PrintPrint CiteCite
Style: MLAAPAChicago Close


Improving the Sovereign Debt Restructuring Process: Problems in Restructuring, Proposed Solutions and a Roadmap for Reform

Authors: Brad W. Setser, Senior Fellow and Acting Director of the Maurice R. Greenberg Center for Geoeconomic Studies, and Nouriel Roubini
March 9, 2003
Council on Foreign Relations


Paper prepared for the conference on “Improving the Sovereign Debt Restructuring Process” co-hosted by the Institute for International Economics and Institut Francais des Relations Internationales, Paris, March 9, 2003. We thank Fred Bergsten, Ted Truman, Bill Cline, Anna Gelpern, and Jean Pisani-Ferry for useful comments and suggestions. The usual disclaimer applies with force: the opinions expressed in this paper are strictly those of the authors and not of any other institution or organization.

Improving the Sovereign Debt Restructuring Process

Nouriel Roubini:; Brad Setser:


The recent debate on reforming the international financial system has focused on the need to improve the sovereign debt restructuring process, and in particular on steps that could limit the risk that litigation could disrupt or delay a sovereign debt restructuring. This debate increasingly has focused on the debt restructuring process in those cases where debt reduction is needed to produce a sustainable debt profile. Less attention has been given to those cases where a sovereign lacks the reserves needed to cover its near term obligations and, absent international support, has a clear need for debt rescheduling to push out near-term maturities.

There is a strong case for seeking to make the sovereign debt restructuring process more orderly, more predictable and more rapid. There is also a strong case that steps to address collective action problem created by the threat of holdout litigation could help to improve the restructuring process, and give all parties more confidence that there is path that can lead a sovereign from the decision that a restructuring is necessary to its successful conclusion. Moreover, reasonable steps to reduce the risk of holdout litigation would not suddenly dilute the sovereign’s incentives to pay: no rational sovereign would prefer default to payment just because it can restructure with a super-majority vote.

It is therefore important to capitalize on the enormous investment that the international community already has made in thinking through the problems that arise in a sovereign debt restructuring. It would be a shame if the debate that began with the G-10’s report on Sovereign Liquidity Crises back in 1996 and that gained new prominence after the IMF proposed a new statutory bankruptcy regime did not ultimately result in the implementation of a series of concrete improvements in the sovereign debt restructuring process.

But it is also important to recognize that there are limitations to what likely can be achieved. The legal ability to bind in a minority alone will not significantly lessen the real economic costs of defaults or make it easy to turn down a country’ request for official support. The case for taking steps to limit the risks of holdouts and to clarify the debt restructuring process with a code of conduct should rest on the ability of these steps to make the restructuring process more transparent and the outcome of the restructuring somewhat easier to predict ex ante -- not on their ability to change dramatically the incentives facing either sovereign debtors or the official sector.

The first half of this paper is organized as follows: an initial section discusses the problems that arise when a sovereign needs to restructure its debts; the second section discusses the three main proposals that have been put forward to improve the sovereign debt restructuring process (clauses, codes and statutes), emphasizing the differences between various contractual and statutory proposals as well as the difference between the contractual and statutory approaches; the third section synthesizes this discussion into four broad approaches to moving forward; the fourth section briefly lays out a set of recommended policy steps; and the fifth section suggests broadening the agenda beyond its current focus on steps to facilitate the restructuring of the sovereign’s external debt in debt reduction cases. These five sections constitute a relatively short, self-contained paper that provides an overview of the issues that arise in a sovereign debt restructuring and potential solutions. The second half of the paper provides a more detailed presentation of the case for the recommended set policy steps, with analysis to help back up the judgments embodied in the proposed course of action.

Section I: Problems, Proposed Solutions and a Roadmap for Reform

A. The problems that arise in a sovereign debt restructuring

Everyone -- or at least almost everyone -- can agree that the sovereign debt restructuring process should be improved. But agreement on the need for improvement masks different conceptions of the fundamental problems in the current sovereign debt restructuring process.

The official sector looks at the sovereign debt restructuring process and sees a process that is too “disorderly” and too costly to the debtor. In part, this is a lament by those in the official sector who must inevitably make difficult choices. A restructuring -- which necessarily involves rewriting contracts to change their financial terms -- will always be more complex and difficult to organize than the most obvious policy alternative: a bailout which allows a country to meet its contractual obligations, at least for a while. But there is also a sense that there is not a good enough map to guide a sovereign from the scary decision that it needs to restructure its debts to the successful conclusion of a restructuring. The lack of a good map, in turn, is one reason why sovereigns may wait too long to initiate a debt restructuring. There is also a sense that the existing process leaves the sovereign too vulnerable to the threat of litigation -- whether immediately after default or after the conclusion of a restructuring.

Private creditors look at the existing restructuring process and tend to see a rather different set of problems. They see a process where there is no agreement -- or rules -- that outline how different creditor groups will be treated, and hence see a constant risk that a sovereign’s external creditors will be treated less well than other creditor groups. They also see a process where -- official sector complaints about litigation to the contrary -- it seems that creditors have far fewer legal rights than creditors of a bankrupt firm. There is no prospect that the creditors of a mismanaged sovereign will ever be able to change the sovereign’s “management.” They cannot even force a sovereign to put a restructuring proposal on the table within a defined time frame. They worry that any further steps to protect the sovereign from litigation would weaken the sovereign’s incentive to pay, and thus undermine the sovereign debt market. Many of the complaints of creditors also are a lament against the immutable realities of sovereign lending. A sovereign, unlike a firm, is never going to go away, and eventually should be able to pay something. But sovereignty also generates at least partial freedom from external legal authority.

Sovereign debtors look at the international financial system and see yet another set of problems. They worry that the official sector’s interest in developing a bankruptcy regime is primarily motivated by a desire to reduce the level of official support provided to debtor countries. Most want a global system that offers financial insurance against market turbulence that can result in potentially avoidable defaults far more than a global system that offers protection from litigation. Guillermo Ortiz memorably criticized the official community for spending too much time trying to build morgues rather than constructing centers for preventive care. Emerging debtors know better than most that capital flows to emerging markets have already dried up and, like creditors, they worry that proposals to introduce new legal protections for debtors would further shrink the market for international sovereign debt. At the same time, many debtors -- particularly those that have had to restructure their debts in the past -- do see advantages in a process that would allow for more restructuring decisions to be taken by majority vote.

Rather than detailing all of the specific complaints that have been made, it makes more sense to step back and consider some of the basic problems that arise in a restructuring.[1] These include:

  • A rush to exit from the sovereign’s own debt. Creditors have good reason not to roll over claims on a sovereign that they believe will need to seek a restructuring in the near future. Indeed, creditors have an incentive -- given all the inherent uncertainties associated with any debt restructuring process -- not to roll over claims on an “illiquid” but still “solvent” sovereign. Indeed, the biggest cost associated with a run is a potentially avoidable default in a liquidity crisis. But it is worth remembering that a run on the sovereign’s own debt is far less disorderly than say a run on a bank, or even a run on a fixed exchange rate. The maturity structure of the sovereign’s debt tells creditors rather precisely where they stand in the queue -- and gives some clues about how orderly the queue is likely to be. Those in the front of the queue may be able to get out before the sovereign runs out of either international reserves or loses the will to spend its remaining reserves to let those creditors at the front of the queue avoid any haircut. A larger stock of short-term debt increases the risk of a disorderly rush for the exits. Those holding long maturity claims are stuck. They can sell their claim to another creditor, but that transfers rather than extinguishes the claim. They do not benefit if a debtor spends its remaining reserves to pay short-term debts in a futile bid to avoid a restructuring. The Bank of Canada and Bank of England, among others, have noted that there is a simple solution to such a run -- suspend payments on the sovereign’s debt (or at least the sovereign’s external debt), and then renegotiate the debt’s terms. The new maturity structure stops the direct drain on the sovereign’s reserves. But the admission that the sovereign cannot pay can itself trigger other types of runs.

  • A rush to the courthouse. Suspending payments stops the drain on reserves associated with a rush to the exits, but without a formal stay on litigation or an informal agreement with its creditors, it also leaves the debtor open to risk of litigation. There is an advantage to being the first creditor to litigate -- so long as the first creditor to litigate has the ability to successfully lay claim to the sovereign’s few remaining assets. In practice, though, a “creditor grab race” has not proven to be much of a problem in the sovereign context. A bankrupt sovereign typically does not have many international assets to begin with, and the sovereign’s international reserves and diplomatic property already enjoy considerable protection from litigation. External creditors have neither the ability to lay claim on the sovereign’s domestic assets nor the ability to go to court and force the sovereign to spend less and tax more to be able to pay more on its external debt. Creditors can litigate to make it difficult for a sovereign to “selectively default” on its external commercial debt, but cannot do much more. A sovereign cannot force its external creditors to give up their legal claims without their consent. At the same time, the sovereign’s external creditors are not likely to recover much by litigating prior to a restructuring agreement, because of the considerable protection a sovereign already enjoys. Ultimately, external creditors need to agree to a restructuring to start receiving payments again.

  • Free riders or holdouts. All creditors would be better off if they all agreed to restructuring that put the debtor back on a sustainable path, but each creditor individually would be better off if it got paid in full and other creditors agreed to the restructuring. Of course, not all creditors can hold out. If too many creditors hold out, there is no deal that makes sense for the sovereign. But if the number of creditors holding out is manageable (or if the legal risk posed by holdouts is judged to be low), the sovereign may opt to go ahead with the restructuring despite some holdouts. These holdouts can then sue for full payment, and seek to convince a court that the sovereign should not be allowed to pay its new debt so long as it is not fully honoring its old contractual commitment. This may give the holdout leverage to extract a favorable settlement. But holding out and litigating is costly and potentially risky strategy. Elliot’s successful litigation hinged on an interpretation of the pari pasu clause that may, or may not, survive further court rests. Most creditors are likely to conclude that the downside of holding an illiquid claim whose value depends on the courts exceeds the potential benefits of holding out. Moreover, the ability to amend the non-financial terms of even those bonds whose financial terms cannot be amended can significantly reduce the attractiveness of holding out. But without the ability to make a restructuring plan approved by a super-majority of creditors binding on a minority, there is no way to fully protect a sovereign from the risk of holdout litigation.

  • The absence of an enforceable priority structure for the sovereign’s own debt that helps to settle questions of equity and the relative treatment of different creditor groups.[2] A bankruptcy regime typically lets creditors know ex ante how they will be treated in relation to other creditors holding similar types of claims. Typically, all unsecured creditors are offered the same basic restructuring terms, because all unsecured debt has the same legal priority. The court can refuse to sanction any restructuring proposal that does not respect the existing rules of priority. In a sovereign debt restructuring, there are some informal rules of priority that are generally followed -- it is hard for a sovereign that needs to reduce its debt to treat one bond issue better than another bond issue,[3] and the IFIs are de facto given priority relative to other external claims. But there are no rules that determine how a sovereign should treat their unsecured domestic debt vis a vis their unsecured external debt, and the existing framework for coordinating the restructuring of external private debt and external debt owed to other governments is under considerable strain. The absence of an agreed priority structure can make it difficult to reach agreement on a restructuring -- the debtor and its creditors have to agree in broad terms on how different creditor groups should be treated before they can get down to negotiating financial terms. This is in no small measure one reason why it would take time to reach agreement on a restructuring even if the sovereign did not face the risk of holdouts. But the ability of the debtor to privilege some creditors without effective protest from others creditors also has some advantages. By paying the IFIs, trade creditors and in some cases domestic debt, the sovereign can sustain access to some new financing even in the absence of any formal bankruptcy regime.

  • Policy conditionality. There is an obvious need to coordinate the steps that creditors agree to take to make it easier for the debtor to pay with the steps that the debtor agrees to take to increase its ability to pay. This too poses challenges of collective action -- after all, creditors are unlikely to have the exact same conception of what steps the debtor should take. It also can be a source of delay. Sovereign debtors in particular can have trouble adopting a coherent set of policies amid the economic and political chaos that follows a default. The current system for sovereign debt restructuring attempts to address this problem largely by linking the debtor’s program of policy reforms to new money (or the refinancing of existing exposure) from the IMF rather than by direct negotiation between the debtor and its external private creditors. External private creditors could refuse to restructure their own claims unless the debtor agrees to additional policy changes, but in practice this has rarely happened. Private creditors point out -- correctly -- that creditors play a much larger role in commercial debt reorganizations. But creditor demands for a larger role in the sovereign process are not matched by a willingness to increase their leverage by putting up new money, and there is no obvious solution to the inherent difficulties in coordinating the different demands for policy reform that come from disparate groups of private creditors.

  • Rush to default. Sovereign borrowers worry that their creditors will rush to the exits at the first hint of trouble, precipitating an avoidable default. Private creditors worry that a sovereign will fail to honor its contractual commitment to pay, preferring an opportunistic default to necessary policy adjustments. Making restructuring too easy risks making default too likely, and therefore making credit for the sovereign too scarce. This risk has to be balanced against the gains from making it easier for a debtor that finds itself in a position where it cannot pay to reach agreement with its creditors on a restructuring that allows it to resume payments.

  • Other runs. Stopping payments on the sovereign’s own debt eliminates a direct source of pressure on the sovereign’s reserves. But it may trigger other runs -- a run on the domestic banking system, a run on the currency and a withdrawal of cross-border bank credit. Domestic residents may seek to trade domestic bank deposits and other local financial assets for foreign bank accounts and foreign financial assets. However, if everyone wants out, in many cases, no one can get out. Prices go into free fall, markets break down, reserves are exhausted, banks cannot meet their commitments to return deposits on demand and both bank deposits and currency markets end up frozen -- at least temporarily.

There are obvious similarities between corporate bankruptcy and sovereign debt restructuring. But the analogy also has its limits. The absence of a formal bankruptcy process does create real complications for a sovereign debt restructuring -- for example, a sovereign that reaches agreement with a large majority of its creditors still needs to worry about a few holdouts, while a firm does not. However, the absence of bankruptcy style protection poses less of a problem than one might think. Litigation immediately after default is less of a threat, as the IMF’s most recent proposal for an SDRM has recognized. Delay (with continued protection from litigation) is costly, but it is always an alternative to a restructuring that leaves the sovereign too vulnerable to holdout litigation. The existence of the IMF, an international institution that links new money for the sovereign to policy conditionality, provides a means of providing new money in the absence of any established or enforceable system of priorities. It is also a substitute (creditors might argue an imperfect substitute) for court supervision of the debtor while the debtor is negotiating new payment terms with its creditors.

It is also important to remember that the sovereign debt restructuring process differs from corporate bankruptcy more because of intrinsic differences between a sovereign and a firm than because a sovereign lacks “bankruptcy” protection. No firm issues its own currency, or indirectly backstops the banking system.[4] Sovereign debt is a typically a far more important asset in a country’s financial system than the debt of even a very large local firm, and it plays a far more crucial role in the local financial system. There is no way to avoid the fact that a sovereign default is going to be more disruptive than the default of a firm. There may be room at the margins to improve on the existing sovereign debt restructuring process, but it is hard to change many of the basic realities that make sovereign default unpleasant.

B. The range of proposed solutions

No single proposal realistically could be expected to provide a comprehensive solution to the full range of problems that arise in a sovereign debt restructuring. Solving some problems may make other problems worse -- offering the debtor too much protection after default may dilute the debtor’s incentives to reach an agreement to resume payment, if not the debtor’s incentives to take policy steps to avoid default. There are problems, like the holdout problem, that lend themselves to a legal solution, and problems which are unlikely to be ameliorated by legal change. Making it harder to litigate against the sovereign after it concludes a successful restructuring is not likely to change the behavior of either domestic bank depositors or cross border lenders immediately after a default. A sovereign that needs to reduce the value of its debt by half is not able to credibly protect bank depositors from the risk of losses, even in the unlikely event that the banking system itself does not hold much sovereign debt. Depositors run for a good reason.

Three general proposals have been put forward to solve some of the problems that arise in a sovereign debt restructuring: the introduction of new contractual provisions into new external debt contracts; the development of a code of conduct for a sovereign (and perhaps also its creditors) to follow during a debt restructuring; and the creation of a new statutory regime to provide bankruptcy-style protection for a sovereign. There are different variants within each option -- both creditors and the official sector have put forward “contractual” proposals, and the IMF has put forward three different proposals for a statutory SDRM. The debate is not just clauses v. a statutory SDRM v. a code. It is also over what kind of clauses, what kind of SDRM and what kind of code.

Contractual proposals. All contractual proposals seek to change the restructuring process by changing the provisions found in sovereign debt contracts. Contractual change is best suited to either retarding a rush to the courthouse, or making it easier to address the holdout problems by allowing a supermajority to vote on restructuring terms. All contractual proposals would only have an impact if a large share of sovereign debt contracts contained the provisions; the existing $200 billion plus stock of external law sovereign bonds is a constraint on the ability to any proposal to change the restructuring process immediately -- for the better or for the worse.

There are four basic options that are implicitly on the table: using current New York law documentation in other jurisdictions; using current English law documentation in other jurisdictions; adopting the new clauses proposed by official groups like the G-10; and adopting the new clauses proposed by the “Group of Six” creditor organizations.[5]

  • New York law documentation. A standard New York law contract requires the unanimous consent of all creditors to change “key financial terms”, which are defined narrowly as payment dates and amounts. All other terms typically can be amended with the support of one-half or two-thirds of the outstanding bondholders. Some but by no means all New York law bonds also require that 25% of the bondholders agree before litigation can be initiated -- creating an effective litigation retardant.

  • English law documentation. A standard English law contract allows a super-majority of bondholders (typically 75%) present at a meeting that meets quorum requirements to amend all the bond’s terms, including the bond’s payment dates and amounts. Many English law bonds also have provisions that make it more difficult for an individual bondholder to initiate litigation.

  • The G-10’s draft clauses. The G-10 recommended following the English law convention and allowing a bond’s financial terms to be amended with a 75% vote. The G-10 also sought to broaden the use of trustees: a trustee provides a creditor representative that could facilitate communication between the debtor and its bondholders. Adoption of a trustee structure also would have the effect of making the decision to initiate judicial proceedings a collective decision.

  • Group of Six clauses. Private creditors have proposed allowing 85% of bondholders to amend a bond’s financial terms, so long as no more than 10% of the bondholders object. That means it effectively takes 90% of bondholders to overcome the opposition of 10%. The definition of financial terms would be expanded beyond payment dates and amounts, the remaining non-financial terms could be amended only with the support of 75% of the bondholders, and certain provisions that relate the ability of creditors to sue to collect on their bonds could not be amended at all. The debtor would have to meet additional financial disclosure requirements, and pay the expenses of a committee selected by creditors to represent their interests. In sum, provisions for amending financial terms would be tighter than the provisions now found in English law bonds, and provisions for amending non-financial terms would be tighter than the provisions now found in New York law bonds.

Deciding on the “right” set of clauses, however, is only part of the problem. The real challenge is convincing debtors to change the provisions used in their New York law bond issuance. There are a range of options, all of which are discussed in detail in the second half of the paper.

A code of conduct

A code of conduct cannot give the sovereign formal protection against litigation or the ability to restructure with a majority vote. No matter how much it might improve debtor behavior, a code cannot offer a definitive solution to the holdout problem. A code is better suited to addressing the other coordination problems that arise in a restructuring -- coordination problems that stem fundamentally from the absence of an agreed system of priorities and the challenge of linking debt restructuring to policy change. Specifically, a restructuring requires:

  • Coordinating the restructuring of an individual bond issue, i.e. limiting the risk that a minority will not go along with the restructuring of that bond’s terms.

  • Coordinating the restructuring of different bond issues -- something that is now typically done by an exchange offer.

  • Coordinating the restructuring of external bonds with other types of sovereign debt. The Paris Club’s comparability requirement is one example of mechanisms to deal with this kind of coordination.

  • Coordinating the debt restructuring with the changes in the debtor’s economic policies.

  • Creditors also need to coordinate amongst themselves if they want to respond in a unified rather than disparate way to various proposals. The holders of a single bond can organize to represent the interests of that bond, the holders of all bonds can organize to represent the common interest of all bondholders, and all creditors with exposure to the sovereign can organize to represent the common interest of all creditors.

A code conceivably could lay out a roadmap describing how a debtor and its creditors should try to coordinate the restructuring of individual debt instruments so as to produce an overall change in the sovereign’s debt structure that restores sustainability. A code could lay out a set of general principles, or it could introduce detailed procedural requirements that a debtor would need to meet to qualify for IMF lending. A code could lay out a process for restructuring a sovereign debtor’s external private debt (largely bonds), or it could lay out a broader process for restructuring the debtor’s overall debt stock and coordinating that restructuring with policy change. In sum, a code could be relatively modest, or it could seek to offer a comprehensive solution to many of the problems that arise in a debt restructuring.

No matter what the code aims to do, particular attention needs to be given to the set of incentives that will lead all parties to have an interest in abiding by a non-binding code. A code could be designed to be self-enforcing -- it could just lay out a set of minimal requirements that all parties should want to follow. Alternatively, a code could set out the requirements a debtor would need to meet to gain access to new money (presumably new money from the IMF, unless creditors are willing to commit new money to a debtor that abides by the code.)[6] It is more difficult to see how a code could also much from creditors in the absence of a mechanism that lets creditors act collectively. Even if most creditors are willing to refrain from litigation if the debtor follows the code, these creditors cannot protect the debtor from a small set of creditors intent on litigating.

Existing proposals include:

  • The Banque de France has suggested developing a code that would set out both general principles and best practices for meeting these general principles.[7] The best practices could evolve in light of experience.

  • The IIF has proposed a code which is long on requirements for the debtor and short on credible commitments by creditors. In effect, it is a very extensive code of debtor conduct to be enforced by IMF conditionality.

  • Glenn Hubbard’s call for a forum that would help a debtor contact its creditors, and help creditors organize to represent their interests in a restructuring.

  • The IMF’s lending into arrears policy can be considered a code of sorts. It links access to IMF financing to two general principles -- “good faith” and “good policies” -- and sets out some basic guidance on how “good faith” will be interpreted by the IMF.

A code of conduct potentially could help to facilitate a restructuring well before most bond contracts contain collective action clauses. Most proposals are not intended to substitute for efforts to introduce of new clauses into bond documentation.

Statutory proposals

Like contractual proposals, statutory proposals primarily aim to address the “rush to the courthouse” problem and the holdout/ free rider problem, though the creation of a new statute creates at least the potential of solving other types of problems as well.

All statutory proposals offer two basic advantages over contractual proposals. First, a statutory regime would create the capacity to override existing sovereign debt contracts, and thus to immediately allow majority voting. There is no need to wait for the existing stock of roughly $250 billion in sovereign debt governed by German and New York law to be retired. Second, a statutory regime would replace a process that requires amending the financial terms of each and every bond with a single aggregated vote by all bondholders on the debtor’s restructuring proposal. It is a lot harder to buy up a blocking position in the debtor’s entire debt stock than to buy a strong, and perhaps blocking, position in a single bond issue. Aggregated voting makes holding out much more difficult.

However, statutory proposals differ in other ways. The IMF has put forward three different proposals for addressing the “rush to the courthouse.”

  • Anne Krueger’s November 2001 speech suggested giving the IMF the ability to provide a debtor with temporary legal protection.

  • The March 2002 IMF proposal suggested allowing a super-majority of creditors to vote to determine whether or not to give the debtor legal protection. The IMF muddied up this proposal by suggesting that the IMF be able to protect the debtor while the creditor vote is being organized, but the basic logic of the proposal can best be understood as simply granting all authority to creditors.

  • In January 2003, the IMF proposed dropping a stay altogether, and relying instead on the deterrent value of the ability to bind in holdouts and perhaps other litigation retardants.[8] In its most pure form, this proposal would rely entirely on the protection national laws already provide a sovereign to deter litigation prior to a restructuring, and majority voting to avoid any holdouts.

Statutory proposals can differ in other ways as well. The scope of debt instruments that can be part of the binding vote on the restructuring proposal can be broad (including either Paris Club debt or domestic debt) or narrow (limited to external law debt held by commercial creditors). The debtor can have substantial freedom to select which particular instruments it wants to include in the restructuring plan, or it can be obligated to restructure all instruments that meet certain criteria. And finally, the protections of a statutory regime can be linked to a number of other debtor policies -- including respecting a system of priority that might be designed to offer new money greater assurance of repayment -- or the protections of a statutory regime can simply be made available to any debtor that can convince a majority of creditors to support its restructuring proposal.

Four additional points are worth emphasizing.

  • First, there is no way to force a sovereign to make use of a statutory regime. The sovereign will always have the ability to opt to restructure its debts outside the statutory regime, using the “current” non-system. At least some creditors, in contrast, can be forced into the statutory regime against their will.

  • Second, upfront legal protection after a default is not likely to radically transform the debt restructuring process, for the simple reason that sovereigns already enjoy considerable legal protection. Creditors are right -- there has not been a rush to the courthouse in recent cases. Argentina’s experience is far from over, but so far litigation by external creditors has not been a major source of difficulty.

  • Third, giving the IMF the power to impose a stay would transform the IMF’s role in the system more than it would transform the existing debt restructuring process, as the IMF would be able to link policy conditionality to legal protection as well as to the provision of new money.

  • Fourth, not all statutory proposals are the same. At one extreme, a “light” statutory regime could generate a restructuring process that closely resembles the restructuring process found in existing English law clauses, but with aggregation and without the need to wait until the existing stock of New York law and German law bonds is retired. The IMF’s most recent proposal is in this vein. At the other extreme, a statutory regime could create a restructuring process that differs radically from the process that exists now. An ambitious statutory regime could try to draw in a wide range of different debts -- including domestic debt -- into a single framework, and condition IMF sanctioned legal protection on an extensive set of policy guidelines. Indeed, the policies a debtor would need to adopt to qualify for statutory protection could be set in advance to provide more “predictability”; they would not necessarily need to be left to the discretion of IMF staff and management.

C. The broad policy choices

Clauses, a code and statutory change are not mutually exclusive options. A code of conduct could be combined with contractual change, or embedded in a statutory regime. But there are some combinations that make more sense than others. Indeed, there are really four broad options on the table.

  • Live with the status quo. There is an existing debt restructuring process. It has its flaws, but at the end of the day, the current system allowed Ecuador to restructure its New York law bonds, Ukraine to restructure its English and German law bonds and Russia to restructure over $28 billion in “London Club” loans into $21 billion of new Eurobonds. In the near term, all contractual proposals will result in a restructuring process that is far closer to the existing process than to any idealized model, given the large stock of outstanding sovereign debt that will lack any contractual innovations.

  • Majority restructuring clauses and a non-binding code. The introduction of contractual changes that allow a majority of bond holders to bind in a minority would gradually change the contractual terms in the existing debt stock. This would, over time, reduce the risks associated with holdout litigation and hopefully make it somewhat easier for debtor to reach rapid agreement on restructuring terms with its creditors. Such clauses could be supported by a code that clarifies the existing debt restructuring process without imposing major new constraints on the debtor.

  • The Group of Six’s clauses and the IIF’s code. Changes could be introduced into the existing process to give external private creditors increased leverage over a sovereign debtor. Contractual changes could limit the debtor’s ability to amend non-financial terms, set high thresholds for amending financial terms, and force the debtor to finance the creation of creditors committees. A parallel code of conduct could set out major new requirements that a debtor would need to meet to gain access to IMF financing.

  • A statutory SDRM. A new statutory regime could be created to give the debtor the advantages of a single aggregated vote on its restructuring plan. Access to this new statutory power could be linked to “a code of debtor conduct” embedded in the new statutory regime. An ambitious SDRM has the potential to radically change the current debt restructuring process.

D. Recommended next steps

Our analysis of the sources of difficulties in the current sovereign debt restructuring process lends support to the second broad option -- clauses and a non-binding code. The magnitude of the set of problems that can be solved by introducing a new legal regime is too small to justify imposing a radical new regime on reluctant creditors and debtors, with unknowable consequences. But there is scope to make improvements in the current system. And even modest improvements are unlikely to happen without impetus from the official sector.

We specifically recommend:

  • Contracts. Build on the momentum created by Mexico’s decision to introduce collective action clauses in its New York law bonds and strongly encourage other issuer to make majority amendment clauses the new norm in the U.S. market. It is important not to gain the ability to amend the key financial terms of New York law bonds at very high thresholds -- 90 to 95% -- by accepting other changes that would in practice make it easier to sue a sovereign or harder to restructure a bond (such as severe limitations on the ability of the debtor to amend non-financial terms). Change would no doubt be facilitated if U.S.-based investors were willing to accept in principle what they already buy in practice -- dollar denominated debt governed by English law with provisions that allow 75% of the outstanding principal to bind in a minority of 25%.[9] If other New York law issuers do not follow Mexico’s lead, the official should be prepared to go beyond jawboning to arm-twisting, and eventually to seek regulation or legislation that would require the use of clauses. This really does not require G-7 coordination, or changes in IMF policies. It does require a willingness on the part of U.S. authorities to impose change on the market if those countries that typically issue in dollars using New York law do not move on their own.

  • A code of conduct. Develop a non-binding code that sets out a basic process for sovereign debt restructurings. This code should contain four core elements:

    • New disclosure principles. The debtor should be expected to provide full and accurate information about its debt profile and restructuring plans to its creditors. This should include publishing a full accounting (detailed and disaggregated) of its outstanding debts soon after defaulting, and informing creditors of any significant changes to its debt stock. At the time it is ready to put forward its initial restructuring proposal, the debtor should also provide, as in the IMF’s most recent SDRM proposal, a list of claims that are being restructured through the exchange/ the initial restructuring proposal, a list of claims that are being restructured through other processes, and a list of claims that are not being restructured. It should also indicate how its overall restructuring proposal would apportion available near term cash flow across different creditor groups, and as well as how each creditor group is contributing to the creation of a viable medium term debt profile.

    • An outline for how to move from imminent default to a successful restructuring. Debtors will need to work with creditor representatives to develop a restructuring proposal. Creditors will need to organize themselves so that they can efficiently provide constructive input.[10] Debtors should not be required to reach agreement with every member of the committee before launching an exchange -- such a requirement could introduce a new source of delay into the existing process, without assuring that the exchange will attract broad based participation. Creditors will have the legal right to initiate litigation while the debtor is developing its restructuring proposal, though hopefully most creditors will refrain from litigating. Debtors are within their rights to seek to use their existing contractual powers -- the ability to amend the financial terms of English law bonds, the ability to amend through exit consents the non-financial terms of New York law bonds -- to limit the risks posed by holdouts.

    • Realistic expectations about inter-creditor equity. It would facilitate agreement if external creditors would recognize that external and domestic debt is unlikely to be restructured on the same terms. Perfect equity may not even be in the interest of external creditors: a domestic debt restructuring that triggers a bank run that could ultimately result in lower recovery levels for external creditors. At the same time, debtors should not expect that domestic debt can be entirely left out of any restructuring. In many cases, both domestic debt and external debt will need to be restructured. But domestic debt and external debt are likely to be restructured at different times, using different restructuring processes and on different terms. Absent ex ante consensus on what is a “fair” allocation between domestic and external creditors, little is gained from promising too much. A code should not do more than insist that the debtor lay out its plans for all to see and assess.

    • Clauses in new bonds. There should be an expectation that the bonds that emerge from the restructuring will contain clauses that allow the amendment of the bond’s financial terms.

  • The SDRM. Accept that there is not currently the political consensus needed to create a statutory SDRM by amending the IMF’s Articles. But it is still worth seeking broad G-7 and G-20 consensus on what any future statutory regime should aim to do. Key questions that have not been resolved include: Is there a need for a stay on litigation, or does the prospect of being bound in by a super-majority vote at the time of the restructuring suffice to restrain a litigious minority? Is there a need to give legal priority to new private money, or should the IMF/ MDBs remain the only sources of “debtor in possession” financing in the system? What role should the IMF play in determining whether a country should have access to the protections of an international bankruptcy regime? This means stepping back from the current push to reach quick agreement on a detailed, operational design.[11]

We have no illusions about what can be achieved by these proposals. A code of conduct combined with the gradual introduction of collective action clauses will not suddenly make a sovereign debt restructuring fast, painless or easy. Such changes offer no guarantees that debtors will decide to seek necessary restructurings more quickly. Indeed, it is doubtful that debtors delay right now primarily because some existing bond documentation fails to provide for a majority vote to amend key financial terms. Steps to lay out a more defined and transparent process for debt restructuring -- even if buttressed with clauses -- will not make it substantially easier for the international community to deny a country access to large scale emergency financing at early stages of a crisis. The immediate costs of initiating a restructuring will always be higher than the immediate costs of providing a country with time to see if it can avoid a restructuring with large scale financing. A more transparent restructuring process does not even guarantee more rapid agreement on restructuring terms.

These proposals will strengthen, not replace, the existing process for sovereign debt restructuring. In the near term, most bonds will lack clauses allowing the amendment of financial terms. Debtors with outstanding New York law bonds will need to rely on the second best solution of amending the bonds’ non-financial terms (through exit consents) to make holding out less attractive. Debtors also will need to continue to use multi-instrument exchange offers to coordinate the restructuring of different bond instruments. Such exchanges lack the elegance of a single aggregated vote, but do allow the debtor to know the number of holders of each bond who have accepted its proposal (and thus the size of any holdout problem) before it decides whether or not to go forward with the exchange. There are limits to what the IMF can be expected to impose on debtors through its lending into arrears policy -- the IMF must balance a series of different objectives after a default and will not focus its conditionality exclusively on enforcing an extensive code of conduct. A more modest code that focuses on making the existing process more transparent is therefore likely to be more robust than a more ambitious code that aims to do more than is possible.

E. Broadening the work program

Finally, it is important to broaden the official sector’s work program beyond the current focus on limiting the risks of litigation in debt reduction cases. There are a range of problems in the international financial architecture that fundamentally do not stem from the risk of litigation in those cases where a sovereign needs to reduce the value of its debt. There is no doubt a long list of such problems. But our short list would include:

  • How best to handle those cases where the country confronts an immediate shortage of liquidity that stems in part by fears about the country’s future solvency? The easiest answer is to provide the needed liquidity through large scale official financing and to link the financing to policy adjustments to address fears about solvency. If the country’s short-term needs for liquidity are large and the risks to its solvency are great, this implies official sector must assume large risks. There is a risk that the upfront liquidity supplied by the official sector may not be matched by the sustained adjustment needed for solvency. There is also the risk that the official sector will base its financing on overly optimistic assumptions about future market access, and the agreed combination of financing and adjustment will not result in private financing on a scale that allows quick repayment of the official sector or on terms that are consistent with the country’s ultimate solvency.[12] Since there will be cases where the risk of lending to a potentially insolvent country will preclude the provision of sufficient official financing to address the country’s immediate liquidity problems, it would be worthwhile to work seriously to develop a broader set of approaches for those cases that fall in the messy middle between pure illiquidity and pure insolvency. The option of using of more limited amounts of official financing to support countries that are pursuing targeted debt restructurings that aim to extend maturities at a reasonable cost should be put squarely on the table.[13]

  • How best to handle non-sovereign balance of payment crises? The current focus on sovereign debt restructuring is hard to square with the difficulties that Asian emerging markets experienced with non-sovereign debt. The sovereign’s external debt usually has a longer maturity structure than the private sector’s external debt, in part because emerging market sovereigns typically raise external funds by issuing long-term bonds while banks and firms often rely more on short-term cross-border bank lending. The fall in cross border bank exposure in Korea, Turkey -- and most likely in Brazil last year— was faster than the fall in the stock of external bonds outstanding.[14]

  • Are there ways to restructure the sovereign’s debt without triggering a banking crisis? And, when a restructuring of the sovereign’s liabilities also requires a restructuring of the banking sector’s liabilities, what is the best way of allocate unavoidable losses among the bank’s equity investors, the bank’s local depositors and external banks with cross border exposure? Domestic banks, not external bondholders, are the largest creditors of most of the world’s most indebted emerging market economies. In some cases, it may be possible to insulate the banking system from the impact of a sovereign default and restructuring. But it is hard to see how the debt relief some countries would need if they had to seek a restructuring could come exclusively from external bondholders. Standard bank restructuring techniques offer little useful advice -- standard banking crisis management is all about avoiding the need to restructure bank liabilities by using the strength of the sovereign’s balance sheet to protect depositors from losses, not how to pass the losses on the banking system’s sovereign lending onto bank depositors.

  • Are there creative ways to limit the risk that a sovereign debt crisis will also turn into a currency crisis, and the risk that a currency crisis will trigger a sovereign debt crisis? The decision to seek a restructuring can trigger a run on the currency and a rush to sell all liquid domestic financial assets. The large real depreciation in turn impacts on the balance sheet of local residents who have borrowed in foreign currency. A currency crisis -- particularly if there is severe overshooting -- can also trigger a sovereign debt crisis if the sovereign has large foreign currency debts. Options for limiting this risk include a greater willingness to provide new official financing in the context of a restructuring, and a greater willingness to sanction targeted exchange and currency controls -- including selective standstills on some types of external payments

  • How to unwind large scale IMF lending in those cases where the country simply not in a position to be able to repay the IMF as quickly as promised? After default, the speed with which the IMF reduces its exposure will have a major impact on the cash available to be “spent” on the resumption of payments on private debt, or other “goods.”

Section II. An Agenda for Reform

This section of the paper provides a more complete rationale for the proposals put forward in the first half of the paper.

The analysis is informed by the description of the problems that arise in a debt restructuring that laid out at the beginning of the paper. It is also informed by a sense of what is achievable. There is agreement on the need to improve the sovereign debt restructuring process -- and consensus at least within the official sector that steps to limit the risk of holdout litigation could play a role in catalyzing that improvement. But there is little prospect of developing an international consensus on many other major issues in the near term.

  • There is not agreement on when a sovereign debt restructuring, or indeed any form of debt restructuring, is the right response to a balance of payments crisis. Some believe debt restructuring should be reserved for only those cases where debt is unsustainable, others believe that some form of restructuring is an option when financing needs are large but the overall debt is sustainable (i.e. there is a liquidity problem). It is easier to agree on how a country should restructure its debt than to agree on when a sovereign should restructure its debt.

  • There is not agreement on the right level of IMF assistance. The G-7 has consistently expressed its desire for a tougher, more restrictive access policy. The G-7, and others, also have consistently supported giving exceptional access to most large emerging market economies and a few small emerging economies as well. Exceptional is the norm for major emerging economies in trouble. But there is no desire to harmonize rhetoric with reality -- adjusting the rhetoric to match the reality of large packages has its costs; cutting back the level of access to match the rhetoric also has its costs. It is easier to live with inconsistency.

  • There is not agreement on the need to act now to create a SDRM. No matter how clever the design proposed by the IMF, the broad-based political will needed to impose an unwanted statutory regime on issuers and investors does not currently exist. This all may change if Argentina experiences difficulties with holdouts -- after all, a complex multi-instrument restructuring is precisely the kind of case where a SDRM should be most useful. But there is no willingness to fight a preemptive war to “liberate” Latin America -- and perhaps others -- from the threat of rogue creditors.

There is still scope to move forward, however, if one accepts that the impact of any change is likely to be rather modest in the near term. The dual-track approach should not mean more analytic work on both the statutory and contractual approach approaches and concrete action on neither. Contractual changes should be introduced to make it harder for individual creditors to holdout and litigate after a restructuring deal has been completed. A code could try to set out a framework that makes the sovereign debt restructuring process more transparent. U.S. Under Secretary Taylor was right last April: the time for action is now.

The contractual approach

The bottom line: Market participants currently do not see holdout litigation as a major threat to their interests. If the official sector believes that it is important for systemic reasons to change documentation to make holdout litigation more difficult and market participants do not share a similar concern, the official sector cannot realistically expect such change to emerge spontaneously from the market. Either sovereign debtors will have to insist on such provisions in their new issues, or official action will be needed to achieve the official sector’s goals. Mexico’s decision to take the lead and to issue a New York law bond with provisions that allow amendment of its financial terms should be followed by a serious effort to convince other debtors to follow Mexico’s example. If this does not work, and additional arm twisting also fails, a serious effort to change SEC regulation is needed.

The challenge is actually quite simple: changing market practice in bonds governed by New York law.

Bonds governed by English law typically allow a qualified majority to amend key financial terms. One major emerging market economy (Russia) traditionally has used English law for its dollar-denominated bonds, and a second major issuer (Mexico) just announced that it would introduce English style provisions into its new New York law bonds. A wide range of issuers -- including a few Latin countries that were major issuers until a few years ago -- use English law for their Euro-denominated issues.[15]

English law documentation is also accepted by investors in the secondary market -- consider the popularity of Russia’s Eurobonds. Indeed, any U.S. fund manager that avoided those dollar-denominated bonds that have clauses in 2001 and 2002 would almost certainly be out of job by now. In the market for dollar bonds, going “underweight” bonds with clauses meant going underweight star performers Russia and Ukraine and going overweight Argentina and Brazil.[16] It remains to be seen how the market will react to Mexico’s initiative.

However, investor groups have shown far more enthusiasm for contractual changes that would make bonds more difficult to restructure than for changes that would make it easier to avoid holdout litigation.[17] Issuers that traditionally have not used English law also have been reluctant to change their issuance pattern. Prior to Mexico’s decision, those emerging markets that typically issue dollar bonds governed by New York law were reluctant to change the documentation they use in New York or to shift their dollar issuance to London -- presumably because of fears that any change would be perceived as signaling a reduced commitment to pay.[18]

Movement on clauses ultimately requires two things: agreement on a set of changes to the documentation “boilerplate” that is now used in most New York law bonds, and willingness on the part of issuers to insist -- despite the resistance that will come from some in the market -- that the new documentation be used in their New York law bonds. Both the G-10 and a group of six organizations representing private sector investors have been working on a set of draft clauses. But the challenge of deciding on a reasonable set of clauses is ultimately not that complex a problem.

  • There is little need to deviate from the template used in English law bonds. Bonds governed by English law are widely accepted by the market. English law bonds make up 20% of the EMBI and a higher percentage of the less important Euro denominated indexes. There is little reason to believe that moving to English style documentation would kill the market, or lead flows to dry up. No one seriously doubts Russia’s market access right now -- and Russia traditionally has used English law for its dollar bonds.

  • There is no evidence that the use of English governing law provisions changes an issuer’s incentives to pay, or that issuers discriminate against bonds that contain clauses in a multi-instrument restructuring. It is not necessary to offset the introduction of majority amendment provisions with other provisions to strengthen the debtor’s incentive to pay, because there is no evidence that English style clauses dilute a debtor’s incentive to pay. Rather, the evidence to date strongly suggests that governing law does not drive decisions about which bonds to restructure. Ecuador defaulted on its New York law Brady/ Eurobonds. Russia opted to exclude its English law Eurobonds from its restructuring. Argentina’s bonds governed by New York law have not been treated better -- or worse -- than Argentina’s English law bonds. Ukraine offered English law bonds with clauses and German law bonds without clauses the same basic restructuring terms -- it exchanged old bonds for new bonds at par, with cash settlement of accrued interest.

  • There is room to make some small changes to English law documentation. Restructuring a bond’s financial terms with a vote of 75% of the bond’s outstanding face, not 75% of those who show up at a meeting, provides investors with a degree of protection without significantly compromising a debtor’s ability to make use of these provisions. There also are the advantages to being able to amend the bonds with a written vote; it is cumbersome to have to call a formal meeting after a successful exchange offer.

  • There is a case for excluding bonds held directly by government/ central bank from the vote. It is also worth thinking through those cases where the domestic banking system holds large quantities of the sovereign’s international bonds (Lebanon, Turkey). If bonds held directly by the central bank and the government are excluded, it also makes sense to exclude those bonds held by banks that have been formally intervened and taken over by the government. But bonds held by regulated domestic banks that have not been closed down should be allowed to vote. The additional protection excluding these bonds would provide international investors is not worth the risk to the domestic financial system that would be created if a minority of international investors could determine the restructuring terms of bonds held primarily by the domestic bank system.

  • It is vital for debtors that a majority of creditors be able to rescind the acceleration of a bond, and it helps if exit consents can amend the bond’s non-financial terms to gut the ability of holdouts to reaccelerate the bond. This increases the cost and reduces the gains from holdout litigation.[19]

  • Poison pill clauses that allow amendment of a bond’s financial terms at high thresholds (90 to 95%) in exchange for a host of changes that make it more difficult to amend non-financial terms and otherwise make restructurings more difficult [20] will not achieve the official sector’s core objective of making it easier for the debtor to avoid holdout litigation. Certain creditor concerns can be accommodated, but the overall package should clearly make it easier than it is now, not harder, for the debtor to deal with holdouts.

  • 90% majority action (or 85% unless 10% objects, which effectively is 90%) sets too low a threshold for a typical emerging market bond. Many new bonds are issued in denominations of $500 million, and relatively few new issues are larger than $1 billion. With 90% majority action, a creditor needs to purchase $50 million face to block any amendment of a $500 million bond. If the bond trades at 20 cents on the dollar, that only requires a $10 million investment. Elliot, it is worth recalling, spent around $11.4 million to buy around $20 million (principal value) in debt, and ultimately settled for around $58 million (principal, interest arrears, and legal fees).[21] Consequently, 90% majority action does not offer the debtor much practical protection against smart, motivated holdouts -- particularly for relatively small bond issues. In most circumstances, debtors would be able to defend themselves against holdouts more effectively with the current New York law documentation, which defines non-financial terms broadly and gives the debtor the right to amend non-financial terms with the support of 2/3s of the bondholders.[22]

  • However, there is room to make it more difficult to make use of provisions to amend non-financial terms in the context of a broader package that clearly give the debtor a real prospect of avoid

More on This Topic

Analysis Brief

Bailouts Revisited

Author: Lee Hudson Teslik

The U.S. Treasury's refusal to save Lehman Brothers suggests its willingness to assume risk to protect private institutions may be at an end....