August 23, 1996

Collective Action Difficulties in Foreign Lending: Banks and Bonds

Mark M. Spiegel

One of the major barriers to resolving the Latin American debt crisis of the 1980s was the “collective action” difficulties among creditors–that is, the difficulty of getting the lenders to take actions that would benefit them as members of the group but that might not be in their individual interest. In the case of sovereign lending, for example, where enforceable legal mechanisms are absent, collective action difficulties arise for two reasons. First, because creditors differ markedly in their characteristics and incentives, they have problems negotiating as a group with their debtors. Second, in “free-riding” situations, where individual creditors can benefit from the efforts of others, they are likely to undertake less effort than would be optimal for the group as a whole.

Recent changes in the composition of foreign borrowing indicate that foreign debt claimants are likely to become more widespread and diverse. Figure 1 shows that developing countries in Latin America and East Asia have seen a rapid increase in the share of foreign borrowing through public and private issues of bonds–from 5 percent in 1986 to 32 percent in 1994; in contrast, the share financed through commercial bank lending has exhibited a slow, but secular decline–falling from a recent high of 44 percent in 1987 to a low of 28 percent in 1993. Since bondholders are both less concentrated and more heterogeneous than banks, their increasing share of disbursements enhances the potential for collective action difficulties in the future.

This Economic Letter examines the role of collective action difficulties in foreign lending disputes. It first reviews the experience of banks during the Latin American debt crisis of the 1980s. It then discusses the mechanisms that did prove successful in resolving collective action difficulties among these creditors, arguing that lending disputes were better resolved through agreements flexible enough to accommodate creditor heterogeneity. Finally, it reviews the covenants contained in public and private international bonds which are designed to facilitate the resolution of potential future foreign lending crises.

Collective action difficulties among banks

The Baker Plan, first implemented in 1986, illustrates the nature of the difficulty banks have experienced in coordinating their lending efforts. The Baker Plan was designed to address the debt problems that arose in 15 developing nations after Mexico suspended its external debt payments in 1982. The general belief underlying the Baker Plan was that the payments difficulties of these countries reflected their “illiquidity” rather than insolvency; that is, they appeared to have the resources and growth prospects necessary to service their debt obligations in the long run, but they were experiencing temporary problems in maintaining their current payments.

Therefore, the Baker Plan called for commercial banks and international financial institutions to extend loans to these 15 countries over a three-year period. Commercial banks were called on to extend approximately $7 billion annually (roughly 2.5 percent of their total exposure) for a total disbursement of approximately $20 billion, and the IMF and the World Bank were called on to provide between $20 and $25 billion. The logic behind this call for additional lending was that if the debtor countries could receive adequate financing to allow them to maintain their debt payments through a temporary difficult period, they would be more likely to meet all of their debt obligations over time.

In studies of the efficacy of the Baker Plan, estimates of the magnitude of the disbursements commercial banks actually made over the period differ; however, most do agree that they fell far short of their goals. For example, Husain (1989) estimates that the highly indebted countries received only $4 billion in net new long-term financing. Moreover, he found that if payments on private non-guaranteed debt are taken into account, commercial banks actually received more in repayments than they extended in new money.

If one believes that highly indebted nations in fact faced liquidity problems, then the shortfall in new disbursements to them could be described as a problem in collective action across bank creditors. While it was generally perceived that it was in the interest of creditors as a group to lend more in the short run to enhance their total discounted stream of payments, individually each bank had the incentive to “free-ride” on the efforts of other creditors by not extending the amount required under the Baker Plan. This explains why the level of new money extended to the highly indebted nations under the Baker Plan may have been below the optimum.

The Brady Plan, which began in 1989, allowed creditors to convert their existing debt claims into a variety or “menu” of new claims. The plan involved two rounds. In the first round, creditors bargained with debtors over the terms of these new claims. Loosely interpreted, the options contained different mixes of “exit” and “new money” options. The exit options were designed for creditors who wanted to reduce their exposure to a debtor country. These options allowed creditors to reduce their exposure to debtor nations, albeit at a discount. The new money options reflected the belief that those creditors who chose not to exit would experience a capital gain from the transaction, since the nominal outstanding debt obligation of the debtor would be reduced, and with it the probability of future default. These options allowed creditors to retain their exposure, but required additional credit extension designed to “tax” the expected capital gains. The instruments included in the Brady deals differed in other aspects as well. The principal of many instruments was collateralized, as were “rolling interest guarantees,” which guaranteed payment for fixed short periods. The first round negotiations thus involved the determination of the effective magnitude of discount on the exit options together with the amount of new lending called for under the new money options.

In the second round, creditors converted their existing claims into their choice among the “menu” of options agreed upon in the first round. The penalties for creditors failing to comply with the terms of the deal were never made explicit. Nevertheless, compliance was not an important problem under the Brady Plan. Banks wishing to cease their foreign lending activities tended to choose the exit option under the auspices of the deal.

The Brady Plan therefore achieved better results than the Baker Plan since it was designed to incorporate creditor heterogeneity. Creditors faced a variety of options into which they could convert their claims. By offering a “menu” of options, the Brady Plan permitted credit restructurings to be tailored to the heterogeneous preferences of creditors. The terms achieved under these deals indicate that debtors used the menu approach to reduce the cost of debt reduction (see Diwan and Spiegel 1994).

Collective action difficulties among bondholders

The experiences of the Baker and Brady deals illustrate the importance of allowing creditors to negotiate with debtors and make decisions among themselves as a group. The intransigence of a relatively small number of creditors whose incentives diverged from those of the larger group could preclude the closing of a globally desirable deal. In the Baker Plan, the uniform call for new lending was unacceptable to some creditors, who extended less than their allocation of new funds. Although the Brady Plan offered more flexibility, it still left some creditors dissatisfied with the outcomes. In the Brazilian Brady deal, for example, a number of lawsuits have been filed by creditors unhappy with its terms.

Because bond holders represent an even less concentrated and more heterogeneous group than commercial banks, collective action difficulties among bondholders are likely to be even more prevalent. One response is the inclusion of contractual provisions which could help to facilitate the resolution of a liquidity crisis. The Group of Ten reports that these provisions, which are already included in many syndicated loan agreements, are also being included in some bond contracts [Group of Ten (1996), pg. 14].

These provisions fall loosely into one of three categories: First, there are provision that allow for the collective representation of bondholders in the event of a liquidity crisis. These provisions~typically designate the rules under which bondholder representatives are selected. However, in practice international bond issues rarely delegate extensive power to such representatives. Typically, they designate an agent who can call meetings and issue notices, but who is not mandated to make significant decisions on behalf of individual bondholders. Second, there are “qualified majority voting clauses,” which allow for changes in a bond contract to be made without the unanimous consent of the bond holders. These clauses are designed to preclude a small group of minority creditors from preventing the resolution of a liquidity crisis. Third, there are “comparable treatment” and sharing provisions, which discourage individual creditors from taking individual actions with or against the debtor. These provisions are designed to maintain contractual seniority among debtholders. For example, they preclude the debtor from negotiating a deal with an individual creditor which dilutes the value of a second creditor’s claim.

In practice, there are limitations to the effectiveness of all types of contractual provisions. Since the international bond issues rarely delegate extensive power to an agent, the ability of delegates to facilitate desirable outcomes may be limited. Bond prices indicate that the market does not believe that majority voting clauses have a large impact on voting outcomes among bondholders. For example, English bonds which include qualified majority voting clauses are not priced very differently by the market from bonds without such clauses. The effectiveness of sharing clauses in international bond issues in facilitating the resolution of financial crises is also likely to be limited. For example, a country’s foreign bonds are issued by both public and private agents and at different points in time. The myriad classes of bonds that can be issued leaves the identification of the relative seniority of claims unclear. Bondholders of these disparate claims are likely to question any decision concerning relative seniority among classes of issues. Consequently, these clauses are only likely to be effective across holders of an individual bond issue.


Since the strength of contractual provisions in alleviating collective action problems is limited, it is unclear how useful these provisions will be in addressing future international liquidity crises. Nevertheless, the movement to bond finance in international borrowing does not necessarily threaten the stability of the financial system. Indeed, there are a number of reasons to consider this change in the composition of lending desirable. If debtors are moving from financing through commercial banks to financing through bonds, it must be the case that this form of financing is available on better terms. This implies that the market is relatively unconcerned about the increase in collective action difficulties associated with bond finance. Moreover, creditor countries may find it desirable that bonds have no explicit government distortions analogous to fixed-premium deposit insurance.

However, the increased diffusion of creditors, which makes it more difficult to address liquidity crises once they arise, makes the avoidance of these crises even more imperative. The Group of Ten report offers a number of recommendations. These include a renewed emphasis on enhancing the circulation of information necessary to establish accurately and clearly the creditworthiness of sovereign debtors, and a call for greater surveillance by the international financial institutions.

Mark Spiegel
Senior Economist


Diwan, Ishac, and Mark M. Spiegel. 1995. “Are Buybacks Back?: Menu-Driven Debt-Reduction Schemes with Heterogeneous Creditors.” Journal of Monetary Economics 34, pp. 279-293.

Group of Ten. 1996. “The Resolution of Sovereign Liquidity Crises: A Report to the Ministers and Governors Prepared under the Auspices of the Deputies.” BIS, Basle, and IMF, Washington, D.C. May.

Husain, Ishrat. 1989. “Recent Experience with the Debt Strategy.” Finance and Development 26, pp. 12-15.

Opinions expressed in FRBSF Economic Letter do not necessarily reflect the views of the management of the Federal Reserve Bank of San Francisco or of the Board of Governors of the Federal Reserve System. This publication is edited by Sam Zuckerman and Anita Todd. Permission to reprint must be obtained in writing.

Please send editorial comments and requests for reprint permission to

Research Library
Attn: Research publications, MS 1140
Federal Reserve Bank of San Francisco
P.O. Box 7702
San Francisco, CA 94120