Reforming the International Monetary and Financial System

Comments: Private Sector Participation in Crisis Prevention and Resolution

Alexander Swoboda, and Peter Kenen
Published Date:
December 2000
  • ShareShare
Show Summary Details
William R. Cline

Barry Eichengreen’s paper provides a generally balanced review of the issues for this session, and highlights important institutional facts (such as the structure of British-style bond contracts) and country experiences. Rather than focus my comments directly on his paper, I will attempt to set forth the framework that I consider most useful in addressing this topic.1 It should be apparent where there are significant divergences from Eichengreen’s approach.

External finance for economic development in middle-income economies has come primarily from the private sector in the 1990s, and it is likely to do so in the coming decades as well. From 1992 through 1998, of $1.6 trillion in net capital flows to 29 major emerging market economies, $1.4 trillion came from the private sector (Institute of International Finance, 1999d). The challenge of improving international financial architecture is to adopt those changes that will help revive these flows from their recent low levels and help channel them in productive directions, and to reject those changes that will tend to do the opposite. Despite the recent improvement in sentiment in emerging capital markets, they remain in a state of fragility rather than of exuberance like they were in 1996 and early 1997. It is important that permanent changes in the architecture not be made on the premise that the normal state of these markets is one of excessive capital inflows that can usefully be curbed, as the opposite is more likely to be the case for some time.

A fundamental question in examining the role of the private sector is whether the financial crises have primarily arisen from domestic policy mistakes, or from systematic behavioral distortions on the part of private international lenders that require government intervention to correct a market distortion. The fact that the majority of emerging market economies did not experience severe financial crises in the last two years suggests that domestic distortions played a leading role in those that did.

Avoiding Crises

There is, by now, a relatively widely agreed agenda of domestic policies in emerging market economies needed to help ensure against financial crisis, including a realistic exchange rate (though the new conventional wisdom blessing currency boards and floats but condemning everything in between is at best premature); fiscal balance (underscored by Russia and Brazil); moderate credit expansion; and a sufficiently strong financial system.

Short-Term Debt

There is also broad agreement that excessive buildup of short-term external debt can invite a liquidity crisis even where overall debt is moderate. Korea, in particular, ran up its short-term debt to almost $100 billion and ran down its usable reserves below $10 billion, in considerable part because in the liberalization of its capital market it started with short-term flows while retaining restrictions against the safer long-term flows (especially direct investment). Chile’s experience shows that a prudent approach to the level of short-term debt can help avert crisis, and that this is possible in a relatively market-friendly price-based framework that essentially levies a tax if inflows are excessive rather than prohibiting them. In general, the best defense against excessive short-term debt will typically be appropriate economic policies in the borrowing country, including avoidance of the dangerous combination of a fixed exchange rate with high domestic interest rates and complete capital mobility. It is also evident that domestic distortions biasing capital flows away from longer-term forms—direct investment, medium-term bank lending, and bonds—should be removed.

At the international level, one proposal to address short-term debt is to reduce or remove its preferential risk-weighting (20 percent) for capital required against short-term bank lending (Greenspan, 1998; IMF, 1999). Requiring full risk-weighting might boost spreads by, say, 100 basis points.2 It is unclear, however, what the impact would be, as the rates could still look relatively attractive to emerging market borrowers who face high domestic interest rates. For interbank lending, there is the important question of whether and to what extent liquidity in the banking system would suffer from such a change (both internationally and domestically within industrial countries), or alternatively whether it would be appropriate to apply a change only on loans to emerging market banks. For trade credit, there are the questions of whether the existence of merchandise collateral should not warrant lower risk-weighting, and whether a higher risk-weighting would complicate expansion of trade and hence economic growth. More fundamentally, the design of capital requirements under the Basel accord is under much broader review.

Risk Assessment

The private sector can contribute to crisis avoidance by sharpening its risk assessment capacity (Institute of International Finance, 1999c). Experience in the recent crises has shown that greater emphasis needs to be placed on scenario analysis and stress testing in the use of value-at-risk modeling, and on the link between credit risk at the level of the borrowing firm and macroeconomic conditions. Emerging market economies can help substantially in this process, however, by providing more timely and complete data (see Institute of International Finance, 1999d). Some of the most severe shocks in the recent crises arose because of severe market surprises about, in particular, the amounts of unencumbered reserves.

Private Sector-Borrowing Country Dialogue

A concrete step that most emerging market economies could take to help reduce volatility of private short-term capital flows would be to establish investor relations systems such as Mexico’s quarterly briefings on economic policy by financial authorities (Institute of International Finance, 1999e). Conference calls or investor meetings on a regular basis would help avoid adverse market surprises on policies and data. They would also provide a vehicle for feedback from private market participants to financial authorities that could signal growing market concern when certain policies need to be corrected. In precrisis situations, the authorities could usefully enter into more intense consultations with private firms to gauge likely market response to policy changes, although presumably not for some prospective actions for which advance information could be sensitive (such as an imminent devaluation).

Crisis Resolution

Confidence-Restoring Adjustment Programs

Crisis resolution in the late 1990s has been based on the objectives of renewed growth and prompt restoration of capital market access through policy adjustment, official sector support, and, in some cases, private sector financing arrangements. There have been six major international support programs. Four have been broadly successful: Mexico in 1995; and in 1997–99, Thailand, Korea, and Brazil.3 In two cases, domestic political disarray has caused collapse (Russia) or extreme delay (Indonesia). The strategy of restoring confidence has broadly avoided the seizing up of capital markets that likely would have occurred if approaches such as comprehensive debt rescheduling had been pursued instead. Considering that the six countries account for half of total external debt of emerging market economies, there should be little doubt that together their crises posed a systemic threat.

Moral Hazard and Burden Sharing

The large official support mobilized for these programs has, however, raised the critique of moral hazard, the argument that the private sector will have learned that investing in emerging markets is low risk because of official bailouts. The reality is rather different. As a result of the East Asian and Russian crises, there were losses amounting to about $60 billion for international banks, $50 billion for other private creditors including bondholders, and some $240 billion for equity investors (Institute of International Finances, 1999a, p. 61). After the Russian default, in particular, the extremely low flows of capital and high lending spreads showed that the private sector had been shocked by this experience, not lulled into complacency by the official support programs. Recent empirical tests conducted at the Institute of International Finance show that in the post- Mexico decline in lending spreads through mid-1997, the principal forces were generalized international liquidity (as captured by falling high-yield U.S. spreads) and country economic variables, and the hypothesis that there was a moral-hazard-induced spread reduction from the Mexico support program is statistically rejected (Zhang, 1999). The main instance of moral hazard seems likely to have been the case of Russia for geopolitical reasons.

Even if the recent support programs did not set the stage for excessive lending because of moral hazard, one can still ask whether they represented inadequate burden sharing by the private sector. At the broadest level, burden sharing should be evaluated in light of the fact that almost 90 percent of capital flows to emerging markets in recent years has been from the private sector. In practice, private financing arrangements have become increasingly important in the series of support programs, and the central question is not whether the private sector should be involved but what is the most productive way to obtain its participation, and, conversely, what approaches are counterproductive and should be avoided.

Forms of Private Sector Involvement

The central principle in answering this question is: What approach is the closest to a voluntary, market-based arrangement that will still provide good chances for the program’s success. The first program, in Mexico, did not have a private sector component, in considerable part because of the challenges posed by potential rescheduling of short-term domestic government debt held by foreigners. Thereafter, however, there were varying degrees and modalities of private sector participation.

At the most voluntary end of the spectrum, one vehicle that would seem promising for private sector participation is the private contingent credit line, of the type maintained by Argentina, Mexico, and Indonesia. The commitment fees and spreads on these lines are likely to be more costly, and the terms more closely linked to contemporary market conditions, than before the experience of Mexico’s drawdown in late 1998, but this vehicle is nonetheless potentially a cost-effective form of crisis-abatement insurance, and one that emerging market economies would seem well advised to pursue. Another purely voluntary modality, albeit perhaps one of limited potential scope, is the mobilization of private credit through risk mitigation by the official sector, as in the Asian Development Bank’s guarantee of about $1 billion in private bank lending to the Export-Import Bank of Thailand in early 1998.

Still in a relatively voluntary range of the spectrum, arrangements by major international banks to maintain their short-term interbank and trade credit lines, as in the case of Brazil in early 1999, also provide considerable promise. The impact of such arrangements extends beyond the fraction of total debt covered by these claims, because once there is a sense that the larger bank claims will not run off rapidly, other creditors including holders of short-term government obligations are much more likely to judge that the situation can be controlled and hence to remain present in the country rather than seek to exit.

Modestly less voluntary is the conversion of short-term bank claims into longer-term claims, as occurred with about $22 billion in bank claims on Korean banks, at spreads higher than the original claims but below the market level at the time of the crisis. Also in the relatively voluntary range is surgical rescheduling of a modest part of the debt, as occurred in Thailand with the debt of finance companies.

The more involuntary modalities, such as the comprehensive reschedulings of Latin American debt in the 1980s, or the recent broad debt rescheduling in Indonesia, have much more adverse effects and should be avoided unless the country’s indebtedness is so extreme that such measures cannot be escaped. Even in the most severe cases, unilateral actions such as Russia’s default are likely to be counterproductive for the country itself.

Private Sector Collective Action

Analytically perhaps the most interesting case is that of the Brazilian voluntary maintenance of short-term lines. Essentially this represented a case of collective action. Yet it was primarily collective action undertaken by private firms in their mutual self-interest, rather than imposed by government action.4 It seems useful to think about this type of behavior as a “repeat game.” The principal creditors are broadly aware of what their counterparts are doing. They know that if they all act jointly, the situation can be resolved to their mutual gain. If a player does not cooperate, it should expect that the other players may be unprepared to cooperate in turn in a future episode where the player in question may have greater stakes. To be sure, it will require, at the very least a request by the borrowing country’s authorities, and more likely some moral suasion by the G-10 central banks (or finance ministries) and the IMF, to help catalyze this collective action.

To recapitulate, the best approach for involving the private sector in crisis resolution is on a voluntary or quasi-voluntary basis rather than in mandatory forms such as comprehensive reschedulings (or, worse, forced debt reductions), as the voluntary approach is most likely to permit prompt return to market access. This is perhaps best illustrated by the following thought experiment. Suppose that in September 1998, the IMF and Brazil had forced foreign banks and bondholders to accept a multiyear rescheduling of debt. Would Brazil have been able to return to the global bond market for $2 billion by April? The answer is, not likely.

Bond Rescheduling Clauses

A great deal of attention in the discussion on international financial architecture has focused on potential problems caused by the rising relative weight of bonds in the system, and the greater difficulty of rescheduling bonds than bank loans. While this is a legitimate issue that can be important in a few cases, the bulk of the financial crisis problem has not stemmed from bond obligations in the recent past and is unlikely to do so in the future. The bonds for which the rescheduling clause discussion has been relevant are medium-term Eurobonds, not the short-term treasury bills that got Mexico and Russia into trouble. More fundamentally, the recent financial crises have been driven by the runoff in short-term claims, not an immense bunching of bond obligations.

The fact that bonds are long-term instruments should already be a signal that they are not at the heart of the problem. The amortization numbers are even more eloquent. At present, the outstanding stock of short-term external debt of emerging market economies is about $500 billion. In contrast, the amount of Eurobond amortization coming due annually is only about $30 billion.5 This means that for the Korea- and Brazil-type financial crises, the availability of bond rescheduling clauses would largely be irrelevant. This conclusion is reinforced if one considers that if the crisis resolution had progressed to the stage of formal debt rescheduling, whether of bonds or of bank claims, the objective of prompt restoration of capital market access would have been defeated already.

There is ample room for firms and sovereigns to experiment with inclusion of rescheduling clauses in their bonds, on a voluntary basis and with the market determining how much of a risk premium they will have to pay in the spreads. Indeed, an important implication of Eichengreen’s paper is that countries seeking this flexibility can largely obtain it already by borrowing in London using British-style covenants rather than in New York, Tokyo, or Frankfurt. The right way to think about bond rescheduling clauses, then, is that they are neither urgently needed to address the classical liquidity crises that have marked the late 1990s, nor should they be pursued in a manner that is mandatory, for example by regulatory requirement. The latter approach would tend to send the signal that the official community seeks to facilitate rescheduling. It should always be kept in mind that sovereign lending is inherently difficult because of the absence of collateral, and the impact of adverse reputation and other sanctions upon default is the only substitute for collateral. If international official policy forces a facilitation of rescheduling, it erodes this collateral substitute.6 Moreover, to mandate rescheduling clauses only for emerging market bonds would represent a formal discrimination against them; yet it is highly unlikely that industrial countries will institute rescheduling clauses in their own treasury bonds or, for that matter, that their taxpayers would be well served if they did.7

Forced Bond Reschedulings for Paris Club Comparability

Recent cases, especially those of Pakistan, Ukraine, and Romania, have raised the prospect that the Paris Club would insist that a country reschedule its Eurobond obligations as part of “comparable treatment” parallel to the rescheduling of bilateral claims. Clearly, no particular class of private sector claim should be given senior status over others and, in some cases, bond reschedulings may be necessary, with or without rescheduling clauses. It seems questionable, however, that the Paris Club would mandate that the cash-flow drain represented by bonds must be addressed by rescheduling them. Instead, the country’s access to capital markets would seem best served if the country has the flexibility to seek its own alternative financing arrangements, which may include an exchange of assets or some other operation that does not involve formal rescheduling of bonds. Again, the principle is to remain as voluntary and market-based as possible in the design of crisis resolution, in this case involving a more protracted debt problem than one of short-term liquidity.

Lending into Arrears and an IMF Stay of Litigation Amendment

I have suggested that the best way to involve the private sector in resolving financial crises is to maximize the voluntary and market-based nature of the participation of private creditors. If this principle is accepted, then it will be rare that the appropriate strategy is instead to “bind in” creditors through highly involuntary approaches. IMF lending into arrears, or making new IMF loans to countries that have not reached broad agreement with their private creditors on resolving unpaid interest and principal, has been endorsed by the various official entities as a promising approach to securing private sector participation. Yet, in most circumstances it would likely violate the principle of choosing the most voluntary approach feasible. The only way to turn lending into arrears into a market-friendly instrument would be to limit it to cases where the bulk of creditors are already in agreement with the debtor and consider that IMF lending would have a positive effect, and where only a recalcitrant few opposed the proposed resolution of arrears.

It has also been suggested that the IMF Articles of Agreement could be amended to permit the IMF to determine that private debt contracts temporarily cannot be enforced—a stay of litigation. Because it would be difficult to persuade most market participants that such a power would be exercised only against the recalcitrant few, instead of against the bulk of creditors for purposes of altering the balance of power in the debt negotiations, this idea also would seem to hold more potential for harm than for good in improving the system’s ability to respond to crises.


Brazil’s rapid and promising resurgence from the financial crisis in early 1999 is perhaps the best evidence yet that given today’s international capital markets, a voluntary or quasi-voluntary and market-based approach to liquidity crises is the best strategy for crisis resolution. The extent to which private sector participation in crisis resolution can be kept more voluntary will vary from case to case in future crises. What is important, however, is to recognize that keeping participation as voluntary as possible, taking into account the potential for private sector collective action, is crucial when designing proper crisis responses, if the prospects for robust and sustainable emerging capital markets are to be maximized. Those who argue instead that “adverse selection” or some other externality require a public sector intervention on a mandatory basis8 would seem to bear a considerable burden of proof that such action will make borrowing countries and the international financial system better off, rather than worse off, in the long run.


    GreenspanAlan1998remarks before the 34th Annual Conference on Bank Structure and CompetitionChicagoIllinois Federal Reserve Bank of ChicagoMay 7.

    Institute of International Finance1999aReport of the Working Group on Financial Crises in Emerging Markets (Washington: Institute of International FinanceJanuary).

    Institute of International Finance1999bReport of the Working Group on Transparency in Emerging Markets Finance (Washington: Institute of International Finance, March).

    Institute of International Finance1999cReport of the Task Force on Risk Assessment (Washington: Institute of International FinanceMarch).

    Institute of International Finance1999dSteering Committee on Emerging Markets Finance“Involving the Private Sector in the Resolution of Financial Crises in Emerging Markets” (Washington: Institute of International Finance) April.

    Institute of International Finance1999eSteering Committee on Emerging Markets Finance“Strengthening Relations Between Emerging Market Authorities and Private Investors/Creditors: A Country-Focused Approach to Crisis Avoidance” (Washington: Institute of International Finance) April.

    International Monetary Fund1999Involving the Private Sector in Forestalling and Resolving Financial Crises (Washington: International Monetary FundApril).

    Organization for Economic Cooperation and Development1999OECD Economic Outlook No. 65 (Paris: OECDMay).

    ZhangXiaomingAlan1999“Testing for ‘Moral Hazard’ in Emerging Markets Lending,”IIF Research Paper No. 99-1 (Washington: Institute of International Finance).

The views expressed here should not be attributed to the Board of Directors of the Institute of International Finance.

For earlier statements of the main arguments here, see Institute of International Finance (1999a, 1999d).

With 8 percent capital requirement, full risk-weighting increases the requirement by 6.4 percentage points. With a capital opportunity cost of 15 percent, this requires an extra 96 basis points return.

Thus, in the case of Mexico, net private capital flows fell from an average of about $30 billion annually in 1993–94 to an outflow of $5 billion in 1995, but rebounded to positive inflows of about $20 billion annually in 1996–97.

In particular, there was no imposition of exchange controls to prevent servicing of debt to those banks that decided not to cooperate, and the Brazilian government and IMF monitored the market-based “rollover ratio” rather than seeking to force it to 100 percent by such alternative restrictions.

Even after plausible adjustment for bond “put” clauses, increasing potential short-term amortization, the sharp asymmetry in the amounts as compared to short-term debt would remain.

Eichengreen recognizes this “bonding role of debt” but minimizes its importance by noting that domestic bankruptcy arrangements nonetheless exist for orderly workouts. This argument, however, does not address the fact that in domestic bankruptcy, physical assets can ultimately be seized and management can be replaced by creditors, “bonding” mechanisms not present in sovereign lending.

It is unclear that Italy, whose net public sector debt exceeds 100 percent of GDP, or Japan, whose corresponding ratio is set to rise from 40 percent today to 100 percent by 2008 (OECD, 1999, p. 37), could announce the introduction of rescheduling clauses with no increase in the interest rate paid.

Including, for example, obligatory UDROPs and other such devices reviewed but not necessarily endorsed by Eichengreen.

    Other Resources Citing This Publication