Toward a Statutory Approach to Sovereign Debt Restructuring
Lessons From Corporate Bankruptcy Practice Around the World
  • 1 https://isni.org/isni/0000000404811396, International Monetary Fund

Contributor Notes

Author’s E-Mail Address: pbolton@princeton.edu

This paper provides an overview of key elements of Corporate Bankruptcy Codes and Practice around the world that are relevant to the debate on Sovereign Debt Restructuring. It highlights four components common to most bankruptcy reorganization institutions: a stay on debt collection efforts to prevent a costly run for the assets, broad enforcement of absolute priority, majority voting among creditors on the proposed reorganization plan, and new higher priority financing to keep the firm going while its liabilities are restructured. The paper argues that these components ought to be present in any sovereign debt restructuring procedure.

Abstract

This paper provides an overview of key elements of Corporate Bankruptcy Codes and Practice around the world that are relevant to the debate on Sovereign Debt Restructuring. It highlights four components common to most bankruptcy reorganization institutions: a stay on debt collection efforts to prevent a costly run for the assets, broad enforcement of absolute priority, majority voting among creditors on the proposed reorganization plan, and new higher priority financing to keep the firm going while its liabilities are restructured. The paper argues that these components ought to be present in any sovereign debt restructuring procedure.

I. Introduction

The IMF’s (2001, 2002) recent proposals for the introduction of a sovereign debt restructuring forum to facilitate the resolution of sovereign debt crises, led by Anne Krueger, the First Deputy Managing Director of the IMF, draw an important analogy between corporate bankruptcy and sovereign debt restructuring. These proposals build on policy reassessments following the Mexican crisis of 1995 (e.g. the Rey Report, 1996) and, more importantly, the Asian crisis of 1997. They also build on perceptive early policy proposals by academics, most notably Oechsli (1981) and Sachs (1995) (see Rogoff and Zettelmeyer, 2002 for a survey of proposals).

The Mexican and Asian crises revealed two major problems with the established policy approach towards emerging market balance of payments crises: IMF programs providing temporary bailouts to financially stricken member countries. Even if a policy of granting massive financial aid could succeed, as it did for Mexico, it was likely to increasingly strain the IMF’s own financial resources. In addition, a policy based mainly on bailouts was likely to result in careless future lending (“moral hazard in lending”) if it had not already done so.

To address these problems a change in orientation has been proposed to allow for debt renegotiations and for more “private sector involvement” (e.g., the Rey Report, 1996). This change of policy direction reached a culmination point when proposals for a sovereign debt restructuring mechanism (SDRM) were recently aired by Anne Krueger (2001, 2002). These proposals do not call for a radical shift away from existing policies based on IMF programs but rather for the addition of a last resort debt reduction option to a policy based mainly on bailouts.

The ideas for the establishment of an SDRM draw an important analogy with Chapter 11 of the U.S. Bankruptcy Code. They suggest that a financially distressed sovereign nation should be allowed to declare a standstill on debt payments and to put itself under the protection of an international bankruptcy forum. To be sure, the analogy is not perfect. For example, creditors cannot easily foreclose on a sovereign’s assets. Despite important differences, this paper will argue that the history of U.S. corporate bankruptcy law and the varied practice of corporate bankruptcy around the world contain useful insights for the policy debate on sovereign debt restructuring.

Not all corporate bankruptcy codes around the world are equally relevant and some may have no relevance at all. Nevertheless it is useful to look into corporate bankruptcy practice outside the United States to get an idea of where the greatest differences lie and why. Thus, the main contribution of this paper is to bring in other countries’ bankruptcy experience as well as the history of U.S. corporate bankruptcy, which is particularly enlightening, and to draw a few lessons for the relevance of a statutory SDRM relative to less intrusive proposals such as collective action clauses.

The paper begins by describing the basic economic problem corporate bankruptcy law is designed to address. It then discusses key elements of the different legal responses in the United States over its history. Congress has passed three major bankruptcy acts at 40-year intervals: the first in 1898, the Chandler act in 1938, and finally the bankruptcy act of 1978. Each of these acts delineates different approaches to corporate bankruptcy reorganization. The U.S. history of corporate bankruptcy is by far the richest of all countries and provides a natural first reference point.

The paper then discusses other countries’ bankruptcy approaches to the extent that they offer other examples or models not seen in the United States. Other countries’ bankruptcy procedures differ along at least two main dimensions: (i) how they balance creditor and debtor interests; and (ii) how much court or regulatory involvement they mandate. At one end of this latter spectrum are pure “contract-based” procedures and at the other pure “administrative” procedures. U.K. bankruptcy practice is one of the main alternative approaches to the U.S. approach. At the risk of oversimplification, other countries’ approaches can generally be seen as mainly a combination of elements from the United Kingdom and past and present U.S. models. In fact, most of the legal experimentation with bankruptcy reorganization in the world can be found in the United States, the United Kingdom, and continental European and Japanese experiences. There are a few exceptions, of course. Most notably in recent years, in the transition economies. These will also be discussed briefly.

The paper continues by discussing the limits of the analogy with corporate bankruptcy as well as the relative merits of a statutory, contractual, or market-based approach to sovereign debt restructuring. It raises some important open issues and, finally, offers a few concluding comments.

II. The Corporate Bankruptcy Reorganization Problem and U.S. Bankruptcy Law

As many commentators have pointed out, Chapter 7 of the 1978 U.S. code, which regulates how a bankrupt firm’s assets should be liquidated and how the proceeds from liquidation should be divided up between creditors, is not particularly relevant for sovereign debt restructuring, as sovereign states never get liquidated. Accordingly, the discussion below focuses entirely on the reorganization part of corporate bankruptcy.

A. Brief Historical Overview

U.S. bankruptcy reorganization law grew out of so-called “equity receiverships.” These were designed to deal with the numerous railroad failures that arose at a time when the United States did not have any formal bankruptcy institutions. When a railroad failed in the 1850s, it could not turn to the states or the federal government for a rescue, as there was little political support at the time for bailouts. The railroads had to turn to courts, which gradually transformed a procedure designed to close down and liquidate a debtor’s assets into a full-fledged reorganization procedure preserving the going-concern value of the railroad. There were three important steps in this transformation:

  1. In an effort to preserve the liquidation value of the firm, courts were led to also try and maintain the railroad’s going-concern value. It was easy to see that there was not much value in selling track piecemeal.

  2. Preservation of the going-concern value required the appointment of a “receiver” to run the failed railroad while liabilities were being restructured; it also required a de facto stay on individual creditor suits.

  3. Investment banks formed bondholder committees to represent dispersed bondholders in restructuring negotiations. To obtain maximum leverage, these banks encouraged bondholders to give the committee proxy rights. When there were multiple issues and multiple committees a super-committee was formed, which formally purchased the railroad in a liquidation sale and worked out a reorganization deal. The new shell company would then distribute new claims to the old bondholders in accordance with the terms agreed by the reorganization committee.

  4. While this negotiation process was ongoing, railroads needed new funds to keep running and pay suppliers. Initially the new financing was provided by the existing claimholders (shareholders, creditors). Eventually this form of lending was transformed into what became known as “debtor-in-possession” financing (DIP financing).

  5. Despite investment banks’ best efforts, there were always “hold-out” bondholders who refused to deposit their bonds with the creditor committee. The courts’ initial response to this “hold-out” problem was to specify a minimum bid (or “upset price”) in the formal liquidation sale. Any bondholder who did not want to exchange his old bonds for new ones under the proposed restructuring terms could exchange his old claim for cash at that price. Generally the upset price was set at sufficiently low levels to make it unattractive to sell. Later, the practice evolved and reorganization deals often excluded unsecured creditors altogether, induced shareholders to contribute more cash in exchange for a stake in the reorganized railroad, and otherwise divided up the spoils among secured creditors. The Supreme Court put an end to this practice in Northern Pacific v. Boyd in 1913 by spelling out for the first time the outlines of an absolute priority rule: shareholders should not be entitled to any new claims if creditors had not been paid in full. In an interesting twist, later practice found an imaginative interpretation of the absolute priority rule, requiring that both unsecured creditors and shareholders pay cash to be able to participate in the reorganized firm (see Skeel, 2001).

The creation and evolution of equity receiverships is interesting because it highlights the role of the key features of modern reorganization law that are taken for granted today, such as the stay on individual lawsuits, the creation of creditor committees, and the importance of DIP financing. It also points to an interesting way of dealing with holdouts, which may be relevant for sovereign debt restructuring. Holdouts must either sell their claims for cash at a low price or agree to the exchange terms negotiated by the creditor committees. Equity receiverships also illustrate that the creation of a super-committee is necessary when many different bond issues are involved.

Interestingly, the early model of equity receiverships, where a receiver appointed by bondholders ran the failed railroad, was later transformed into a debtor-in-possession model, with managers obtaining court approval for “pre-emptive” reorganizations before bondholder committees could step in and appoint a trustee to run the railroad.

The great depression brought major changes to this basic template with the Chandler Act of 1938 (and the Trust Indenture Act of 1939). In response to widespread criticism of the generous fees bankers and lawyers obtained in some reorganization cases, as well as the length of the renegotiation process, it was felt that the “judicial” model of corporate reorganization did not provide adequate supervision of the negotiating parties and their lawyers. In other words, it was deemed that the people in charge of debt renegotiations under the existing model were not always acting in the best interest of all parties involved and that greater court supervision was required. Thus, a new “administrative” model was introduced under Chapters X and XI of the Chandler Act.

In contrast to the later equity receiverships, where management remained in control of the distressed firm and negotiated a restructuring agreement with creditor committees, Chapter X mandated the appointment of an independent trustee to run the company. In addition, Chapter X removed the power to propose a reorganization plan from the debtor and gave it to the trustee. Severe restrictions were also placed on who could qualify as a trustee. Basically the trustee had to be an independent agent with no connection to the firm. Once a plan had been formulated by the trustee it first had to be approved by the bankruptcy court before creditors could put it to a vote. The new plan had to respect the absolute priority rule and the SEC closely monitored the whole reorganization process. In short, the new bankruptcy reorganization code was designed to give an independent administrator a lot of power at the expense of the contracting parties (debtor and creditors) or their representatives, who were no longer trusted to come to an efficient reorganization solution. Relative to equity receiverships the new procedure involved a much more heavy-handed bureaucracy.

Anticipating that this rigid administrative approach would create incentives for the contracting parties to circumvent or avoid the new formal bankruptcy reorganization procedure, Congress went as far as passing the Trust Indenture Act of 1939 requiring the unanimous agreement of all bondholders to restructure a bond outside of bankruptcy court.2 Thus, only a bankruptcy court (under a Chapter X filing) could bind a dissenting minority to a reorganization plan.

The immediate effect of the new Bankruptcy Act was to encourage managers of publicly traded firms to avoid a bankruptcy filing under Chapter X at all cost. While the number of Chapter X filings in 1939 was over 500, and remained close to 300 in the subsequent two years, they remained around 100 per year until the replacement of the Chandler Act by the 1978 code. Partly this small number of filings was the result of greater efforts by firms to avoid financial distress altogether. But mainly the small number of filings under Chapter X was the result of firms’ greater and greater ability to file under Chapter XI of the Chandler Act.

The Chandler Act intended that all publicly traded firms file under Chapter X and smaller privately held firms file under Chapter XI. But no explicit formal restrictions were put on Chapter XI filings (except for the absence of secured lending). In contrast to Chapter X, under Chapter XI management retained control and there was no oversight by the Securities and Exchange Commission (SEC). Not surprisingly, publicly traded firms increasingly tested the gray area delineating the limits of Chapter XI filings. Over time larger and larger firms successfully filed under Chapter XI of the Chandler Act, which became the natural precursor for today’s bankruptcy reorganization procedure under Chapter 11 of the 1978 code.

This very brief historical overview of U.S. bankruptcy reorganization law contains some useful lessons for the sovereign debt restructuring mechanism. Mainly it vividly illustrates how the institution of choice in corporate reorganization is a court supervised renegotiation procedure where the debtor remains in possession of the firm while negotiations are ongoing. When firms had a choice between two alternative procedures, they showed that they preferred the procedure with a debtor in possession (Chapter XI of the Chandler Act) to that with an independently appointed trustee and regulatory oversight by the SEC.

In addition, U.S. bankruptcy history shows that a “spontaneous” creation of a bankruptcy-reorganization institution—equity receiverships3—has three main elements: (1) a stay on individual debt-collection efforts and possibly a suspension of debt repayments; (2) new financing to preserve the going-concern value of the firm; and (3) delegation of negotiations to creditor committees and the possibility to bind a dissenting minority. All three elements appear to be important for an effective bankruptcy reorganization procedure and are also likely to be essential for a sovereign debt restructuring mechanism.

B. U.S. Bankruptcy Reorganization Under the 1978 Code

We have already described the underlying economic principles of bankruptcy reorganization that have given rise to equity receiverships. We now turn to a more detailed description of the reorganization process under current U.S. bankruptcy law.

The right to file a bankruptcy petition resides mainly with the debtor. Although courts can reject a petition if the debtor has not acted “in good faith” they have (almost) never done so. Creditors can also file but only under very limited conditions. In particular it must be the case that the debtor has defaulted on its debts.4 Creditors can also ask the court to deny the bankruptcy petition.

When the petition has been granted there is a general stay on debt-collection actions. In addition, interest payments (on unsecured debt) are suspended. Once in Chapter 11, the debtor continues to run the firm5 and has the exclusive right to propose a plan of reorganization during the first 120 days. Courts can extend this period of exclusivity and often do so, but creditors have the right to object. The court oversees the reorganization process but generally is not involved in running the firm on a day-to-day basis or in the elaboration of the reorganization plan. Major decisions such as new investment or asset sales and new debt issues are subject to court approval. Again, creditors have the right to object to new debtor-in-possession financing or asset sales, although they rarely do so in practice. Creditors can also petition to liquidate the firm.

When a reorganization plan is proposed it is put to a vote. Approval of the plan requires approval by all creditor classes. Approval by one class requires a majority of all creditors, owning at least two-thirds of the value of the debt in the class, to vote in favor. When one of the classes rejects the plan one or several new plans can be proposed by management and/or creditors. In the event that none of the new plans is approved by all classes it is still possible for the judge to confirm the reorganization through a “cram down” procedure. This involves enforcing absolute priority in a way that classes ranked lower than the dissenting class do not get any new claims unless the dissenting class is paid in full. As dramatic as the “cram down” may sound it is rarely used in practice, and even then only after a lengthy negotiation process.

This completes the broad-brush description of the reorganization process under Chapter 11. Besides overseeing the renegotiation process courts are also responsible for other important and more technically involved decisions. An essential responsibility is to identify all assets and liabilities. This is often the most time-consuming part of bankruptcy proceedings. Courts also have the authority to void certain payments by the debtor prior to bankruptcy if the main motive behind these payments was to fraudulently prevent debt collection (or if these payments unfairly compensate one set of claimholders at the expense of other creditors).

Following the introduction of the new bankruptcy act in 1978 important new developments in corporate debt markets took place, which may have been responses to the new bankruptcy regime. The 1980s have witnessed a sharp increase in corporate bond issues, the spectacular creation of junk bond markets, and the appearance of leveraged buy-outs. While these developments have largely been seen as unrelated to bankruptcy reform they may actually have been facilitated by the new debtor-friendly law and may be directly attributable to the new regime. Prior to the 1978 act, SEC guidelines pushed firms with public bond issues into the costly Chapter X procedure. Firms that had both equity and bond issues could file for the more debtor friendly Chapter 11 procedure only on an ad hoc basis. After 1978, however, financially distressed firms with outstanding bond issues were assured of gaining protection under the new Chapter 11 proceedings and remaining in control while the firm’s debts were reorganized. The new law significantly reduced the cost of financial distress for the issuer and may have been an important factor behind the growth in the corporate bond market. There is no other country in the world in which corporations rely as much on bond issues. Even in the United Kingdom, corporate bond issues are only a small fraction of total corporate debt. The reason may again be the treatment of debtors and bondholders in bankruptcy. As will be explained below, U.K. bankruptcy law is heavily tilted towards protecting banks and may be as much of a disincentive to issue bonds as the Chandler act.6

Some commentators have argued that the 1978 act may be too debtor-friendly and that it encourages firms that should be liquidated to prolong their life inefficiently in Chapter 11. A number of empirical studies have found evidence that can be interpreted as corroborating this view. Thus, for example, systematic deviations from absolute priority in Chapter 11 reorganization plans have been documented (Franks and Torous, 1989). Also, several studies have found that close to 70 percent of firms that file for Chapter 11 end up being liquidated (Lopucki, 1983; White, 1984; and Flynn, 1989). But among larger firms this percentage drops to 40 percent. Other studies have found that managers who are forced into bankruptcy by their creditors tend to file for Chapter 11 even though their firm is eventually liquidated (Lopucki, 1983). Finally, several studies have found that the average time spent in Chapter 11 is close to or over two years (Flynn, 1989; and Franks and Torous, 1989). More recent scholarship, however, suggests that the typical large firm Chapter 11 case is considerably faster and generally lasts less than six months (see Baird and Rasmussen, 2002).

Perhaps the main elements of Chapter 11 that are relevant for the SDRM are the timetable for proposing a reorganization plan and the voting procedure. By tightening the majority requirement and/or allocating the exclusive right to propose a plan the SDRM can to some extent control the allocation of bargaining power between debtor and creditors. Another potentially important role for a sovereign debt restructuring body is the enforcement of absolute priority. Recently some economists and legal scholars have proposed a reform of Chapter 11 to make it more market-based. The main suggested change is to lift the exclusivity in making proposals and to allow for competing reorganization plans that creditors could choose from (see Aghion, Hart and Moore, 1993). Again, to let creditors vote over competing plans would shift the bargaining power more in their favor. Finally, the expansion of corporate bond markets following the introduction of the 1978 act highlights that both parties to the bond contract must have an interest in issuing bonds. A bankruptcy procedure that is too creditor friendly or too costly discourages bond issues as much as a procedure that is too debtor friendly.

C. Political Economy of U.S. Bankruptcy Law

Lessons can also be learned from the political history of U.S. bankruptcy legislation. If personal and corporate bankruptcy law is now generally accepted as an integral part of the legal and institutional arsenal underpinning an advanced market economy it has not always been this way. Even though the U.S. Constitution gave Congress authority to pass a federal bankruptcy law, there was no such law in place throughout most of the 19th century. The absence of a law was not merely due to a lack of interest in bankruptcy legislation or to a general perception that there was no need for any political intervention in debt contracts. It was also due to major ideological differences and conflicts between debtor and creditor interest groups on the basic orientation of a bankruptcy law as well as widespread skepticism on the efficiency of an administered insolvency procedure.

At least seven attempts to introduce bankruptcy legislation were made in Congress from 1789 to 1898. Each time legislative activity was prompted by a major economic crisis but bankruptcy laws were successfully enacted only when Republicans or their forerunners controlled both houses of Congress and the Presidency. Economic crises called for debt relief whether at the state or federal level. Republicans were more closely aligned with interstate commercial and industrial interests and were more likely to respond to pressures to introduce bankruptcy legislation at the federal level. This happened four times, in 1800, 1841, 1867, and 1898. The first three laws were repealed within a few years of their adoption, in respectively 1803, 1843, and 1878 (see Warren, 1935; Domowitz and Tamer, 1997; Skeel, 2001; and Berglof and Rosenthal, 2002).

As Skeel (2001) and Berglof and Rosenthal (2002) have highlighted, bankruptcy was a very controversial and politically divisive issue with a strong ideological divide mostly between the Democratic left, which represented debtor constituencies and was in favor of state law, and the Republican right, which supported creditor rights and a federal law. The most favored option for the Republican right was an “involuntary” bankruptcy law giving only creditors the right to declare bankruptcy and thus initiate debt collection actions. At the other end of the political spectrum the Democratic left’s preferred law was a “voluntary” bankruptcy law giving only the debtor the right to declare bankruptcy and to seek protection from creditors. In the absence of any federal law several southern and frontier states, including Florida and Texas, enacted state laws granting various forms of protections to indebted farmers and other debtors such as homestead exemptions or stay laws granting temporary or permanent relief (see Bolton and Rosenthal, 2002) for a political economy analysis of these legislative actions following the panic of 1819). Representatives and senators from these states were particularly opposed to any Republican-sponsored efforts to introduce legislation at the federal level, which might repeal these protections.

Data on roll call voting indicate that bankruptcy legislation occupied an important part of legislative activity in Congress in the first half of the 19th century. For example, as many as 68 out of 974 roll calls in the House of Representatives and 23 out of 822 in the Senate were devoted to the passing and repealing the 1841 Act. Later in the century, however, bankruptcy became a less ideologically loaded issue as is reflected in the decline in floor activity (Berglof and Rosenthal, 2002). At last, at the turn of the century a new Republican administration controlling both houses and the executive branch was able to pass a bankruptcy bill in 1898, but not without making significant concessions to debtor-oriented legislators by leaving states free to enact exemptions (Skeel, 2001).

The history of U.S. bankruptcy legislation draws attention to fundamental conflicts of interest between debtor and creditor interests, to important ideological divisions, and to major difficulties in building a majority in support of a law, which nevertheless after the fact has established its usefulness. While important political divisions remain today as to the orientation of personal bankruptcy there has now been a consensus for a long time on the need for a bankruptcy law for both individuals and corporations. While similar ideological divisions on the necessity and orientation of a statutory sovereign debt restructuring procedure exist today, the day may well come when such a procedure is seen as an essential building block of the international financial architecture. However, it is likely that sufficient support for such a procedure may only be found if exemptions or opt-out clauses are introduced to mollify the most ardent opponent debtor countries.

Interestingly, there is one major difference in the politics of U.S. bankruptcy and the politics of the SDRM. While pro-debtor constituencies in the United States long opposed bankruptcy law for fear of granting excessive powers to creditors, the opposition to an SDRM in debtor countries is based rather on the fear of excessive powers given to debtors to cancel or reduce their debts. Such powers, would, it is feared, undermine debtor countries’ ability to commit to repay their debts and thus lead to higher borrowing costs.

D. Other Relevant Aspects of U.S. Bankruptcy Reorganization Practice

So far the discussion has centered on corporate debt restructuring under some form of court supervision. But there are at least three other aspects of U.S. corporate debt restructuring procedure that are potentially relevant for sovereign debt restructuring.

Workouts, Bond Exchanges, and Prepackaged Bankruptcy Procedures

Chapter 11 only defines the rules of the end game of corporate debt restructuring, and most financially distressed firms who restructure their debts actually never end up in bankruptcy court. Before entering Chapter 11 firms attempt a workout or an exchange offer and often successfully restructure their debts that way. Some firms have gone as far as designing their own “prepackaged bankruptcy procedure” to be applied in a workout and, if needed, enforced in Chapter 11. That is, the court is only used to confirm the reorganization plan elaborated in the workout. This allows the parties to speed up negotiations. Prepackaged bankruptcies typically last only a few months. Obviously, firms and their creditors attempt a workout first as a way of economizing on the significant costs and delays of a court-led restructuring. But also in some cases the reason for going through a prepackaged bankruptcy procedure is to attempt to waive the stay on retrieving collateral. The extent to which this can be achieved is, however, severely restricted.

Even though prepackaged bankruptcy procedures are cheaper, firms cannot avoid Chapter 11 completely. One reason is that when (junk) bonds are involved and exchange offers are necessary to restructure the debt outside Chapter 11 there is a significant risk that the restructuring fails due to insufficient tendering (and gaming by “vulture funds”) in the exchange offer. This is true despite the fact that corporate exchange offers can impose a penalty on holdouts and partially “disfigure” their claims by giving higher priority ranking to the new claims over the old ones. For example, between 1977 and 1990 only 73 exchange offers have been successful out of 156 cases of distressed bond issuers. Of those 73 at least 23 subsequently filed for Chapter 11 (Gertner and Scharfstein, 1993). Another reason why firms cannot avoid Chapter 11 entirely is that they may need to have a stay imposed to avoid a run on their assets, or they may need DIP-financing.

Workouts and prepackaged bankruptcy procedures can also be envisioned in the sovereign context even if a sovereign debt restructuring body is in place. That is, the sovereign debt restructuring body should not necessarily be seen as “crowding out” market-based collective action procedures. The two procedures may well be complementary. An important issue for an international bankruptcy institution in this respect is whether debtors and their creditors should be allowed partially or fully to opt out of a court-led international debt restructuring procedure. Opt-outs of Chapter 11 are not possible as U.S. bankruptcy law is mandatory. But in the sovereign context it is possible and perhaps desirable to allow for opt-outs.

Government Bailouts

A second relevant aspect of U.S. bankruptcy practice is the role of the government in the largest bankruptcy cases. The government is always an important party, as the corporation may owe taxes and, more importantly, the government has a stake in maintaining employment and economic activity. U.S. government authorities have intervened in some of the largest bankruptcy cases with the domestic equivalent of a rescue package. Intervention has ranged from simple guarantees on new debt to bailouts as in the case of Lockheed, Chrysler, and more recently the major U.S. airlines and insurance companies.

To the extent that government bailouts can be granted even in the absence of any institutionalized mechanism and even in a framework where bankruptcy is the only institutional mechanism in place for the resolution of corporate debt crises, it is to be expected that similar bailouts will continue to be relevant in the sovereign context even when a formal international bankruptcy mechanism is in place. These bailouts serve if anything a more important economic role in overcoming liquidity crises and contagion than domestic ones.

Jurisdiction shopping

A third relevant aspect of U.S. corporate bankruptcy practice is “jurisdiction shopping.” A corporation can file for bankruptcy in any state where it has its “principal place of business.” This means that a firm that is incorporated in Delaware but has its corporate headquarters in another state has a choice between filing in that state or in Delaware. Since most large U.S. corporations are incorporated in Delaware they can engage in “jurisdiction shopping.” Over the 1990s Delaware has emerged as the jurisdiction of choice for Chapter 11 filings. So much so, that the Delaware bankruptcy court is increasingly unable to handle all cases. Why did Delaware emerge as the preferred jurisdiction and why does jurisdiction shopping matter at all given that bankruptcy is a federal law? The answer seems to be that the Delaware court is more efficient and expeditious. In particular, the Delaware court has shown that it is able to respond quickly to firms’ needs for debtor-in-possession financing and promptly grant “first-day-orders” when new cash is needed to preserve the going-concern value of the firm.

A form of jurisdiction shopping could be contemplated under an international bankruptcy procedure as well. To encourage courts to respond to the needs of the contracting parties, it may be desirable to allow for jurisdiction shopping, say, by letting existing bankruptcy courts also handle cases of sovereign defaults. Thus, sovereigns and/or their creditors could file for bankruptcy protection in U.S. bankruptcy courts (say, New York or Delaware), in U.K. courts, or in an ad hoc sovereign debt restructuring body akin to the International Center for Settlement of Investment Disputes (as Schwarcz, 2000 has proposed). As always, an important issue with jurisdiction shopping concerns the bias that competition between jurisdictions might introduce. Is competition going to lead to a “race to the bottom” or a “race to the top”? Put differently, if the right to file for bankruptcy protection resides with the debtor would jurisdiction shopping lead to excessively debtor-friendly courts? The allegation has been made about Delaware courts in their treatment of corporate bankruptcies, although there is no clear supporting evidence.7 If such a bias were seen to be a problem, a solution might be to let creditors and/or the IMF object to a particular choice of court.

III. The U.K. and Other Countries’ Bankruptcy Law

A. U.K. Bankruptcy Reorganization Practice

As the brief overview of U.S. corporate bankruptcy law highlights, bankruptcy in the United States is governed to a large extent by statute. In contrast, the U.K. bankruptcy procedures are based much more on a contractual approach.

The dominant bankruptcy-reorganization procedure in the United Kingdom is the “administrative receivership”. It is a creditor-controlled procedure with little if any court involvement. Under this procedure, a creditor who is generally the holder of a “floating charge” appoints a receiver. The floating charge is a form of collateral on the firm’s residual assets that have not otherwise been collateralized. Generally the floating charge holder is a large bank (or group of banks). The absence of court involvement and the power of the floating charge holder, both make for a much quicker debt restructuring process than in the United States. This is the main selling point of the U.K. approach to corporate bankruptcy. Another benefit that is sometimes mentioned is that debtors in the United Kingdom have stronger incentives to try to avoid bankruptcy.

Thus, a fundamental difference between U.K. and U.S. practice is that in bankruptcy the debtor remains in possession in the United States but not in the United Kingdom. This key feature of U.K. bankruptcy practice, unfortunately, makes it much less relevant for the SDRM, as it is inconceivable to think of an SDRM without a debtor-in-possession.

In the United Kingdom, the receiver can choose to liquidate or sell the firm immediately or to run and reorganize it before selling it later. The receiver’s main concern is to get maximum repayment on the debt of the floating charge holder. Any proceeds beyond the floating charge holder’s claims are distributed according to absolute priority to the other claimholders. If new financing is required to preserve the going-concern value of the firm while it is in receivership, then this funding generally comes from the floating charge holder. Finally, following the sale of the firm as a going concern the old management team often ends up back in charge as it often emerges as the highest bidder. U.K. law provides some safeguards to protect other creditors against the receiver’s actions. Despite these safeguards, however, receivership is widely seen as favoring the interests of the floating charge holder at the expense of other creditors and also at the expense of preserving the going-concern value of the firm. In short, U.K. receivership is biased in favor of liquidation. In contrast, U.S. bankruptcy is probably biased in favor of continuation.

Most small and medium-sized firms have a floating charge holder and are therefore restructured under administrative receivership when they are financially distressed. The largest firms, on the other hand, may not always have a floating charge holder. When that is the case, another bankruptcy reorganization procedure applies: administration. This procedure was introduced in 1985 as an alternative to straight liquidation, which was the almost inevitable outcome in the absence of a floating charge, and is seen as the U.K. counterpart of Chapter 11. The key difference between receivership and administration is that the bankruptcy (or companies) court, rather than the main creditor, gets to appoint the administrator (or trustee). The debtor or a creditor can file for the appointment of an administrator. The administrator has the right to impose a stay on debt collection actions and debt repayments for a limited period. He can also raise new funds to preserve the going-concern value of the firm. A majority of creditors can object and vote to end administration at any time. A majority of creditors must also approve the reorganization plan proposed by the administrator.

Administration is rarely used even for large firms. A key reason is that even under administration management is removed following a bankruptcy filing. The main creditors, bank syndicates, often prefer to avoid this outcome and therefore attempt to restructure the firm’s debts in an informal workout. This practice is known as the “London approach.” It is confined to bank lenders, conducted in secret negotiations, and generally involves both a temporary stay on bank debt payments and a pro rata haircut. Unfortunately, given the informal and secret nature of these reorganizations little systematic information is available about this practice (see, however, Armour and Deakin, 2001 and Armour, Cheffins and Skeel, 2002).

B. Corporate Bankruptcy in Other EU Countries

Much of continental European practice provides for the appointment of a receiver or trustee—as in the United Kingdom—to run or oversee the management of the firm in bankruptcy. However, relative to the United Kingdom, courts generally tend to be much more involved in supervising the firm. There is no analogue of administrative receivership and the powers of the floating charge holder in continental Europe. Bankruptcy institutions tend to be more debtorfriendly and reorganizations generally take longer to resolve in continental Europe than in the United Kingdom.

In France the main stated objective in bankruptcy is to save the firm and to maintain employment. When a firm files for bankruptcy protection a heavily administrative procedure is initiated: (i) the court appoints a receiver to oversee the firm’s management; (ii) the receiver represents the state and not the creditors; (iii) at the end of a period of observation that can last up to a year and a half it is the receiver who decides whether to let the firm continue or to liquidate its assets; and (iv) it is also up the receiver to propose a reorganization plan. In short, the French model is, if anything, even more interventionist than Chapter X under the Chandler act. It leaves little room for negotiation between the debtor and creditors.

In Germany bankruptcy reorganization is also based on an administrative model, the court appointing an administrator to oversee or run the firm. Until 1999, when a new insolvency code was enacted, it was very rare for a firm to file for bankruptcy reorganization. There are two important reasons for this. First, the old reorganization law did not provide a stay on payments or on collection of secured assets. Second, the law required repayment of at least 35 percent of unsecured creditor claims. The new law aims to make reorganization a more attractive option and introduces a stay. It also removes the 35 percent minimum requirement on unsecured debt.

In sum, continental European practice allows for more administrative involvement, as under Chapter X of the Chandler Act. Generally the debtor does not have as many protections as under current bankruptcy law in the United States. For that reason continental European practice is perhaps less relevant for sovereign debt restructuring.

C. The Japanese Model

There are three different bankruptcy reorganization procedures in Japan: a formal corporate reorganization law and two so-called composition procedures. The most commonly used by far are the composition procedures. Interestingly, the reason why firms avoid the corporate reorganization institution is again that this procedure involves a heavy-handed administrative intervention as under Chapter X of the Chandler act, on which it is based. In contrast the composition procedures appear to be very flexible and if anything are even more tilted towards continuation than Chapter 11 in the United States. A unique feature of composition is that when a firm files for protection it must submit a reorganization (or composition) plan at the time of filing. Also, composition does not provide an automatic stay. Instead, firms themselves must apply for a stay and a suspension of debt payments. The stay cannot be all-encompassing and does not cover secured creditors’ attempts to foreclose on their secured assets.

When a firm applies for composition, the court appoints both an examiner to determine whether the firm is eligible for composition and a trustee to oversee the firm’s operations while the firm is in reorganization. Both a majority of unsecured creditors and a supermajority of claims at least equal to three-quarters of all unsecured claims must approve the reorganization plan.

Another unique feature of Japanese corporate debt restructuring is the practice by a substantial number of financially distressed firms filing for composition (up to 39 percent according to Eisenberg and Tagashira, 1994) of seeking the protection of a stay only as a temporary relief from creditors and pulling out before a reorganization plan gets approved. For the remainder of firms applying for composition, the rate of successful reorganizations is also significantly higher than in U.S. Chapter 11, again according to Eisenberg and Tagashira (1994). However, part of this difference may be due to the larger average size of firms applying for composition.

Perhaps one of the most relevant special features of Japanese composition for sovereign debt restructuring is the submission of a reorganization plan at the time of filing and the examination of the plan by a court-appointed examiner before protection is granted. Such a requirement for sovereigns seeking protection under a sovereign debt restructuring body could substantially accelerate renegotiation proceedings. It could also offer basic protections to creditors, with plans being rejected if they excessively favor one group of creditors at the expense of others. The IMF is in a natural position to play the role of an examiner, as it is the best repository of information about sovereign debtors and as its participation in sovereign debt restructuring is essential.

D. Corporate Bankruptcy in Other Asian Countries

The Asian crisis of 1997 has put the spotlight on bankruptcy institutions in the countries hardest hit by the crisis. These institutions generally appeared to be archaic, inefficient, and highly inadequate for handling the high volume of business failures produced by the crisis. As a result, much corporate debt restructuring has taken place outside the courts through workouts. Indonesia, Korea, and Taiwan Province of China have bankruptcy reorganization institutions that are closest to the Japanese model, with composition as the generally preferred procedure. In contrast, Malaysia and Thailand have institutions that are closer to the U.K. receivership model, with a strong bias towards liquidation and protection of secured creditors. Not surprisingly, these latter countries have been able to handle liquidations more expeditiously and have relied less on potentially inefficient and politically influenced courts (Hussain and Wihlborg, 1999).

E. Latin American Bankruptcy Procedures

Argentina has the most up-to-date bankruptcy institution of all Latin American countries, with a new—and largely untested until the current crisis—insolvency law passed in 1995. The new reorganization procedure (concurso) is a unique combination of elements from bankruptcy laws of different countries. As under Chapter 11, it leaves the debtor in possession, and provides for an exclusivity period during which the debtor can propose a reorganization plan. As under the old German law, however, the reorganization plan must compensate unsecured creditors for at least 40 percent of the face value of their debts. And, as under Japanese composition, the law requires the court to appoint a trustee to oversee the debtor’s management. Finally, unlike Chapter 11, the law grants an exclusivity period to creditor proposals after the initial exclusivity period has expired without an agreed reorganization plan.

Brazil’s and Mexico’s bankruptcy laws are much more archaic. In addition, courts are underfunded, inefficient, and sometimes corrupt. Moreover, debt collection can be undermined by weak enforcement of court rulings. Brazil’s reorganization procedure (concordata) leaves the debtor in possession and is extremely debtor friendly in practice. In contrast, Mexico’s reorganization procedure (convenio) requires the court to appoint both a trustee to take over the firm’s management and a creditors’ committee to oversee the firm’s operation while debts are restructured. The procedure is also more heavily biased in favor of protecting worker claims by extending claims beyond due wages and giving worker claims a higher priority. An interesting unique feature of Mexico’s reorganization procedure is the specification of a sliding scale for creditor approval of a reorganization plan. The bigger the “haircut” (or percentage of the face value of debt that is forgiven), the higher the supermajority required to approve the plan: for a haircut no larger than 25 percent only a simple majority is required, for a haircut between 25 percent and 55 percent a supermajority of 75 percent of claims is required (Rowat and Astigarraga, 1999).

F. Other Relevant Bankruptcy Practice

Some of the most recent institutional innovations in corporate bankruptcy have taken place in transition economies, not always to great effect. Most relevant for sovereign debt restructuring is the approach to corporate debt restructuring in Poland. The government encouraged firms to restructure their debts through out-of-court workouts, with the added incentive of diluting its tax claims if firms achieved a workable debt-for-equity swap.8 The Polish approach may be relevant for the sovereign debt restructuring to the extent that it points to a potential role for the IMF and other multilateral lenders. By offering to let go of the seniority of its claims, conditional on a successful restructuring, the IMF could potentially increase the efficiency of a sovereign debt restructuring body.

G. Taking Stock

As this brief overview of corporate bankruptcy practice around the world indicates, there are three common underlying principles and three common elements to all bankruptcy reorganization procedures. The three general principles are that bankruptcy institutions are designed to:

  1. Address a “common pool” or “run for the assets” and “race to the courthouse” problem arising when multiple creditors have conflicting claims over a distressed firm’s assets. Bankruptcy law intends to preserve the “going-concern value” of the firm and to avoid an inefficient breakup of the firm. In some countries the law’s stated objective also extends to the preservation of employment.

  2. Enforce the payment of creditors according to “absolute priority.” Claimants with a security interest or higher priority get paid first and lower priority claimants only get the residual value of the reorganized or liquidated firm. To be sure, there are deviations from the absolute priority rule but the overall stated objective remains the implementation of this rule.

  3. Mandate the cancellation of all or most unpaid debts following liquidation of the firm to allow the bankrupt firm’s owners and managers the option to have a “fresh start.”

Together with these three general principles there are also three basic elements common to virtually all bankruptcy procedures: (1) a stay on part or all debt payments and collection actions to prevent a run on the firm’s assets, (2) some form of DIP-financing authorized by the receiver or the bankruptcy court to preserve the going-concern value of the firm, and (3) if a reorganization is attempted, a debt restructuring agreement that is approved by some form of majority voting among creditors organized in different classes. The outcome of the vote is generally binding on the dissenting minority.

Beyond these common traits there are major differences, however. The various laws differ mainly in the extent of court or administrative involvement in the management of the bankrupt firm, in the balancing of creditor and debtor interests, the representation of other stakeholders interests, the limits put on debtors’ liability, and in the coverage of the debt standstill. In terms of administrative involvement, the United Kingdom and the United States stand at one extreme, with minimal court intervention in the management of the bankrupt firm while its debts are restructured. Almost everywhere else in the world there is greater court involvement and supervision. The United Kingdom also stands out in the extent of creditor protection its laws provide. Nowhere else do creditors have the powers of the “floating charge” holder in “administrative receivership.” However, Japan and until recently Germany also provide unusually strong protections to secured creditors. Countries like France and India (for large companies) provide unusually strong protections to other stakeholders, in particular to employees. Japan and until recently Germany also stand out in the limits they set on debt standstills and the exclusion of part or all secured debt from the court imposed stay on debt collection efforts. Finally, the United States stands out in the generous debtor protections granted by some states through homestead exemptions.

This great diversity in legal rules raises at least three important questions for the future orientation of an SDRM:

First, how deep should the administrative involvement in sovereign debt restructuring be? Should the decision to restructure be left to the debtor country and its creditors (as under U.S. bankruptcy law), or should a sovereign debt restructuring body have the authority to determine whether a debtor country is eligible for restructuring (as under Japanese composition law and Chapter 9 of the 1978 code in the United States)? Should the government in the debtor country remain in control (as under U.S. law), or should the international body appoint a receiver (as under French law)?

Second, how comprehensive should the debt standstill be? Should there be an automatic stay on all debt payments (as under U.S. law), possibly combined with capital controls, or should some debt payments be excluded from the standstill (as under Japanese and German law)?

Third, how much involvement of other stakeholders like labor unions, chambers of commerce or other business and social organizations should there be? Should debt restructuring be left entirely to the debtor country government and its creditors (as under U.S. and U.K. law), or should labor unions (and possibly other business and social organizations) also participate in negotiations (as under French law)?

IV. The Sovereign Debt Restructuring Problem and The Case for A Statutory Approach

The recent proposals for the creation of a sovereign debt restructuring body are based on an important analogy between corporate and sovereign debt restructuring. We shall argue in this section that this analogy is imperfect, but that despite the important differences between sovereign and corporate debt a strong case can be made for a statutory approach to sovereign debt restructuring. The question is not whether an international debt restructuring body should be set up but what the contours of a sovereign bankruptcy procedure should be. We shall argue that because of some key differences between sovereign and corporate debt a sovereign bankruptcy statute may require major adaptations of a Chapter 11-type procedure.

A. The Imperfect Analogy Between Corporate and Sovereign Debt Restructuring

As appealing as the analogy with corporate bankruptcy may be there are important differences between sovereigns and corporations, which limit the applicability of institutions designed for corporations. Some specific features of sovereign lending tend to tilt the balance more in favor of the debtor than in the corporate context; others tend to tilt the balance in favor of creditors.

A key difference between sovereign debt restructuring and corporate bankruptcy is that sovereign states never get liquidated. Some commentators have argued that, as a result, sovereign debtors are too strong and that, unlike for corporations, it is the creditors that are in need of protection and not the debtor. The fact that a sovereign debtor cannot be liquidated may indeed have important consequences for how debt-restructuring negotiations play out. What is the sanction for failing to reach an agreement that induces debtor and creditors to strike a deal in a timely fashion? There may be a greater risk of prolonged negotiations. Deadline effects like the expiration of an exclusivity period for proposing a restructuring plan may be weaker than in the corporate context if there is no threat of liquidation disciplining the parties.

Another important difference is that in contrast to corporations, where creditors are often able to impose a change in management team, sovereign governments cannot be replaced by their creditors following a default. It is also more difficult for sovereign lenders to impose conditions on governments that would make a debt restructuring more acceptable.

A further difference is that sovereigns do not have to protect themselves against creditors racing to grab the debtor’s assets. In this respect the going-concern value of a sovereign is more easily preserved and a stay on debt collection actions is less critical.

On the other hand, sovereigns are more vulnerable than corporations to capital flight and exchange rate and banking crises. It is these risks, more than creditors’ debt collection efforts that threaten the “going-concern value” of the sovereign. These threats are not adequately addressed by a suspension of debt payments and collection actions. They may require instead a much wider standstill (including on bank deposits) and the implementation of capital controls. The analogy with corporate bankruptcy here has more to do with debtor-in-possession financing than with debt standstills.

The preservation of a firm’s going-concern value usually requires debtor-in-possession financing (DIP-financing) to make sure that trade creditors remain on board. Such financing may be even more critical in the sovereign context for bank depositors as well as other “trade creditors” to continue to have confidence in the country. Standstills on withdrawals from bank deposits undermine the sovereign’s entire payment system and destroy the sovereign’s “going-concern value” when they are extended beyond a very short time period. Therefore, to the extent possible, such drastic measures should not be taken and instead the viability of the banking system should be maintained through some form of debtor-in-possession financing.

To the extent that such debtor-in-possession financing is not forthcoming, a sovereign debt crisis coupled with an exchange and banking crisis can result in substantially higher costs than a situation of financial distress for a corporation. It is the fear of these costs that induces sovereigns to meet their debt obligations and to postpone debt-restructuring efforts.

Finally, another critical difference between sovereign and corporate debt restructuring is the presence of the IMF and other multilateral lending institutions. The IMF has a special role to play that has no counterpart in the corporate context, except perhaps the relationship banking role played by a corporation’s main bank. The IMF can avoid or postpone debt restructuring by lending. It has information about the sovereign’s creditworthiness that no one else has. It can also impose conditions on sovereign borrowers that no other institution or lender can. The IMF thus, has a critical role to play in any sovereign debt restructuring. This role could range from provision of debtor-in-possession financing, to overseeing and intervening in the debt restructuring process in the capacity of a trustee or administrator, or to playing the role of the judge granting protection and supervising bankruptcy court proceedings.

B. Contractual and Market Solutions

Many participants in sovereign bond markets and some issuing country governments are wary of a heavy-handed statutory approach to sovereign debt restructuring and have expressed a strong preference for “contractual” or “market-based” solutions to sovereign debt restructuring. A common denominator among all opponents to a statutory approach is the belief that an administrative intervention in sovereign debt restructuring is bound to be misguided, will undermine sovereign bond markets, and will raise the cost of borrowing. Beyond this common ideology, however, there are radically different perspectives, which are not easily reconciled. Many advocates of a contractual approach argue that it already delivers most of the benefits a statutory approach could bring (Roubini, 2002a, b). Other supporters argue the exact opposite: under laissez-faire sovereign debt restructuring is likely to be highly inefficient ex post, but this is desirable from an ex ante perspective since it imposes discipline on the debtor country (Dooley, 2000).

Two contractual procedures are typically cited, exchange offers and collective action clauses (CAC). As we have already pointed out, an exchange offer is the only way of restructuring a sovereign bond issued under New York law. Indeed, under the 1939 Trust Indenture Act any renegotiation of the payment terms of a bond issued in the United States requires unanimous agreement of all bondholders.9 This is virtually impossible to achieve and therefore the only way forward is to offer bondholders to exchange their old bonds for a new claim with longer maturity and/or higher priority. The exchange offer is conditional on a sufficient fraction of old bonds being tendered. The incentive to tender for any bondholder ensues from the risk of seeing the old claims “disfigured” by the new higher priority claims should the tender offer succeed.10

As Detragiache and Garella (1996) and Roubini (2002a, b) among others have pointed out, exchange offers can be an efficient way of restructuring a distressed bond issue. Detragiache and Garella show that the ex post efficient outcome can be achieved under full information when all bondholders have sufficiently large holdings for their individual participation in the exchange to be able to affect the likelihood of success of the tender offer. Roubini argues that exchange offers combining “carrots” and “sticks” can successfully address the holdout problem, and have proved to be successful in several instances.

However, Detragiache and Garella also point out that if bondholders are not all equally well informed about the long-run value of the old and new claims then the ex post efficient outcome can no longer be achieved. Furthermore, they provide an example where a better outcome can be achieved with an alternative majority voting renegotiation procedure on whether to waive a seniority covenant of the bond issue or not. Similarly, Gertner and Scharfstein (1991) show that when ownership of a bond issue is widely dispersed then an exchange offer can only be successful if the new claim has higher priority than the old claim. Otherwise, a best response for an individual small bondholder is to just hold on to the old claim. Thus, a potential problem for sovereign debt restructuring through exchange offers is that the offer may fail if the new claim is perceived to have equal priority to the old one. These observations raise the difficult issue of how absolute priority can be enforced for sovereign debt payments. In the absence of such enforcement how effective can exchange offers ultimately be?

The recent successful completion of several exchange offers in the Ukraine, Pakistan, and Ecuador is often cited as providing support for such a “market-based” approach to sovereign debt restructuring (see Eichengreen and Ruhl, 2000 and Roubini, 2002). It is argued that these successful exchange offers provide compelling evidence that the problem of coordinating dispersed bondholders and reducing the risk of free-riding can be adequately handled with exchange offers. However, two notes of caution are in order.

First, these exchange offers took place in small countries with extremely simple debt structures. For these countries the IMF’s new stated policy, that it would not consider any bailouts without some concessions from other creditors and bondholders, may well have been credible. The ownership structure of these bonds may also have facilitated the exchange offer.

Second, in another instance in Russia in July 1998 a proposed exchange offer engineered by Goldman Sachs seriously went awry. Goldman Sachs failed to win sufficient support for its exchange offer of high yield GKOs for longer maturity lower yield Eurobonds. The main apparent reason was that too many bondholders were banking on an IMF bailout and therefore saw no need to exchange their claims. The failure of this particular exchange offer was one factor that precipitated the Russian crisis (see Blustein, 2001). This latter episode provides a particularly vivid illustration of the potential critical weakness of market-based solutions like exchange offers. These solutions may not work well if the market expects that the debt crisis will be resolved with an IMF-led bailout. When exchange offers fail, or are postponed because they are expected to fail, the delayed debt restructuring increases the pressure for a bailout. As a result, market-based procedures may do little to reduce incentives to provide bailouts and thus may do little to reduce moral hazard in lending.

The other widely favored contractual solution involves collective action clauses (CACs). These clauses allow for the renegotiation of the terms of a bond issue using a majority voting procedure among bondholders and currently apply mainly to bonds issued in London under English law. Typically, corporate or sovereign bonds issued in London provide for the appointment of a trustee monitoring the borrower and representing the interests of bondholders. The trustee and/or a group of bondholders can convene a meeting of bondholders and propose changes to the bond agreement. If at least 75 percent of the bonds in an issue approves the changes, the renegotiation offer is accepted and binding on any dissenting minority (see Buchheit, Gulati, and Mody, 2002 for a description of a typical CAC).

Advocates of CACs argue that this form of creditor-initiated renegotiation sanctioned by majority voting provides most of the benefits a sovereign debt-restructuring forum would offer (see Eichengreen, 2002; Taylor, 2002a, b; and Buchheit, Gulati, and Mody, 2002). They argue that CACs forestall a race to the courthouse and that they solve free rider and holdout problems, since the majority decision is binding on all holders of the bond issue. To be sure, even the advocates of CACs recognize that a fully operational contractual procedure is not yet in place and that difficult open issues remain. They acknowledge that many existing sovereign bonds do not contain any such clauses and that important coordination problems remain across different bond issues and other sovereign debt claims. They believe, however, that the path of least resistance to reform is one of encouraging the widespread adoption of CACs and otherwise to leave sovereign debt restructuring to market participants. One often invoked important “tactical” benefit of this course of action is that it would not require major changes to existing international treaties or to the IMF’s Articles of Agreement and that therefore this path to reform would meet minimal political resistance.

Interestingly, CACs could have been used in Pakistan and the Ukraine, but an exchange offer was preferred. Several reasons have been given for this choice, among which a concern that a meeting of bondholders might trigger acceleration of other outstanding debt and the perceived extreme caution of bond trustees, who are seen to be reluctant to propose reductions in bond repayments for fear of litigation. Whatever the case may be, this reluctance in making use of CACs suggests that this renegotiation procedure may not be as straightforward as it is made out to be.11

To recapitulate, one group of advocates of a contractual or market solution, which reflects to a large extent the opinions of market participants, argue that collective action problems among creditors are exaggerated, and to the extent that they do exist they can be adequately addressed with CACs or exchange offers.

These proponents are united in their opposition to a statutory approach with other commentators, who hold, however, the strictly opposite view on the efficiency of a laissez-faire approach to debt restructuring (possibly augmented by collective action clauses). The latter commentators argue that widely held sovereign bonds are actually quite difficult to restructure, but that the ex post inefficiency of debt restructuring under laissez-faire is desirable from an ex ante perspective. The high cost of debt restructuring under laissez-faire is a necessary disciplining device to induce the sovereign debtor to repay its debts (see e.g., Dooley, 2000). This view is close to the position of some issuing country governments, who are mainly concerned that a procedure which facilitates debt restructuring would undermine the commitment value provided by the current high restructuring costs and would therefore lead to higher ex ante costs of borrowing.12

The observation that higher ex post debt renegotiation costs impose greater discipline on the debtor and hence reduce the cost of borrowing is certainly well taken. It is debatable, however, whether sovereign debtors would pick an ex ante efficient level of debt renegotiation costs that optimally trades off ex ante costs of borrowing and ex post costs of financial distress. There are at least five reasons why sovereign debtors might choose to overborrow by committing to excessively high ex post debt restructuring costs: first, administrations that build up debt are typically no longer around when the time comes to repay and may not fully internalize the future costs of financial distress; second, governments have private goals besides the welfare of their country and may choose to borrow mainly for their own benefit even if this can impose heavy debt restructuring costs on their country; third, creditworthy countries may accept excessively high debt restructuring costs as a way of signaling their creditworthiness (Eichengreen, 2002); fourth, creditors may insist on high restructuring costs mainly as a way of guaranteeing a form of priority against other lenders; fifth, excessively high restructuring costs may be chosen as a way of encouraging ex post bailouts (see Bolton and Jeanne, 2002 and Bulow, 2002). For all these reasons it is likely that equilibrium sovereign debt structures under laissez-faire would be excessively hard to restructure ex post.

As Haldane et al. (2002) have shown, an important concern with sovereign debt structures that are difficult to renegotiate ex post is that they are more prone to debt rollover crises. If individual debt holders anticipate that there will be a long and costly restructuring phase following default they will be less willing to roll over their debts. This is likely to result in both more frequent and more severe liquidity crises.

Another important caveat to the view that sovereign debt structures are deliberately designed to be hard to renegotiate as a way of disciplining borrowing governments is that debt-restructuring costs alone are unlikely to be the main disciplining device anyway. The loss of reputation and the ensuing increase in the cost of future borrowing when debt repayments must be rescheduled are perhaps even more important deterrents. In addition, a sovereign debt crisis is likely to trigger a wider financial crisis, social dislocation, and a run on the country’s currency, which the sovereign borrower will be even more eager to avoid.

Finally, an unspoken concern that is also likely to unite all opponents to a sovereign debt restructuring body, whatever their views on the ex post efficiency of a contractual approach, is that by facilitating debt write-downs the sovereign debt restructuring body would also make it easier to hold back on IMF-led bailouts. Both borrowing governments and lenders benefit from potential bailout subsidies. In this respect it is not surprising that they would be opposed to a policy aimed at reducing this subsidy.

C. The Case for Statutory Approach

Even if the analogy between corporate and sovereign debt restructuring is imperfect, it still provides valuable insights into the question of the advantages of a statutory approach relative to exchange offers or collective action clauses. As the overview of corporate bankruptcy practice in this paper underscores, there could be at least four fundamental aspects of corporate reorganization procedures that may not be adequately addressed by contractual or market approaches to sovereign debt restructuring.

First, as already alluded to, the contractual or market approaches that are currently envisioned do not guarantee a comprehensive restructuring agreement of all bond issues and other debt claims. Collective action clauses only apply bond issue by bond issue and it is unclear how a contractual debt restructuring process would play out in the presence of multiple issues and other debt claims. Similarly, the recent successful debt exchanges only involved one or two separate bond issues. Whether they would work successfully in the presence of a dozen or more issues is unclear and untested. One obvious concern with CACs when there are multiple debt claims is that the holders of any one bond issue may want to wait to see how restructuring plays out for other issues. There may be holdout behavior across issues and inefficient delays may ensue. Under the proposed contractual approaches there is no analog for the creditor super-committee found in equity receiverships when multiple corporate bond issues had to be renegotiated and there is no way of compelling the renegotiation of all claims at once unless they are contractually connected through cross-default and debt acceleration clauses.13 As Roubini (2002a) aptly points out, it is obviously always possible to broaden CACs to ensure proper coordination across debt claims “via the use of super-clauses, arbitration and other meta-clauses, [but] such a beefed-up contractual approach ends up becoming very close to a creditor-centered statutory one.”

Second, contractual and market approaches do not guarantee the enforcement of absolute priority. However, the enforcement of some form of priority is necessary to both implement an ex ante efficient level of borrowing and to facilitate exchange offers. Most sovereign bond issues contain pari passu clauses, guaranteeing equal priority with other debt claims. But equal priority does not eliminate a borrower’s incentives to dilute existing debt claims, by adding new debt to the outstanding stock. To eliminate or further reduce such dilution, new debt issues would have to have lower priority than older debt claims in the event of a default. Similarly, the success of an exchange offer may depend critically on a higher priority ranking of the new claim over the old one. By leaving debt restructuring to the market there is no clear way of guaranteeing that agreed upon priority arrangements will actually be enforced when a sovereign is financially distressed.14 As some prominent legal scholars on corporate bankruptcy have argued, the primary and only function of corporate bankruptcy procedure ought to be the enforcement of absolute priority (see Baird, 1986 and Bebchuk, 1988). By their logic an important benefit of a statutory approach to sovereign debt restructuring would be possible enforcement of absolute priority and in this way the systematic introduction and enforcement of different priority rankings of sovereign debts.

Third, and perhaps most importantly, contractual and market approaches do not explicitly deal with DIP-financing. As the discussion of corporate bankruptcy practice in Section II has emphasized, an essential part of corporate reorganization practice is the possibility of obtaining DIP-financing to preserve the going-concern value of the firm. Indeed, Ayotte and Skeel (2002) have recently shown that Delaware bankruptcy courts’ efficiency in handling requests for DIP-financing is a primary reason why many large distressed companies choose to file there. As we have already hinted at, DIP-financing may be even more critical for sovereigns since they may be vulnerable to capital flight and exchange rate crises. In many ways, some form of DIP-financing already takes place in the shape of IMF lending into arrears. However, under the current regime such lending is not directly tied to a debt restructuring agreement between the sovereign and its creditors. As a result, rather than facilitating a debt restructuring agreement such lending encourages creditors to delay completion of negotiations in the hope of benefiting from an IMF transfer (see Bulow and Rogoff, 1990). A major reason for the introduction of a sovereign debt restructuring forum is precisely to correct the incentives for lender moral hazard generated by IMF lending into arrears under the status quo.

Finally, a sovereign’s renegotiated debt obligations under contractual and market restructuring procedures may still leave the country with an excessively high debt burden. Creditors will trade off the efficiency benefits of debt reductions against the costs in terms of reduced expected debt repayments and therefore a debt restructuring procedure which is too creditor friendly may result in inefficiently low debt forgiveness (see Helpman, 1989 and Ayotte, 2002). A statutory approach can be balanced to be more debtor-friendly should such an inefficiency be a concern.

A statutory debt restructuring procedure can also be designed to address other issues of potential concern such as the risk of protracted negotiations under the contractual approach and uncoordinated legal actions by groups of creditors. By specifying a strict timetable for making restructuring proposals a restructuring forum may be able to speed up negotiations. Similarly, by imposing a stay on litigation while negotiations are ongoing the statutory approach can easily limit legal uncertainty and reduce the deadweight cost of litigation. Perhaps most importantly, under a statutory regime a unified case law will develop over time, which will gradually refine and improve the process of debt renegotiation.

D. A New Role for IMF

In some early proposals on a statutory regime for sovereign debt restructuring it was suggested that the IMF might be a possible forum. But, as many critics of the idea have argued, one major drawback in having the IMF guide and supervise debt restructuring proposals is that it would not be a fully independent body and it would also be in the awkward position of being both judge and interested party. For these reasons it would be desirable to set up a separate independent forum. Several suggestions have been made (see Schwarcz, 2000 and Rogoff and Zettelmeyer, 2002) for other candidate forums, but the analogy with U.S. corporate reorganization practice highlights the potential benefits to be obtained from allowing some scope for jurisdiction shopping. To the extent that some bankruptcy courts have developed expertise in handling large corporate reorganization cases it may be worth exploring whether these courts could also supervise sovereign debt restructuring cases.

Even if an independent debt-restructuring forum is created, the statutory approach will inevitably change the IMF’s role in sovereign debt crises. It may further enhance the IMF’s role in providing DIP-financing or coordinating the supply of DIP-financing by other creditors. It may also enhance the IMF’s role in certifying the sustainability of a country’s debt. If a sovereign debt restructuring forum significantly reduces the costs of renegotiating the debt and if the high restructuring costs under laissez-faire partly serve the purpose of imposing discipline on the debtor then it may be desirable from an ex ante efficiency perspective to exclude from the forum solvent countries that are unwilling to repay all of their debts and are attempting to default strategically (see Bolton and Jeanne, 2002). The IMF has the greatest expertise available to make a determination on the solvency of a debtor and the filing for bankruptcy protection may be made conditional on IMF certification. Similarly, the IMF may be well placed to monitor and enforce a restructuring agreement by making future access to IMF lending conditional on diligent enforcement of the restructuring agreement. One major potential drawback, however, in relying on the IMF to make such an assessment is that decisions to exclude a country from the SDRM or not, just as current decisions to extend an IMF program, are in danger of becoming highly politicized.

E. Political Constraints and Possible Transitory Regimes

One important reason why contractual and market solutions are favored by many commentators is the perception that a more ambitious statutory approach will not be available in the near future as there are likely to be too many political obstacles to its implementation. Concretely, the imposition of standstills and the use of majority voting for debt restructuring proposals, which could be binding on a dissenting minority, require an amendment to the IMF’s Articles of Agreement. Such an amendment, in turn, requires the approval of 85 percent of member country votes. Since the United States has more than 15 percent of the votes, this means that the U.S. Congress must approve the change and, so far at least, there is no indication that there would be sufficient political support for the SDRM in Congress. It is already apparent that Wall Street will resist a move towards a statutory approach to sovereign debt restructuring even if it has the strong backing of the U.S. Treasury.15 Therefore, if the better and more ambitious reform is likely to be bogged down in political wrangling why not explore less daring but more feasible options?

Again, the analogy with the political history of corporate bankruptcy in the United States may provide useful insights in building political support for the implementation of a statutory sovereign debt restructuring procedure. As we have highlighted, the ultimate success of the 1898 Bankruptcy Act is due to a considerable part to a willingness by Republican legislators to reach a compromise with the pro-debtor opposition to a federal bankruptcy law. One critical part of the compromise was to leave authority to the states to enact more or less generous (homestead) exemptions. Similarly, one can envision reaching a compromise with debtor nations that have expressed strong reservations about the proposed new statutory regime by letting them opt out of part or all of the new procedure ex ante.16 Such an opt-out clause ought to satisfy debtor nations that are skeptical of the SDRM, but may still not placate Wall Street interests. The hope, of course, is that if the new procedure produces its expected benefits then the skeptics will be gradually turned around. Some commentators favor even stronger incentives and propose that debtor countries that agree to an SDRM procedure should have easier access to IMF programs17 (see Kenen, 2001).

V. Open Issues

Even if the outlines of a statutory approach sketched in this paper are compelling many difficulties remain in defining the details of the procedure and in addressing economic issues specific to sovereign debt crises. The recent analyses and discussions of the SDRM—including this paper—have focused to a large extent on the parallel with Chapter 11 of the 1978 code and on the merits of a statutory approach. Relatively little research has yet been devoted to the many details of the actual operation of an SDRM and a number of important features of the procedure still need to be determined.

A first important feature that is not clear-cut is how bankruptcy protection will be triggered. Should a filing for protection be “voluntary,” that is entirely under the discretion of the debtor? Or should it be left to the initiative of creditors, or both? Similarly, should the granting of protection be conditional on an assessment of the sustainability of the country’s debt or not? As we have argued above, it may be desirable from an ex ante efficiency perspective to make access to an SDRM procedure conditional on the insolvency of the debtor. But a number of ex post considerations may weigh against such a move. It is well recognized that the assessment of debt sustainability is currently more an art than a science. In addition, if access to a sovereign debt-restructuring forum is conditional on being judged to be insolvent then the announcement of a filing may have dramatic effects on capital markets.

A second critical design issue is the statutory time period permitted for working out a debt restructuring. Related to this issue is the question of who can make restructuring proposals to be put up for a vote. As we have seen, large corporate reorganization cases in the United States can last anything between three to four months and over two years. One would hope that a swifter resolution could be found in the case of sovereign debt restructuring. Part of the delay in corporate reorganization cases comes from the exclusivity period of three months given to the firm to come up with a reorganization proposal to be put to a vote. A quicker resolution might be available if the exclusivity given to the debtor was lifted and if competing offers were allowed, as Aghion, Hart, and Moore (1993) have suggested. Another reason why distressed corporations remain in Chapter 11 for a long time is that the costs of staying under the court’s protection and of dragging out negotiations may not be so high if the firm is able to operate normally. Unfortunately, there may be additional factors in the sovereign context, which could contribute to further delays. The debtor country may be going through a political crisis, or decision making may be paralyzed due to upcoming elections. There may also be substantial incentives to delay an agreement if more financial support can be squeezed out of the IMF and other multilateral institutions. It is also not at all obvious that a sovereign may be able to “function normally” while it is renegotiating its debts, even if it has access to DIP-financing. It would, indeed, be especially important to come to a quick resolution if along with the standstill the SDRM also required the imposition of capital controls or even limits on bank deposit withdrawals.

This latter point raises perhaps the most important and difficult design question for the SDRM. Should a financially distressed sovereign be granted an automatic stay on all debt payments when it is granted bankruptcy protection? Should it be allowed to impose capital controls, and should it go as far as imposing a suspension of bank deposit withdrawals to avert a run on its banks? The recent crisis in Argentina has made abundantly clear that a suspension of bank deposit withdrawals is no panacea. It has seriously impaired Argentina’s payment system and has fatally damaged the reputation of Argentina’s banks. The crisis in Argentina has uncovered the Achilles heel of the policy of a temporary suspension of bank deposit withdrawals: such a policy is only likely to postpone a run on the banks to the moment when the suspension is lifted. As our overview of corporate bankruptcy reorganization around the world has brought to light, the automatic stay of a bankrupt firm is by no means a universal policy. Neither in Germany (until recently) nor in Japan is the stay on payments allencompassing. Also, an automatic stay is not necessarily what one would expect to emerge as an optimal regulatory response based on first principles. Accordingly, an automatic standstill is not as clear-cut a benefit as has generally been assumed in the policy debate on the SDRM. There may be ex ante benefits in clearly isolating some debts as outside the reach of a standstill, for example. But most importantly of all, the experience in Argentina suggests that everything should be done to avoid a temporary suspension of deposit withdrawals.

If a sovereign debt crisis runs the risk of triggering a banking crisis with a run on bank deposits then a more appropriate response may be to make sure that the central bank acts and can act as a lender of last resort. As is well recognized, the main problem with this policy prescription is that the central bank may not be in a position to act as a lender of last resort if the banking sector’s foreign currency liabilities exceed the country’s foreign currency reserves (as was the case in Argentina). One, admittedly unorthodox, way of addressing this problem, without imposing any suspension of deposit withdrawals may be to allow banks to meet depositors’ demands for withdrawal of foreign currency deposits in local currency only, but at the going market exchange rate. Such a move could preserve the credibility of a lender-of-last-resort policy unless the banking sector is clearly insolvent. It also creates incentives for depositors to keep their deposits in the bank. Indeed, if the lender-of-last-resort policy is expected to give rise to inflation then the foreign currency deposits become a form of inflation-indexed deposits, as long as depositors keep their money in the bank. Admittedly, if the banking sector is clearly insolvent (dollar liabilities exceed total asset values in dollars) then some form of “haircut” may need to be imposed on depositors and some form of suspension may be needed while the restructuring plan is put in place. However, given the cost of such a move it is important to investigate the feasibility of limited and short forms of suspensions, which would preserve the payment system. Also, unless a credible lender-of-last-resort policy is in place the banking system will remain vulnerable to a run once the temporary suspension is lifted.

Closely related to the issue of the scope of a standstill is the question of the possible role of an SDRM in an emerging market debt crisis like the Asian crisis in 1997, which is not mainly or only a sovereign debt crisis. As currently envisioned in the IMF proposals (IMF, 2001, 2002) the SDRM would apply only to a sovereign debt crisis. But if IMF bailouts are also seen as an important response to balance of payments crises involving mainly private debt, shouldn’t the reach of the SDRM extend even to such crises? And if so, how? An interesting proposal by Buiter and Sibert (1999) and Kenen (2001) based on an augmented contractual approach is to include collective action clauses allowing for an automatic rollover option of 90 days in all debt contracts, private and sovereign, to be triggered in the event of a crisis. Such a clause will undoubtedly help but is unlikely to be a sufficient response to emerging market debt crises on the scale of South Korea’s debt crisis.

VI. Conclusion

A statutory debt restructuring procedure may bring substantial benefits over a contractual or market-based approach. It may encourage financially distressed countries to file early. It may reduce the costs of restructuring complex sovereign debt structures, involving multiple bond issues, bank loans, and other liabilities. It may also speed up the restructuring process, reduce the amplitude of the economic crisis that is likely to go along with a default, and last but not least, provide elements of a “fresh start” for the distressed country.

These important benefits, unfortunately, may still pale in comparison with the economic dislocation costs that are likely to follow a default. The SDRM cannot realistically be seen as an alternative to current policy based on IMF bailouts. It is more appropriate to think of the SDRM and IMF programs as important complementary policy instruments. That is, the effectiveness of the SDRM is likely to be greater if the distressed country can also count on significant IMF programs. These programs can help the country rebound and return more quickly to its full production capacity. They also continue to have a role to play in preempting liquidity crises and in restoring confidence in volatile debt and foreign exchange markets. If the crisis following a default is likely to take the dramatic proportions seen in Thailand, Indonesia, or Argentina then the fear of default may be so great that much of the benefit of the SDRM will be wasted. Countries will continue to delay debt restructuring until it is far too late, even if they can get greater protection from creditors.

Thus, the ultimate success of the SDRM is likely to depend as much on how well the debt restructuring forum is designed as on the continuation and expansion of existing IMF policies based on bailouts. Similarly, the effectiveness of IMF programs in preventing and responding to crises will be enhanced if an SDRM is in place, which facilitates private sector involvement and makes it easier to reduce a country’s debt burden to more manageable levels.