New Approaches to Sovereign Debt Restructuring: An Update on Our Thinking, Anne Krueger, First Deputy Managing Director, IMF

April 1, 2002

Anne Krueger
First Deputy Managing Director
International Monetary Fund
Conference on "Sovereign Debt Workouts: Hopes and Hazards"
Institute for International Economics
Washington DC, April 1, 2002

1. Introduction

Thank you, Fred. And good evening ladies and gentlemen.

It is a great pleasure to join you here this evening and especially to be speaking in the IIE's elegant new home. For 20 years the institute has been undertaking important and interesting work on a wide variety of international economic issues, combining intellectual rigor and policy relevance. I hope that your new surroundings are proving suitably inspirational and that we can expect more of the same in the future.

It is four months now since I made the case for a new approach to sovereign debt restructuring in a speech to the National Economists' Club. I said then that there were many difficult questions that had to be tackled before we would have a detailed blueprint. We are some way further down that road now, thanks in no small part to the many useful inputs we have received from academics, the private creditor community, non-governmental organizations, and our member countries. They have enriched our understanding both of the weaknesses in the current system and the possible solutions. Doubtless your deliberations tomorrow will provide us with further food for thought.

To that end, I thought it would be useful to set the stage for your discussions by explaining briefly tonight how our thinking has progressed since last November.

I will have three key points to make:

  • First, I believe from the reactions we have received over the past four months that the need for better incentives to ensure the orderly and timely restructuring of unsustainable sovereign debts is now widely accepted.

  • Second, more ambitious use of collective action clauses could contribute significantly to more efficient sovereign debt restructuring. But a purely contractual approach cannot resolve all the weaknesses of the current system.

  • Third, this implies that any new framework to encourage more orderly and timely debt restructuring is likely to require a statutory basis. But this need not mean a significant extension of the IMF's legal authority. Where possible, we should give control over the major decisions in the restructuring process to the debtor and a super-majority of creditors, not to the Fund.

To expand on these points, let me begin by outlining the case for reform and the core features that any new approach to sovereign debt restructuring would have to possess. I will then turn to the roles of the Fund and private creditors in the restructuring process, the question that has dominated discussion of the proposal to date. Finally, I will discuss some of the key questions that remain outstanding, before concluding.

2. The Case for Reform

Over the last two to three decades, greater integration of capital markets and a shift from syndicated bank loans to traded securities have both had a profound impact on the way that emerging market countries finance themselves. Sovereign borrowers increasingly issue debt in a range of legal jurisdictions, using a variety of instruments, to a diverse and diffuse group of creditors. Different creditors have different time horizons for their investments, which means they respond in different ways to economic shocks that affect a borrowing country's ability to service its debt.

These developments have enhanced the efficiency of international capital markets. They have broadened the investor base from which emerging market sovereigns can seek finance and helped investors diversify risk. But, at the same time, an increasingly diverse and diffuse creditor community also creates problems of coordination and collective action when a sovereign's scheduled debt service exceeds its ability to repay and a rescheduling or restructuring of its debt becomes necessary. It also hampers restructuring by exacerbating concerns about inter-creditor equity and by making it more difficult to establish a collaborative relationship between debtor and creditors.

During the debt crisis of the 1980s, collective action problems were limited by the presence of a relatively small number of large creditors, by contractual provisions in syndicated loans that deterred litigation, by the desire of banks to maintain good relations with the debtor to secure future business, and, on occasion, by moral suasion from supervisory authorities. Restructuring took time, but was in most cases orderly.

The move from commercial bank loans to bond issuance in the 1980s and 1990s has made creditor coordination much more cumbersome. This in turn has made it more difficult for all concerned to predict how the restructuring process will unfold. Many creditors now have no ongoing business relationship with the debtor to protect. Their interests are more diverse. They are less subject to moral suasion. And some now specialize in buying distressed debt cheaply and suing for full payment. Typically, a comprehensive sovereign restructuring may require coordination across many bond issues, as well as syndicated loans and trade financing. The task is further complicated by the repackaging of creditor claims, for example through mutual funds that separate the lender of record from the end-investor who holds the economic interest.

A sovereign borrower with an unsustainable debt burden and a diffuse group of creditors may find it very difficult to get them to agree collectively to a restructuring that reduces the net present value of its obligations to a manageable level. Even if the restructuring would be in the interests of creditors as a group, some may consider that their interests would best be served by trying to "free-ride" in the hope that ultimately they can secure repayment in line with their original contracts. In addition, credit derivatives may provide investors with incentives to hold out in the hope of forcing a default, thereby triggering a repayment under the terms of the derivative contract.

Prior to a default, individual creditors have an incentive to hold out in the hope that restructuring by others will allow the debtor to continue to pay the free-riders. Once a default has occurred, collective action issues may still be a problem, especially if a large reduction in the net present value of the debt is required.

Some creditors may be tempted to sue, although litigation has never been a particularly attractive route for most creditors seeking repayment by a sovereign - in part because of difficulty identifying assets that could realistically be seized to enforce a judgment. As the historian Max Winkler observed in 1933:

"In the case of a private default the lender can follow up the defaulter to his very fount and origin, and discover for himself his prospect of repayment. When a government defaults, the creditor must seek his way through myriad miles of tape of all colors, must track and backtrack across a road obscured by the prints of a thousand red herrings, before he can even come to the surface of the facts".

But although litigation remains costly and difficult, new legal strategies have been developed in recent years that have made the outcome of post-default restructuring more uncertain. For example, recent legal action against Peru shows that holdouts can try to extract full payment from the sovereign by threatening to interrupt payments on the restructured debt, rather than by trying to seize its assets. This possibility may make potentially cooperative creditors more reluctant to participate in a restructuring. Currently there is pending litigation in another case, regarding the Democratic Republic of Congo, that raises somewhat similar issues. But as it is still before the courts, I do not want to comment further tonight.

Having initially focused on coordination problems and the threat of litigation, I have been struck - after many discussions with the private creditor community - by the importance of two other factors that hamper necessary restructuring. First, concerns about inter-creditor equity in case debtors attempt to pay some favored creditors ahead of others or to subordinate one class of creditors to another. Second, the diversity of the creditor community encourages debtors to opt for take-it-or-leave exchange offers rather than engaging in a collaborative dialogue with all creditors. Any new approach should address these barriers to orderly restructuring too.

The bottom line is that far-reaching developments in capital markets over the last two or three decades have not been matched by the development of an orderly, predictable framework for creditor coordination, in which the roles of the debtor, the creditors and the international community are clearly spelt out. This is more than a matter of academic interest. It imposes significant costs on all the parties involved.

Like a patient with toothache avoiding a trip to the dentist, a debtor country will all too often delay a necessary restructuring until the last possible moment, draining its reserves and increasing the eventual cost of restoring sustainability. Creditors suffer too, as the fear that some may be unfairly favored in a disorderly workout depresses the value of claims on the secondary market and, at worst, may block agreement on a necessary restructuring. All this can leave the international community with the unpalatable choice of accepting a disruptive and potentially contagious unilateral default, or bailing out private creditors and thereby contributing to moral hazard.

Our goal therefore should be the creation of better incentives to encourage the orderly and timely restructuring of unsustainable sovereign debts, while protecting asset values and creditors' rights. As well as reducing the costs of disorderly workouts, a more predictable framework would help private investors and lenders distinguish between good and bad risks. By reducing moral hazard and making borrowing costs better reflect true risks, this should help countries with good policies attract capital more cheaply, and should help prevent countries with weak policies from building up excessive debts that might leave them vulnerable to potential crises.

All in all, this would result in a better allocation of global capital and make the international financial system stronger, more efficient, and more stable. In so doing, it would complement the valuable reforms that have already been undertaken in crisis prevention and crisis management in response to the market turmoil of the late 1990s.

Some commentators fear that alleviating the collective action problem will make default an easy way out. But the prospect of economic dislocation, political upheaval, and possible long-term loss of access to international capital markets will still make countries loath to default on their debt service obligations in all but the most extreme circumstances. We would also seek to structure any new framework so that countries only have an incentive to use it if their debts are truly unsustainable and the Fund is therefore been compelled to stop providing financial support until there has been a restructuring. As a result, it seems hard to argue that facilitating orderly debt workouts would significantly weaken the credit culture or create moral hazard.

3. Core Features of a Sovereign Debt Restructuring Mechanism

From what we have heard over the last four months, this diagnosis seems to be shared quite widely among informed observers of the international financial system - certainly among our member countries, but also in the private financial community. Of course, it is one thing to agree on the diagnosis. It is another to agree on a cure.

If we are to create a better framework for the restructuring of unsustainable sovereign debt, the central feature would have to be a mechanism enabling a super-majority of creditors - across the broad range of credit instruments - to make the terms of a restructuring binding on the rest. From the point of view of the creditors, this would ensure that any forbearance exercised by the majority would not be abused by free riders. From the point of view of the debtor, it would make an early agreement more likely and eliminate the threat of disruptive litigation thereafter.

Hopefully, this provision would help restructuring take place before the debtor had defaulted on the original claim. But in case this proved not to be possible, the majority restructuring provision would require the support of three other features:

  • First, the debtor would need protection from legal action after the suspension of payments and while negotiations were taking place. The majority restructuring provision might in itself make legal action more likely during this period, as it could not be resorted to once the restructuring was agreed. A stay on legal action would be of fixed duration, but potentially renewable.

  • Second, creditors would need to be given adequate assurances that their interests were being protected during the period of the stay. Without them, there is a serious danger that the framework would unduly deter capital flows to emerging markets. Two sets of assurances would be needed. First, the sovereign would be required not to make payments to nonpriority creditors, thereby avoiding the dissipation of resources that could be used to service the claims of creditors as a whole. Second, the debtor would have to conduct its economic policies in a way that would help put the country back on the road to growth and viability. Implementation of an IMF-supported program would be one way to provide these assurances. Creditors would have an interest not only in monetary, fiscal and exchange rate policies, but also in bank restructuring, the integrity of the domestic payments system, the operation of the domestic bankruptcy regime, and the nature of any exchange and capital controls.

  • Third, there would need to be a guarantee that any fresh financing provided by private creditors after the introduction of the stay would not be involved in the restructuring. The provision of new money would aid the debtor in a number of ways. It could help limit economic dislocation, provide trade credit, finance payments to priority creditors, and provide resources for a return to generalized debt servicing. In the current environment, individual creditors have no incentive to provide new money in such circumstances. Not only could it be caught up in the restructuring, but any benefits of a return to generalized debt servicing would have to be shared among all creditors as a group.

To make these features operational, the framework would also require independent arrangements for the verification of creditors' claims, the resolution of disputes, and the supervision of voting. This would help reassure investors that they need not worry about the potential for fraud, for example if a country were deliberately to issue debt to friendly creditors in sufficient quantities to give them a super-majority to impose a big haircut on all creditors.

Taken together, these elements would provide an environment within which timely and orderly restructuring of unsustainable sovereign debt could take place. Most importantly, it would address the collective action problems that currently make the cost of restructuring excessive for debtors and creditors alike.

Use of the mechanism would be for the debtor to request, not for the Fund to impose. If a debtor can reach agreement with creditors of its own accord, then it would be under no pressure to resort to the mechanism. Indeed, the existence of a predictable framework should in itself help catalyze agreement without the need for formal activation. This is what happens in well-designed domestic bankruptcy regimes. Most restructurings take place "in the shadow of the law" rather than in court.

4. The Role of the Fund and Private Creditors

Now let me turn to the question of who would make the key decisions under the framework and what the legal authority for those decisions would be.

If the problem is essentially one of collective action, it seems logical that the features I have just described should be activated through decisions taken jointly by the debtor and a super-majority of its creditors. Once taken, they would be made binding on the entire creditor body. In principle, this would apply not only to decisions regarding the terms of the restructuring, but also to the stay on legal enforcement and the offer of seniority for new financing. Under this approach, the Fund would not be empowered to make decisions that would undermine the enforcement of creditor rights. Rather, as I will explain shortly, it would rely on its existing financial powers to create the right incentives for debtors and creditors to use the mechanism appropriately.

How would decisions taken by the debtor and the super-majority of creditors be made binding on all creditors?

One approach would be to focus on more ambitious use of the collective action clauses found in some sovereign bond contracts, notably those issued under English law. After all, these clauses already typically embody two key features of the framework we have proposed. First, a provision allowing a super-majority of creditors to block legal action by a minority to force repayment. And, second, a provision allowing a super-majority to bind the minority into the terms of a restructuring.

But exclusive reliance on collective action clauses raises two difficulties:

  • First, it would be difficult to establish such a framework. Even if one could ensure that collective action clauses would be included in all new debt instruments - which seems unlikely given recent experience - one would still be faced with the large stock of outstanding bonds (many with long maturities) that do not contain such clauses.

  • Second, and perhaps more importantly, even if these clauses existed in all bonds, they would not provide a comprehensive and durable solution to the collective action problem. These clauses traditionally only bind holders of the bond issue in question. Typically, a country with an unsustainable debt burden will require a comprehensive restructuring across a broad range of indebtedness, potentially including different bonds issued under different jurisdictions, bank loans, trade credits, and some official claims.

To address this problem one could develop a "super collective action clause" that would provide for restructuring of a given instrument on the basis of an affirmative vote by a super-majority of all creditors, not just those of the instrument in question.

But there are a number of difficulties with this approach:

  • First, it would be hard to persuade debtors and creditors to include super collective action clauses in all debt instruments. Indeed, it is hard enough to get them to include traditional collective action clauses. Certain creditor groups would be even more reluctant to agree voluntarily to have their claims aggregated for voting purposes with all other present or future creditors.

  • Second, emerging market countries that borrow heavily typically issue bonds in a variety of different legal jurisdictions with different laws. Even if all debt instruments contained identical restructuring language - which would be difficult to achieve - there would be no guarantee that they would be interpreted and applied in a uniform way.

  • Third, it may not be possible to guarantee the integrity of the voting procedure if we were to rely entirely on a contractual approach. There is no international equivalent of the bankruptcy court found in a domestic insolvency regime.

  • Fourth, it is not clear that super collective action clauses would be consistent with the existing legislation of all the Fund's members. In some jurisdictions, it would appear that the absence of collective action clauses arises, in part, from the lack of a clear statutory basis that allows for the rights of minority of creditors to to be modified witout their consent. This problem would be amplified if different types of claims were aggregated.

Collective action clauses can make a useful contribution to the resolution of debt problems, especially in cases of illiquidity where a smoothing-out of the debt service profile is required rather than a reduction in the net present value of the sovereign's overall obligations. The international community has been urging emerging market countries to adopt collective action clauses for the past five years, with very limited success. Doubtless the task would be easier if more large industrial countries joined the UK and Canada in starting to include them in their own foreign currency denominated sovereign bonds.

But while the wider use of collective action clauses - in either their traditional or an enhanced form - would help resolve some debt problems, it would only take us part of the way. As a result, it would seem essential to give the new approach a statutory underpinning. This would enable claims to be aggregated for voting purposes. But aggregation would not result in the equalization of all claims for restructuring purposes. Safeguards would be needed to protect the seniority of certain claims.

Achieving a statutory basis through a universal treaty obligation, rather than via piecemeal amendments to domestic legislation, would have a number of advantages:

  • First, it would prevent creditors from shopping around for jurisdictions in which they could enforce their legal claims through the courts.

  • Second, it would ensure uniformity of text and allow a single institution to be given the authority to ensure uniformity of interpretation.

  • Third, it would avoid a free-rider problem, in which countries would be reluctant to introduce legislation until other countries had done so too.

  • Fourth, it would facilitate the creation of a single international judicial entity that could arbitrate disputes and oversee voting.

A treaty obligation could be established by amending the IMF's Articles of Agreement. This would require acceptance by three-fifths of our members, carrying 85 percent of the total voting power. This means that although the approach would rightly require very broad support in the international community to be implemented, we could achieve universality without the need for unanimity.

But the following point should be emphasized: while an amendment to the Fund's Articles would be used as the tool to give the mechanism legal force, it would not entail a significant transfer of legal authority to the institution. Under the approach that I have in mind, the essential decision-making power would be vested in the debtor and a super-majority of its creditors - not the Fund. This is consistent with the decision making process that is included in collective action clauses.

Let us examine how this would work, bearing in mind the four key decisions that have to be taken during the course of the restructuring process: the initial activation of the stay on creditor legal action, the extension of the stay (if necessary), the provision of new financing during the stay, and the approval of the final restructuring.

  • First, the initial activation of the stay. This is the most difficult decision to leave to a creditor majority, because it might take three months or so after a debtor requested a stay before it was possible to verify the creditor claims and arrange a vote. The delay could be reduced by establishing a standing organization to register claims and facilitate organization of creditors if a restructuring proved necessary. Reduced delay would make a useful contribution to prudent debt management in its own right. The debtor might not be too vulnerable during the period before a vote, because it would take some time for potential dissident creditors to pursue their legal claim for repayment. And, if necessary, capital controls could be imposed temporarily to prevent capital flight. If it were felt essential for the stay to have legal force from day one, the IMF could have the power to approve it for a limited period, say 90 days. A less attractive option would be to give the debtor the right to declare a legally-binding stay unilaterally, but this would clearly be open to abuse.

  • Second, the maintenance of the stay. If the stay came to an end without agreement on a restructuring, a super- majority of creditors could vote to extend it for further negotiations. Two important judgments are involved here: whether the debtor is negotiating in good faith and protecting creditor interests; and whether it is adopting appropriate economic policies. The creditors can judge the first for themselves. On the second, they will no doubt be guided by whether the Fund is supporting the program. Conceivably, creditors might withhold an extension in a deliberate attempt to extract more financing from the Fund or excessive adjustment from the debtor. Under these circumstances, the Fund's policy on lending into arrears would allow it to support the member's program even though it had interrupted payments to its private creditors.

  • Third, the encouragement of new private financing. A super-majority of creditors could be given the power to subordinate the claims of all private creditors to claims arising from financing provided after the stay took effect.

  • Fourth, approval of the final restructuring agreement. Again, this decision could be left to the debtor and a super-majority of creditors. But the decision would clearly be influenced by the Fund's assessment of whether the terms of the agreement adequately reduce the debt burden to a sustainable level. If they did not, then the Fund would be compelled to withhold further financing, creating pressure for another restructuring further down the road.

5. Outstanding Questions

Whichever approach - or combination of approaches - we adopt in improving the sovereign debt restructuring framework, there are difficult practical and analytical questions that we need to do more work on. Let me briefly mention five of them:

  • First, should we include debt owed to the official sector in the restructuring process? Some commentators have argued that debt owed to the Fund should be included. But we are not a commercial organization seeking profitable lending opportunities. We lend at precisely the point at which other creditors are reluctant to do so - and at rates well below those that would be charged by the private sector. In so doing, we help countries avoid disorderly adjustment and policies that would harm themselves, private creditors, and other countries. Putting outstanding loans to the Fund together with commercial claims in a workout would fundamentally undermine our capacity to play that vital role in future. Members of the private creditor community have frequently pointed out that the case for excluding debt owed to other sovereigns is less clear cut. Typically, these claims began life as export credit guarantees, which were priced with the risk of default in mind. If Paris Club claims were to be included in the restructuring process, the framework would have to be constructed to pay due regard to the different nature of these claims. We will be discussing how to proceed here with the Paris Club and the private creditor community.

  • Second, how should we treat sovereign debt owed to domestic creditors? In some circumstances, it may be necessary to restructure domestic debt if the overall burden is to be reduced to a sustainable level. Nonresident investors may only be willing to provide substantial debt reduction if they know that domestic creditors are shouldering a fair share of the burden too. But these judgments would have to be made on a case-by-case basis, taking careful account of the different nature of these claims, as well as the possible impact on the domestic banking system and capital market.

  • Third, how can we promote a collaborative interaction between debtors and creditors in developing restructuring proposals? As I mentioned earlier, creditors have expressed concern that they are too often confronted with take-it-or-leave-it exchange offers rather than engaged in genuine dialogue. We need to consider how best to strike a balance here, and whether there might be some role for an updated version of the creditor committees that were so important in the 1980s. Improving the collaborative nature of the process would help make emerging market debt more attractive as an asset class.

  • Fourth, what role should exchange controls play? Temporary controls may be necessary if a sovereign default threatens capital flight, undermining the country's ability to return to generalized debt servicing. But the advantages of controls have to be weighed against the risk that they might broaden a sovereign crisis to potentially solvent private firms. In principle, the mechanism could be designed to provide limited legal protection to otherwise viable domestic firms that might be prevented from servicing their debts solely by the controls. But this raises a number of thorny issues - not least how to distinguish between viable and nonviable firms. In the end, it may be necessary to limit the mechanism to sovereign debt, although this would still make it possible to use controls to limit capital flight. Timing is another question: if controls are to be used, should they be implemented at the outset or only when resources have been exhausted? In any event, controls should be accompanied by policies that would allow them to be lifted as soon as possible. Countries should not be encouraged to leave them in place longer than they are needed.

  • Fifth, what arrangements should be made to adjudicate disputes and oversee voting? The Executive Board could not play this role - it would be hard to avoid the perception that its decisions would be guided by the IMF's interests as a creditor, or by the debtor country's membership of the Fund. We would need an independent judicial organ, insulated from the Board and from IMF staff and management. One could imagine judges being appointed for limited periods and selected from a list drawn up by a qualified and independent panel.

6. Conclusion

From what I have said this evening, I hope you will agree that we have come a long way since November - although we still have some way to go. The development of our ideas has benefited enormously from the comments and suggestions we have received from outside the Fund and we are looking forward to a continued dialogue as we move forward. I am glad to say that our Executive Board, which discussed the issue last month, shares our diagnosis of the problem and is keen for us to continue working on the various approaches that could contribute to a solution.

Examining the pros and cons of the contractual approach and the possible extension of the Fund's legal authority that we considered last November has led us to a richer understanding of the issues involved and of the challenges we have to overcome.

I believe that using a statutory approach to empower a debtor and a super-majority of creditors could combine the strengths of both approaches, while addressing the weaknesses of each. Combined with a further effort to encourage the use of collective action clauses, this offers us a real hope of helping countries resolve unsustainable sovereign debt problems in a timely and orderly way, to the benefit of debtors, creditors, and the international community.

Lord Jenkins, the Chancellor of Oxford University, said of his recent biography of Winston Churchill: "This is a long book, but not - I hope - a monstrous book." Well, this has been a long speech, but not - I hope - a monstrous speech. The subject is certainly a complex and multi-faceted one. But I believe that tackling it is vital to the future strength and stability of the international financial system. Your deliberations tomorrow will contribute to that goal. I look forward to hearing your conclusions.


Public Affairs    Media Relations
E-mail: E-mail:
Fax: 202-623-6278 Phone: 202-623-7100