The Need to Improve the Resolution of Financial Crises: An Emerging Consensus? Address by Anne Krueger, First Deputy Managing Director, IMF

March 27, 2003

The Need to Improve the Resolution of Financial Crises:
An Emerging Consensus?

Address by Anne Krueger, First Deputy Managing Director
International Monetary Fund
To the Harvard University Business School's Finance Club
Boston, March 27, 2003


1. It is a great pleasure to be here this afternoon at the invitation of the Harvard Business School Finance Club. I hope that I can in some small way help fulfill one element of the mission of the Finance Club, by providing members with information about one strand of the work of the International Monetary Fund, surely one of the world's premier financial institutions.1 It remains to be seen whether we can also fulfill another element of your mission this afternoon-"to provide unique and exciting social opportunities." But I am by nature an optimist!

2. The past year has witnessed a vigorous and constructive debate regarding the need to improve arrangements for the resolving financial crises, and in particular the tools for restructuring sovereign debt. To be sure, this remains a controversial topic. But the debate has served to help define the issues, and to build understanding on possible ways to strengthen the international financial system. In particular, how to address the hopefully rare cases in which sovereign debtors and creditors must confront debt burdens that have become unsustainable. At times, however, the intensity of the discourse may have tended to mask the extent to which there has indeed been a convergence of views regarding the nature of the problem, and the desirability of taking actions to strengthen the system.

3. Bertrand Russell is reported to have said that the impact of education is to raise confusion to a higher level. On that basis of that definition, I think that we can safely say that the debate has been educational!

4. Today I would like to step back from the fray and look at the convergence of views on the diagnosis of the problem, and on some of the factors that will have to be part of the solution.

Importance of policies

5. A first point on which there is general agreement concerns the importance of efforts to reduce vulnerabilities to crises. And the debate about how to improve mechanisms for sovereign debt restructurings must never be allowed to divert attention from the importance of prevention. Substantial progress has been made in recent years. Moves toward more flexible exchange rate regimes, strengthening of domestic financial systems, and rebuilding official reserves have contributed to making economies more robust and less vulnerable to crises. We should not forget that the crises in Asia, Russia and Brazil of the late 1990s were typically associated with pegged exchange rates, a mix of monetary and fiscal policies that attracted short-term capital hoping to benefit from high domestic interest rates, and generally inadequate banking supervision.

6. For our part, we have put enormous emphasis on prevention. In particular, we have worked on strengthening macroeconomic policy frameworks, assessing and managing vulnerabilities, strengthening surveillance over financial systems, and improving debt management. Moreover, with the aim of improving the environment for private sector decision taking, we have promoted transparency, and have disseminated—and encouraged adherence to—standards and codes. Of course, we must avoid complacency, and must recognize the strengths and limitations of transparency. The last couple of years have witnessed acute financial difficulties by major U.S. corporations despite adherence to the letter (if not always the spirit) of generally accepted accounting standards.

7. Stronger economic and financial policies have been combined with improving the environment for private sector decision taking in ways that should facilitate the assessment and management of risk. This offers the best prospect of allowing countries to reap the potential gains from globalization, while minimizing the likelihood, and potential severity, of crises. This ambitious work program is central to our agenda for the prevention and resolution of crises.

8. But we must recognize that despite best efforts at prevention, crises will occur. And in a hopefully limited number of cases, sovereigns' debt burdens—as a result of some combination of bad policies and bad luck—may become unsustainable. Which is to say, countries may reach a point at which there is no feasible set of policies that are likely to be politically sustainable that offer a realistic prospect of a return to viability.

Assessments of sustainability

9. That said, I do not want to minimize the difficulty of making the judgments that underpin assessments of sustainability. This is not an exact science. Judgments need to be based, inter alia, on assessments of authorities' ability to mobilize and sustain support for adjustment efforts. Careful analysis is also required of the likely response of the economy to policies—including likely developments in real interest rates and the real exchange rate, as well as the prospects of the external environment. The analysis must also reflect a sober assessment of the likely fiscal implications of resolving difficulties in the domestic financial sector. Beyond that, account also needs to be taken of the economic vulnerability to future shocks. Here there is a need for careful analysis of the interlinkages between balance sheets in the economy, and the ways in which developments in one sector may spillover to others—including the fiscal accounts and the implications for the prospects for medium-term sustainability.

10. Of course, judgments that a country has reached the point of being unsustainable are not taken lightly. For the Fund, such judgments have immediate implications for our ability to provide financing. It implies that we can no longer support a member's adjustment efforts without seeking an adjustment to the level and profile of the sovereign's debt-service obligations. These judgments are reached fully cognizant of the likely costs to the member of such a restructuring, which include:

  • Its reputation in, and prospects for access to, capital markets.

  • A possibility that the restructuring of debt—or even the threat of a restructuring—will trigger important ramifications in the domestic economy—particularly in the domestic banking system. As we have seen, these can exacerbate economic dislocation and accelerate both capital flight and the withdrawal of foreign capital.

11. But continued Fund lending in the face of unsustainable debt burdens would be no panacea. Perhaps it could buy a little time. But increasing the burden of debt that benefits from the Fund's preferred creditor status must only increase the magnitude of the debt adjustment that must eventually be borne by private creditors in situations where there is no underlying improvement in the country's capacity to service its debt.

Pre-crisis window of opportunity

12. There is typically at least a brief period between the recognition that a member has an unsustainable debt burden and the outset of a full-blown crisis. But in such circumstances, time is the friend of neither a debtor nor its creditors. There is likely to be a window of opportunity in which it may be possible to reach agreement on a restructuring that—in conjunction with the sustained implementation of appropriate policies—offers the prospect of restoring sustainability while limiting the scale of economic dislocation and preserving assets' economic value. The challenge is to establish a system that increases the likelihood that the window will be utilized to avert an even worse outcome.

13. The potential benefits of moving rapidly to restructure debt while there is money on the table and before economic dislocation has taken hold is widely recognized in credit markets. Of course, we must be cautious in drawing parallels between domestic bankruptcy regimes and the restructuring of sovereign debt. In contrast to the position of a corporation, in the sovereign context there is no ultimate sanction of liquidation, a bankruptcy court does not supervise policies, and creditors cannot insist on a change in management. Nevertheless, it is worth noting that Chapter 11 of the U.S. Bankruptcy Code, and the move in a number of European countries to adopt statutory rehabilitation frameworks, reflects recognition of the potential benefits of moving rapidly to protect economic value and the interests of stakeholders. This prescription carries over to the sovereign context. Countries and their creditors would be well served by moving rapidly to put debt into a sustainable basis. Market failures that introduce unnecessary delays into this process lead to debtors and their creditors bearing costs that are unduly large.

Impediments to rapid progress toward restructuring

14. The debate on sovereign debt restructuring has highlighted a number of key issues that determine the pace at which agreement can be reached on restructurings, and shortcomings of the existing system.


15. Let me first reiterate the importance of the sustained implementation of appropriate policies. Proposals for improving arrangements for restructuring are complementary to—and not substitutes for—appropriate policies. In the absence of appropriate policies—and the associated increased clarity with regard to medium-term prospects—it is unlikely that either a debtor or its creditors would wish to enter into a restructuring agreement. Debtors will want to avoid deals that may need to be reopened. At the same time, investors would typically retain the option value associated with their original claim, rather than make concessions in the face of great uncertainty concerning the economic prospects for a debtor and its payments capacity.

Predictable and transparent process

16. A second set of concerns regarding the existing arrangements for sovereign debt restructuring relates to the absence of a predictable and transparent process. Concerns have highlighted, in particular, the absence of procedural clarity regarding the conduct of debtors and creditors.

Code of Good Conduct

17. These concerns have led to calls for a voluntary Code of Good Conduct. The various proposals that have emerged from the public and private sectors are constructive, and could, in the view of the Fund, help provide greater predictability to the restructuring process under any legal framework.

  • A Code could be made applicable to a broad set of circumstances, ranging from periods of relative tranquility to periods of acute stress, and could constitute an established set of best practices. In contrast, proposals for strengthening arrangements for debt restructuring have a more limited scope and purpose—to facilitate the resolution of financial crises.

  • By its very nature a voluntary Code, while potentially helpful, could not resolve collective action problems. This is a subject to which I shall return in a moment.

  • Finally, a Code could only be effective to the extent to which it is able to attract broad support among debtors and their creditors. Accordingly, the most promising approach to developing a code that could form the basis of a consensus would be for it to be developed jointly by debtors, their creditors, and other interested parties (including the Fund). Conversely, it appears unlikely that a Code designed by the Fund or other organizations would attract broad support.

The Fund's Lending into Arrears Policy

18. Recent adaptations in Fund policies could be complementary to a Code. Last year the Fund's Executive Board adopted a modification of the lending into arrears policy—the policies that govern the circumstances under which the Fund can provide financial support for a member's adjustment program during the period in which it has arrears to private creditors, and is attempting to reach agreement on a restructuring. This policy establishes expectations regarding the behavior of debtors that are receiving financial support from the Fund in such circumstances:

19. The debtor should engage in an early dialogue, which should continue until the restructuring is completed:

  • The debtor should share relevant, non-confidential information with all creditors on a timely basis. This would include an explanation of the adjustment program and the financial circumstances that justify a restructuring, as well as a comprehensive picture of all domestic and external claims on the sovereign.

  • The debtor should provide creditors with an early opportunity to give input on the design of the restructuring strategy. This could help address the specific needs of different types of investors, thereby increasing the likelihood of a high participation rate.

20. In addition, in cases in which creditors have organized a reasonably representative committee on a timely basis, there is an expectation that the member would negotiate with such a committee. Our policy suggests a number of principles that should guide the debtor's conduct during negotiations. In formulating these principles, we have drawn on the expertise of workout specialists reflected, for example, in the report by the Council on Foreign Relations (CFR), and efforts by International Federation of Insolvency Professionals (INSOL) to distill best practice for nonsovereign workouts.

Collective action difficulties

21. A third shortcoming of the existing arrangements for the restructuring of sovereign debt relates to the failure of collective action. This market failure provides a potential rationale for public intervention. It complicates the process of reaching agreement on a restructuring. There is a danger that individual creditors will decline to participate in a voluntary restructuring in the hope of recovering payment on the original contractual terms, even though creditors—as a group—would be best served by agreeing to a restructuring.

  • The problem of collective action is most acute prior to a default, where individual creditors may have some reasonable hope of continuing to receive payments under the terms of their original contracts. A debtor that had reached agreement with the bulk of its creditors on a restructuring would doubtless hesitate to default on a small amount of the original debt simply to secure unanimity in the restructuring. But if a numerically significant proportion of creditors elect not to participate in a restructuring, each hoping to be part of a small minority, the restructuring will inevitably fail.

  • Following a default, the options facing creditors—particularly those who have no interest in litigation—are more limited and so the problems of collective action may be less acute. But a more formal mechanism would still make sense in such cases, as it would provide greater clarity as to the predictability and transparency of the process by which agreement could be reached, while of course eliminating the possibility of at least some creditors hoping to eventually secure better terms through litigation.

But we should not fall into the trap of believing that default is a good solution to collective action difficulties. Of course, there is no doubt that following default, agreement on a restructuring would eventually be reached. But default—and the associated uncertainties regarding creditor-debtor relations—tends to be associated with widespread economic dislocation. Proposals for strengthening arrangements for debt restructuring are intended to increase the likelihood that early agreement can be reached on restructuring that can restore viability. And that neither debtors nor their creditors must bear costs that are unduly large.

Proposals for strengthening arrangements for sovereign debt restructuring

22. On occasion, the debate over alternative approaches to strengthening arrangements for sovereign debt restructuring has tended to become polarized. This is regrettable. The development of sound public policy demands dispassionate analysis of the strengths and limitations of alternative approaches. It must not be allowed to become the advocacy of one approach to the exclusion of others. Our objective is to develop effective tools that allow a country that has an unsustainable debt burden to reach agreement with its creditors on a restructuring that—in conjunction with the sustained implementation of appropriate policies—offers the prospect of restoring sustainability while limiting the scale of economic dislocation and preserving assets' economic value. With this in mind, I would like to offer a brief survey of the main proposals that are under active consideration.

23. Most proposals for improving sovereign debt restructuring have, at their heart, recognition of, and measures to address, collective action difficulties. They fall into two broad categories. The first are voluntary in character, and rely on contract. The second rely on a statutory framework.

The contractual approach

Collective Action Clauses (CACs)

24. As you know, we continue to support actively efforts to develop, and promote the adoption of, collective action clauses in sovereign debt contracts. There is no doubt that the widespread adoption of CACs would be a major step, and could be helpful in facilitating agreement on restructurings. Two recent developments provide grounds for cautious optimism that we are starting to witness a shift in market practice toward more use of CACs. For all too long, this has been elusive. While one swallow does not make a summer, these developments reinforce my view that we must redouble our efforts to ensure that the recent momentum is maintained.

25. Let me applaud Mexico for the successful placement of a $1 billion bond governed by New York law that includes collective action clauses. In particular, (i) investors holding 75 percent of outstanding principal can vote to modify the payment terms; and (ii) acceleration of the bonds following an event of default requires the consent of investors holding 25 percent of outstanding principal. In contrast, previous Mexican bond issues in the New York market have included no provision for majority action with regard to the modification of payment terms or of acceleration. In a number of important respects, the clauses in this bond follow the recommendations of a G-10 Working Group.

26. What price did Mexico have to pay be to incorporate CACs into a bond? The spread at issue was inline with the Mexican yield curve, suggesting that any premium paid for CACs was negligible. I might add that the bond has continued to trade well in the secondary market. The question, now that the first mover problem has been solved, is how best to ensure that this bond issue serves as a guidepost for the development of market practice.

A second development with respect to collective action clauses has been the proposals presented by a number of financial industry trade associations, the so-called "Gang of 7." We welcome this initiative, which proposes model clauses for bond contracts in all jurisdictions, including those, such as the New York market, that currently contain no such clauses. However, on balance, we consider the proposals problematic, and there are questions as to whether we could recommend their adoption of the proposed clauses to our emerging market members.

  • The draft clauses include a number of features that may limit their ability to resolve collective action problems. Principal among these is the provision for a de facto 90 percent voting threshold for majority restructuring (through the blocking ability of a 10 percent voting minority).2 This could largely defeat the objective of resolving collective action difficulties.

  • They are a step backwards from the established practice in the sovereign debt market in London and Luxemburg by recommending a much higher voting threshold than the 75 percent of a quorum typically found in bonds issued in these markets.

The proposals include new transparency requirements by issuers. While some could make a positive contribution to crisis prevention, others are problematic. The latter include covenants for the provision of information that may raise an issue of confidentiality, and could apparently give rise to events of default in cases where debtors are reluctant to disclose information requested by a small minority of creditors.

27. Although we see many benefits to the widespread adoption of CACs, the limitations of this approach are well known. While they provide for collective action among holders of individual bond issues, they do not provide for the aggregation of voting across instruments. Moreover, CACs can be introduced in new instruments, but do not address the problem of the outstanding stock of bonds and other debt instruments that do not include these clauses.

JP Morgan proposal

28. A contractual approach has been suggested by JP Morgan, which is based upon a variation of use of collective action clauses. This proposal recognizes—and seeks to address—the problems of aggregation across debt instruments and the stock-of-debt instruments that do not include collective action clauses. The approach involves a two-stage process. In the first stage, investors holding international bonds and a wide range of credit instruments would exchange their existing instruments for new instruments. These would preserve repayment terms, but would be linked by collective action clauses, which would provide for an aggregation across instruments of reaching for the second stage—the agreement on the terms of the final restructuring.

29. As the collective action problem reflects a divergence between what may be optional for investors acting individually as opposed to in a group, how could the collective action problem associated with agreement on the first stage be overcome? JP Morgan has suggested a system of carrots and sticks. The carrot would take the form of an up-front cash payment for participation, while the stick would take the form of the aggressive use of exit consents, which—by eroding the contractual rights of investors who elect not to participate in the first stage—would reduce the attractiveness of a holdout strategy.

30. The proposal is ingenious. And it warrants further consideration. But there are some doubts regarding its feasibility. In a pre-default case where, as I have argued, collective action problem may be most acute, the tight timetable for restructuring before resources available for debt service are exhausted suggests that the two steps would need to be conducted back-to-back in the same timeframe. This would negate any advantage of getting creditors into a collective framework. Decisions regarding participation in the first and second stage would be taken at the same time, and so, of course, the operation would be equivalent to a straightforward exchange that, by its nature, does not resolve collective action problems.

  • In a post-default case, in circumstances where there may be substantial doubts about the future course of economic policies and the terms of an eventual restructuring proposal, would investors be willing to surrender their individual contractual rights and enter into a collective process? Indeed, would such a move be consistent with portfolio mangers' fiduciary responsibilities to their end investors? Is there a danger that they would decide to preserve their rights under their existing instruments while they wait and see what will happen?

  • Without an effective and transparent mechanism for resolving disputes—whether between creditors or between a debtor and its creditors—would investors consider that the first stage of the process provides adequate protection against the inclusion of fraudulent claims? Here too we must be cognizant of investors' fiduciary responsibilities. There is perhaps a question of whether the procedure could be made subject to independent arbitration.

31. Before I turn to the next topic, let me just say that aggregation of voting for decision taking by investors is generally considered to be an essential feature of statutory frameworks for corporate rehabilitation. The potential benefits of aggregation appear to carry over to sovereign workouts. The JP Morgan proposal is one approach. But I think that there is a need for further exploration of approaches to contractual arrangements for aggregation in the context of new sovereign bond issues.

Statutory approaches

32. Now let me turn to statutory approaches to sovereign debt restructuring. There has been a range of proposals. I would like to highlight two: The first seeks to use established statutory and judicial processes, and the second envisages the establishment of a new universal legal framework.

Use of a class action framework

33. An approach suggested by two American lawyers involves using the existing statutory rules of civil procedure in the countries whose laws are typically chosen to govern sovereign debt obligations to affect a restructuring of those instruments when necessary. In the United States, these are the so-called "class action" rules; in England, they are called "representative actions." In a nutshell, a group or committee of representative creditors, with the acquiescence of the sovereign debtor, would commence in domestic court a class or representative action on behalf of all similarly situated debt holders. Under this approach, there would be a judicial mechanism for resolving disputes among creditors and between a debtor and its creditors. A proposed restructuring—in the form of a settlement of the class action—that enjoyed widespread creditor support could then be made binding on all affected creditors through a court judgment.

34. The proposal involves complex legal analysis. But the most important elements are recognition of the need for aggregation across debt instruments, and the need for a statutory framework to bind dissident creditors.

35. The authors of this innovative approach have highlighted legal uncertainties regarding, for example, whether domestic court would accept a mandatory class action covering sovereign debt instruments. There are also uncertainties regarding the criteria that a judge would apply to assessing the acceptability of a restructuring proposal. The authors have suggested that these and other issues may need to be addressed through new legislation in order to provide some measure of predictability and procedural clarity to the approach.

36. A second set of difficult questions relates to the application of the approach to cases in which a debtor has issued debt in a number of jurisdictions. Perhaps innovative solutions could be found to minimize these difficulties. But discussions with a range of legal practitioners, however, suggest that it would to be difficult to capture the benefits of aggregation without a uniform—and unified—legal framework.

37. But overall I am confident that there is a growing recognition that the integration of capital markets will require a strong, centralized legal framework to support its operation. This is not a new issue. The drafters of the U.S. Constitution grappled with the question of which essential powers in the economic arena would need to be conferred on the Federal Government, in order to support the inevitable growth of interstate commerce. Interestingly, the framers of the Constitution decided that the Federal Government should have the authority to enact a unified federal bankruptcy law. How—they must have wondered—could a fragmented, state-by-state insolvency framework support an increasingly interdependent economy? The analogy is loose but—at a certain level of abstraction—is apt. Credit markets can only develop in a stable manner if the legal and institutional frameworks that support them also evolve.


38. As you know, we consider that a statutory framework could most effectively be established through the Sovereign Debt Restructuring Mechanism, using an amendment to the Fund's Articles. The Fund has been asked by the International Monetary and Financial Committee—which is made up of 24 Ministers of finance—to develop a concrete proposal for a statutory mechanism, and to report to the Committee's next meeting in April. Our proposal has been developed with the enormous benefit of extensive dialogue with the private sector, workout professionals, academics, and the official community.

39. In developing the proposal, we have been guided by a number of principles. In particular:

  • First, the mechanism should only be used to restructure debt that is judged unsustainable. It should neither increase the likelihood of restructuring nor encourage defaults.

  • Second, any interference with contractual relations should be limited to those measures that are needed to resolve the most important collective action problems

  • Third, the framework should be designed in a manner that promotes greater transparency in the restructuring process, and to encourage early and active creditor participation during the restructuring process.

  • Fourth, the integrity of the decision making process under the mechanism should be safeguarded by an efficient and impartial dispute resolution process.

  • Fifth, the framework should be designed to catalyze early and effective dialogue between the debtor and creditors—it should not increase the role of the Fund in this dialogue.

  • Finally, the SDRM should not give rise to an expansion of the Fund's legal powers.

40. There is not enough time to describe the proposal for the SDRM in any detail. But let me just highlight three key features.

  • The principal feature of the mechanism is that it would allow a sovereign and a qualified majority of creditors to reach an agreement that would then be made binding on all creditors that are subject to the restructuring, paying due regard to seniority among claims and the diversity of creditor interest. Giving creditors the ability to make this decision on an aggregated basis will not shift the legal leverage from the creditors to the debtor; rather it would increase the leverage of creditors over potential holdouts and free riders, enabling an agreement to be secured more rapidly.

  • Second, we are not proposing that there be an automatic stay on creditor enforcement or a general suspension of contractual provisions during the period between activation and the reaching of a restructuring agreement. This has perhaps been the most controversial feature of the proposal. But after extensive discussions, we have concluded that a generalized stay would not be a measured response to the threat of creditor litigation, and would represent an unwarranted interference with creditors' rights. Instead, we envisage much more limited means for addressing the possibility that that the prospects for early agreement could be undermined by creditor litigation.

  • Third, an independent dispute resolution forum would be established to verify claims, ensure the integrity of the voting process, and adjudicate disputes that might arise following activation of the SDRM.

41. Where do we go from here? We believe the proposal is balanced, and will provide a credible basis for reporting to Ministers at the forthcoming April meeting of the International Monetary and Finance Committee on a concrete proposal for a statutory approach to a sovereign debt restructuring mechanism. It is perhaps no surprise that we do not yet have the very high level of support that would be required for an Amendment to the Fund's Articles. So this will not be the end of the story.

42. When I reflect on the current position of the SDRM, I am reminded of Benjamin Disraeli's maiden speech in the House of Commons in 1837: "Though I sit down now, the time will come when you will hear me." In any event, we will be looking to Ministers for their guidance as to how we take this initiative forward.


43. In conclusion, let me summarize the five key areas where there appears to be broad agreement.

  • First, there is a need to improve the existing arrangements for restructuring sovereign debt. The intention is not to make sovereign debt restructuring an easy option. But rather to allow debtors with unsustainable debt burdens to reach agreement on a restructuring without unnecessary dislocation and loss of asset values. Of course, nobody believes that this is the complete answer to the difficulties of resolving financial crises. The restructuring of sovereign debt is likely, as a practical matter, to need to be complemented by other measures. Including those to stabilize and restructure the domestic financial system, and perhaps resolve balance sheet difficulties in the corporate sector. It may also need to be conducted against the background of temporary resort to some combination of exchange controls and a deposit freeze. But that is a subject for another day.

  • Second, the key market failure relates to collective action. This is the main pillar of both the contractual and statutory approaches to improving the arrangements for sovereign debt restructuring.

  • Third, progress is being made with the incorporation of collective action clauses into sovereign debt contracts. This is encouraging. Now that the first mover problem has been resolved, we must redouble our efforts to promote the widespread adoption of such clauses.

  • Fourth, there is recognition in many quarters that there would be substantial benefits to aggregation for the purposes of decision taking by creditors. The variation of CACs suggested by JP Morgan, the possible use of the so-called class action rules, and the SDRM, represent different approaches to seeking aggregation across instruments, as well as dealing with the stock of bonds that do not include CACs.

  • Finally, there is broad agreement that the restructuring process would be enhanced through greater transparency and predictability. Here I see strong complementarities between a possible Code of Good Conduct, some of the proposals for CACs, the recent revisions to the Fund's lending into arrears policy—which establishes expectations regarding debtor's behavior vis-à-vis their creditors—and a number of key features of the SDRM proposal. But we have a lot further to go.

Thank you very much.

1 See:
2 This is an improvement over earlier proposals suggesting a 95 percent threshold (Model covenants for new sovereign debt issues, EMCA, May 3, 2002).


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