Bulletin – August 2002 Recent Proposals for Reform of Sovereign Debt Restructuring
The emerging market crises of recent years have prompted widespread calls for reform of the ‘international financial architecture’. These have included calls for measures to reduce the frequency of crises and reforms to the way that crises are managed.
There has already been important progress in reforms aimed at preventing crises. These have included initiatives for stronger financial systems and greater transparency at the government and corporate levels, and greater attention to the currency risk inherent in foreign debt and the rollover risk of short-term debt. Furthermore, there is now greater awareness that fixed or near-fixed exchange rate regimes can increase the risk of crisis and the severity of crises when they do occur.
However, there has been relatively little progress in reforming the way that sovereign debt crises are managed once they occur. In particular, there is no framework for speedy resolution. The result is that incipient crises are not addressed at an early stage by policy-makers or creditors and this in turn leads to large falls in GDP, downwards overshooting of asset prices, and possible contagion to other markets.
The debate on possible reforms has been invigorated recently by a call for a sovereign debt restructuring mechanism by Dr Anne Krueger, the First Deputy Managing Director of the International Monetary Fund (IMF) and by various comments by US officials. This article outlines the recent IMF initiative and various responses, and assesses the main issues involved.
Recent Emerging Market Crises
Recent emerging market crises have been associated with large depreciations in previously fixed or near-fixed exchange rates and also debt servicing problems. However, there are substantial differences between different crises, with implications for the role that a sovereign debt restructuring mechanism can or cannot play.
In the case of the Asian crisis countries (especially Korea, Thailand and Indonesia) the problems were largely problems of servicing unhedged private sector foreign currency debt following the depreciations of the domestic currency. In the case of Korea, the problem of financial sector debt became a public sector problem because of a government guarantee, but the rollover problem was then resolved with foreign banks agreeing to extend the maturity of their loans. In the case of Thailand and Indonesia, the debt servicing problems were pure private sector ones, and the claims of foreign creditors on domestic debtors were largely dealt with in the same way as claims of domestic creditors, i.e., through the domestic courts and bankruptcy systems.
In the case of Russia in 1998 and the current crisis in Argentina, the debt problems were largely (but not exclusively) public sector ones. In these cases there was a substantial quantity of foreign currency debt issued in international markets, with contracts specifying that foreign creditors are entitled to take legal action against the sovereign in courts in international financial centres. However, there is no international bankruptcy system that sets out the rights of debtors and creditors in such cases. Hence, some of the protections that domestic bankruptcy systems offer to debtors – in the interest of maintaining the value of the claims of all creditors – do not exist in the international context. It is this latter type of crises – i.e., sovereign debt crises – that are the subject of the recent international debate. In particular, the IMF and several international groups have argued that the resolution of sovereign debt crises would be easier if there were clearer ‘rules of the game’ for such cases.
The current debt restructuring proposals therefore would not directly address the main type of crises that beset the east Asian region in 1997/98 and have been of greatest concern to Australian policy-makers. Fortunately, there has been significant progress is addressing the types of vulnerabilities that led to the Asian crises, and there is also greater awareness of appropriate policy responses to the onset of crises, including the possible imposition of temporary capital controls. Furthermore, the increased focus on cooperative solutions in dealings between sovereigns and their creditors may also be helpful in suggesting procedures or contractual terms by which private sector debt crises can also be addressed more effectively.
The IMF Proposal
Proposals for a sovereign debt restructuring mechanism (SDRM) are based on the view that there is a deficiency in the international financial architecture arising from the absence of a framework for sovereign nations to restructure their debt burden in a prompt and orderly manner when the debt becomes unsustainable. The IMF and others have argued that an orderly path for sovereign debt work-outs is in the interest of all parties. For the sovereign, the uncertainty over the outcome in the current system may induce it to delay approaching its creditors until it is forced to do so, probably after substantial capital flight and the depletion of official reserves. By contrast, a predictable process is more likely to result in it approaching creditors earlier and adjusting its economic policies sooner, thereby minimising the output losses and ensuring an earlier recovery. Creditors, meanwhile, are likely to benefit from a reduction in the magnitude of losses on the value of their claims on the sovereign. And the presence of a restructuring framework would reduce the incidence of bail-outs by the international financial institutions and the moral hazard that results when creditors continue lending in the expectation of such official support.
Proponents of a formal process for debt restructuring argue that the main problem with the current system is its inability to ensure collective action by creditors and agreement on restructurings that are in the interest of creditors as a group. Whereas sovereigns previously relied predominantly on syndicated bank loans for their funding requirements (involving a relatively small number of creditors with similar interests), there has been a shift towards the issuance of bonds traded on international financial markets (involving a diverse and diffuse creditor community). As a result, it may be more difficult to coordinate among creditors and achieve a restructuring. In addition, debt restructurings that are in the interest of creditors as a group may be prevented if individual creditors consider that their individual best interests are served by not participating in the debt restructuring (i.e., choosing to hold out) in the hope of subsequently receiving full repayment in line with their original contracts. A more extreme form of hold-out action is where certain creditors decide to pursue litigation to recover the full value of their contract. Together, these problems are referred to as the collective action problem. In the absence of collective action, individual creditors that might have been willing to participate in a collective solution may have incentives to seek early repayment or take legal or other action against the debtor, leading to a deepening of the debtor's problems.
The IMF's proposal to overcome these and other related problems has drawn on the mechanisms and institutions that exist in many national bankruptcy systems. The IMF's proposal for a SDRM would include allowance for:
- Majority restructuring: A qualified majority (say 75 per cent) of all creditors would be able to agree to a restructuring of the sovereign's liabilities that would be binding on all creditors. Importantly, the voting process would include the claims of all creditors regardless of the type of instrument (bonds, bank loans, etc) so that a comprehensive restructuring of all debt would be possible.
- A stay on creditor enforcement: A temporary stay (or standstill) on creditor litigation could be enforced while negotiations for a restructuring agreement were underway. This would help prevent a ‘grab-race’ by individual creditors seeking to have their debts repaid ahead of others.
- Protection of creditors interests: During the period of the stay, there would be limitations on the debtor's actions to prevent it from harming the interests of creditors, for example by paying non-priority creditors or undertaking adverse policy measures.
- Priority financing: The provision of new financing from private creditors (say for new trade financing) could help preserve the value of claims of existing creditors, and hence could be given priority over the claims of existing creditors.
The purpose of the measures would be to provide sovereign debtors with immunity from legal action during a temporary standstill on debt repayment, during which time the sovereign would undertake restructuring negotiations with creditor representatives and simultaneously implement appropriate economic policies designed to safeguard creditors' interests. The SDRM, as originally outlined, envisaged a central role for the IMF in endorsing the standstill and subsequent restructuring, based on assessments of debt sustainability and the country's economic policies. The Fund's position has since evolved on this issue. It now endorses the view that key decisions be placed in the hands of the debtor and the majority of creditors. The Fund's main role under the mechanism would instead be to develop structures to support the SDRM, perhaps including the creation of an international body to oversee the framework, verify creditor claims, etc.
The implementation of a SDRM would require coordinated action by all countries. One possibility would be for all individual countries to separately modify their relevant national legislation, although this would likely encounter coordination problems. An alternative would be for changes to be made in the IMF's Articles of Agreement to mandate the SDRM. Following the adoption of the new Articles (which would require a 60 per cent majority of all IMF members, including members accounting for 85 per cent of the voting power) the SDRM would then have universal application for all IMF members.
An Alternative ‘Contractual’ Approach
The IMF proposal has been followed by an alternative proposal from Dr John Taylor of the US Treasury. Since the IMF proposal would require changes to legal frameworks, it has been referred to as a ‘statutory’ approach to the problem, whereas the US has proposed an alternative ‘contractual’ approach.
The US proposal would largely retain the current market-oriented approach towards restructuring but utilise some additional tools that would allow a more systematic approach. The approach calls for sovereign borrowers and creditors to insert a package of new ‘contingency clauses’ into future bond contracts, which would describe the process that would be followed if a restructuring proved necessary. The clauses would provide for an initial standstill period to initiate the restructuring talks, provide for collective decisions by a majority of bond holders that would be binding on all bond holders, and cover processes by which debtors and creditors convene and are represented in discussions.
Just as the legal infrastructure for a statutory approach has only been discussed in the abstract, the precise contingency clauses described above also do not yet exist. The US proposal envisages that debtors and creditors and their lawyers would work together and gradually come to a set of clauses that would become the standard clauses for contracts, building on existing ‘collective action clauses’ (discussed below) which are already used in some jurisdictions. Unlike the SDRM, the US Treasury approach would focus on separately restructuring different types of debt and different bond issues. However, there is flexibility in the proposal for it to use so-called ‘super collective action clauses’ whereby such clauses could be placed in all forms of debt, allowing restructuring by a qualified majority of all creditors.
At a broader international level, the Group of Seven (G7) nations has also endorsed a contractual solution that shares many elements in common with the US Treasury model. However, G7 members have also suggested work should continue on developing the statutory-based approach, which could be pursued in a complementary fashion to the contractual approach. At a more specific level, the G7 has endorsed ongoing work on aggregation issues (e.g., how to bring together creditors in different instruments) and the treatment of new private lending.
Private Sector Responses
The views of different private sector groups towards proposed reforms to the international financial architecture have been evolving over time. The response of private sector market participants to reform was initially fairly negative, although it has recently evolved to support for a version of the Taylor proposal.
The initial resistance to any change can be characterised as having four main points:
- First, it has been argued that the current system is actually working quite well. In particular, many private sector market participants have noted that the current system has not prevented sovereigns restructuring their liabilities via exchange offers, in which bondholders voluntarily tender their existing securities for new securities that modify (and effectively reduce) the sovereign's payments structure into a sustainable stream of payments. Pakistan, Ecuador, Ukraine and Russia have all managed to restructure their debts under such exchanges, even in circumstances where there were large numbers of diverse creditors in existence (see Box A).
Box A: Recent Sovereign Debt Exchanges
Ecuador, Pakistan, Russia and Ukraine provide recent examples of countries that have had debt servicing problems that resulted in the restructuring of external debt obligations. In each case they were able to do so via debt exchanges, whereby creditors agreed to exchange their existing bonds for new bonds that reduced debt service obligations. Although there were differences between the cases, they each provide examples of successful market-driven approaches to debt restructuring.
In each exchange the market value of the debt had fallen to low levels at the time of negotiations, and the mark-to-market gains that resulted from each deal provided a ‘sweetener’ to investors that (in addition to cash payments in some cases) facilitated agreement. These gains presumably reflected the perception that the restructuring enabled the sovereign to return to a sustainable debt profile, and are an example of the gains that accrue to creditors from successful debt restructurings, as opposed to messy defaults where the creditors have little prospect of successful legal action to recover the value of their claims.
There were some unique features in each exchange. In Pakistan's case, the debtor had not yet defaulted on the Eurobond issues that were exchanged for longer maturity bonds. Furthermore, although the Eurobonds contained collective action clauses (CACs), these were not invoked to call a meeting of creditors to restructure the terms of the existing bond: creditors were instead persuaded to tender their bonds in a voluntary exchange, rather than risk default. Indeed, there is only one recent case – the exchange of Ukraine's Eurobonds – where CACs have been used to facilitate a deal.
In the case of Ecuador, the defaulted bonds that were involved did not have CACs and required unanimous consent for changes in the bonds' payment terms. However, the US style bond contracts actually allowed a qualified majority of bondholders to change other terms of the contract. Creditors agreeing to the exchange were required to tender their bonds and agree to restructure the non-payment terms in such a way as to make the old bonds particularly unattractive for minority hold-out creditors. This was the first time such ‘exit amendments’ (or ‘exit consents’) had been used in a sovereign restructuring. Some commentators have, however, questioned whether this will be a reliable precedent for possible further use of exit amendments, since US courts might not allow such amendments if they were viewed as excessive.
In the case of Russia, the defaulted US dollar securities were not obligations of the Russian Federation but were the obligations of the (quasi-sovereign) Vnesheconombank that had resulted from the London Club settlement of commercial payment obligations of the former Soviet Union.
Accordingly, one of the factors that helped the debt exchange was that creditors were given new Eurobonds that were true sovereign obligations.
These four cases provide examples of how sovereign debtors have been able to work with their advisors and with creditors to come up with exchange offers that were acceptable to the vast majority of creditors. One interpretation would be that both sides have strong incentives to come to some form of agreement, regardless of the particular terms of the contract that governs their relationship. However, it should be noted that in all cases negotiations took many months and may well have resulted in greater costs to the debtor's economy than might have occurred under an alternative framework. Furthermore, in each case there were a small number of different securities involved. It is unlikely that the favourable outcomes seen in these cases would be easily transferred to more complicated ones, for example the case of Argentina which had more than 80 different external bond issues outstanding at the time of its default in late 2001.
- Second, proponents of the status quo have argued that the perceived problems that have motivated recent calls for reforms are ‘non-problems’. Based on the experience in exchanges that have occurred so far, they argue that collective action problems have not been material and that the problems of hold-out or litigious creditors have not eventuated in practice. They also note that the suggestion that widely held bonded debt is far more difficult to restructure than narrowly held bank debt might not be confirmed by the 1980s debt crisis where banks took up to a decade to agree to debt restructurings.
- Third, it has been argued that there are good reasons for not trying to replicate domestic bankruptcy frameworks in the international system. Behind much of the resistance to change is the notion that attempts to make defaults smoother are misguided, because default is not meant to be an easy process for debtors. Underlying these arguments is the premise from theoretical models that since sovereigns cannot be forced to repay, default must be made so costly (via output losses that result from loss of access to international financial markets) that those who can repay will indeed choose to repay rather than default. (The output losses imposed on those who indeed cannot repay are an unfortunate by-product that is implicitly ignored.) According to this line, sovereign debtors already have certain rights – sovereign immunity, the ability to determine domestic policies with no input from creditors, etc – that do not exist in domestic bankruptcy systems, so there are good reasons not to also give sovereigns some of the protections available to debtors in domestic systems. The proponents of this type of argument have suggested that if any changes are to be made to the international financial architecture, they should be in the direction of strengthening creditor rights, rather than making restructuring easier for debtors.
- Fourth, some market participants argue that changes to the current system could reduce the willingness of investors to provide financing to emerging market countries, with negative implications for economic growth in those countries.
Despite these various reservations about proposed changes, a consensus has recently emerged among major private sector groups. In particular, groups representing investors, international banks and those involved in bond issuance and trading have endorsed a market-based solution along the lines of the US proposal. While endorsing the use of collective action clauses, they have opposed any contractual provisions that would allow for standstills, and have argued more generally for changes that would strengthen creditors' rights. They have also endorsed greater transparency by borrowers and more frequent consultation between debtors and creditor representatives.
A Statutory or a Contractual Approach?
Although debt exchanges have been feasible in recent years, the problems identified by the IMF suggest that changes to the current system could improve the chances of smoother restructurings in future. However, the strong opposition of many in the private sector to the SDRM proposals has been noteworthy, given that the framework is intended to benefit creditors as well as debtors. This raises the question of whether a major change such as this might indeed upset the ‘delicate balance’ between the rights of debtors and creditors.
Fortunately, many of the goals of the SDRM can also be achieved via a contractual approach. Given the choice between a contractual approach and the statutory alternative that would override contracts, it may be desirable to opt first for the contractual approach, particularly since a market-driven contractual approach might be less likely to have undesired impacts on the cost and availability of financing to emerging markets. If the contractual route proved difficult to implement it might then prove desirable or necessary to consider the alternative of a formal SDRM. Accordingly, work on the SDRM could proceed in parallel with the work on a contractual approach.
The general principle guiding changes should be in giving creditors and debtors tools that will not restrict their rights (with the possible exception of ‘rogue’ creditors) but will facilitate agreements that are in both sides' interests. Any changes should address the collective action problems that have been highlighted, yet not impede market incentives (e.g., that investors bear the risk of their investment decisions, and countries that can repay do so).
Greater Use of Collective Action Clauses
A core element in a contractual approach would be greater use of collective action clauses (CACs) to address the collective action problems discussed above. Such clauses are already used in many international bond issues in the Euromarket, and most emerging market sovereign borrowers have bonds outstanding both with and without CACs (see Box B). The most important clauses in this regard are those allowing for a qualified majority of bondholders (say 75 per cent) to modify payment terms, and for greater use of trustees to restrain individual bondholders from seeking repayment of their own claims at the expense of other bondholders.
There has, however, been substantial opposition among the private sector in the United States to wider use of CACs. Indeed, official groups such as the G7, G10 and G22 and some private sector groups have been calling for the use of CACs for several years, with no success in the case of the US market.
Box B: What Are Collective Action Clauses? Are they Costly for Borrowers?
Collective action clauses (CACs) include clauses that allow for:
- collective representation – procedures for bondholders to organise and designate a representative to negotiate on their behalf with the debtor;
- qualified majority voting – which enables changes to be made in the terms of a bond contract without the unanimous consent of bondholders, and thus prevent a small number of dissenting bondholders from blocking an agreement beneficial to the majority; and
- sharing among bondholders – which requires bondholders (generally through a trustee) to share the proceeds of litigation against a debtor with all other creditors, thus reducing the incentive for individual creditors to take independent legal action against the debtor.
‘British-style’ bonds issued in the Euromarket under English governing law almost invariably contain CACs. In particular, they allow a qualified majority (often 75 per cent) of bondholders to vote to make changes to the terms of the bond contract and make these changes binding on all bondholders.
However, bonds issued into the US market typically do not contain CACs, nor do bonds targeted at the German market. Global bonds, which are issued simultaneously into several markets have also followed the US convention and excluded CACs. The contractual terms of ‘American-style’ international bonds typically require unanimous consent before the payment terms of bonds can be changed, and provide few limitations on the ability of individual bondholders to initiate and benefit from legal action on their claims.
Table B1 shows the distribution of sovereign bond issuance over the period January 2001–April 2002.
|Global bonds (without CACs)||45|
|– with CACs||22|
|– without CACs||26|
Source: International Monetary Fund
Since there exist a large number of international bonds issued into the Euromarket with CACs, it is possible to examine if the use of CACs in the Euromarket has influenced borrowing costs until now. The most comprehensive study on this question is a study by Becker, Richards and Thaicharoen (2002) that examines the pricing of bonds with and without CACs in several data sets:
- yields on around 300 bonds trading in the secondary market in June 1998;
- yields on around 490 bonds trading in the secondary market in June 2000; and
- yields on 1,520 bonds at the time of issuance, over 1991–2000.
The evidence suggests that there has been no significant impact of CACs on bond yields, which is consistent with the fact that financial market participants have not hitherto focused on CACs. As the authors note, the use or non-use of CACs has typically been a question of market convention rather than a deliberate decision by issuers, and many sovereigns switch between using or not using them in different bond issues. Similarly, investment banks have not focused on CACs in their analysis of pricing, ratings agencies have not considered CACs in rating decisions, and market reporting services have never reported on CACs as being a factor in the pricing of new bonds.
Looking ahead, it is possible that CACs might yet affect borrowing costs in the event of major changes to the international financial architecture. However, Becker, Richards and Thaicharoen (2002) look at a recent snapshot of yields in the secondary market and find no evidence that CACs affected yields as of 30 April 2002. This analysis is based on 130 bonds from 14 sovereigns with bonds outstanding with and without CACs. Given that this snapshot follows the substantial discussion of the Krueger and Taylor proposals, it suggests that market participants have still not focused on CACs as being an important factor in determining yields.
The opposition of some US market participants to CACs in part reflects inertia. In particular, although there are currently no legal restrictions to the issuance in New York of sovereign bonds with CACs, the market convention has been that bonds sold into the US market (Yankee bonds and global bonds) do not contain CACs. In addition, there is an element of ignorance about CACs. For example, many US investors appear to be unaware that they already hold bonds issued into the Euromarket that contain CACs – for example the Russian Federation bonds included in JP Morgan's EMBI+ benchmark index all contain CACs.
Furthermore, some investors appear to believe that the inclusion of CACs would weaken their rights, when a strong case can be made that the use of standard Euromarket-style CACs could actually strengthen their rights. In particular, bonds issued in the US market may require unanimous approval to change the payment terms on the bond, but they often require only a 50 per cent vote to change the non-payment terms, making it relatively easy for the latter terms to be changed in ways that can be quite harmful to creditors' interests. By contrast, bonds with CACs in the Euromarket have a similar threshold (often 75 per cent) for both sets of terms, which may better serve the interests of creditors.
There is also some reluctance on the part of borrowers to issue bonds with CACs in the US market. In particular, borrowers are concerned that the use of CACs may result in higher borrowing costs. This fear appears to be linked to the misperceptions by investors about the nature of CACs. However, empirical research suggests that there has hitherto been no difference in yields between bonds issued with and without CACs, which is consistent with borrowers having frequently switched between the two types of contractual terms as they switch between different markets.
Although there may be no rational reasons for either borrowers or creditors to oppose the use of CACs in bonds issued in the US market, the use of CACs would represent a change in market convention, and could present ‘first mover’ problems. If these can be overcome, and CACs included in global and Yankee bonds, the share of new emerging market issuance with CACs would rise from about 20 per cent to nearly 70 per cent. The prospects for converting all issuance in the Euromarket to CACs would presumably then be very good, and essentially all new issuance of bonds would then contain CACs. The problem of the stock of outstanding bonds without CACs would not have been addressed, but the share of bonds with CACs would gradually rise, and a well-established framework for dealing with sovereign distress would have been established.
Other Possible Measures
There are a number of other areas where changes to contractual terms may be useful in reducing the risk of debt crises or minimising the costs when they occur. Just as covenants (e.g., that limit the behaviour or aggregate indebtedness of borrowers) are useful in the domestic corporate framework to constrain the behaviour of borrowers, so too could covenants be useful in the sovereign framework. In addition, clauses requiring sovereigns to provide more information about their macroeconomic policies and overall indebtedness may also be useful for both creditors and debtors. The provision of more information about indebtedness may have parallels with the benefits of domestic credit registries that exist in many industrial countries and may decrease the cost of funds to borrowers by removing some of the uncertainty about their financial status.
Moving away from contractual terms, there are clearly other areas where progress can be made in reducing the probability or costs of sovereign debt problems. One particular concern expressed by creditors is the unwillingness of sovereigns to maintain an ongoing dialogue with their creditors, especially as they are heading towards debt problems. Accordingly, a stronger mechanism for creditor–debtor exchanges of information may be useful on an ongoing basis and not just as an element of a formal restructuring mechanism.
The continuing lack of progress in dealing with sovereign debt problems in recent years has been disappointing. In part this may be due to the earlier use of the terms ‘burden-sharing’ or ‘bailing-in’, which evoked strong negative connotations for the private sector. Fortunately the debate appears now to have moved towards finding ways of giving debtors and creditors new tools that may facilitate outcomes that are in both sides' interests.
The recent IMF proposal for a SDRM has been useful in reinvigorating the debate on the problems of sovereign debt restructuring. Given that debate has been going on for more than five years with no actual progress, it will be important to take advantage of the current momentum for change and ensure that some tangible results are achieved. Since a number of private sector groups have now endorsed a version of the Taylor proposal, it will be a good starting point in setting down some of the avenues that will be open to creditors and debtors in resolving sovereign debt problems. However, the alternative statutory approach also warrants ongoing work. Indeed, if modest changes like those in the Taylor proposal cannot be made, then the statutory route may be necessary.
The reforms envisaged in the Taylor proposal and discussed above are only a subset of the changes that may be desirable. Nonetheless, they would be a useful first step in forcing creditors and debtors to be explicit in setting down some of the avenues that will be open to them in resolving sovereign debt problems. If modest changes like these cannot be made, then the statutory route may be necessary.
Although the reforms addressed above should be useful in resolving sovereign debt problems when they occur, they represent only an element of the reforms necessary for reducing the costs of sovereign debt crises. In particular, the output costs that have been seen in sovereign debt crises are presumably more the result of deeper macroeconomic problems than of the lack of a good restructuring mechanism. Episodes such as those in Latin America in the 1980s, Russia in 1998, and Argentina in 2001/02 are crises to which various macroeconomic factors contributed. Debt restructuring in these circumstances may never be easy, because debtors are unlikely to be able to put forward proposals that will be acceptable to creditors until they have re-established some form of macroeconomic stability. Real progress in reducing the costs of sovereign debt crises will require that all players – the international financial institutions, the major industrial countries, policy-makers in emerging markets, and creditors – address unsustainable situations more quickly. In particular, there will need to be greater domestic willingness to admit problems (especially with exchange rate regimes), and policy changes by the international financial institutions and major countries that make it more credible that they will not provide funding in unsustainable cases. This would reduce the incentives (the ‘moral hazard play’) for markets to continue providing financing in such cases.
Finally, although the recent debate has only covered cases of sovereign debt problems, it is possible that it may have implications for dealing with private sector debt problems. The cases of the coordinated rollovers of Korean and Brazilian bank debt provide examples of how debt rollovers can be in the interests of both debtors and creditors. However, in both cases there were collective action problems that required some public sector coordination. Indeed, in the case of the Korean crisis, some private sector participants have commented that they wished that the official sector had been quicker to intervene to help the private sector resolve its collective action problems. It is to be hoped that the experience in drafting new contractual clauses for sovereign debt problems may suggest new ways of overcoming private sector debt problems too.
This article was prepared by Anthony Richards, Darren Flood and Mark Gugiatti, International Department. 
See AO Krueger, ‘A New Approach to Sovereign Debt Restructuring’, International Monetary Fund, April 2002. 
In the case of Brazil in early 1999, there was also a coordinated rollover of bank debt. 
Of course, in all three cases the crises did affect the public sector substantially through the large run-down in foreign reserves from exchange market intervention, and the deterioration in the fiscal position from lower tax revenues and the costs of recapitalising the banking sector. 
See, e.g., JD Sachs, ‘Do We Need an International Lender of Last Resort’, Graham Memorial Lecture, Princeton University, 1995. 
See JB Taylor, ‘Sovereign Debt Restructuring: A U.S. Perspective’, Remarks at the Institute for International Economics Conference on ‘Sovereign Debt Workouts: Hopes and Hazards’, Washington, 2 April 2002. 
One creditor group has already proposed a set of model covenants, albeit with voting thresholds that appear excessively high. In addition, a G10 working group is developing model clauses. 
See, e.g., MM Chamberlin, ‘Revisiting the IMF's Sovereign Bankruptcy Proposal and the Quest for More Orderly Sovereign Work-Outs’, Remarks at the Institute for International Economics Conference on ‘Sovereign Debt Workouts: Hopes and Hazards’, Washington, 2 April 2002. 
Litigation was not a factor in the debt exchanges undertaken in recent years by Ecuador, Pakistan, Russia or Ukraine. However, an important exception is a court ruling in 2000 in Elliott Associates v Peru. In that case, a creditor held out from participating in an earlier restructuring of debt guaranteed by the government of Peru, and instead pursued litigation in order to enforce the original contractual obligations. While the matter was ultimately settled privately, a European court did issue a legal interpretation in favour of the hold-out creditor that argued that the sovereign was legally prevented from paying one group of creditors (those involved in the restructuring) ahead of others (the hold-out). While the implications of the case are unclear (as no legal precedent was set and the interpretation has been widely criticised), the decision has added an element of uncertainty to how future debt restructurings may evolve, by potentially strengthening the position of hold-out creditors. 
See the 3 June 2002 joint letter from representatives of the Emerging Markets Traders Association, the Institute of International Finance, the International Primary Market Association, the Bond Market Association, the Securities Industry Association, and the Emerging Markets Creditors Association to US Treasury Secretary O'Neill, available at <http://www.emta.org.ndevelop/oneill.pdf>. 
See T Becker, AJ Richards and Y Thaicharoen (2002), ‘Moral Hazard and Bond Restructuring: Are Collective Action Clauses Costly?’, forthcoming in the Journal of International Economics. An earlier version of this paper was published as IMF Working Paper WP/01/92. 
Footnotes Box B
Some of the material in this Box is taken from T Becker, AJ Richards and Y Thaicharoen (2002), ‘Moral Hazard and Bond Restructuring: Are Collective Action Clauses Costly?’, forthcoming in the Journal of International Economics. An earlier version of this paper was published as IMF Working Paper WP/01/92.