisis in Argentina in the 1950 s. 7 The Paris Club of creditor countries was formed in It is an informal group of creditor governments from major industrialized countries. There are 19 members of the Paris Club: Australia, Austria, Belgium, Canada, Denmark, Finland, France, Germany, Ireland, Italy, Japan, Netherlands, Norway, Russia, Spain, Sweden, Switzerland, the United Kingdom and the United States of America. 8

12 organization challenges surrounding debt restructurings. In fact, during the history of bond financing several attempts were made for organizing bondholders in a more permanent fashion in the 19 th century in England and in the early 20 th century in the US. 8 So, if bondholder s committees existed before in history, there is no reason to believe that such an organization will not re-emerge in the 21 st century, if we continue to see sovereign defaults. In fact, a bond committee is in the process of defining itself with the recent constitution of the Emerging Markets Creditors Association (EMCA). Chronology of a modern attempt at a bondholders association: EMCA The first attempt at consulting creditors in a bond debt restructuring (post Brady) came with Russia in However, in this case the country had defaulted on notes that as Buchheit argues are functionally like bonds (Buchheit, 2000) that had been the result of restructured loans of commercial banks to the former Soviet Union only a few years back ( ). When it found itself in a default situation in August 1998, Russia decided to use the same mechanism that it had followed when it restructured its Soviet-era debt. As a result, a type of London Club commercial bank advisory committee was recreated to carry out Russia s debt restructuring given that these notes were for the most part expected to be in the hands of these banks. In Ukraine s and Pakistan s case, the approach was mostly one of informal discussions with major bondholders and the use of Wall Street firms to carry out unilateral debt exchanges as it were another debt swap or bond issuance. Until recently, the arguments against a bondholders association were that such committees were not needed given that Ukraine and Pakistan could reach an agreement without the need for a formal mechanism or organization of bondholders. However, as some had pointed out, debt exchanges are easier when there is no discount or write down of debt involved, which were the cases of Ukraine and Pakistan (Eichengreen and Portes, 1997). It was not until Ecuador s debt restructuring in 2000 that a first step was taken from the part of the country to contact a bondholder-based organization and from the part of bondholders to organize themselves into one. Ecuador first attempted to informally contact an organization that had been around since the early 1990s called 8 As early as 1868 England saw the emergence of a corporation of bondholders formed by a consortium of loan houses and brokers (Portes, 2000). With the debt crises of the interwar period and the growing role of the United States as a bond lender in international markets came the US version of the UK bondholder organization: the foreign bondholders protective council. This council was sponsored by the US government: under the Securities Act of 1933, a provision was made for the creation of a Corporation of Foreign Security Holders with a similar role than the British Corporation of Foreign Bondholders. The act states that the corporation shall be created "for the purpose of protecting, conserving and advancing the interests of the holders of foreign securities in default." However, with the significant decrease of bond lending to emerging markets after the 1950 s, these bond committees lost momentum (Buchheit, 2000). 9

13 the Emerging Markets Traders Association (EMTA). 9 This never became more than very informal preliminary attempts at a discussion with bondholders and eventually Ecuador ended up following a more unilateral approach to debt restructuring, although it did form a consultative group with its major bondholders for information purposes. Ecuador hired Salomon Smith Barney as its financial advisor and Cleary, Gottlieb, Steen and Hamilton as its legal advisor to come up with a one-time exchange proposal which included a write down of almost 40% of the face value of bonds. EMTA is an organization that represents the sell-side. And for bondholders this is an inconvenience. They argue that a body that represents the sell side, which is mainly formed by investment banks, is ill fitted to negotiate solely with the interest of bondholders in mind given that they have other commercial interest in the country. 10 Thus the emerging markets creditors association (EMCA) was created as an organization of bondholders from the buy-side. EMCA is still at a very incipient organization stage and thus it is not yet possible to clearly envisage the role it could play in providing more orderly debt restructurings or in leading debt workouts. The rationale behind the creation of a bondholders committee is the refusal of take-or-leave-it propositions that have been the characteristic of bond restructurings until now. It has been the realization of their inability to influence the terms of bonds restructurings that led to the need for some kind of organization to ventilate bondholders positions in future debt restructurings. So far, this bondholders association has not formally engaged in debt restructuring negotiations with a sovereign: Argentina could be the first case. However, it seems that given the nature of the market, the organizational dynamics will be based on ad hoc committees organized with the major bondholders of a specific defaulted country. This organization could be linked to EMCA, which is supposed to have a more permanent status. In fact, the organizational attempts of Argentina s international bondholders seem to indicate that ad hoc bondholder s committees to treat specific countries is the preferred option for bondholders. 11 The Argentina Bondholder s Committee (ABC) was created in November 2001 and claims to currently represent approximately 60% of the outstanding foreign bonds. If indeed bondholders follow through with this idea, it will be a departure from their previous position of non-organization and it will mean that a bondholders creditor club would have finally asserted itself. 9 From an interview with representatives of EMTA on February 20 th, 2002 in New York. 10 From an interview with representatives of the Argentina Bondholders Committee on March 12 th, 2002, in New York. 11 From an interview with representatives of the Argentina Bondholders Committee on March 12 th, 2002, in New York. 10

14 A bondholder s committee: is this good news for the reformers of the international financial architecture? An organization of bondholders could be part of the solution for achieving more orderly debt workouts and propositions to resuscitate such type of organizations have existed since mid-1990 s (Macmillan, 1995 and Eichengreen and Portes, 1995). However, some have argued that a bondholder s club has limitations as it can only offer non-binding recommendations to bondholders who can either accept or reject such propositions (Portes, 2000). From this perspective, a bondholder s committee might not be able to guarantee orderly workouts. Some policymakers of multilaterals and G7 countries believe that restructurings are problematic because of the intrinsic rules governing bond contracts. In fact, a bondholder s committee will face a more complicated situation than it did before the last bond debt crisis of the 1930 s, as during that time a major change in the rules surrounding bond contracts occurred and collective action clauses were eliminated with the intention to protect creditor s rights from abuses. 12 It is exactly this inflexibility of NY-law based bond contracts to changes in the financial terms of bonds that leads policymakers to believe that bond debt restructurings are potentially more distressful and prone to disorder than commercial or official lending. Thus, it is these underlying rules of bond contracts that government officials and policymakers in multilateral institutions want to reform. VI. Why are bond restructurings an issue? The international bond markets are, despite electronic registers held by the likes of Euroclear and Cedel, still characterized as bearer markets. Bond holders are not easily identifiable nor do they have or seek the same degree of intimacy with the issuer that a commercial bank has with its borrower" Euromoney, May 1999 Sovereign bond restructurings are believed to be more complicated than official or commercial restructurings because of collective action problems, which are exacerbated by the underlying rules of bond contracts: a) Collective action a problems: It is argued that the important number of bondholders means that collective action problems are an issue in bond restructurings. Bondholders have different incentives when entering into bond restructurings, unlike commercial banks and official creditors who generally have a long-term commitment in debtor countries. The bondholder is more often than not a short-term and 12 Buchheit (2002) argues that only a small percentage of US bonds included collective action clauses before the 1930 s. However, in 1939, the Trust Indenture Act explicitly stated as a rule the unanimity clause in US-bond contracts to payment terms as opposed to the majority clause used in UK-bond contracts. 11

15 sometimes one-time investor in the country. Moreover, bonds change hands very frequently, which render difficult debt restructurings given that the sovereign has to start by identifying its creditors. This is not to say that restructuring commercial bank loans is easy, however, in past experiences there were never more than 750 banks involved in a sovereign debt rescheduling and bank advisory committees rarely had more than 15 members (Eichengreen and Portes, 1997). b) Legal rules: Irrespective of the nature of the creditors and of the bond market, the most controversial aspect is the legal framework that governs bond contracts. Bond contracts normally have London or New York law as their legal jurisdiction and each jurisdiction commands different rules of the games for engaging in bond restructurings. Most notably, the law of New York does not permit amendments to the payment terms of the instrument (due dates and amounts) without the consent of each bondholder. This makes it almost impossible to modify the current financial characteristics of the bonds. Conversely, if the English law were applicable such amendments could be possible with a majority of bondholders accepting the terms of the exchange. Under such a framework, all bondholders will be bound by the decisions taken by the majority of the bondholders. 13 If bond covenants cannot be modified under the New York law, does this mean that bonds cannot be restructured? Not exactly: under the New York law a debt restructuring of sovereign bonds is, nonetheless, possible through a debt exchange, with the new bonds reflecting the conditions of the debt to be restructured. However, there is a problem: the exchange is not binding on all bondholders. Therefore, as Buchheit and Gulati (2000) argue, there are potential legal risks from creditors that hold out on the exchange and that can pursue legal action to claim a better deal from the sovereign. This has already occurred in Peru after its 1997 Brady exchange in the case known as the Elliot case (to be treated later). In syndicated bank loans, for example, this risk is minimized by the fact that these loans have sharing clauses that require individual creditors to share with other bondholders any amounts recovered from the borrower. This rule discourages individual mavericks from resorting to lawsuits. Some multilateral officials and academics argue that the potential risk for legal action and the complications surrounding bond restructurings have effectively led to fewer debt workouts that would have otherwise been the case had the IMF not intervened in the different emerging market crises since Mexico in Therefore, the IMF has found itself bailing out bond creditors and in the process setting the precedent for moral hazard and creating distortions in financial markets. Propositions for breaking 13 This majority can go from 66 2/3% to 75% of bondholders. 12

16 this perceived vicious cycle are varied starting from the inclusion of collective action clauses into US-law governed bonds to minimize the legal risks in restructurings (embraced by John Taylor, US Undersecretary of Treasury) and the use of exit amendments to deter holdouts as used in the Ecuador bond debt exchange 14 to the most radical proposed by Anne Krueger from the IMF which sees the creation of a sort of international bankruptcy court. V. A proposal for an international proxy for a domestic bankruptcy court: Does Anne Krueger have a point? The proposal advanced by Ms. Krueger would represent a step backward in the area of crisis resolution, because it would threaten legitimate creditor rights and severely interfere with the voluntary and market-based resolution of sovereign financial crises A group of private investors 15 Discussions about reforming the rules governing the international financial markets date back to the Mexican crisis and the presence of IMF bail-out packages. 16 Proponents of reform argue that expectations of IMF s bail-outs have created moral hazard and thus have impeded the true assessment of lending risks in emerging markets. Moreover, some argue that far from guaranteeing the roll over of debt from the private sector, multilateral bail-outs have indeed financed the exit of private capital (Eichengreen and Mody, 2001). Anne Krueger, First Deputy Managing Director of the IMF, argues that indeed such expectations of IMF bail-outs have made the international financial system less stable and less efficient (IMF, 2001b). Table 4. Amount of bail-out packages* Countries Packages Mexico US$50 billion South Korea US$57 billion Thailand US$34 billion Indonesia US$40 billion Brazil US$42 billion 14 For a more comprehensive view of exit amendments, see Buchheit and Gulati (2000). Mr, Buchheit is a partner at the law firm Cleary, Gottlieb, Steen and Hamilton that advised Ecuador on its sovereign bond debt restructuring. 15 From a letter addressed to the Managing Director of the IMF, Mr. Horst Kohler signed by the presidents of the Securities Industry Association, the Emerging Markets Traders Association, the International Primary Markets Association, the International Securities Market Association and the Bond Market Association posted on February 6 th The Mexican crisis with the immediate collapse of the peso and of the government s dollar-linked securities posed an unprecedented challenge for modern international financial management: how to restructure bonds when your pool of creditors is not only numerous but highly unknown and unstable (Eichengreen and Portes, 1997). 13

17 Russia US$20 billion Argentina US$40 billion 17 *Some of these bail-out packages include private sector participation. Source: Eichengreen and Portes (1997) and Portes (2000). Proponents of reforms believe that if the international financial system provided clear and simple rules for more orderly bond debt workouts countries will not dangerously wait before confronting inevitable solvency problems, and the IMF will not be as pressured to bail out private investors from crisis in emerging markets. Argentina is cited as a case at hand given that prolonging the inevitable led to unnecessary value destruction in the domestic corporate and financial sector making harder the prospects of recovery and thus longer the restoration of repayment capacity in the country. As a result, Ms. Krueger has proposed a formal mechanism that would allow a country to request a temporary standstill on its debts, during which time the country will negotiate a restructuring with its creditors (Krueger, 2001a). This workout mechanism seeks to achieve for sovereign restructurings what a domestic bankruptcy court provides in corporate restructurings. 18 However, even if mimicking these procedures, certain limitations will always be attached to such a bankruptcy scheme given the particular nature of the debtor: a sovereign. In domestic bankruptcy procedures a stay is imposed on all legal actions against the debtor (i.e. a freeze on creditor lawsuits), a reorganization plan is dictated, legal incentives are provided for those who give new credit, and restructuring plans agreed by the majority are binding to all creditors. Therefore, Krueger argues that the following conditions should be assured in sovereign workouts: a) the prevention of maverick creditors from seeking payments through the court system; b) the implementation of proper policies by the debtor country and the fair treatment of creditors; c) the provision of new money from private lenders; and, d) the binding of all creditors to the terms agreed by a large majority of investors (Krueger, 2001a). This proposition has faced strong opposition from different sectors and has led to certain revisions to the original blueprint. Here a some of the most important arguments against this scheme: a) On the content of the proposition: a bankruptcy model for sovereigns faces important shortcomings given that, as opposed to corporates, the management of a 17 This refers only to the initial bail-out package announced by the IMF on December 2000, which included also private sector and official participation. 18 This is not the first time the sovereign bankruptcy idea had been put forward. Eichengreen and Mody (2001) argue that it had been first presented by Williamson (1992) and Sachs (1995) and had also been evaluated in 1996 by the G10 in the Rey report in the aftermath of the Mexican crisis. For a background paper on the history behind the idea of a sovereign bankruptcy framework, see Rogoff and Zettelmeyer (2002). 14

18 government cannot be removed by its creditors, total liquidation of the assets to recover value is not a possibility, reorganization or contingency plan dictated by creditors is not imposable on a sovereign. These shortcomings could limit the accountability of defaulting for a sovereign and thus, the interest of creditors in such a scheme. b) On the role of the IMF as mediator: some question the appropriateness of the IMF to act as mediator between the debtor countries and its creditors given its conflicting interests as a creditor in its own right. Some critics point to the IMF s role in even providing the assessment of the solvency status of the debtor, which Krueger s proposal had initially define as a main role of the IMF in such a scheme. 19 However, in the most recent version of the proposed sovereign debt restructuring mechanism, Krueger provides for a less active role of the IMF in the decision to go into a debt restructuring, on the activation of the stay (i.e. protection against litigation) and on the maintenance of the stay. Krueger now proposes that these decisions be left to the debtor country and the majority of the creditors (Krueger, 2002c). c) On the consequences for the emerging markets: Private creditors argue that any proposition that weakens creditor s right, shifts the balance of power to debtors and makes it easier for countries to restructure their debt will actually increase borrowing cost and further reduce the flow of private capital flows to emerging markets (Chamberlain, 2002, p.3). Krueger believes that on the contrary providing such types of mechanisms will increase investor s discrimination between good and bad credit. 20 She argues that, in fact, it will not necessarily lead to increasing borrowing cost for good credit countries but it is likely that it will increase borrowing costs to bad credit countries. 21 On the whole, emerging borrowers have not manifested their support or desire for mechanisms or provisions that facilitate debt workouts precisely because supporting such propositions will give a bad signal to market participants about the willingness of a country to pay. However, Krueger claims that it is also in the interest of bondholders to support such a proposition because it will avoid the huge fall in the value of bondholders claims that has characterized bond defaults. The unnecessary loss in value will beavoided, Ms. Krueger argued, by giving certainty to the restructuring process itself. 19 In her words The Fund s involvement would be essential to the success of such a system. We are the most effective channel through which the international community can reach a judgment on the sustainability of a country s debt and of its economic policies, an whether it is doing the necessary to get its balance of payments back into shape and avoid future debt problems (Krueger, 2001a). 20 From a speech to the Indian Council for Research and International Relations, Delhi, India, December For example, Eichengreen and Mody (2001) argue that the inclusion of collective action clauses into bond contracts will not necessarily increase borrowing costs. 15

19 d) On the feasibility of the project: an international bankruptcy court will require reform of domestic legislations. This requirement makes Krueger s proposition politically very difficult to implement since as JP Morgan-Chase argues to think that so many national governments legislatures will adopt the proposal, let alone agree on a common definition of bankruptcy is unrealistic (Emerging Markets Today, March 5 th, 2002). 22 This is why Krueger proposes its implementation through an amendment of the IMF s Articles of Agreement, which is binding on all members once the 60% of IMF members holding 85% of the IMF s voting power approves it. Getting 85% of voting power to agree necessarily requires the consent of the US whose support is not yet completely clear (Mr. Taylor s own proposition for the inclusion of collective action clauses into bond contracts is a signal to the type of more market-oriented solution that the US would be willing to advocate). Moreover, even with the passing of this amendment on the IMF s Articles of Agreement it is not clear that this will be sufficient to override domestic bankruptcy procedures (JP Morgan-Chase Emerging Markets Today, March 5 th, 2002). e) On the lack of clarity on some major points regarding the scheme: Which creditors will be involved in an international bankruptcy scheme? How will seniority be determined? How will the need for a write down be defined and which creditors should assume it? How these questions are answered will be of utmost importance for the future of the Paris Club and London Club since an international bankruptcy court will have the capacity to overrule any decisions taken by a particular creditor s club given that its decisions are meant to be binding on all creditors. In Krueger s most recent version of the bankruptcy proposal (Krueger, 2002c), the question of how to treat bilateral official debt is explicitly raised but a solution is not proposed: we will need to explore further whether it will be feasible to include bilateral official debt under the sovereign debt restructuring mechanism and, if so, how this would be done in a manner that pays due regard to the special features of these claims. 23 f) On the need for such a reform: Many private investors believe that the IMF is overstating the legal risks involved in debt restructurings and in return argue that the cost for emerging markets will outweigh the benefits. Two arguments have been put forward: The rogue creditor argument has been exaggerated: Private investors believe that the markets have actually been very effective in dealing with restructurings: debtor countries such as Russia, Ecuador, Pakistan and Ukraine have successfully been able to restructure all or a substantial portion of their external bonds through the use of exchange offers Experience has shown that so-called rogue creditors have not 22 The IMF currently has 183 member countries. 23 See Anne Krueger (2002c, p. 18). 16

20 obstructed these restructurings The most widely cited (and possibly only) example of rogue behavior, a single creditor in fact failed to disrupt Peru s restructuring in 1997 and only collected payment few years later. Accordingly, we believe that the danger of rogue creditors has been highly exaggerated. 24 To these claims, the IMF does admit that fears of legal action might be exaggerated due to the costs and time involved in pursuing such a course of action, but the fact that it is always a potential risk can create distortions and uncertainty in debt workouts. Bondholders have actually important incentives to participate in an exchange because they mark their positions to market. Thus, the objective of bondholders is to return value to the paper they hold as soon as possible. A delay in completing a restructuring program will only leave the paper on a default status on the creditor s book (Buchheit, 2000). From this discussion some questions arise: if bond debt restructurings are supposed to be disorderly, why have bondholders until now participated in bond exchanges in such a significant number? Why aren t legal actions against the countries that restructure more frequent? Is the Elliot case an exception? 24 On a letter addressed to the Managing Director of the IMF, Mr. Horst Kohler, posted on February 6 th 2002 and signed by the presidents of Securities Industry Association, the Emerging Markets Traders Association, the International Primary Markets Association, the International Securities Market Association and the Bond Market Association. 17

21 TEXT BOX 1: Peru debt restructuring and the Elliot Case: -In October 1995, Peru announced a Brady restructuring in the context of an IMF-supported program. Some 18 months after this announcement, a so-called vulture creditor fund, Elliot Associates, purchased US$20.7 million face value of Peruvian commercial loans that had been guaranteed by Peru. Unlike most other creditors, Elliot did not accept the Brady bonds in exchange for Peruvian debt. -Elliot held out on the exchange and tried to negotiate better terms with the sovereign. It then proceeded to file a lawsuit in New York for recovery of the full face value plus interest on the loans that it held. After several years of legal proceedings, Elliot, in June 2000, obtained a judgment against Peru for US$56 million and an attachment order against Peru s assets abroad which will have diverted debt service payment of its regular debt. Given the possibility that it could end up being forced to default on its debt, the country did not appeal the decision and decided to settle. The fact that Peru did not fight this leaves the door open for vulture funds to pursue legal action and for countries to want to settle. - The legal case against Peru lasted 4 years. In the end, Peru was obliged to pay Elliot the full face value of the bonds plus all the legal fees involved. Source: IMF, 2001 VI. Understanding recent bond restructurings: What have we learned? How have collective action clauses and exit amendments performed so far? Since 1998 four countries have restructured international sovereign bonds/notes: Russia, Ukraine, Pakistan and Ecuador (see Table 5 for a comparison between sovereign bond/note restructurings). There is not a standardized way in which countries have defaulted. Yet, it seems that there are certain commonalities in the way they have restructured. It is interesting to note that there has been a certain learning process with each default and with each restructuring. Russia and its selectivity of default Russia defaulted on its domestic ruble-denominated debt that was held by both residents and non-residents. It also defaulted on its Soviet-era debt both to the Paris Club and to London Club creditors: In August 1998 and September 1998 it missed a payment to Germany as part of the Paris Club debt, and defaulted on December 1998 on its Soviet-era debt to commercial banks. Russia selectively decided to continue its timely payments on its US$16 billion Eurobonds issued after

22 As mentioned before, Russia used its London Club advisory committee to go about restructuring its debt (Russia s Soviet-era debt had already been through a restructuring in 1997). Russia obtained a write down of 37.5% on its already restructured principal notes (PRINs) and 33% on its restructured interest notes (IANs). The most important precedent set by Russia was its selectivity of default, which led creditors in subsequent negotiations to seek comparability of treatment. 25 Pakistan: a first sovereign default on Eurobonds pressured by the Paris Club In the debt restructurings of the 1980 s, the comparability of treatment principal was mostly applied to commercial banks loans. International bondholders were for the most part left unscathed. The almost sacrosanct position of international bonds with respect to other debts was also respected with Russia but was nonetheless rapidly challenged with Pakistan. In the case of Pakistan, the restructuring of sovereign bonds came after pressure from the Paris Club who had agreed on February 1999 to restructure US$3.3 billion of Pakistani debt on the condition of equality of treatment with other Pakistani creditors. This meant, basically, that Pakistan had to also restructure its Eurobonds, which it had continued to service even when it went in default with the Paris Club (The Economist, May 8 th, 1999). The case of Pakistan served to set a peaceful and uneventful precedent in inter-creditor relations mainly because of the small fraction of the country s bond external obligations in absolute terms (US$608.3 million) but also in relation to Pakistan s Paris Club debt, at $3.3 billion (total debt to the official sector was about $26 billion). Ukraine and the use of collective action clauses: Ukraine s debt restructuring in February 2000 introduced an important novelty: the use of collective action clauses to change the payment terms and amortization schedule. Collective action clauses allow changes on the provisions of bond contracts with only the agreement of a qualified majority of bondholders: the agreed changes are binding on all creditors. Actually two restructuring methods had to be used given that only 3 of the bonds to be restructured were under Luxembourg law (another sovereign bond was German-law based which does not incorporate collective action clauses into bond documentation). For the non-luxembourg bonds a simple debt exchange was proposed. For the Luxembourg-based bonds a change in the original payment structure was first enacted using collective action clauses and then a debt exchange was proposed for new bonds with the same terms as those that had just 25 For comparison purposes, this paper will only focus on the restructured international notes and not on the controversial domestic debt restructuring of GKOs. A comparison of domestic debt restructurings is beyond the scope of this paper. 19