The Benefits and Limits of Contract

The inclusion of collective action clauses in all international sovereign bonds would represent an important improvement in the international financial architecture. As has been discussed in earlier sections of this paper, and has been demonstrated in recent cases, collective action clauses include two provisions that can facilitate an orderly restructuring of sovereign indebtedness: (i) a provision that enables a qualified majority of bondholders to bind all bondholders of the same issuance to the terms of a restructuring agreement and (ii) a provision that enables a qualified majority of bondholders to prevent all bondholders of the same issuance from enforcing their claims against the sovereign.

The insertion of collective action clauses in all future international sovereign bonds would not require wholesale statutory reform. For example, although such provisions are not typically found in international sovereign bonds governed by New York law, they could be introduced without any legislative changes.

Moreover, it should also be noted that, even if a sovereign debt restructuring mechanism was established through legislation, as discussed below, such clauses could still play an important role. For example, since a statutory mechanism would only apply in circumstances where the member’s debt is unsustainable, collective action clauses could facilitate restructurings in circumstances where the problems facing the member arise from illiquidity.

However, relying exclusively on contract as the legal basis for a sovereign debt restructuring mechanism would limit the effectiveness of such a mechanism.

First, it would be difficult to establish a purely contractbased framework.

There is, at the outset, the problem of incentives for the adoption of traditional collective action provisions in all new indebtedness. By definition, a contractual approach would require the sovereign and its creditors to agree to the inclusion of these provisions in all future international sovereign bonds, and also in other debt and debt-like instruments that the market developed. Recent experience demonstrates that sovereign debtors facing financial difficulties actually prefer to exclude such provisions as a way of demonstrating their firm intention to avoid a restructuring. Neither have creditors pressed for their inclusion, notwithstanding the fact that they may make an unavoidable restructuring more prompt and orderly. The advantage of giving the framework for sovereign debt restructuring a statutory basis is that the collective action provisions that it would contain would effectively override the restructuring and enforcement terms set forth in the underlying agreements, as is the case with the collective action provisions contained in domestic insolvency laws.

Another barrier to the establishment of such a framework is the transitional problem. Even if all new bonds make use of the needed contractual provisions, a large portion of outstanding bonds with long maturities, including bonds governed by New York law, do not contain such provisions.³ While this problem could conceivably be addressed by a series of exchanges that retired existing bonds, it is not clear how debtors and creditors would be persuaded to take such action. It is also possible that use could be made of existing provisions that allow for amendment of terms not related to payment to facilitate debt restructurings in the interim. For example, Ecuador recently made use of “exit consents,” to overcome the problem of holdout creditors generated by the absence of provisions allowing a majority to amend payment terms in outstanding bonds governed by New York law. Under this technique, bondholders who accepted the exchange voted to amend non-payment terms in ways that made holding “old bonds” less attractive. However, this technique has been somewhat controversial and it may not be immune from legal challenge in the future.

Second, even if a contract-based framework could be established, it would not provide a comprehensive and durable solution to collective action problems.

Collective action clauses traditionally only bind bondholders of the same issue. In contrast, the collective action provisions of a statute would be designed to apply across a broad range of indebtedness (potentially including international and domestic debt, bank loans, trade credit and official claims, if applicable). This is one of the reasons why the collective action provisions of insolvency laws are so effective. To address issues arising from the relative seniority of certain indebtedness, insolvency laws often provide for the classification of debt for both voting and distribution purposes. As discussed earlier, similar safeguards would need to be established under the mechanism.

To address the above limitation, one could conceive of the introduction of contractual provisions that provide for the restructuring of the instrument in question on the basis of an affirmative vote of creditors holding a qualified majority of all private credit. While further study on the feasibility of developing such clauses should be encouraged, such an approach would raise its own set of issues.

• First, such a provision would exacerbate the incentive problem: if it is difficult to convince a sovereign and the purchasers of one bond issue to agree to the inclusion of a collective action clause in that issue, it would be even more difficult to persuade debtors and creditors to include such provisions in all forms of debt instruments in a uniform manner. Indeed, a sovereign facing financial difficulties would come under pressure from certain creditors to exclude such provisions as a means of giving such creditors effective seniority. Moreover, it can be expected that certain creditor groups would be particularly reluctant to agree voluntarily to an arrangement whereby, for voting purposes, their claims were aggregated with all other present or future creditors.

• Second, even if all debt instruments contained identical restructuring texts, which would be difficult to achieve, there would be no assurance of uniform interpretation and application unless they were governed by the same law and subject to the same jurisdiction. In the present environment, emerging market countries that have borrowed heavily often have a variety of bond issuances outstanding which are governed by the laws of different jurisdictions.

• Third, it may not be feasible to establish a process by contract that would effectively guarantee the integrity of the voting procedure. Under the statutory framework that governs the domestic insolvency process, a court oversees this process, including the verification of claims, so as to guard against fraud. In the absence of an independent party to verify the true value of claims, a debtor could, for example, inflate its debt stock by establishing matching credit and debt positions with a related party. That entity—which could hold a qualified majority of all debt—could vote to reduce the value of all creditor claims.

• Fourth, it is not clear that such provisions would be consistent with the existing legislation of all members. The fact that traditional collective action clauses are not included in international sovereign bonds in some jurisdictions arises, in part, from the absence of a clear statutory basis that allows for the rights of a minority of creditors to be modified without their consent. This issue would be amplified if contractual provisions attempted to aggregate claims for voting purposes.

• Finally, and more generally, the financial markets have consistently demonstrated the ability to innovate. A statutory regime is therefore likely to provide a more stable background than contractual provisions even if it were feasible to overcome all of the other difficulties referred to above.

Implementing a Statutory Framework

If a statutory approach that creates the legal basis for majority action across all sovereign indebtedness offers the best method of achieving the objectives of a sovereign debt restructuring mechanism, the question arises as to how best to implement a change in the statutory regime.

There are a number of reasons why the statutory approach could be more effectively implemented through the establishment of universal treaty obligations rather than through the enactment of legislation in a limited number of jurisdictions.⁴ First, it would prevent circumvention: if the statutory framework is only in place in a limited number of jurisdictions, creditors could ensure that future instruments enable them to enforce their claims in jurisdictions that have not adopted such jurisdictions but whose money judgments are recognized in key jurisdictions under treaties or local law.⁵ Second, an international treaty would ensure both uniformity of text and (if there is an institution given interpretive authority) uniformity of interpretation. Third, it would address a potential “free rider” problem: without a treaty, countries would be reluctant to adopt legislation until they were assured other countries had also done so. (A treaty could be designed that would enter into force at the same time for all signatory countries.) Finally, the establishment of a treaty facilitates the establishment of a single international judicial entity that would have exclusive jurisdiction over all disputes that would arise between the debtor and its domestic and international creditors and among such creditors. Moreover, such an entity would also have responsibility for the administration of a unified voting process, including the verification of all creditor claims. If one relied exclusively on domestic legislation in a variety of jurisdictions, the process for dispute resolution and claims verification would be fragmented one, with different claims being subject to the jurisdiction of different courts, depending, inter alia, on the governing law of the instrument.

what would be the advantages of establishing the treaty framework through an amendment of the IMF’s Articles? This would be a means of achieving universality in the absence of unanimity: an amendment of the Articles can be made binding upon the entire membership once it is accepted by three-fifths of the members, having 85 percent of the total voting power. Moreover, given the considerable benefits of IMF membership, it is very unlikely that a member would wish to opt out of IMF membership in order to avoid application of the SDRM. It should be emphasized that, if an amendment of the Articles were merely to provide the legal basis for the “majority action” decisions, as described in the previous section of the paper, it would not give the Executive Board any additional legal authority. Rather, it would give a majority of creditors the legal authority to bind a dissenting minority.

Notwithstanding the above, relying on the IMF’s Articles as a means of providing the statutory basis for majority action decisions to be taken by sovereign debtors and their creditors will require the resolution of an important institutional issue. As noted above, a treaty framework will require the establishment of a verification of claims and dispute resolutions process. However, the IMF’s existing institutional infrastructure would not accommodate it playing such a role. Specifically, the IMF’s Executive Board would not be perceived as impartial in this process since the IMF is a creditor and also represents the interest of the sovereign debtor and other bilateral creditors.

One way of addressing this institutional issue would be to rely on the same amendment of the Articles that would be used to establish the collective action framework, described above, as the basis for establishing a new judicial organ that would carry out these very limited functions. Clearly, a key question is whether there would be adequate safeguards to ensure that such an organ operated—and was perceived as operating—independently from the Executive Board and the Board of Governors.

As a legal matter, the independence of the organ could be established by the text of the amendment itself. The amendment would provide that decisions of the judicial organ would not be subject to review by any of the IMF’s other organs and that, more generally, the judges appointed to this organ would not be subject to the interference or influence of the staff and management of the IMF, the Executive Board or any IMF member. The text of the amendment could also specify in some detail the qualifications of the judges to be selected and, to ensure security of tenure, the grounds for their dismissal. One way of ensuring that the judges serving on the organ maintain some distance from the staff and the Executive Board would be to appoint them for a limited but possibly renewable period. Moreover, a procedure could be established whereby the judges appointed by the Managing Director (or the Board) would be derived from a list of candidates that would have been selected by a qualified and independent panel.

It should be emphasized that the role of this judicial organ—wherever it is located—would be a limited one. Specifically, the organ would have no authority to challenge decisions made by the Executive Board regarding, inter alia, the adequacy of a member’s policies or the sustainability of the member’s debt.

Exchange Controls

In the context of financial crises, exchange controls may need to be relied upon in at least two circumstances. First, in circumstances where a sovereign defaults on its own indebtedness, it is likely that such a default will trigger capital flight, particularly where the restructuring will also embrace claims on the sovereign held by the domestic banking system and the member maintains an open capital account. Second, even where the external debt of the sovereign is not significant, a financial crisis can arise because of the overindebtedness of the banking and corporate sectors which, when coupled with a loss of creditor confidence, leads to a sudden depletion of foreign exchange reserves. In these circumstances, there may be a case for the authorities to impose exchange controls for a temporary period.

The possible resort to exchange controls raises a number of complex issues that would need to be addressed on a case-by-case basis. Inevitably, difficult judgments will need to be made against the background of considerable uncertainty regarding the ways in which events may unfold. Nevertheless, two broad sets of issues would need to be considered: first the timing of the imposition of controls, and second, their coverage across different types of transactions.

As regards timing, there is a question of whether it would be appropriate to impose controls at an early stage of capital flight with a view to stanching the hemorrhage of reserves, thereby preserving the resources available to the economy, including for debt service. This would have the effect of reducing the difference in the ability of investors holding claims of various maturities to exit early, and from this perspective permitting a broader degree of equity in the treatment of various types of investors. It is worth noting, however, that differences in the ability of investors to exit early stemming from the relative maturity of claims is presumably reflected in the market pricing of the instruments concerned and compensates investors for the relative risks. Moreover, a shift toward a presumption that exchange controls would be imposed at an early stage of capital flight could reduce the ability of domestic banks to attract and intermediate domestic savings and foreign capital, as residents would be more likely to hold savings abroad and foreign creditors would raise the cost of short-term capital.

An alternative approach of waiting until resources are exhausted before resort to controls would lean in the direction of respecting the contractual rights of investors holding short-term claims. It would also keep open the possibility that if confidence stabilizes resort to exchange controls could be avoided. It has the drawback, however, that once controls are imposed the resources available to the economy have been depleted, which will have adverse effects on the pace of recovery and capacity to generate resources for debt service.

A second question relates to the scope of the controls. In cases where a member has the institutional capacity to implement exchange controls, it may be possible to arrest capital flight without an interruption in debt service and other contractual obligations. But this will depend on the severity of the crisis and the institutional capacity of the member. In circumstances where it is necessary to interrupt external debt service, it would be important for the authorities to put in place a framework for the eventual normalization of creditor relations by nonsovereign debtors, in order to minimize the long-term impact on corporations’ market access. Such a framework could include two key features. First, the facilitation of an out-of-court workout mechanism operating in the shadow of domestic bankruptcy. Second, a specification of the minimum terms under which foreign exchange would be made available to service restructured debts.

The question arises, however, as to whether an SDRM should be designed to provide limited legal protection (in the form of a stay) during the period of renegotiation to domestic enterprises that might otherwise be subject to litigation as a result of the default arising from the imposition of controls.

It should be noted at the outset that, even if the decision were made to exclude nonsovereign debt from the coverage of an SDRM, exchange controls would still provide considerable legal protection in at least two respects. First, any restrictions imposed on the ability of residents or nonresidents to make transfers abroad would still be enforceable within the territory of the sovereign. Second, in the event that the controls give rise to payments arrears, foreign creditors would be precluded from enforcing their claims against a resident debtor in the territory of the sovereign. The legal protection that may not be provided by the controls would be protection against the enforcement of claims by nonresidents with respect to a resident debtor’s assets that are located overseas. It is this latter category of protection that an SDRM could be designed to provide.⁶

Among the complex issues that would arise if an SDRM were to apply to exchange controls is the feasibility of making a distinction between those debtors that, except for exchange controls, would be able to service their debt, and other debtors that are not healthy and need to be restructured. While it would be reasonable for the former category to enjoy some temporary legal protection under an SDRM, it would be preferable to make the latter category subject to the local insolvency law. A second difficulty relates to the protection of creditor interests. During the period of the stay on litigation, what measures could be put in place that would give creditors the assurance that the debtor is not using the stay as a means of facilitating asset stripping?

A final question relates to the role of the IMF. As discussed above, in the context of sovereign indebtedness, it is possible to design a framework where the key decisions are made by the majority of creditors rather than the IMF. However, in the context of exchange controls that gives rise to the default of a multitude of debtors (each with their own group of creditors), such an approach would not be feasible. In these circumstances, the legal authority to approve a temporary stay, if that were deemed an eventual feature of a new statutory mechanism, would need to reside with the IMF.