FRBSF Economic Letter
2002-19 | June 28, 2002
Towards a Sovereign Debt Restructuring Mechanism
- The decentralized approach to the collective action problem
- Problems with a purely decentralized solution
- Details of a formal workout mechanism
- Remaining issues
Over the 1990s, public and private international borrowers shifted the composition of their external financing—instead of relying primarily on loans from a syndicate of a few banks, they turned to issuing bonds. This has resulted in many more creditors of various kinds holding claims on sovereign debt in the forms of different debt instruments with different time horizons.
Under usual circumstances, this shift has some desirable consequences. With a wider creditor base, the risk of lending to a country is spread more widely, so the country can borrow at more favorable terms. Moreover, reducing banks’ exposure to borrowers’ risk lessens the risk of financial contagion when financial difficulties arise in any individual borrowing country.
However, the experiences in recent financial crises suggest that when a country does require a debt restructuring, the outcomes are likely to be less orderly than those obtained in the days of bank financing. Under bank financing, it was possible to assemble the major claimants of a problem debtor and work out an acceptable rescheduling. In the case of bond finance, in contrast, the set of claimants is much more numerous, widespread, and heterogeneous. As a result, it can be very difficult to work out an agreement because a small group of “holdout” creditors can veto any package that they fail to find acceptable. This difficulty is commonly known as a collective action problem. It is particularly severe for bonds written in the United States, where changes to the original terms of a bond require agreement by all bondholders. In response, there has been a call for reforming the institutions governing sovereign debt restructuring.
In this Economic Letter, I examine two proposals that have been put forth for reforming the debt restructuring process. One advocates a decentralized approach, and it is currently associated with policymakers in the U.S. Treasury. The other, sponsored by the International Monetary Fund (IMF), proposes a formal workout mechanism. Despite their differences, the proposals are not mutually exclusive, and at this point it appears that a set of new policies on debt restructuring will include elements of both proposals.
The decentralized approach to the collective action problem
Many policymakers, most notably officials of the U.S. Treasury (see Taylor 2002), advocate addressing the collective action problem by including three new types of clauses in sovereign bond contracts issued in the United States. This is a decentralized approach in the sense that it does not require a central authority to manage the process.
The first type of clause calls for a majority voting rule. This would allow a “super-majority” of bondholders, say 75%, to agree on the terms of a debt restructuring. Eichengreen and Mody (2000) studied the impact of majority voting clauses on debt issued in the United Kingdom, where such clauses are common, and found evidence suggesting that they do not result in increased borrowing costs.
The second type of clause establishes rules governing the renegotiation process. These clauses would specify the terms of creditor representation, as well as the information that the debtor would be required to provide in a renegotiation. The designated “creditor representative” would have a formal role in these renegotiations, with the authority to negotiate on behalf of the creditors and to initiate litigation against the debtor nation, and it would act according to the dictates of a majority of bondholders.
The third type of clause specifies the terms for launching restructurings. In particular, the Treasury advocates a “cooling-off period” between the date that the debtor announces its intention to restructure and the date the creditor representative is chosen. The cooling-off period is envisioned to last about 60 days (Taylor 2002).
Problems with a purely decentralized solution
While the IMF welcomes the inclusion of collective action clauses in bond issues as a fundamental component of addressing problems in sovereign borrowing (see, for example, Krueger 2002), it has argued that a policy based solely on the inclusion of such clauses is not sufficient.
One concern is, what country would be willing to go first; that is, since collective action clauses have not been a regular feature in bonds issued in many countries, including the U.S., there is a perception that no individual country would want to be the first to include such clauses in a major issue. The problem is that, if investors saw that one country was leading the way in including such clauses, they might think this country is abnormally concerned with debt renegotiation procedures and assume that it perceives a higher than average probability of finding itself in a renegotiation. The investors would then punish that country with inferior credit terms. Anecdotal experiences from sovereign issues in the U.S. bond market suggest that sovereign borrowers have been discouraged by their underwriters from including such clauses because of this concern.
A solution to this problem would be to alter the incentives faced by sovereigns issuing bonds in the U.S. such that all issuers immediately jumped to the inclusion of collective action clauses. For example, the Treasury has discussed the possibility that the IMF could withhold assistance from any country that failed to include such clauses. Alternatively, the IMF could offer superior borrowing terms to issuers who include such clauses. With a sufficient combination of sticks and carrots, all developing nations would perceive it as in their interest to include collective action clauses in their bond issues. Consequently, there would be no stigma against any individual nation doing so.
Of course, the IMF may not wish to provide such a pledge. One could envision a scenario where a government that failed to include such clauses could find itself in economic difficulty after the fact. Under those circumstances, it may be hard for the IMF to resist providing assistance to a country that is taking positive steps to address its financial difficulties. The IMF may then violate its pledge and damage its credibility.
Another limitation of collective action clauses as a coordination solution is that while they may mitigate coordination problems among heterogeneous bondholders, they would fail to address differences across different classes of debt holders, such as claimants on syndicated bank loans. Treasury proposes that decisions be made on an issue-by-issue basis through majority voting, with inconsistencies across different types of debt claims to be handled through an “arbitration process” (Taylor 2002). Even so, there is a large stock of debt instruments currently outstanding that do not contain such clauses. To deal with this difficulty, the IMF proposes a “super collective clause” that would allow for restructuring a given instrument after an affirmative vote by a super-majority of all creditors, not just those of the instrument in question.
However, countries typically issue debt in a number of legal jurisdictions, so there is no guarantee that such a clause would be interpreted the same way across jurisdictions, even if it were worded identically. Consequently, the IMF maintains that, for collective action clauses to work, they must be accompanied by a formal workout mechanism. This mechanism could be established through an amendment to the IMF Articles of Agreement.
Details of a formal workout mechanism
The mechanism proposed by the IMF would allow the debtor to request a temporary standstill from the Fund. During this period, the debtor would negotiate a rescheduling or a restructuring, with IMF approval, with its creditors. Capital controls would be used to ensure that reserves did not flee the nation during this negotiation period. Krueger (2002) has noted that creating a formal workout mechanism need not significantly expand the IMF’s legal authority. Ultimate approval of the terms of the restructuring could remain in the hands of a majority of a super-majority of creditors, across a broad range of credit instruments, and the sovereign debtor. If the standstill ended without an agreement, a super-majority of creditors could extend it.
The IMF’s proposed international workout mechanism has four main features. First, there would be a stay on creditor enforcement during the negotiation period. Second, debtor behavior would need to be constrained during the negotiation period to ensure protection of creditor interests. In particular, the debtor would be prohibited from allocating funds to non-priority creditors, and the debtor government would be required to pursue policies consistent with maintaining its capacity to service its debt obligations. It is envisioned that the latter goal could be achieved under an IMF-supported adjustment program. Third, private creditors would be encouraged to provide new financing, perhaps through some kind of explicit seniority mechanism favoring new money over old claims. This provision also could be mandated by a super-majority of creditors.
Finally, creditors would be bound by the terms agreed upon by the super-majority through collective action clauses. In this way, the IMF need not be making decisions that undermine creditor rights. However, the IMF would assess whether the terms of the agreement would adequately reduce the debt burden to a sustainable level. If it felt that it did not, the IMF could withhold further financing.
There is some fear that a more orderly workout mechanism would have the undesirable impact of easing the pain of debt restructuring and thereby increase the probability of default. The IMF has two responses to this fear. First, because restructurings currently are so costly, countries wait as long as possible to ask for debt restructuring. This can delay movement towards serious reform and it also can result in reduced creditor payoffs. Second, even though the cost of requesting a restructuring will be reduced by a workout mechanism, restructurings still will be sufficiently disruptive and costly to ensure that debtors will not enter into them if they can be feasibly avoided.
Another issue is the treatment of domestic creditors. It is clear that the poorer is the treatment of domestic creditors, the greater will be the amount of funds remaining for servicing foreign debt. This is, of course, a highly controversial issue. The IMF has raised it without taking any formal stance on this point, merely stating that judgments about sharing the burden between domestic and foreign creditors would need to be made on a case-by-case basis that incorporates the impact of such decisions on the domestic financial market.
Finally, there is the issue of adjudicating disputes. Throughout most of the issues raised above, there is a tension between the stance that decisions are to be left in the hands of a super-majority of creditors and the perceived need for a rapid and orderly workout that disrupts the domestic economy as little as possible. For example, this tension arises in determining the fairness of restructuring terms after approval by the creditor super-majority, as well as in determining the appropriate burden-sharing both among different classes of foreign creditors and between foreign and domestic creditors. The IMF has acknowledged that it would be difficult for its Executive Board to adjudicate disputes of this kind. Instead, it envisions an independent panel of judges insulated from IMF management to play that role.
While there has been a spirited discussion about the proper mechanisms to deal with problems in the international debt markets, one should not underestimate the degree to which a consensus is emerging. Officials at the IMF have backed away from their initial calls for an IMF-run international bankruptcy court in favor of an independent review process. The U.S. Treasury continues to advocate a more decentralized approach, but also acknowledges that at the end of the day some sort of arbitration procedure will be necessary to reconcile claims across different classes of sovereign creditors; it seems likely that this arbitration procedure will need to operate at a multilateral level. As such, it appears that the emerging consensus will be a set of policies designed to encourage collective action clauses, as the U.S. Treasury is advocating, while ultimate arbitration across different classes of claimants will fall to a multilateral entity, as is being advocated by the IMF.
Mark M. Spiegel
Eichengreen, Barry, and Ashoka Mody. 2000. “Would Collective Action Clauses Raise Borrowing Costs?” NBER Working Paper No. 7458.
Krueger, Anne O. 2002. “A New Approach to Sovereign Debt Restructuring.” Speech delivered to the Indian Council for Research on International Economic Relations in Delhi, India December 20, 2001. International Monetary Fund.
Taylor, John B. 2002. Conference on “Sovereign Debt Workouts: Hopes and Hazards?” Institute for International Economics, Washington, DC.
Opinions expressed in FRBSF Economic Letter do not necessarily reflect the views of the management of the Federal Reserve Bank of San Francisco or of the Board of Governors of the Federal Reserve System. This publication is edited by Sam Zuckerman and Anita Todd. Permission to reprint must be obtained in writing.
Please send editorial comments and requests for reprint permission to
Attn: Research publications, MS 1140
Federal Reserve Bank of San Francisco
P.O. Box 7702
San Francisco, CA 94120