Historical Experience with Bond Financing to Developing Countries
  • 1 0000000404811396https://isni.org/isni/0000000404811396International Monetary Fund

The paper reviews the historical experience of developing countries with bond issues in international markets in order to put the recent wave of bond financing by these countries in some perspective. It examines developments in the early part of this century and during the mid-1970s and early 1980s. The sources and the role played by bond financing during these periods are discussed. The payments problems associated with these bonds that emerged during the 1930s and during the latter half of the 1980s and the ways in which these problems were resolved are also examined.

Abstract

The paper reviews the historical experience of developing countries with bond issues in international markets in order to put the recent wave of bond financing by these countries in some perspective. It examines developments in the early part of this century and during the mid-1970s and early 1980s. The sources and the role played by bond financing during these periods are discussed. The payments problems associated with these bonds that emerged during the 1930s and during the latter half of the 1980s and the ways in which these problems were resolved are also examined.

I. Historical Experience with Bond Financing to Developing Countries

As developing countries have re-established access to international capital markets, bond placements in the five-year period since 1988 have accounted for a substantial portion of the funds raised. Historically, bonds were the predominant means by which countries raised financing internationally. Only during the 1970s did bank lending play a dominant role and external bonds could become again a significant source of financing for developing countries. In order to put recent developments in private market financing for developing countries in some perspective, the experiences during the last major wave of bond financing is reviewed. This review draws on the extensive literature that analyzes the experience with bond financing in the early 1900s (Borchard (1983), Fishlow (1985), Eichengreen (1991), Eichengreen and Lindert (1989)). In addition, the paper looks at some of the bond financing that took place during the 1970s and early 1980s.

1. Experience in the early 1900s

a. Bond financing flows

A wide range of countries engaged in foreign borrowing through the late 1930s. The regions with the largest gross international obligations in 1938 were North America and Asia; Eastern Europe was also an external debtor mainly on account of sovereign obligations. 1/ The United States, however, after World War I became a net creditor, and during the period after 1920 and before the U.S. Stock Market crash in 1929, it was a major source of funds, especially for borrowers in Latin America. In this period, net U.S. purchases of foreign securities amounted to nearly $4.5 billion (equivalent to roughly 1/2 percent of GNP). 2/ Latin American obligations accounted for up to one quarter of new bond issues floated in the United States by foreign entities in the 1920s. The principal borrowers were national governments of major countries, including Argentina, Brazil, and Chile. Between 1920 and 1929, Latin American entities issued $2.2 billion in bonds world-wide, of which $1.3 billion was placed by sovereigns. 3/

Countries issued bonds abroad to finance railroad and road building, the cost of wars or reconstruction, or simply to support their international reserves in the face of a confidence crisis during domestic upheavals (e.g., Mexico 1911). Investment houses often guaranteed the solvency of the issuer, and countries at times also enhanced the bonds through various types of collateral such as revenues from railroads, tobacco, or other industries, and assets like gold or other commodities (e.g., a 1922 British loan to Brazil was backed by coffee). In some cases bondholders assumed the administration of pledged revenues, as was the case in Nicaragua where creditors assumed the management of the country’s steamship and railroad lines under the loan contract of March 1912. 1/

Developing country bonds offered higher returns than bonds issued in the investors’ home country, with maturities often in the 20-30 year range--similar to maturities on domestic bonds. Latin America’s dollar bonds issued in the 1920s, for example, had on average yields generally above those of industrial country public entities and corporate bonds. The average yield at the time of issue (during the 1920s) on national and nationally guaranteed dollar bonds ranged from 6.2 percent (Argentina) to 7.6 percent (Bolivia). By comparison, the Canadian provinces of Alberta and Vancouver issued 40-year bonds in 1927 and 1928, yielding 4 1/2 percent. Moreover, the average yield for five major Latin American countries taken together fluctuated from 8 to 99 percent basis points above that on U.S. low-grade (Baa) corporate bonds, with the exception of Argentine bonds in 1923-24 which yielded less than the Baa rate. 2/

b. Payments difficulties in the 1930s

A steady deterioration in the terms of trade of primary commodity exports during the 1920s and a contraction of world trade in the later part of the decade and into the 1930s combined with the sharp rise in foreign borrowing during the 1920s to leave a number of developing countries vulnerable to external shocks. Thus, the world-wide depression of the 1930s caused severe financial problems for these countries and resulted in a number of countries partially or totally defaulting on their foreign bonds.

These defaults were generally settled through negotiations between the debtor countries and representatives of the bondholders. 3/ To deal with debt servicing problems, the British Corporation of Foreign Bondholders (BCFB) had been established in 1868. In 1933, a similar organization, the Foreign Bondholders Protective Council (FBPC), was set up in the United States with support from the U.S. Government. These formal organizations replaced ad hoc committees which tended to compete for support of the various bondholder groups by making exorbitant promises to bondholders, protected exclusively their own interests, and inflated their expenses. Debtor countries, however, did not always recognize the formal British and American committees as representing the interests of all bondholders, and at times, they did not respond to these committees’ requests for information, since these requests were viewed as potentially infringing on national sovereignty. 1/

After assessing a debtor country’s capacity to pay, the bondholders’ committees would negotiate a rescheduling plan with the country. This plan would be published with a recommendation for approval by all bondholders. Settlements could be temporary or permanent, based on judgments regarding the nature of the shock that affected the debtor at the time of its default. Some settlements included contingency arrangements to reflect an unexpected improvement or deterioration in a debtor’s economic situation. 2/ Debtors preferred to represent that their financial impairment was permanent, but initially bondholders’ committees tended to resist permanent debt adjustments, instead attempting to arrange short-term settlements subject to revision as they fell due for renewal. 3/ In time, permanent settlements became predominant. Less than half of the debt relief plans recommended by the FBPC in the period 1934-39 were permanent. In contrast, of the 19 debt relief plans recommended in the period 1940-67, only two were temporary. The offer of a permanent settlement by the bondholders’ committee was used as a means of extracting a more precise estimate of the debtor’s capacity to pay. Also, permanent settlements were granted when the new instruments issued in exchange for old debt were clearly senior to other issues or carried other enhancements. For example, Buenos Aires Province was able to obtain a permanent settlement in 1935 by offering as security for the new bonds issued revenues collected by the Federal Government and paid directly to the bonds’ paying agents in New York.

The acceptance by bondholders of a settlement was signalled by cashing the first coupon under a rescheduling plan or by exchanging an old certificate for a new one. 4/ Debtor countries would sometimes set a time limit for the acceptance of the rescheduling agreement, and debt service payments made under it were considered to constitute full payment. Bondholders could hold out for a better deal, but the debtor governments involved had little incentive to continue negotiations once a deal had been reached with the relevant bondholders’ committee and their good standing had been restored. 1/ On occasion, debtors made unilateral offers to bondholders which were not based on the completion of negotiations, with marginal success. In 1935, Chile established a unilateral plan to service its external bonds based on the principle that revenues from the nitrate and copper industry would be assigned in equal parts to the servicing and retirement of bonds that had lapsed into default.

Negotiations to settle bond defaults were protracted, with it generally taking more than five years to reach a rescheduling agreement. In the case of Chile, it took 17 years from the default on its bonds in 1931 to reach a final settlement with creditors (Table 1). Agreement was finally reached under pressure from the World Bank, which would not disburse new loans to a country in formal default. 2/ It took Brazil about ten years to complete the restructuring of its bonds after declaring suspension of debt payments in November 1937. Several countries bought back some of the bonds in the secondary market at a discount while they were in the process of negotiating a final settlement (e.g., Chile in 1934-35), but this practice was strongly criticized by the bondholders’ committees.

Bond rescheduling agreements generally included a negative pledge clause which automatically extended to the rescheduled debt the advantages granted by the debtor to any of its outstanding external obligations. Some agreements granted preferences on the timing of interest and principal payments to the rescheduled bonds, the assignment of new security (or collateral), the allocation of certain revenues for servicing them, and the transfer of the service abroad by preferred assignment of foreign exchange.

The most common methods used in providing debt relief generally consisted of replacing defaulted bonds by new bonds carrying longer repayment periods, lower interest rates, or lower principal. 3/ New money was disbursed in some cases by allowing the debtor to float new bonds as part of the deal. In some instances, debtor countries merged existing obligations into a single new debt instrument, to arrive at a uniform interest rate and facilitate administration. Liquidity-constrained countries used this method to reorganize and stabilize their financial situation, even if they had not actually lapsed into default. 1/ Although bondholders regarded the reduction of principal and interest as the most radical method, in some cases this was justified because the face value of debt was judged to be obviously in excess of the debtor’s capacity to repay. 2/ As part of the debt settlements, interest arrears were usually partially or entirely canceled. 3/ Debtors were generally unable to make up-front cash payments, except in instances where new money was provided. In many cases, a reversion clause was included in the agreements which specified that failure of the debtor to comply with the new terms would lead to an immediate reversion to the terms of the original contract.

c. Access following restructurings

According to most accounts, defaults in the 1930s appear to have carried a high price for developing countries in terms of subsequent market access. Following payments suspensions during that decade, it took about 40 years for these countries to regain substantial access to international capital markets. 4/ Lenders’ reactions also were across the board, affecting all borrowing countries, including those which did not reschedule their debt obligations. These contagion effects were particularly evident for Latin American countries. Between 1950 and 1964 (the 15-year period immediately after the time that most settlements of the 1930s defaults were reached), capital flows to these countries remained modest. 5/ Not until the 1970s did developing countries re-establish significant market access. The lengthy period of time before access was restored may have reflected the widespread nature of the bond defaults in the 1930s and the nature of the settlements, leading lenders to a general reassessment of the external bond market. 1/ Moreover, lenders did not entirely forget the experience of the 1930s when market access was re-established in the 1970s. When countries that defaulted on bonds in the 1930s came back to international financial markets, they generally had to pay a higher interest premium on funds borrowed than other developing countries.

2. Experience in the 1970s and 1980s

a. Bond financing flows

Although most developing countries in the 1970s and early 1980s received external financing mainly in the form of syndicated loans, some of them also chose to issue bonds. However, bond issues never reached the annual levels observed in the early 1900s, and bond financing declined in relative importance. International bonds issued by developing countries in the first half of the 1980s amounted to $24 billion, with the stock of bond debt of major indebted countries amounting to only $17 billion in the mid-1980s. Bond financing during this period tended not to be project-specific, rather it was one of several venues that countries used to cover their overall financing needs.

The pool of lenders in the 1970s and 1980s became quite diversified as the Euromarket and the Samurai market competed with New York to float new developing country issues. The terms for the bonds mirrored those of syndicated commercial bank loans ranging from about 1 percent to 2 percent above LIBOR, but in some cases bonds with fixed interest rates of 11-12 percent were issued. Maturities varied from 5 to 12 years--considerably shorter than those of developing country bonds issued earlier in the century. Developing countries in some cases used enhancements to make bonds more attractive to investors. For example, some Mexican issues were indexed peso-denominated petro bonds and silver-indexed bonds. 2/

b. Payments difficulties in the 1980s

Repayments problems in the 1980s followed a period of poor economic policies by some developing countries, but were also triggered in some cases by adverse external shocks, such as lower oil prices or higher world interest rates. The incidence of payment difficulties on bonds was smaller relative to that on other debts, perhaps pointing to some preferential treatment of these instruments by debtor countries. 1/ In some cases, bonds were restructured as they matured (e.g., Costa Rica and Guatemala), while in others defaults took place (e.g., Panama and Nigeria).

A distinctive feature of the restructurings that took place was that no official bondholder committees were involved in the settlements. Rather, debtor countries facing servicing difficulties made unilateral offers to bondholders, who signalled their approval by exchanging instruments carrying new terms for the old instruments. These offers, however, were formulated based on at least informal contacts with the main groups holding the securities. All settlements were final. No debt reduction was involved, and some of the settlements required the debtors’ to make up-front cash payments of interest arrears (see Table 1). Moreover, interest rates on the new instruments were generally higher than those on the old bonds.

In some cases (notably Panama), the formulation of an acceptable settlement was complicated by the diverse group of creditors involved and the different currencies of denomination of the bonds issued (U.S. dollars, yen, and ECUs). All settlements included negative pledge clauses, providing that bonds would be extended any advantage granted to new or outstanding obligations. In addition, some settlements included early redemption options and allowed for debt conversions into equity. In the cases of Costa Rica and Guatemala, it took from six months to a year to complete a bond restructuring, with 90-100 percent of bondholders accepting the offer. The speed with which agreement was reached and the high level of acceptance reflected the fact that these agreements involved the rollover of maturing obligations, more favorable terms were offered on the new instruments exchanged for the old obligations, and in the case of Costa Rica interest arrears were paid in part up-front as part of the settlements. Restructurings of bonds in default by Nigeria and Panama were also worked out in six months to a year with the participation of virtually all bondholders; however, only after these bonds had been in default for extended periods.

c. Access after recent restructurings

The evidence on new access following recent restructurings is mixed. Broad statements are difficult to make because the sample is small, and some of these restructurings are too recent. More importantly, it is difficult to disentangle the influences of the broader problem with bank debt. On the whole, new access has been quite limited for all countries that restructured their bonds in the 1980s and 1990s. Where countries have managed to raise new funds, amounts have generally been small, interest rate premiums high, and maturities short.

Guatemala returned to the international bond markets in August 1993. The country’s Asociación Nacional del Café raised $60 million through a five-year Eurobond guaranteed by the Government and reportedly partially collateralized with coffee export receivables. The issue was priced at 605 basis points over U.S. Treasury securities with comparable maturities. The spread on this issue substantially exceeded the average spread on other Latin American bonds of around 350 basis points which prevailed at that time.

Costa Rica returned to international capital markets in January 1994, after 14 years’ absence, with a three year $50 million Eurobond at a 395 basis point spread issued by the state-owned electricity and telecommunications monopoly (ICE), without any explicit government guarantee. While the bond was relatively short term and the spread on the high end of the spectrum for Latin American bond issues, the deal was considered a success with demand from a wide range of investors.

Table 1.

Selected Bond Restructurings

article image
article image
article image
Sources: Cardoso E, and Dornbusch R. (1989), “Brazilian Debt Crises: Past and Present,” in B. Eichengreen and Lindert, P. (eds), The International Debt Crisis in Historical Perspective. Cambridge: The MIT Press; Jorgensen and Sachs (1989); various issues of the Foreign Bondholders Protective Council, Inc. Report; and Fund Staff.

References

  • Borchard, E. M. (1983), State Insolvency and Foreign Bondholders. New York & London: Garland Publishing.

  • Cardoso E. and Dornbusch R. (1989), “Brazilian Debt Crises: Past and Present,” in B. Eichengreen and Lindert, P. (eds), The International Debt Crisis in Historical Perspective, Cambridge: The MIT Press.

    • Search Google Scholar
    • Export Citation
  • De Cecco, M. (1984), “The International Debt Problem in the Interwar Period,European University Institute, Working Paper No. 84/103.

    • Search Google Scholar
    • Export Citation
  • Eichengreen, B. (1991), “Historical Research on International Lending and Debt,Journal of Economic Perspectives, Vol. 5, No. 2, pp. 149169.

    • Search Google Scholar
    • Export Citation
  • Eichengreen, B. and Lindert, P. (1989), The International Debt Crisis in Historical Perspective, Cambridge: The MIT Press.

  • Eichengreen, B. and Portes, R. (1989), “Dealing with Debt: the 1930’s and the 1980’s,NBER Working Paper No. 2867.

  • Fishlow, A. (1985), “The Debt Crisis in Historical Perspective,” in M. Kahler (ed), “The Politics of International Debt, Ithaca and London: Cornell University Press.

    • Search Google Scholar
    • Export Citation
  • Foreign Bondholders Protective Council, Inc. Report, various issues.

  • Jorgensen, E. and Sachs, J. (1989), “Default and renegotiation of Latin American Foreign Bonds in the Interwar Period,” in B. Eichengreen and Lindert, P. (eds), The International Debt Crisis in Historical Perspective. Cambridge: The MIT Press.

    • Search Google Scholar
    • Export Citation
  • OECD (1981), Financial Market Trends. March.

  • Balance of Payments. Office of Business Economics U.S. Department of Commerce (1963).

1/

The authors would like to thank Steven Dunaway and Robert Rennhack for very helpful comments. The views expressed in the paper and any remaining errors are the authors’.

2/

Balance of Payments, Office of Business Economics U.S. Department of Commerce (1963). During 1921-28, gross purchases of foreign securities were $6.3 billion (3/4 percent of GDP).

2/

The returns on Latin American bonds are those reported by Jorgensen and Sachs (1989) based on five selected countries.

3/

In general, legal action against defaulting debtor countries had to be taken in the debtor country, which had the option of determining whether it would allow itself to be sued. Even when the debtor country could be sued in its own courts by consent, legal recourse could prove ineffective because the bond issue was generally created by legislation (an act of sovereignty) and could be suspended, modified, or repudiated by a similar method, which would be binding on the national courts (Borchard 1983).

2/

Borchard (1983), cites a mid 1930’s agreement between Great Britain and Greece, that contains such clauses.

3/

In the early stages of the debt crisis of the 1980s, commercial bank reschedulings of syndicated loans followed a similar pattern.

4/

The U.S. Government sometimes intervened directly in the negotiations, although not always successfully. According to Borchard (1983), in 1941 and 1942 the State Department approved a Colombian and a Dominican Republic proposal, but the FBPC disapproved the plans because they were judged not to be fair enough to bondholders.

2/

According to the FBPC Report, some bondholders that accepted Chile’s 1935 unilateral plan received interest payments of 1-1 1/2 percent of face value over several years. However, in 1940, the Chilean Government began to divert to other uses a large part of the funds that should have been used for amortization under the unilateral offer to bondholders. Under the final offer made by Chile and agreed with the FBPC in 1948, those bondholders who had not agreed to the 1935 plan were to receive in payment of back interest the amounts that had been paid to participating bondholders under the plan. In addition, those who had assented late to the Government’s unilateral offer--and under the terms of the offer had not been entitled to payments made before the year of acceptance--were to receive payments made in earlier years.

3/

The agreement reached by Peru in 1943 included provisions to stretch debt maturity from 1958-60 to 1997, whereas the maturities of Brazil’s debt were stretched out by 40 to 60 years.

1/

As an example, in 1903 the external and internal debt of Venezuela was converted into a unified issue.

2/

Interest payments were sometimes reduced on the grounds that interest rate levels had fallen significantly since the time of issue of the original bonds, although such settlements often provided for a gradual increase over time. The interest reduction agreements of Brazil In 1934 and the Province of Buenos Aires in 1935 provided for a rising interest rate schedule. Brazil’s settlement in 1943 also included a provision in one of its two plans for principal reduction.

3/

Chile’s settlement of 1948 included payment of about 12 percent of its unpaid interest since default, Peru’s 1943 agreement included a 10 percent payment, and Brazil’s 1943 agreement only a small portion.

4/

World War II and its immediate aftermath contributed in part to this lengthy delay in regaining market access.

1/

To some extent, the pervasiveness of defaults in the 1930s is explained by the reluctance of bondholders to provide new money, and thus mitigate borrowing countries’ illiquidity as noted in Eichengreen (1991). In addition, creditor-country governments appear to have been unwilling to intervene. Some political nuances to this explanation are added by De Cecco (1984), who argues that the post-World War I recognition by the U.S. and Great Britain that German war reparations --required by the Peace Treaty--were unrealistic, and that Germany would be unable to honor them, opened the door for repudiation and default on foreign obligations as a legitimate recourse for other debtor countries encountering financial difficulties in the 1930s.

1/

This experience is often cited as supporting the current market view that bonds are seen as being senior to other, forms of indebtedness.