Information about Asia and the Pacific Asia y el Pacífico
Asian Financial crises

Chapter 36 Bailing in the Private Sector*

International Monetary Fund
Published Date:
January 2001
  • ShareShare
Information about Asia and the Pacific Asia y el Pacífico
Show Summary Details

A particular concern of many critics of the existing international financial architecture is that official support has been used to bail out investors. In Mexico in 1995, Korea in 1997, and Russia in 1998, official funds were used to repurchase and retire short-term debt that private investors were unwilling to hold. Having benefited from high interest rates while their money was in place, creditors were effectively protected from capital losses when it came time to sell. The moral hazard thereby created provided an obvious incentive to engage in even less prudent lending, setting the stage for still larger crises and still larger bailouts. It would be better from a public-policy standpoint, the conclusion follows, were private investors and international banks in particular forced to “take a hit.”

The Mexican crisis illustrates the problem. The government entered its crisis at the end of 1994 with some $28 billion of short-term foreign-currency-indexed government obligations (tesobonos) about to mature but only $6 billion of international reserves. Once confidence was lost, no investor had an incentive to make available additional foreign exchange. Had international assistance not been provided, Mexico would have been forced to suspend redemption of these debts, inflicting significant losses on its creditors and risking damage to its credit worthiness. Instead, the government used its U.S. and IMF loans to retire its tesobonos at full value as they matured.

In Korea the mechanism was more involved but the result was the same. Foreign creditors who had extended short-term loans to Korean banks attempted to withdraw their balances all at once. Those short-term credits far exceeded the government’s foreign reserves. The Bank of Korea’s reserves fell from $31 billion at end October 1997 to $21 billion at the end of December, and more than half of this latter amount was immobilized in the form of deposits with foreign branches of domestic banks.1 Figures vary for the short-term foreign currency obligations of the Korean financial sector, but one seemingly reliable estimate puts them at about $26 billion in December.2 Had no official assistance arrived, the Korean authorities would have been forced to declare a moratorium. Instead, the U.S.- and IMF-led loan enabled the government to inject more credit into the banking system, deposit more reserves at overseas branches of Korean banks, and keep interest rates lower than would have been feasible otherwise, while maturing foreign credits were paid back in full. The result was to replace a significant share of those foreign credits with official funds.3

One rationale for the Korean package sometimes heard in Washington, DC is that the Korean Peninsula is too important geopolitically for the South’s economy to be left unaided. Similar arguments have been made about Indonesia, which sits astride some of the world’s most important ocean-shipping lines. Russia being “Indonesia with nukes,” the Korean and Indonesian precedents gave investors confidence that Russia would receive similar assistance. This rhetoric may seem exaggerated, but the fact of the matter is that in the wake of the IMF’s Asian support programs the international bankers and others who poured money into Russia referred to positions in Russian GKO’s (treasury bills) as the “moral-hazard play.” There were sizable inflows into Russia in the months following the Asian rescue packages, followed by financial difficulties, a crisis, and an IMF package in the summer of 1998.4

In practice, of course, not all investors were shielded from losses. In Mexico, the prices of tesobonos tumbled before it became clear that the Mexican government would be able to retire them, and many investors who scrambled out of the market did so at a loss. In Mexico in 1994 and even more so in Asia in 1997, official support did not avert major declines in stock, bond and real estate prices, and investors in these markets incurred extensive losses. None of this is to deny that moral hazard is a problem and that existing arrangements for handling crises are deficient, but to caution that one should not exaggerate the extent to which investors have been shielded.

The real dilemma is presented by bank creditors. Foreign funding of domestic banks in the form of deposits and deposit-like instruments is highly liquid. Deposits have a fixed face value. Banks being key to the stability of a country’s payments and credit system, governments are understandably reluctant to contemplate any treatment of these claims that might threaten their provision. These facts make it extremely difficult to write down foreign claims on domestic banks. It would be nice if foreign bank creditors could be made to “take a hit”—if they would agree to reschedule and write down their claims. But so long as they have the option of fleeing and bringing down the banking system in their wake, governments will contemplate this option only in extremis.5

If the tendency for official support to shield creditors from losses and to thereby encourage imprudent investor behavior was not sufficient grounds for concern, there is also the question of whether international assistance as currently constituted can protect the recipients from serious damage. All too often, IMF-led rescues are ineffective in containing a panic because the fund’s resources are limited and doled out a drop at a time. Unlike a domestic lender of last resort, the IMF cannot print money, nor would its principal shareholders let it.6 They are reluctant to do so because, compared to a government with regulatory powers dealing with a problem bank, the IMF has less ability to force corrective action on its members. Since the fund does not posses the other legal and supervisory powers of a government, in other words, it lacks the leverage to ensure that it will get its money back. This is why the IMF is endowed with fewer resources, why it lends less freely, and why it relies more heavily on continuing performance criteria when disbursing its funds. It is why the IMF hesitates to front-load its disbursements, requiring evidence that policy reforms are in train before it releases each additional bit of finance.7

For all these reasons, official support is generally less than meets the eye. The international community committed $57 billion to Korea, for example, but released only $13.2 billion by the time the crisis there reached its height.8 As a result, countries receiving international assistance and trying to avoid a debt moratorium, standstill, or “pause” have to hike interest rates both to lure back foreign investors and to satisfy the fund.

Why are governments so willing to put their economies through the wringer? Why in particular do they hesitate to suspend payments and negotiate an agreement with the creditors to restructure the debt? Doing so would certainly discourage imprudent lending. More importantly from the domestic of view, governments could avoid putting their economies through the wrenching deflationary consequences of the adjustment required for the maintenance of external debt service.9

This option is shunned because governments and the international policy community regard the collateral damage as too severe. Countries that suspend payments and attempt to restructure find it difficult and costly to reach an agreement with their creditors. Their reputations are damaged. They find it harder to borrow on international capital markets subsequently.10 However expedient the short-run policy, most governments regard the long-term consequences as unsupportable.11

Thus, “bailing in” the private sector—ensuring that private investors also “take a hit”—presupposes changes in institutional and contractual arrangements that make it palatable for governments to declare a moratorium and restructure their debts. It requires changes in the international financial architecture.


Two classes of measures can be considered under this heading: measures to discourage bank-to-bank lending, and the negotiation of international credit lines.

1.1 Discouraging short-term borrowing

The most direct way to avoid letting foreign creditors off scot free is not to borrow from them in the first place. This is not meant sarcastically. Short-term foreign credits, and the short-term foreign credits of domestic banks in particular, pose a special problem because they are so liquid, making it especially easy for those extending them to scramble for the exits. They pose a special problem because the institutions dependent on them are central to financial stability. The need to preserve the stability of the banking system thus makes it hard to impose on its creditors a share of the adjustment burden. This provides an argument for discouraging reliance on short-term foreign credits to the banking system in the first place. It is an argument for raising the Basle risk weights for foreign bank lending and for keying those weights to the source of banks’ funding as well as the riskiness of their investments. In markets where political pressure prevents capital from being written down, it is an argument for taxes or quantitative ceilings on short-term foreign funding. And where nonfinancial firms can do the borrowing and pass the proceeds on to financial intermediaries, it is an argument for using measures like those employed by Chile, whose government, while applying a tax to all capital inflows, structures it so that it falls most heavily on short-term inflows. If administered successfully, such measures would increase foreign portfolio investors’ reliance on stocks, bonds and other long-term instruments, on which they would automatically suffer capital losses in the event of a financial crisis.

1.2 Standby lines of credit

A second approach would be for governments to negotiate standby lines of credit. Foreign banks would agree to make these credit lines available in return for a commitment fee. Since foreign bank creditors would no longer be able to eliminate their exposure to the country in question, they would be more predisposed to negotiate a restructuring plan. From the standpoint of the borrowing countries, these credit lines would provide additional resources to insure against shocks to investor confidence. Both Argentina and Mexico have negotiated such agreements with foreign commercial banks despite the fact that neither country has an investment-grade sovereign credit rating. Notwithstanding, Mexico’s recent tough negotiations to renew its credit lines, this suggests that other countries could do the same.12

The main weakness of these arrangements is that the banks will be able to hedge their exposure. At the same time they provide additional credits, they can contract to sell short government bills and bonds. The country will then have no additional financial resources for propping up its banking system and coping with the other consequences of the crisis.

This constraint can be relaxed were the IMF and the regional development banks to supplement the contingent facility.13 Like the commercial banks, they too could make credit lines available to governments in return for a commitment fee and an interest charge. They could take guidance on the pricing of the facility from the banks, charging the same fees and requiring the same collateral. Only if a country qualified for a credit from the commercial banks would this be supplemented by the multilaterals. Thus, the conditionality would be negotiated ex ante, and the facility would available only to countries that had already taken reasonable steps to establish and preserve their credit worthiness.14

The Supplemental Reserve Facility (SRF) approved by the IMF’s Executive Board in December 1997 is a step in this direction. The SRF is designed to provide financial assistance to a member suffering from a temporary loss of market confidence and exceptional balance of payments problems (presumably reflecting turbulence in other countries), and where there is the expectation that the problem can be corrected in short order. While access is not negotiated in advance, the idea is that these extra funds, sufficient to offset the impact on domestic markets of a sudden decline in market confidence, would be disbursed quickly (in conjunction with an existing standby or extended arrangement).15

The limitation of this approach is that the resources of the IMF and other multilaterals are small relative to those of the market and that they can be disbursed only following the negotiation of new conditionality.16 The IMF cannot be transformed into a true lender of last resort that can provide unlimited resources to a financially distressed government; its principal shareholders would not let it borrow or print the money needed to underwrite intervention on this scale. Consequently, the fund is concerned to husband its limited resources and would be reluctant to agree to automatically disburse resources without attaching (and therefore having to negotiate) new conditions which maximized the probability of repayment.17 Whether such funds can be disbursed with sufficient speed is therefore an open question. For all these reasons, these kinds of schemes can operate only at a limited scale and speed. They can provide only limited insurance financial shocks.


The most important changes that could be made to facilitate orderly workouts of international debts are the incorporation of new clauses into loan contracts. This section considers the cases of bonds and bank loans in turn. It then discusses IMF lending into arrears and the establishment of standing committees of creditors.

2.1 New provisions in sovereign bond covenants

Money center banks account for so much international lending because they have well-developed capacities to gather information about foreign borrowers and have cultivated long-term relationships with their clients which provide leverage when it comes time to collect on loans.18 That said, changes in technology and market organization suggest that securitized instruments (bonds and derivative instruments based upon them) will account for a growing fraction of international lending over time. Improvements in information and communications technologies tend to undermine the informational advantage of banks. Advances in financial technology enable individual investors to unbundle and hedge credit and currency risks. Meanwhile, the growth of mutual funds, pension funds, hedge funds and other collective investment vehicles creates a demand for securitized investments. Securitization has made great strides in the advanced-industrial countries, where observers speak regularly of the shrinking market for banking services. International bonds are already accounting for a growing fraction of new commitments (bonds plus bank loans). One can confidently predict that bonds will account for a growing share of portfolio investments in emerging markets in years to come.

On balance, this technological revolution encouraging international lending to flow through bond markets rather than banks is probably a good thing. Insofar as it reflects improvements in the information environment, which render the market less dependent on banks as vehicles for surmounting informational obstacles, it implies a more efficient allocation of resources. And insofar as emerging market debt becomes less concentrated in the hands of the major money center banks, it leaves the latter less vulnerable to international debt crises.

But like many good things, the securitization of emerging market debt does not come without costs. Here the costs are additional problems at the renegotiation stage. Securitization means a significant increase in the number of creditors, small creditors in particular, multiplying collective-action problems.19 Moreover, restructuring a sovereign bond issued in the United States (more precisely, under the legal provisions that govern bonds issued in that country) typically requires the unanimous consent of the bondholders, which can be a formidable hurdle. In the event of default, each bondholder has the right to sue the issuer, and no bondholder can be forced to agree to new terms by other bondholders. Unlike syndicated bank loans, there are no sharing clauses requiring individual bondholders to share any amount recovered with other bondholders and thereby discouraging recourse to lawsuits. There are no counterparts to the central banks and regulators that used their powers of moral suasion to encourage cooperative behavior by the members of commercial bank syndicates in the 1980s.20 Neither do sovereign issuers have recourse to a bankruptcy filing, under which they would be protected from the threat of lawsuits and in the context of which terms could be imposed on minority creditors. Agreement being difficult to reach, issuers are understandably reluctant to contemplate restructuring.

In the event they do, “vultures” (offshore hedge funds or large individual investors) then have an incentive to purchase bonds from less patient investors and to threaten lawsuits designed to attach the debtor’s assets. Wishing to avoid expensive and embarrassing litigation, the debtor may then feel compelled to buy them out at full price. Taken to the extreme, this suggests that maverick creditors will buy up all the defaulted debt and litigate to prevent sovereign issuers from settling for less than 100 cents on the dollar. Restructuring that involves writing down principal and interest will then be impossible.

One need not subscribe to this extreme version of the argument to see that the provisions governing the issuance of sovereign bonds complicate the process of renegotiation and restructuring.21 It is hardly a mystery that, under present arrangements, governments are reluctant to go this route.

Fortunately, a solution is at hand, having been suggested two years ago by the G-10 in its report, Resolving Sovereign Liquidity Crises and now having been echoed by the G-22 in The Report of the Working Group on International Financial Crises. G-10 Deputies recommended making it easier to undertake negotiations by altering the provisions of loan contracts to include majority voting, sharing, and non-acceleration clauses.22 This would prevent maverick creditors from resorting to lawsuits and other means of obstructing settlements beneficial to the debtor and the vast majority of creditors. (To their recommendations one might add the idea of minimum thresholds for creditor lawsuits, requiring that a certain minimum percentage of creditors, say ten or twenty-five percent, would be required in order to take legal action against the creditor.23) The addition of such clauses to bond contracts is the only practical way of creating an environment conducive to flexible restructuring negotiations. It is infinitely more realistic than advocating some kind of supernational bankruptcy court empowered to cram down settlement terms.

Some object that such provisions, by making it easier for developing countries to wriggle out of debt contracts, would only increase borrowing costs.24 For those who believe that moral hazard and other market imperfections cause governments to rely excessively on foreign borrowing, this is not undesirable. Others question whether borrowing costs will in fact rise.25 To be sure, majority-voting, sharing, and non-acceleration clauses make it easier to renegotiate defaulted debts, but if this permits a long deadlock to be avoided and renders the majority of investors better off, there is no reason why they might shun bonds with these features. Small bondholders, who lack the resources to sue, might be rendered better off if such clauses averted a long period when interest was not paid and bond prices were depressed while the government and maverick creditors fought their war of attrition. Institutional investors might be better off if, in the absence of this market-based solution, they came under pressure from their governments to cut a deal.

It is important to understand that the normal presumption that “if we see it, it must be optimal” is open to question in the present context. Those who argue that the prohibition on majority voting to restructure the terms of a loan is the market’s way of strengthening the bonding role of debt are ignorant of the measure’s history. In fact, the law was introduced not by proponents of the market but by individuals deeply suspicious of it. William O. Douglas championed the measure in the wake of the debt defaults of the early 1930s as one of a number of laws to protect small investors from victimization by securities houses. This peculiar history underscores that there is no necessary reason to retain these archaic measures. So does the fact that bonds issued under the provisions of the laws of Great Britain, a country lacking this one’s populist tradition, have more renegotiation-friendly provisions. There, bondholders are represented by a trustee and cannot sue individually.26 British bonds also provide for the binding of all bondholders by a majority vote at a bondholder meeting.

According to the G-10 report, new provisions are to be introduced into debt instruments through a “market-led process.” Governments are to trumpet the virtues of new clauses but to otherwise take no action. They are to hope that the markets will see the light.

But if changes in contracts were so easily adopted, the markets would have done so already. That no progress has occurred in the intervening years suggests that there are significant obstacles to market-driven reform. One is the adverse signaling effect. If only some issuers include qualified-majority-voting clauses in their loan agreements, creditors may suspect that those debtors regard it as likely that they will have to restructure in the not-too-distant future. The qualified-majority-voting clause will be regarded as a negative signal that the borrower is less than fully committed to servicing his loan, much like a request for a prenuptial agreement. This can allow inefficient arrangements, put in place for historical reasons long past, to become locked in.27

The G-10 report, perhaps in a desire to look market friendly, says little about this dilemma. At one point it acknowledges the first-mover problem and suggests that official support for contractual innovation should be provided “as appropriate” but fails to elaborate. A more pro-active approach is now required. The IMF should urge all its members to adopt majority-representation, sharing, non-acceleration, minimum legal action threshold, and collective-representation clauses (where these last provisions make provision for an indenture trustee to represent and coordinate the bondholders). It should recommend that members require that all international bonds include such provisions as a condition for being admitted to domestic markets (including under Rule 144A).

To be sure, this is no panacea. Private placements would not be affected. New provisions could be added to existing loans only through a voluntary exchange of old bonds for new ones. Not only might some bondholders resist, but any one country that attempted to be the first to carry out the exchange might be seen as signaling that it was contemplating imminent default and precipitate a crisis.

All this means the incorporation of sharing, majority-voting and non-acceleration provisions into bond covenants will be slow. But slow progress is better than no progress.

2.2 New provisions for bank credits

Short-term credits extended by one bank to another are a more difficult case. Since interbank loans are not governed by formal contracts, renegotiation cannot be eased by altering contractual provisions. This can be gotten around if countries adopt laws limiting the terms and conditions under which short-term loans to their banks might be repatriated. Robert Litan and others urge countries to enact legislation imposing an automatic reduction of the principal of all foreign currency loans extended to banks in their countries that are not rolled over in the event of a crisis.28 Foreign creditors could still get out, but only at a loss. The prospect of that loss would strengthen their incentive to stick around, to address their collective action problem, and to restructure the debt.

If this legislation is only passed when the crisis strikes, such initiatives have no advantage vis-á-vis current arrangements. Nothing now prevents countries from freezing or writing down foreign loans to their banks. In any case, governments’ own behavior suggests that they fear action to freeze bank claims would provoke flight by other foreign and domestic creditors, forcing the imposition of across-the-board exchange controls.29 They regard this as too damaging to their reputations for financial probity and to their countries’ ability to borrow.

If the idea is that such legislation should be adopted in advance of any crisis, then the measure is likely to be much more demoralizing to lenders than the addition of majority voting clauses to bond covenants. If new clauses are added to bond covenants, the decision to halt interest payments will still be in the hands of the individual corporate, financial or governmental borrower. The write-down of principal will be determined on a case-by-case basis in negotiations between the debtor and its creditors. In contrast, the obligatory “haircut” for foreign bank creditors would apply across the board. Foreign creditors would be especially alarmed if, as is likely, the circumstances under which the new law was triggered were left to the government’s discretion. Alternatively, if the trigger was the announcement of an IMF program, as some authors recommend, the merest hint that a government was exploring the possibility of obtaining help from the fund would provoke flight by its foreign bank creditors, precisely the outcome that the measure was intended to prevent. More generally, foreign bank creditors worried about a mandatory haircut will be tempted to flee at the first sign of trouble. Once the provision is triggered, of course, they will have an incentive to stay in, but imagine their incentive to get out before the trigger is pulled. This perverse effect has the potential to transform small crises into big ones. Then there is the familiar “after you, Alphonse” problem, that no country will want to be first to impose such legislation for fear of signaling that it is worried about a pending crisis. Finally, foreign banks are likely to respond to the measure by channeling their lending through the offshore branches of the debtor’s banks. Creditors would then dispute the applicability of the developing-country law and appeal to their own courts in the effort to attach the assets of those offshore branches.

For all these reasons, the best way of dealing with the special problems created by short-term bank-to-bank lending is to discourage excessive reliance on this form of funding in the first place.

2.3 IMF lending into arrears

IMF policy through most of the 1980s was to lend to a country that had fallen into arrears on its external debts only after it had reached an agreement in principle with its creditors. The notion was that the fund should provide assistance only if commercial banks contributed to burden sharing by, in part, clearing away the country’s arrears.30 The debt crisis of the 1980s, a protracted affair, raised doubts about this approach. The banks, their balance sheets strengthening as they drew down their Latin American exposure, hardened their positions. Rather than the IMF using this policy as a lever to encourage burden sharing by the banks, the banks realized that they could use it as a club in their battle with governments. If countries refused to settle on favorable terms, the banks could veto new IMF money in addition to denying their own.

Consequently, in the late 1980s, in a departure from past practice, the IMF contributed to the pool of money used to retire nonperforming bank debts and replace them with Brady bonds. Since 1989 the fund has had a de facto policy of providing support for a member’s adjustment efforts after the emergence of arrears but before an agreement had been reached between the debtor and its creditors, so long as the country in question was engaged in good-faith negotiations and making a serious effort to adjust. In more than three dozen instances the IMF has lent in support of adjustment programs before a member has cleared away its arrears to commercial banks. Lending into arrears can provide working capital for an economy that is making an adjustment effort and—analogous to the “debtor-in-possession” financing provided under U.S. corporate bankruptcy procedures—avert unnecessary damage to its economy. Insofar as collective-action problems, exacerbated by rules requiring the unanimous assent of creditors to the terms of any restructuring plan, render negotiations between governments and their creditors excessively protracted, IMF support to a country in arrears can help bring creditors to the bargaining table.31 Insofar as sovereign debtors and the international community generally see the temporary suspension of payments, followed by negotiations to restructure, too difficult and costly to pursue, it may then be desirable for the IMF to tip the balance in this way, opening up debtor-creditor negotiations as a viable alternative to regular IMF rescues.

The G-10’s 1996 report acknowledged that lending into arrears was a way for the IMF to expedite restructuring and asked the fund to contemplate extending the policy from commercial bank loans to bonded debts.32 While the fund has not made lending into arrears standard policy, it seems prepared to move in this direction.33

The risk is that creditors might sue in an effort to attach the proceeds of the loan. While private debtors can seek shelter from a creditor lawsuit in the bankruptcy court, sovereign debtors have no such recourse. If the creditors are commercial banks, they will be subject to moral suasion by their central banks and regulators and are unlikely to seek to attach IMF assets in this way.34 But if they are bondholders, as will increasingly become the case as securitization proceeds, the danger is greater. The fear if not yet the reality of lawsuits is real.

Perhaps IMF balances could be transferred from the fund’s accounts to the central bank in question and the courts would recognize the central bank as a legally separate entity from the government and therefore not responsible for the latter’s debts.35 Perhaps attempts to attach the proceeds of the loan before the fund has disbursed them would be rejected by the courts on the grounds that its Articles of Agreement make the IMF immune from legal process.36 Perhaps the courts could be swayed by a brief filed by a creditor-country government arguing against attaching IMF resources. Perhaps the creditors, knowing that they would have to do battle with the U.S. government and the IMF, would be reluctant to throw down the gauntlet.37

That said, what will happen is uncertain. This has prompted discussions of whether Article VIII.2(b) of the fund’s articles of agreement should be amended to give official status to a country’s standstill of payments and to shelter its government, and any IMF resources lent into arrears, from legal action. Article VIII.2(b) allows countries to apply exchange controls in response to balance-of-payments problems without violating their obligations to the IMF. That article would have to be given an authoritative reinterpretation by the fund’s executive directors or more likely be amended with the consent of countries commanding eighty percent of the fund’s voting power for it to give sanction to a standstill on external debt as opposed to the imposition of exchange controls.

It is unlikely that the requisite majority would agree to vest such powers in the hands of an international organization.38 Not only would market participants oppose empowering the fund to interfere so extensively with private debt contracts, but the fund would not be seen as possessing the impartiality and detachment of a bankruptcy judge. Among other things, it might have made loans to the country itself. The idea of amending Article VIII.2(b) to empower the fund to declare a standstill would be rejected as soon as it was considered on Capitol Hill.39

Fortunately, there exist more practical alternatives. One is a limited amendment to the fund’s articles of agreement in which its members agree to give immunity in their national courts to the fund’s own disbursements and transactions. It is not obvious why, in order to achieve this limited goal, Article VIII.2(b) should have to be amended to give the fund the power to halt all legal action against a government. A complementary approach would be for countries to amend their own sovereign immunities laws to allow their courts to stay attempts to attach sovereign assets.40 In the United States and United Kingdom, creditors are already prevented from attaching certain sovereign assets even when the sovereign has waived its immunity, as is commonly the practice when governments float international bonds. It would be desirable to clarify these provisions and for other countries to emulate them.

2.4 Standing committees of creditors

A final change in the international financial architecture to help bail in the private sector would be to create standing committees of creditors. Restructuring negotiations are most difficult and protracted when information is least complete. Where the preferences and capacities of all parties are common information, agreement should be immediate.41 The more asymmetric the information environment, the more likely are debtors and creditors to fight a lengthy war of attrition. Establishing a standing committee of representatives of the various classes of creditors—bondholders, banks, and other institutional investors—that meets regularly with borrowers would open lines of communication and help to overcome information problems.42

A standing creditors’ committee would also reduce transaction costs in times of crisis. When a crisis erupts and debt service is halted, negotiations cannot proceed until the creditors have been identified, which is time consuming when the process starts from scratch. The existence of a standing committee in continuous contact with its constituents would ease this difficulty. Next the debtor must decide with whom to negotiate—that is, who speaks for the creditors. The existence of a standing committee would answer this question in advance. Finally there is the need to gain the assent of a majority of creditors to the restructuring plan and to buy out those who refuse. The existence of a standing committee on which various classes of creditors interact regularly would create peer pressure for agreement and facilitate the extension of any required side payments. This last point is important: these committees would only offer nonbinding recommendations to the bondholders, who would then have the right to accept them or reject them. They would play much the same role as bank advisory committees in the syndicated bank debt crisis of the 1980s.

The difficulties created by the absence of these committees are evident in the recent experiences of South Korea and Russia. In Korea, the problem in the last week of 1997 was to get the banks to roll over their maturing short-term loans, to accept a delay in making interest payments, and to agree to the principle of converting those short-term credits into long-term loans. The Korean government and the banks reached that agreement by the skin of their teeth. With the help of Bill Rhodes’ Rolodex the relevant bankers were located, pulled from their Christmas dinners, and thrust into negotiations.43 Russia’s experience in August 1998 following its suspension of payments further illustrates the confusion that can arise when there exists no committee of creditors.44 First, the Russian authorities met with a small group of Russian and foreign banks to discuss the formation of a creditors’ committee. Next, it was decided that the committee would be formed only after the authorities had somehow managed to draw up a full list of creditors. Finally, there were a variety of disagreements over the composition of the creditors’ club.

To be sure, these arrangements will grow more complex with the shift from bank to bond finance. That shift will increase the number of interested parties and vest additional power in the hands of a class of creditors less susceptible to moral suasion by their central banks. But it will erode the effectiveness of Rhodes’ Rolodex even more dramatically. What was possible in Korea will not be possible again. Standing committees will become essential.

One sometimes hears the objection that experience with corporate debt workouts suggests that committees of creditors can be quickly constituted if and when the time comes.45 In fact, the situation for corporate bonds is quite different from that affecting sovereign debts. Most corporate bonds are issued in the United States through an indenture trustee. The indenture trustee is responsible for acting as a communications center to coordinate the bondholders. It must communicate with the bondholders and follow the instructions given by a majority. It is the bondholders’ representative in negotiations with the debtor and the court. However, the Trust Indenture Act of 1939 exempts securities issued by foreign governments, their subdivisions and municipalities. Sovereign bonds are typically issued through a fiscal agent rather than an indenture trustee. The fiscal agent has a much more limited role, and its obligations are mainly to the issuer, not the bondholders. Its responsibilities do not extend to acting as a communications center or attempting to coordinate the bondholders.46

In fact, standing committees of creditors were the channel for disseminating information and organizing negotiations when bond finance was last important, from the late nineteenth century through World War II.47 At first, ad-hoc bondholders’ committees were formed in response to each interruption in debt-service payments. Predictably, these committees had trouble establishing contact with a majority of bondholders and opening lines of communication with foreign debtors. In Great Britain, the leading creditor country of the era, the situation was regularized in 1868 by the creation of the Corporation of Foreign Bondholders. Comprised initially of representatives of banking firms and brokerage houses, its governing body, the Council, was expanded in 1898 to include several individual bondholders and a representative of the London Chamber of Commerce. The Council became the recognized spokesman for the bondholders and their representative in negotiations, working closely with the underwriting banks and the London Stock Exchange. The same evolution occurred elsewhere with the establishment of standing bondholders committees in Paris, Amsterdam and Berlin before World War I and in the United States in the 1930s. These committees fell into disuse after World War II because the international capital market was slow to recover from the debt crisis of the 1930s and then because bond finance was superseded by syndicated bank loans. Now, however, bonds are back, and the creditors are more numerous and heterogeneous than when international lending was the domain of bank syndicates.

The idea of creditors’ committees was resuscitated in the wake of the Mexican crisis by Rory Macmillan and by Richard Portes and myself.48 To date, however, the investor community has been reluctant to act. It fears that standing committees would make it too easy for debtors to initiate restructuring negotiations, making it too tempting for them to suspend debt payments. It is better, in the narrow self interest of the creditors, for there to be no one on the other end of the line to pick up the phone.

For policy makers wishing to create a viable alternative to large-scale bailouts of crisis countries and for whom the difficulties of debtor-creditor negotiations render moratoria and restructuring unacceptably difficult and painful, standing committees of creditors are desirable precisely because they make it easier for debtors to initiate negotiations. Their formation is important for creating a viable alternative to ever-more-costly bailouts and disastrous Russian-style defaults, neither of which is an acceptable option.

The creation of such committees would require moral suasion and lobbying by G-7 governments, central banks, and the IMF. There would be nothing unprecedented about their involvement. The Corporation of Foreign Bondholders received a parliamentary charter and other forms of official support. Its U.S. equivalent, the Foreign Bondholders Protective Council, was formed only with the encouragement and support of the U.S. State Department.49 These are precedents that should be followed. Specifically,

Rory Macmillan suggests the creation of two standing committees, a resurrected Foreign Bondholders Protective Council in New York to represent and coordinate the holders of government bonds issued under New York law and submitting to New York courts, and a resurrected Council of Foreign Bondholders to represent and coordinate holders of government bonds issued under English law. Because the vast majority of bonds are subject to either New York or English courts and law, two creditors councils would go a long way toward solving the problem.


Effectively bailing in the private sector requires changes in institutional and contractual arrangements to make it economically acceptable for governments to negotiate an orderly restructuring of their debts. For those who regard ever-larger IMF bailouts as undesirable because of their escalating cost and the moral hazard they create, but are also prepared to acknowledge that under present arrangements governments regard the costs of a unilateral moratorium as too severe, this is the only realistic alternative.

It is important to emphasize that there is unlikely to be a simple and wholly satisfactory solution to this problem. A moratorium on debt repayments should be unattractive; otherwise, the sanctity of loan contracts would be jeopardized. Contracts and institutional arrangements are structured to make the suspension of debt service painful precisely in order to keep borrowers from walking away from their debts. Were this too easy, the capital market would not function at all. Moreover, it is unrealistic to imagine the creation of an international bankruptcy court with the power to cram down settlement terms on debtors and creditors. And there are good reasons why the IMF cannot be transformed into a true international lender of last resort.

Notwithstanding these difficulties, steps can be taken. The most important of these include limiting short-term foreign-currency-denominated borrowing by banks (by, if necessary, placing a holding period tax on all portfolio capital inflows), adding sharing clauses to loan contracts, establishing creditors’ committees, entertaining the possibility of IMF lending into arrears, and amending sovereign immunity laws. History shows that the market left to its own devices cannot provide a perfect solution to international debt problems, any more than the market can efficiently provide for the liquidation and reorganization of financially distressed domestic companies in the absence of an insolvency law. Markets, to operate efficiently, need institutional support. In the present context, the relevant institutional support must be provided by both governments and international organizations.


Some of the material in this section draw from my book, Toward a New International Financial Architecture: A Practical Post-Asia Agenda (Institute for International Economics, 1999), reproduced here by the permission of the Institute for International Economics.


Shin and Hahm (1998), Table 1.7. Korean financial institutions had foreign currency denominated assets as well, since they were required to limit their open foreign exchange positions. But since their loans were of longer maturity than their liabilities, there was still the possibility of a very serious liquidity problem that the central bank was in no position to address.


At that point, the number and exposure of the foreign bank creditors was reduced to the point where the Korean authorities were able to negotiate a restructuring with their bank creditors, in which the latter agreed to a temporary delay in payments and then to the conversion of their short-term assets into longer-term instruments.


As events transpired, the IMF and the leading industrial countries refused to provide yet more money in the middle of August 1998, and the Russia government responded by devaluing and suspending service on most of its debts. But these actions came as a surprise to many investors, which is the point in the present context.


Thus, the Korean negotiations at the end of 1997 are often cited as examples of how international banks should be “bailed in” during crisis negotiations. Again, however, the fact of the matter is that so long as the Korean government was reluctant to halt service on these and other external debts, the banks still had the option of exiting. Thus, the agreement reached with the government of Korea did not impose significant capital losses on the banks, which only agreed to a delay of service payments and, eventually, to the conversion of their short-term claims into longer-term obligations.


As emphasized by Schwartz (1995).


Unavoidably, the Fund must tranche its assistance rather than following Walter Bagehot’s classic advice for a central bank acting as a lender of last resort, namely to lend freely at a penalty rate. This is the main problem with Meltzer’s (1998) otherwise sound advice that the IMF should act more like a true lender of last resort by lending at a penalty rate. That it will not be able to lend freely at a penalty rate suggests lending at higher interest rates in order to limit moral hazard will not suffice to redress the crisis problem.


For clear statements of these arguments from authors with otherwise very different perspectives, see Meltzer (1998) and Radelet and Sachs (1998).


In addition, there is the fear, well founded or not, that a standstill or moratorium will unleash contagion to other countries and threaten the stability of the international system.


There are exceptions. These include Mexico in 1982, South Africa in 1985, Brazil in 1987, Venezuela in 1988, and Russia in 1998. Several distinctive aspects of the Russian situation help to explain the unusual outcome in this case. For one, the fact that Russia, unlike say Mexico and South Korea, did not show the resolve necessary to rein its budget deficit suggested that providing official funds to retire the existing short-term debt would not solve the problem, since additional debt would soon have to be issued. For another, the government’s failure to make headway on its fiscal problems suggested that capital market access was in any case unlikely to be restored soon. Be that as it may, the aftermath of the Russian government’s action, including full-fledge depositor panic, capital flight, and the suspension of foreign exchange trading, hardly reassured those worried that a government’s unilateral suspension of payments could damage its credit worthiness and demoralize the markets.


Argentina’s contingent repurchase facility with thirteen commercial banks provides for $7 billion in standby credits, while Mexico’s arrangement with thirty-one banks provides for $2.5 billion. Under the provisions of the former, the Argentine Central Bank can swap Argentine government securities for U.S. dollars up to the specified ceiling, at an effective interest rate of LIBOR plus 205 basis points. The commitment fee is 33 basis points. Loan length is two to five years, depending on the commercial bank involved. These agreements omit the no-adverse-material-change clause that would otherwise permit the banks to back out of their agreement in the event of a crisis.


A similar idea is developed by Gavin and Hausmann (1998).


Borrowers are expected to repay within one and a one half years of disbursement, and are charged 300 basis points above the rate on other IMF loans (with still higher rates charged if the loan is not repaid at the end of 18 months).


That the IMF is reluctant to preapproval credits is not surprising. Like any lender, the fund would regret having preapproved a credit line were the economic policies and performance of the borrowing countries to suddenly turn for the worse. In addition, however, were the fund to then revoke its pre-approval, the market might take this as a strong negative signal, precipitating a crisis. For related discussion, see Goldstein and Calvo (1996), p. 271.


The Clinton Administration has recently proposed establishing a new facility through which the IMF would extend standing lines of credit to qualifying countries. While this scheme might be practicable for countries like Argentina, which is relatively small relative to global financial markets and whose financial policies are exceptionally strong, there is reason to doubt that it could be used to meet the needs of countries like Brazil and South Korea, because the size of the requisite credits would outstrip available financial resources and because their costliness would lead the Fund to insist on additional conditions at the time of disbursement.


There was some discussion in 1994 of the creation of a short-term financing facility under which the IMF would automatically disburse funds to countries experiencing temporary payments problems and whose eligibility had been determined in advance. But these discussions were not informed by an appreciation of the magnitude of the funds that might have to be disbursed to restore confidence in private financial markets, given the rapid growth of the liquidity of the latter. Once the Mexican crisis revealed the extent of the commitments that could be required, the idea of automatic credit lines receded from view.


Some observers would add that because banks are critical to financial stability, they can count on the support of the U.S. and other creditor-country governments in the event of debt-servicing difficulties.


Stein’s (1989) finding that bank creditors are more likely than bondholders to support corporate debt workouts negotiated in the shadow of the court is consistent with this implication. Miller and Zhang (1997) provide some theoretical and numerical exercises designed to show that this is precisely the consequence of the switch from bank to bond finance.


To be sure, a non-negligible fraction of foreign bonds are held by commercial banks. And pension funds, mutual funds and insurance companies are also subject to regulatory oversight, if not always with the same intensity as banks. But institutional investors as a group hold only a fraction of the bonds outstanding, in contrast to the earlier situation with syndicated bank loans, rendering moral suasion less effective.


Evidence against the extreme view can be drawn from four recent cases where countries succeeded in restructuring their sovereign bonds through unilateral exchange offers: Costa Rica in 1985, Nigeria in 1988, Guatemala in 1989, and Panama in 1993. In all these cases, the vast majority of bondholders, typically more than ninety percent, exchanged their old debt obligations for new ones rather than selling out to vultures (Pinon-Farah 1996). It can be questioned whether these cases are representative, however, since the terms of the exchange implied little reduction of interest or principal, only a grace period, longer maturities, and unbunching of payments.


In the case of most international bonds, ten to twenty-five percent of the bondholders (more precisely, those holding ten to twenty-five percent of the principal) can vote to require immediate repayment of all principal and interest due in the event of default. In contrast, syndicated bank loans typically require fifty percent of creditors to vote for acceleration. This therefore raises the danger that even when a majority of bondholders prefer an orderly workout that will maximize the value of their claims, an impatient minority can trigger the clause requiring immediate repayment. The debtor would then have to reschedule not only interest payments and amounts due to be paid into a sinking fund but all principal as well.


Note the parallel with the minimum percentages that already exist for activating acceleration clauses.


At this point, there is little evidence of whether or not this is the case. While the Hong Kong Airport Authority is frequently invoked as an example of a borrower which has issued bonds with these provisions without obviously elevating its borrowing costs, but it is not a fully sovereign entity, rendering it a special case. A more systematic analysis of the pricing of bonds issued under the British and American models (with and without the relevant provisions) is clearly called for.


Although a specified minority holding a fifth to a quarter of principal can require the trustee to do so on their behalf.


This path-dependence argument is a theme of Roe (1987).


Indeed, the market reaction to the Russian action freezing foreign credits on August 17, 1998, dried up all credit to the Russian financial system, provoked widespread capital flight, and forced the government to halt all foreign exchange trading several times in the final week of August.


The traditional IMF position on lending into arrears was stated in an Executive Board decision in 1970. See Cline (1996).


Goldstein (1996) argues that the fund’s lending into arrears was critical for driving commercial bank creditors to the negotiating table in the second half of the 1980s and finally clearing away the Latin American debt crisis.


Group of Ten (1996). Spokesmen for the creditors responded that lending into arrears may have been appropriate in the 1980s, when the existence of a generalized debt crisis justified the provision of public-sector enhancements to support debt reduction arrangements, but that it is not warranted in the “normalized” capital market conditions that exist today (Institute of International Finance 1996). Recent experience refutes this contention on two grounds: first, even “normalized” conditions can quickly become abnormal; and second, even if debt problems are limited in their geographical scope, the provision of the equivalent of debtor-in-possession financing by the IMF may be important for preventing the kind of complete economic breakdown that can set the stage for the contagious spread of the crisis, and it can help to expedite a negotiated agreement to restructure.


See for example the comments of IMF Managing Director Michel Camdessus at his joint press conference with Philippe Maystadt, chairman of the Interim Committee, at IMF Headquarters, April 16, 1998,


During the debt crisis of the 1980s, when the principal creditors were international banks, lawsuits were consequently rare (except for precautionary suits filed merely to protect against expiration of the statute of limitations). However, when Argentina defaulted on some of its bonds in 1986, three bondholders sued successfully in U.S. courts. And when Panama defaulted in 1987, at least one bondholder pressed a lawsuit.


When it is not waived, central bank reserves enjoy sovereign immunity in the major financial centers, most notably the United States and the United Kingdom.


Under the provisions of Article IX.8(I). In addition, the fund’s articles require it to deal only with members through their treasuries, central banks, and fiscal authorities. Thus, it could argue that it cannot deal directly with a court-appointed receiver or with the creditors’ fiscal agent.


That there have been no attempts to date to attach fund disbursements is consistent with this view.


In any case, member countries would have to pass domestic legislation ensuring that this measure had effect in their domestic courts.


One need only recall congressional resistance to the dispute-panel provisions of the WTO agreement.


As suggested by Hurlock (1995).


This is a basic premise of bankruptcy theory: In a world of complete information and absent transactions costs, there is no need for a bankruptcy code or bankruptcy court, since debtors and creditors will be able to instantaneously adjust their contracts to any unanticipated contingencies.


Some have suggested that such a committee, possibly with rotating membership, could also interface with the IMF and other official bodies. In practice, this would create more problems than it solved. For one thing, since membership on the committee would be selective, some in the markets would inevitably feel that other participants were getting preferential treatment from the IMF. And insofar as the real problem of information asymmetries arises in negotiations between the lenders and the borrowers, and not between the lenders and the IMF, it is with the debtor that the creditors’ representatively most urgently need to interact. In instances where the IMF was negotiating the extension of financial assistance while the debtors and creditors were at the same time attempting to restructure outstanding debts, there might be occasion for exceptional discussions among the three parties, but it is not obvious why regular meetings between the creditors’ committee and the official sector would be essential. And to the extent that there is a need for the fund and the financial community to exchange information in a time of crisis, this can be done more simply, by asking the central bank or national treasury in each of the creditor countries to identify a representative for their financial community.


Or so the situation is described by Lee (1998). Rhodes is the vice-chairman of Citibank and a veteran of the debt crisis of the 1980s.


Thus, Institute of International Finance (1996) cites the case of AeroMexico as an example of how negotiations can proceed and be concluded swiftly in the absence of a bondholders’ committee.


Macmillan (1997), pp. 9–10.


    ClineWilliam R.1996“Crisis Management in Emerging Capital Markets” in From Halifax to Lyons: What Has Been Done About Crisis Management?PeterB. Kenen(ed.)Essays in International Finance No. 200International Finance Section, Department of Economics, Princeton UniversityOctober.

    • Search Google Scholar
    • Export Citation

    EichengreenBarry and RichardPortes. 1989After the Deluge: Default, Negotiation, and Readjustment During the Interwar Years,in The International Debt Crisis in Historical PerspectiveBarryEichengreen and Peter H.Lindert (ed.) CambridgeMA.: MIT Press pp. 1247.

    • Search Google Scholar
    • Export Citation

    EichengreenBarry and RichardPortes.1996Crisis? What Crisis? Orderly Workouts for Sovereign DebtorsLondon: Centre for Economic Policy Research.

    • Search Google Scholar
    • Export Citation

    GavinMichael and RicardoHausmann. 1998“A Contingent Facility to Promote Financial Stability and Strengthen Market Discipline”Office of the Chief Economist, Interamerican Development Bank, unpublished manuscript.

    • Search Google Scholar
    • Export Citation

    GoldsteinMorris. 1996Avoiding Future Mexicos: A Post-Halifax Scorecard on Crisis Prevention and Management,in From Halifax to Lyons: What Has Been Done About Crisis Management?”PeterB. Kenen (ed.) Essays in International Finance, No. 200, International Finance Section, Department of Economics, Princeton UniversityOctober pp. 5672

    • Search Google Scholar
    • Export Citation

    GoldsteinMorris and Guillermo A.Calvo.1996What Role for the Official Sector?,in Private Capital Flows to Emerging Markets After the Mexican Crisis.GuillermoA. CalvoMorrisGoldstein and EduardHochreiter (eds.) Washington, DC: Institute for International Economics pp. 233282.

    • Search Google Scholar
    • Export Citation

    Group of Ten. 1996Resolving Sovereign Liquidity CrisesWashington, DC: Group of Ten.

    Group of Twenty-Two. 1998Report of the Working Group on International Financial CrisesWashington, DC: Group of Twenty Two.

    HurlockJames. 1995The Way Ahead for Sovereign Debt,Euromoney pp. 7879.

    Institute of International Finance. 1996Resolving Sovereign Financial CrisesWashington, DC: Institute of International Finance.

    International Monetary Fund. 1998International Capital MarketsWashington, DC: International Monetary Fund.

    LeePeter. 1998Korea Stares into the Abyss,EuromoneyMarch pp. 3237.

    LitanRobertet al. 1998“Statement of the Shadow Financial Regulatory Committee on International Monetary Fund Assistance and International Crises” Statement No. 145 Shadow Financial Regulatory Committee.

    • Search Google Scholar
    • Export Citation

    MacmillanRory. 1995New Lease on Life for Bondholder Councils,Financial TimesAugust15 p. 11.

    MacmillanRory. 1997“Towards a Sovereign Debt Work-out System” unpublishedmanuscript.

    MeltzerAlan. 1998“Asian Problems and the IMF” Testimony Prepared for the Joint Economic Committee U.S. CongressFebruary24.

    MillerMarcus and LeiZhang. 1997A Bankruptcy Procedure for Sovereign States,Global Economic Institutions working paper No. 34London: Economic and Social Research CouncilAugust.

    • Search Google Scholar
    • Export Citation

    Pinon-FarahMarco A. 1996Private Bond Restructurings: Lessons for the Case of Sovereign Debtors,IMF working paper WP/96/11.

    RadeletSteven and JeffreySachs. 1998The East Asian Financial Crisis: Diagnosis, Remedies, Prospects,Brookings Papers on Economic Activity 1 pp. 174.

    • Search Google Scholar
    • Export Citation

    Reuters.1998Russia, Western Banks to Form Creditors Club.Monday 24August8:25 am Eastern time from the internet at

    • Search Google Scholar
    • Export Citation

    RoeMark. 1987The Voting Prohibition in Bond Workouts,Yale Law Journal 97 pp. 232280.

    SchwartzAnna J.1995G-7 Countries at Halifax Summit Repeat the Mexican Myth,Communique of the Shadow Open Market CommitteeSeptember1011.

    • Search Google Scholar
    • Export Citation

    ShinInseok and Joon-HoHahm. 1998The Korean Crisis—Causes and Resolution,Korea Development Institute, unpublished manuscriptJuly.

    • Search Google Scholar
    • Export Citation

    SteinSol1989A Feast for LawyersNew York: M. Evans and Company.

    Other Resources Citing This Publication