III: Financing for Developing Countries

International Monetary Fund
Published Date:
January 1986
  • ShareShare
Show Summary Details

Distribution and Terms


Total lending to developing countries through bank and bond markets (net of loan repayments, redemptions, and bank purchases of bonds) amounted to only about $15 billion in 1985, and net repayments of $3 billion were made by developing countries during the first half of 1986. This reversal in private lending to developing countries occurred even though the current account deficits of developing countries are estimated to have increased by $14 billion in 1986 to $54 billion. The slowdown in private lending during 1985–86 was primarily due to a decline in bank lending, although during 1986 new gross bond issues by these countries also dropped sharply.

International bank lending to developing countries continued to slow during 1985–86. During the first half of 1986, developing countries repaid net $7 billion to international banks, while they had borrowed $9 billion in 1985. 1 Thus, bank claims declined by more than 2 percent (at an annualized rate) during the first half of 1986, compared with an increase of 2 percent in 1985 (Chart 2 and Table 7). 2 The continued slowdown in bank lending to developing countries since 1982, combined with the resurgence in lending to industrial countries, has diminished the share of lending to developing countries in total international lending to 1 percent during 1985–86 from 27 percent in 1982. The ratio of claims on developing countries in total international claims dropped to 20 percent in mid-1986 from 25 percent in 1983 (Chart 10). Bank lending to the 15 heavily indebted developing countries has also steadily declined. While these countries had borrowed $6 billion in 1984, they repaid net $2 billion in 1985 and an additional $3 billion during the first half of 1986.

Table 7.Bank Lending to Developing Countries, 1983-First Half 19861, 2(In billions of U.S. dollars; or in percent)
198319841985First halfFirst half
Developing countries235.315.09.12.3-7.1
Growth rate732-1
Of which:
Côte d’Ivoire-0.1-0.30.1
South Africa2.9-1.5-0.20.1-1.0
Of which:
Of which:
Middle East3.5-1.2-2.2-2.3-1.0
Of which:
Western Hemisphere14.96.0-0.1-0.6-4.1
Of which:
Memorandum items
Fifteen heavily indebted countries11.15.4-1.9-1.2-3.4
Total, BIS-based26.411.313.63.3-9.6
Growth rate7231-2
Gross bond issues3.35.310.24.93.3
Sources: Bank for International Settlements (BIS); Organization for Economic Cooperation and Development; International Monetary Fund, International Financial Statistics; and Fund staff estimates.

Chart 10.Concentration of Cross-Border Claims, 1976–85

(In percent)

Sources: Bank for International Settlements, International Banking and Financial Market Developments; International Monetary Fund, International Financial Statistics; and Fund staff estimates.

1Owing to a change in the coverage of the BIS reporting area, there is a break in the series.

2Argentina, Brazil, Indonesia, Korea, Mexico, the Philippines, and Venezuela.

These data for bank lending to developing countries, however, underrecord actual flows. One reason for this underestimation is that official statistics record only partially banks’ holdings of bonds. Other factors, inter alia, are loan write-offs, which reduce recorded claims without a repayment, the sale of claims by banks to nonbank investors, and the exercising of guarantees. In light of these factors, actual bank lending to developing countries in 1985 is estimated to have been larger than $9 billion, and was probably in the range of $11-15 billion.

This adjusted flow of $11-15 billion does not coincide with the change in banks’ risk exposure to developing countries because a significant part of bank lending has been officially guaranteed. A joint publication of the BIS and the Organization for Economic Cooperation and Development (OECD) 3 gives information on the stock of banks’ officially guaranteed claims on developing countries relative to their nonguaranteed claims, although it does not give an exchange rate adjusted flow. 4 Allowing for this factor, growth in banks’ risk exposure to developing countries in 1985 is estimated to have been 1½ to 2½ percent (based on a stock of nonguaranteed bank claims of about $515 billion). Since reliable regional estimates are not available for all these factors, the discussion of bank lending that follows does not incorporate any estimate based on underrecording of claims or increases in official guarantees.

One reason for the slowdown in bank lending has been the decline in spontaneous lending. Such lending reached $4 billion in 1985, but turned to net repayments of $9 billion during the first half of 1986. Bank lending to most developing countries in Asia and Europe remained spontaneous in 1985 and accounted for more than the total lending to developing countries in that year. There was also a modest increase in 1985 in bank claims on African countries, while bank claims on developing countries in the Western Hemisphere were virtually unchanged, despite concerted lending of almost $5 billion. Developing countries in the Middle East made net repayments of international bank loans in 1985 of $2 billion. During the first half of 1986, net repayments to banks were recorded by developing countries in all regions except Europe, with the largest repayments experienced by developing countries in the Western Hemisphere. Bank lending to the U.S.S.R. and non-Fund members in Eastern Europe fell to less than $1 billion during the first half of 1986, from $4 billion in 1985.

A second reason for the slowdown in bank lending has been the lower level of disbursements under concerted lending packages during 1985 and the first three quarters of 1986 (Table 45). Such disbursements fell by half in 1985 to $5 billion and amounted to only $1.7 billion during the first three quarters of 1986. During 1985–86, over 90 percent, or about $6.5 billion of these disbursements, went to six countries in the Western Hemisphere (Argentina, Chile, Costa Rica, Ecuador, Mexico, and Panama); the remainder was directed to Côte d’Ivoire and the Philippines.

New long-term bank credit commitments to developing countries, including the agreement in principle with Mexico, amounted to $20 billion during the first three quarters of 1986, compared with $18 billion in 1985 and $31 billion in 1984. This pattern in bank credit commitments can be entirely explained by changes in the level of bank commitments for concerted lending as spontaneous commitments reached about $15 billion in 1984 and 1985 and remained virtually unchanged at $16 billion (at an annualized rate) during the first three quarters of 1986 (Table 8).

Table 8.Long-Term Bank Credit Commitments to Developing Countries, 1981-Third Quarter 1986(In billions of U.S. dollars)


Developing countries48.144.634.931.017.914.020.1
Capital importing47.042.632.629.916.113.218.7
Spontaneous lending54.
Concerted lending50.10.1
Spontaneous lending512.812.610.
Concerted lending50.9
Spontaneous lending54.
Concerted lending50.6
Middle East0.
Western Hemisphere25.223.015.315.
Spontaneous lending525.
Concerted lending513.3614.
Sources: Organization for Economic Cooperation and Development, Financial Statistics Monthly; and Fund staff estimates.

New concerted packages were arranged for six countries in 1985 (Chile, Colombia, Costa Rica, Côte d’Ivoire, Ecuador, and Panama). The only new concerted lending commitment during the first three quarters of 1986 was an agreement in principle in late September 1986 between Mexico and its bank advisory committee on a new money package for 1986–87 consisting of three parts: (1) a concerted loan of $6 billion net, of which $1 billion would be cofinanced with the World Bank and half of that cofinanced loan would receive a World Bank guarantee; (2) a contingent loan of $1.2 billion to support public and private sector investment; and (3) a contingent loan of $0.5 billion to support economic growth; this loan would be entirely cofinanced with the World Bank and half of the loan would have a World Bank guarantee.

Restructuring agreements were reached in principle with six developing countries during the first three quarters of 1986 covering their medium-term bank debt amounting to $56 billion, compared with nine agreements in 1985 for a total of $13 billion. MYRAs were agreed with five countries between January 1985 and September 1986 to restructure $51 billion in medium-term debt falling due. The MYRAs in 1985 and the first three quarters of 1986 included those for Côte d’Ivoire (the first for an African country), the Dominican Republic, Mexico, Uruguay, and Yugoslavia. Data on bank debt restructurings are given in Tables 46 through 49.

Gross international bond issues by developing countries (including offshore centers) during the first three quarters of 1986 (at an annualized rate) were about $6 billion, or 60 percent of the amount issued during 1985 ($10 billion). Nevertheless, gross international issues by these countries during the first three quarters of 1986 (at an annualized rate) were nearly 40 percent above the average level in 1983–84. Eighteen developing countries utilized the international bond market during the first three quarters of 1986, compared with 22 countries in 1985 (Table 9).

Table 9.International Bond Issues by Developing Countries, 1981-Third Quarter 19861(In millions of U.S. dollars)


Hong Kong123.871.762.8185.6934.4438.3403.9
Papua New Guinea20.620.320.3
Saudi Arabia200.0
South Africa92.0314.1532.51,013.9777.9777.9
Sri Lanka11.3
Trinidad and Tobago50.0107.4133.951.0
United Arab Emirates25.0
Source: Organization for Economic Cooperation and Development, Financial Statistics Monthly.

Developing countries from Asia and Europe continued to dominate this market, accounting for 75 percent of the total during 1985 and the first three quarters of 1986. Several developing countries (e.g., Algeria, Malaysia, South Africa, and Thailand) did not issue bonds, or issued only a fraction of the amount, during the first three quarters of 1986 ($0.1 billion) that they issued in 1985 ($4.1 billion). Of the $15 billion issued during 1985 and the first three quarters of 1986, about 60 percent was in the form of U.S. dollar floating rate notes and may have been purchased by banks to a large extent. Certain banks did not report them as part of their international banking claims (e.g., banks in the United Kingdom, the United States, and Switzerland).

Since the end of December 1985, banks in the United Kingdom have reported a country breakdown of their holdings of bonds. Data on the outstanding stock of U.K. banks’ holdings of bonds at the end of December 1985 indicate that about 85½ percent of bonds held by U.K. banks were issued by industrial countries and about 6½ percent by developing countries (Table 10). This distribution is roughly the same proportion as average gross bond issues during 1978–86, implying that U.K. banks have not purchased a disporportionately larger share of bonds issued by developing countries than bonds issued by industrial countries. Total holdings by U.K. banks of bonds issued by developing countries amounted to $2.3 billion, or slightly more than 2 percent of U.K. banks’ credit claims on developing countries at the end of 1985 (Table 50). Considerable variation exists in U.K. banks’ holdings of bonds issued by developing countries, with bonds issued by Asian developing countries accounting for about 42 percent of the total.

Table 10.Bond Holdings of U.K. Banks, December 1985(In millions of U.S. dollars, unless otherwise indicated)
Bank Claims Excluding BondsBank Claims Including BondsBond HoldingsPercentage Distribution of Bond HoldingsBond Holdings in Percent of Bank Claims Excluding Bonds
Industrial countries342,823372,41429,59185.68.6
Developing countries102,510104,7912,2816.62.2
Middle East10,72710,800730.20.7
Western Hemisphere40,76041,3305701.71.4
Centrally planned economies12,58112,5832
Offshore centers83,89985,3631,4644.21.7
International organizations1,5082,4729642.863.9
Source: Bank of England, Quarterly Bulletin, March 1986.

During the first three quarters of 1986, developing countries also arranged $5 billion (annualized) in other long-term external bank facilities, compared with $4 billion during all of 1985. Developing countries also improved their access to these facilities during 1986; such countries in all regions except Africa obtained long-term bank facilities during the first three quarters of 1986, while in 1985 commitments were limited to developing countries in Asia and Europe. In 1985, 13 developing country borrowers arranged note issuance facilities amounting to $3 billion, while 10 developing countries arranged such facilities for a total amount of $1 billion in 1984. During the first three quarters of 1986, 12 developing countries arranged note issuance facilities for $4 billion.

Developing countries withdrew $18 billion in deposits with the international banking system during the first half of 1986, compared with depositing of $24 billion in 1985 and $23 billion in 1984 (Table 42). This turnaround reflects the more difficult external financing situation of developing countries, especially the fuel exporting countries. There was also a large shift in the composition of depositing during 1985 and the first half of 1986. Interbank deposits were drawn down during the first half of 1986, after a small increase in 1985, reflecting the sharp deterioration in the gross official reserves position of these countries. During the first half of 1986, nonbanks from developing countries reduced their deposits by $3 billion, after having increased them by $21 billion in 1985.

Regional Pattern of Private Flows

The regional pattern of banking flows to developing countries became more pronounced during 1985 and the first half of 1986. This development was evidenced on both the debtor and creditor side. 5

Western Hemisphere

Countries in the Western Hemisphere repaid $4 billion to private creditors during the first half of 1986, compared with borrowings of $0.1 billion from international bank and bond markets in 1985. International banks were repaid $4.1 billion net by developing countries in the Western Hemisphere, compared with net repayments of only $0.1 billion during 1985. Disbursements under concerted lending packages fell to $1.5 billion during the first three quarters of 1986 from $4.9 billion in 1985 (including $0.5 billion in a trade deposit facility to Argentina). In Argentina, Chile, and Mexico, the increases in bank claims during the first nine months of 1986 were less than the disbursements of concerted loans.

U.S. banks reduced their consolidated claims on countries in the Western Hemisphere by $6.9 billion (annualized rate), or 7.5 percent, during the first half of 1986 and by $4.5 billion, or 4.7 percent, in 1985 (Tables 11 and 12). 6 Nearly half of the decline in 1985, however, represented the sale of international loans by one U.S. bank (see below). Large declines in U.S. bank claims were recorded on Brazil and Mexico, while Argentina was the only developing country for which U.S. banks significantly increased their claims. The consolidated claims of U.K. banks on developing countries in the Western Hemisphere fell by $0.1 billion (½ percent) on an unadjusted exchange rate basis during 1985.

Table 11.Changes in Bank Claims on Developing Countries, 1982–851(In billions of U.S. dollars and in percent; unadjusted for exchange rate changes)

of U.S.



of U.S.



of U.S.



of U.S.

Growth rate
Developing countries
BIS semiannual39.310.
U.S. claims data11.
U.K. claims data6.411.22.33.7-0.9-1.4-0.42-0.62
Capital importing developing countries
BIS semiannual38.
U.S. claims data11.
U.K. claims data6.
BIS semiannual5.510.83.15.5-1.0-
U.S. claims data1.312.41.08.5-0.8-6.0-2.8-22.4
U.K. claims data2.931.10.54.1-0.2-2.0-0.82-6.42
BIS semiannual10.417.18.712.
U.S. claims data3.814.21.44.5-3.0-9.5-3.4-11.9
U.K. claims data1.416.50.43.6-0.3-
BIS semiannual2.737.51.515.
U.S. claims data0.624.20.619.9-0.2-5.0-0.6-18.8
U.K. claims data0.563.60.326.02.1-0.1-3.0
BIS semiannual3.316.
U.S. claims data2.
U.K. claims data0.310.8-0.2-7.4-0.1-2.5-0.1-2.0
BIS semiannual2.423.50.64.8-1.4-
U.S. claims data0.
U.K. claims data0.
BIS semiannual-0.3-
U.S. claims data-0.8-
U.K. claims data-0.2-
Middle East
BIS semiannual3.519.70.73.3-0.7-
U.S. claims data0.
U.K. claims data0.532.0-0.2-11.2-0.2-13.32.0
Western Hemispnere
BIS semiannual19.
U.S. claims data6.
U.K. claims data1.
BIS semiannual0.
U.S. claims data-0.2-
U.K. claims data-0.3-
BIS semiannual8.
U.S. claims data3.621.
U.K. claims data1.
BIS semiannual5.810.
U.S. claims data2.913.
U.K. claims data0.
BIS semiannual1.35.0-0.3-1.1-0.9-
U.S. claims data1.110.5-0.3-2.8-0.4-4.0-0.7-6.7
U.K. claims data-0.1-4.3-0.2-5.4-0.1-4.2-0.1-2.9
Sources: Bank for International Settlements, The Maturity Distribution of International Bank Lending; Federal Financial Institutions Examination Council, Country Exposure Lending Survey; and Bank of England, Quarterly Bulletin.
Table 12.Changes in Claims of U.S. Banks on Developing Countries, 1982–861(In billions of U.S. dollars; and in percent)
Billions of U.S. dollarsGrowth rateBillions of U.S. dollarsGrowth rateBillions of U.S. dollarsGrowth rateBillions of U.S. dollarsGrowth rateBillions of U.S. dollarsGrowth rate
Developing countries
All banks11.
Nine major banks7.
Next 15 banks2.911.
Capital importing developing countries
All banks11.
Nine major banks7.
Next 15 banks2.811.
All banks1.312.41.08.5-0.8-6.0-2.8-22.4-2.5-25.4
Nine major banks0.78.10.910.2-0.8-8.1-1.6-18.3-1.6-22.2
Next 15 banks0.534.50.315.40.211.9-0.9-38.2-0.6-40.7
All banks3.814.21.44.5-3.0-9.5-3.4-11.9-6.6-26.1
Nine major banks2.714.80.31.3-2.0-9.3-2.6-13.5-3.9-23.4
Next 15 banks0.
All banks0.624.20.619.9-0.2-5.0-0.6-18.8-0.1-5.0
Nine major banks0.526.90.521.0-0.3-8.8-0.5-16.7-0.1-4.8
Next 15 banks10.
All banks2.
Nine major banks1.526.4-0.5-6.8-1.0-15.5-0.5-9.7-1.8-34.5
Next 15 banks0.418.90.520.50.4-0.5-17.0-1.1-47.1
All banks0.
Nine major banks0.
Next 15 banks1.72.7-0.1-4.6-0.1-6.8-0.3-25.5
All banks-0.8-
Nine major banks-0.3-4.60.812.8-0.4-4.9-0.5-6.5-0.2-2.8
Next 15 banks-0.1-
Middle East
All banks0.
Nine major banks0.
Next 15 banks0.
Western Hemisphere
All banks6.
Nine major banks4.
Next 15 banks1.912.11.37.3-0.1-0.6-2.5-13.0-2.3-13.5
All banks-0.2-
Nine major banks-0.1-
Next 15 banks0.
All banks3.621.
Nine major banks2.725.02.518.8-0.3-1.6-0.6-4.1
Next 15 banks0.930.70.410.30.410.0-0.8-16.6-0.5-12.7
All banks2.913.
Nine major banks1.311.
Next 15 banks0.818.70.24.0-0.2-0.6-
All banks1.110.5-0.3-2.8-0.4-4.0-0.7-6.7-0.8-7.4
Nine major banks0.811.3-0.2-2.2-0.2-2.6-0.3-4.1-0.4-5.8
Next 15 banks0.318.81.3-0.1-5.3-0.3-15.9-0.2-10.8
Source: Federal Financial Institutions Examination Council, Country Exposure Lending Survey.

During the first three quarters of 1986, new long-term bank commitments to developing countries in the Western Hemisphere, apart from the concerted lending commitment to Mexico, amounted to only $0.1 billion, reflecting a commitment to The Bahamas; such commitments totaled $2.5 billion in 1985, of which $2.4 billion was on a concerted basis.

Developing countries in the Western Hemisphere did not have significant access to the international bond market during 1985–86. In 1985, bond issues totaling $202 million were made by Barbados, Mexico, and Trinidad and Tobago. The bond placement by Mexico (a foreign yen placement) was its first since the debt crisis of 1982. Mexico also arranged $0.1 billion in a note issuance facility, reportedly the first of a series designed to transform interbank lines into marketable securities. During 1986, Brazilian banks also arranged a note issuance facility for their interbank lines. There were bond issues of $0.5 billion during the first three quarters of 1986 by Barbados, Bermuda, Colombia, and Mexico.

Residents from countries in the Western Hemisphere withdrew $4.8 billion in deposits with international banks during the first half of 1986, after they had deposited abroad $5.6 billion in 1985. Interbank deposits, on the one hand, fell by $8.2 billion during 1985 and the first half of 1986, in line with a decline in gross official reserves; reductions in interbank deposits by banks in Brazil and Mexico were particularly large (Table 43). Nonbanks in the Western Hemisphere, on the other hand, deposited $8.5 billion in 1985, but increased their deposits with international banks by only $0.5 billion during the first half of 1986. New deposits of nonbanks in 1985 were again particularly large for residents of Mexico; during the first half of 1986, for the first time in more than two years, nonbanks from Mexico drew down their international deposits. In 1985, depositing by nonbanks from Venezuela were 50 percent above the 1984 level, but these deposits were also reduced during the first half of 1986 (Table 44).


International lending through bank and bond markets to developing countries in Asia declined to only $1 billion during the first half of 1986 from $12.8 billion in 1985, as a result of reduced borrowing in the bond market and net repayments to international banks. Bank lending to developing countries in Asia declined to $6.9 billion in 1985 from $8.2 billion in 1984, partly because of increased borrowing in the bond market. During the first half of 1986, however, bond issues decreased to $2.3 billion and banks were repaid net $1.3 billion.

The largest bank borrowers during 1985 were China, Korea, and India, while China and Malaysia recorded the largest net repayments to banks during the first half of 1986. The Philippines experienced a net outflow because of net repayments to banks of $0.5 billion in 1985, despite having received $0.4 billion in concerted lending, partly because the branches of Philippine banks abroad (which were not covered in the restructuring agreements) reduced their claims on the Philippines; during the first half of 1986, liabilities of the Philippines vis-à-vis international banks remained unchanged, notwithstanding concerted lending disbursements of $0.2 billion. Malaysia repaid $1.4 billion in bank debt in 1985, while issuing $2.0 billion in international bonds, of which 65 percent were in the form of U.S. dollar floating rate notes. During the first half of 1986, Malaysia continued to repay bank debt, but did not issue any international bonds.

The consolidated claims of U.S. banks on developing countries in Asia fell by $3.4 billion (12 percent) in 1985, while those of U.K. banks rose by $0.2 billion (2 percent). During the first half of 1986, the consolidated claims of U.S. banks on developing countries in Asia continued to decline, falling at an annualized rate of $6.6 billion, or 26 percent.

New long-term international bank credit commitments to developing countries in Asia were $7.9 billion (at an annual rate) during the first three quarters of 1986, slightly above their level in 1985; all of these commitments were spontaneous. During 1983–86, the composition of bank commitments changed, with a decline in credit commitments to about $7.5 billion during 1985–86 from an average of $10.3 billion in 1983–84 and a tripling of the use of other bank credit facilities to $3 billion during 1985–86. The major borrower in the credit market continued to be Korea; however, credits of $3.4 billion were arranged for China during 1985–86. Hong Kong was the most active user of other bank credit facilities, arranging commitments of $1.8 billion during 1985–86. Korea also arranged $1.3 billion in various other bank credit facilities during 1985–86.

Asian developing countries reduced their reliance on international bond markets during the first three quarters of 1986 from the peak level experienced in 1985, but nevertheless international bond issues by these countries remained about 50 percent above the average level of 1982–84. Bond issues by these developing countries totaled almost $3 billion at an annual rate during the first three quarters of 1986, compared with $6 billion in 1985 and $3 billion in 1984. In 1985, Malaysia borrowed $2.0 billion, Korea $1.7 billion, and China almost $1 billion. During the first three quarters of 1986, China, Korea, and Indonesia were the principal borrowers in international bond markets, while Thailand, which was a large issuer in 1985, did not issue any international bonds during the first three quarters of 1986 and Malaysia—another large issuer in 1985—issued less than $0.1 billion during the first three quarters of 1986.

Countries in Asia deposited $8.5 billion in international banks in 1985 (mainly interbank transactions), somewhat less than in 1984; during the first half of 1986, deposits by these countries were reduced slightly. Deposits held by residents in China fell by about $5.6 billion in 1985 and continued to decline by $1.3 billion during the first half of 1986.


International lending through bank and bond markets to developing countries in Europe amounted to about $4.7 billion in 1985 and $1.1 billion during the first half of 1986. Bank lending was $3.2 billion in 1985 and $0.5 billion during the first half of 1986. During 1985, bank lending was mostly directed to Greece, Hungary, and Turkey, while banks further reduced their claims on Romania. During the first half of 1986, only Hungary and Turkey experienced net lending from international banks, while all other developing countries in Europe made net repayments to banks.

During 1985 and the first three quarters of 1986, all new long-term international bank credit commitments to developing countries in Europe were on a spontaneous basis. These bank commitments rose from $4.4 billion in 1985 to $4.9 billion at an annualized rate during the first three quarters of 1986. The major borrowers were Greece, Hungary, Portugal, and Turkey. Just as in the Asian region, developing country borrowers in Europe have increased their issuance of other bank facilities. In 1985, other facilities totaled $0.9 billion, while during the first three quarters of 1986, commitments under these other facilities reached $1.9 billion at an annualized rate.

During the first three quarters of 1986, developing countries in Europe issued $0.9 billion in international bonds, with Portugal the principal issuer. These countries issued $1.6 billion in international bonds in 1985. Almost all of this decrease was accounted for by Greece and Hungary; bond issues by Greece declined to $205 million in 1986 from $745 million in 1985, while Hungary issued $236 million during the first three quarters of 1986, compared with $447 million in 1985. International bank depositing by developing countries in Europe dropped to $2.0 billion in 1985 from $3.6 billion in 1984, largely attributable to a withdrawal of deposits by Romania and Turkey. This trend continued during the first half of 1986, as deposits increased by only $0.2 billion, with Yugoslavia recording a substantial drawdown.


International lending through bank and bond markets to developing countries in Africa was $2.7 billion in 1985, but during the first half of 1986, $1.1 billion was repaid net. Banks lent $1.4 billion to African countries in 1985 and received net repayments of about $1.2 billion during the first half of 1986. Excluding loan repayments by Nigeria and South Africa, bank lending to African developing countries was $2.3 billion in 1985 and zero during the first half of 1986. New long-term international bank credit commitments to developing countries in Africa declined to $1.1 billion during the first three quarters of 1986 after having increased to $1.4 billion in 1985 from $0.5 billion in 1984.

Bond issues by developing countries in Africa were only $0.1 billion during the first three quarters of 1986, down from $1.3 billion in 1985. South Africa and Algeria issued $0.8 billion and $0.5 billion in bonds, respectively, in 1985, but issued zero and $0.1 billion, respectively, during the first three quarters of 1986. Depositing by residents of Algeria and South Africa in 1985 accounted for almost half of the $3.8 billion increase in bank deposits from African countries in 1985; deposits were reduced by $0.8 billion during the first half of 1986.

Middle East

International lending through bank and bond markets to developing countries in the Middle East totaled $1.0 billion in net repayments during the first half of 1986, compared with net repayments of $2.1 billion for all of 1985. Bank claims on these developing countries dropped by $2.2 billion in 1985 and by a further $1.0 billion during the first half of 1986. The cumulative decline was due to repayments of bank debt by Egypt, Israel, and Kuwait. New bank commitments amounted to $0.4 billion in 1985, about the same as in 1984, but fell to $0.1 billion during the first three quarters of 1986. Developing countries in the Middle East also issued $0.1 billion in international bonds in 1985, but did not issue any international bonds during the first three quarters of 1986. Banks received $4.0 billion in deposits from residents in the Middle East in 1985, as depositing by nonbanks in the region rose to $6.2 billion in 1985 from virtually zero in 1984. During the first half of 1986, residents from this region withdrew $11.6 billion in deposits from international banks. This reduction in deposits was to a large extent attributable to a fall in international reserves of Kuwait and Saudi Arabia.

Distribution of Lending

As regards the regional distribution of lending by different nationalities of banks, there was a sharp decline in U.S. banks’ consolidated claims on developing countries during 1985 and the first half of 1986. These claims fell by 8½ percent in 1985 and by 13 percent at an annual rate during the first six months of 1986 (Table 12). However, the decline in 1985 may be overstated; U.S. claims for the end of 1985 were reduced as the result of the sale by Crocker National Bank of $3.1 billion of its international loans to Midland Bank. 7 In 1985, Crocker’s claims on Latin America fell by $2.2 billion, amounting to nearly one half of the total reduction of U.S banks’ claims on that region. 8 U.S. banks’ claims fell on all developing country regions during 1985–86, with the largest percentage declines for developing countries in Africa, the Middle East, and Asia. After adjustment for guarantees and other risk transfers, the decline in U.S. banks’ risk exposure to developing countries was 6½ percent in 1985 and 11 percent at an annual rate during the first half of 1986.

Consolidated claims of U.K. banks expressed in U.S. dollars declined by $0.4 billion (0.6 percent) in 1985 with the fall in these claims concentrated on countries in Africa (6.4 percent); however, this decline was wholly due to a statistical discontinuity in the data for South Africa. Excluding U.K. banks’ claims on South Africa, U.K. banks’ claims on developing countries rose by $0.5 billion in U.S. dollar terms (1 percent). U.K. banks’ claims on countries in Asia, Europe, and the Middle East increased by 2–3 percent in 1985. In the Western Hemisphere, U.K. banks’ claims expressed in U.S. dollars declined by $0.1 billion, or nearly 1 percent; this recorded decline was not affected by the transfer of claims from the Crocker National Bank. These data probably understate the decline in U.K. banks’ claims during 1985, due to exchange rate movements. Guarantees of U.K. bank claims declined slightly in 1985.

German domestic banks showed an increase in claims on developing countries of $1.2 billion (4 percent) during the first half of 1986, after an increase of $2.7 billion (12 percent) in 1985 (allowing for an approximate adjustment for exchange rate changes). The increase in 1985 may largely reflect a transfer of claims to domestic banks from their branches and subsidiaries abroad, whose claims on developing countries fell by $0.5 billion (7 percent) and $1.9 billion (16 percent), respectively. Overall, German banks may have increased modestly their claims on developing countries from Asia during 1985 and on developing countries in Europe during 1985 and the first half of 1986. No geographical analysis of Japanese banks’ claims is published. However, a comparison of lending by other major groups of banks indicates a substantial increase in Japanese banks’ claims on Asian countries, including purchases of bonds. A modest increase may have occurred in Japanese banks’ claims on developing countries in the Western Hemisphere.

Terms of Lending and Restructuring

Terms on new bank credit commitments and bank debt restructuring have generally improved during 1985–86. Terms data on new bank credit commitments supplied by the OECD do not include restructuring terms. Average spreads on new bank credit commitments to developing countries continued to decline during 1985 and the first three quarters of 1986. Spreads for these countries fell to 62 basis points during the first three quarters of 1986 from 94 basis points in 1985. These spreads represent a significant drop from 135 basis points in 1984 and from a peak of 151 basis points in 1983. The difference between spreads on developing countries and industrial countries narrowed to about 28 basis points during the first three quarters of 1986, from 53 basis points in 1985 and 81 basis points in 1984. The average maturity of new bank credit commitments remained unchanged during 1985 at eight and three-fourths years but shortened to eight years during the first three quarters of 1986. Nevertheless, average maturity for bank credit commitments to developing countries remained almost one year longer than similar commitments to borrowers from industrial countries.

The improvement in spreads for developing countries during 1983–86 has been mainly due to lower spreads for concerted lending packages. OECD data on the spreads on spontaneous bank loans to developing countries indicate that, after having dropped by 31 basis points in 1983, they declined by only a further 9 basis points between 1983 and 1985 and by another 6 basis points during the first half of 1986. By contrast, the average spread on concerted lending packages has dropped from 200 basis points in 1983 to 99 basis points in 1985; there were no concerted lending packages recorded by the OECD in the first three quarters of 1986.

In September 1986, the bank advisory committee for Mexico reached an agreement in principle on a concerted loan with a spread of 13/16 of a percentage point (81 basis points) over LIBOR or domestic cost of funds. The difference between average spreads on spontaneous bank and concerted lending to developing countries decreased to 31 basis points in 1985 from 122 basis points in 1983. The spread on the concerted loan for Mexico would be about 20 basis points above the average spread for spontaneous bank loans during the first three quarters of 1986. Maturities of concerted loans have lengthened by 3–4 years between 1983 and 1985 (Table 13). The concerted loan agreed in principle for Mexico in September 1986 represents a further lengthening of maturities; it has a term of 12 years.

Table 13.Terms of Selected Bank Debt Restructurings and Financial Packages, 1983-Third Quarter 19861
CountryYear of AgreementType of TransactionGrace Period (In years)Maturity (In years)Interest Rate (In percent spread over LIBOR/U.S. prime)
Argentina1983New financing32¼-2⅛
1984Restructuring310 to 121⅜-1⅜
1984New financing3101⅝-1¼
1983New financing82⅛-1⅞4
1984New financing592-1¾
Chile1983New financing472¼-2⅛
1984New financing591¾-1½
1985New financing101⅝-1¼
Costa Rica1983Restructuring6½ to 7½2¼-2⅛
Côte d’Ivoire1984Restructuring6271⅞-1⅝
1984New financing371⅞-1⅝
Dominican Republic1983Restructuring152¼-2⅛
1983New financing62⅜-2¼
1984Restructuring 83121⅜
1984New financing2101⅝-1¼
1983New financing362¼-2⅛
1984New financing101½-1⅛
14 ⅞ in 1985–86
1984Restructuring80 to 11⅛ in 1987–91
1¼ in 1992–98
1986New money51213/16
1986New money971213/16
1986New money104813/16
1984New financing591¾-1⅜
1983New financing262¼-2⅛
1986Restructuring8, 113121⅜
1986Restructuring8, 123121⅜
1983New financing361⅞-1¾
Sources: Restructuring agreements; and press reports.

Average spreads on restructuring agreements fell by about 70 basis points between 1983 and 1985 (Table 13). During the first three quarters of 1986, spreads on restructurings declined by another 45 basis points. An examination of spreads on individual restructuring agreements shows that the declines in spreads were similar for different sizes of borrowers. The average spread on restructured bank debt for the four developing countries with the largest bank debt has been consistently ⅛—¼ percent less than the spread on similar debt for other developing countries. MYRAs represent the clearest example of an overall improvement in terms for developing countries, as spreads were reduced, fees eliminated, and repayment periods extended. The restructuring agreements for Brazil and Mexico in 1986 continue the improvement in spreads.

Association with Policy Reform

Overview of Linkage to Reforms

Since the onset of widespread debt-servicing difficulties in 1982, commercial banks have generally sought to associate new financing and restructuring with debtor countries’ progress in implementing macroeconomic policy reforms. Banks have generally phased their disbursements under new concerted lending packages in line with purchases under a Fund arrangement, thus linking their financial contribution to debtor countries’ implementation of macroeconomic policies. Typically, bank debt restructuring agreements have also been conditioned on the existence of an arrangement to use Fund resources or enhanced surveillance.

One important recent development is that commercial banks have sought to link their financing to structural and sectoral reforms implemented with the support of the World Bank, in addition to their traditional association with the Fund. Banks have thus tied some disbursements of concerted loans and some restructuring agreements to World Bank involvement. In Chile and Uruguay, the World Bank assisted in catalyzing commercial bank financing through formal cofinancing arrangements. A second recent development has been the wider application of bank restructuring agreements, such as MYRAs, which cover a consolidation period beyond the period of an existing Fund arrangement. In these cases, banks have sought to ensure that the restructuring of maturities falling due in the later years of the consolidation period would be conditional on satisfactory policy implementation. They have thus included in their agreements conditions relating to future Fund involvement. These developments are discussed in detail below.

There has been considerable diversity in the types of linkage to World Bank activities that have been incorporated in commercial bank financing packages. The linkage may be broadly defined as commercial banks requiring evidence of “progress” in negotiation or implementation of structural reforms. In other cases, the linkage to World Bank loans has been based on certification by the World Bank that debtor countries have borrowed a specified amount by a particular deadline. In some cases, all commercial bank disbursements under a new money package have been tied to performance under both Fund- and Bank-supported programs, while in still other cases linkage to the World Bank applied to only one disbursement from a new money package.

In Chile’s new money agreement signed in November 1985, in addition to certification from the Fund, banks required a notification from the World Bank before each disbursement confirming that Chile’s Structural Adjustment Loan—SAL (or “comparable facility”) was in effect, and that Chile had drawn the full amount (which needed to occur by specified dates) expected to be available to it under the SAL.

In addition to linkage to the Fund, one commercial bank disbursement under Colombia’s new money agreement of December 1985 was contingent on confirmation from the World Bank that Colombia would have access to the second tranche of an IBRD trade policy loan.

In 1985, commercial banks conditioned their restructuring and new money agreements with Costa Rica on the implementation of a World Bank SAL, as well as on a Fund program. Successive delays with implementation of the SAL and the Fund stand-by arrangement meant that the second disbursement of new bank money—expected originally in mid-1985—was made in November 1985.

For Côte d’Ivoire, disbursements under a 1985 new money agreement were conditional on a statement by the World Bank on the eligibility of the borrower for drawdowns under a SAL and indicative statements about future loans. The restructuring agreements for 1984 and 1985 maturities also contained refinancing conditions involving both the Fund and the World Bank. The World Bank condition was not met, as the 1985 SAL was not signed as expected, and banks waived that condition during 1985 in order to continue the refinancing.

In addition to conditions involving the Fund, Panama’s new money agreement for 1985–86 required certification from the World Bank regarding progress in negotiating and implementing a SAL. The double linkage to the Fund and the Bank created difficulties in this case. Although the Fund program was on track, commercial banks did not make the first disbursement when it was scheduled, because adequate progress had not been made with the World Bank. Some months later, progress had been made with the World Bank, but Panama was no longer in compliance with the Fund program, and so the initial disbursement by commercial banks was again delayed.

Monitoring arrangements for MYRAs have raised the issue of bank linkage to policy implementation in future years. In some cases, where banks have restructured more than a single year’s maturities, the consolidation period has been covered by a multiyear Fund arrangement (e.g., Chile). In other cases, banks have been willing to restructure maturities beyond the period of an existing Fund arrangement, but they have sought to include in the loan agreements some continuing Fund involvement. As discussed in the following section, in four of these cases (Ecuador, Mexico, Venezuela, and Yugoslavia), enhanced surveillance has been proposed to the Fund’s Executive Board. Thus, commercial banks have been able to link their restructuring to Fund involvement in the form of enhanced surveillance. In some other cases, banks have concluded MYRAs with countries for which the Fund Board has not approved the use of enhanced surveillance.

Enhanced Surveillance

In 1984 and early 1985, when MYRAs were discussed between creditor banks and debtor countries, it became evident that some form of Fund monitoring could be helpful in restoring normal market access for certain indebted countries. The procedure of enhanced surveillance was therefore developed to improve a country’s capacity to design, implement, and monitor economic policies and to provide information about those policies to creditors; to support banks’ risk evaluation through timely and comprehensive information and through the Fund’s forward-looking assessment of domestic policies; and to foster a shift in responsibility for lending decisions back to commercial banks by avoiding on/off financing indications from the Fund. Enhanced surveillance was conceived as an exceptional and temporary adaptation of Fund procedures and practices for countries with a good record of adjustment and in a position to present an adequate quantified policy program in the framework of consultations with the Fund; it was not intended to become a substitute for stand-by and extended arrangements.

The criteria for the adoption of the enhanced surveillance procedures have been set out in the summing-up of the discussion of “The Role of the Fund in Assisting Members with Commercial Banks and Official Creditors” on September 4, 1985. Enhanced surveillance may be undertaken when the following four conditions are met:9

First, at the request of a member country, who must initiate the procedures;

Second, in cases where a good record of adjustment has been shown;

Third, in cases in which a MYRA is needed to normalize market relations and to facilitate the return to voluntary or spontaneous financing;

Fourth, in cases where the member is in a position to present an adequate quantified policy program in the framework of consultations with the Fund staff, which are part of the procedure of enhanced surveillance.

The duration of enhanced surveillance was also discussed, and the following conclusion was reached:

Directors thought that, on the whole, the early cases of enhanced surveillance had covered rather too long periods. They felt that in the future the Fund should try to limit the procedure to about the consolidation period of a MYRA.

Experience with the four cases of enhanced surveillance—Ecuador, Mexico, Venezuela, and Yugoslavia—has been limited. Thus far, Fund staff reports have been distributed to creditor banks only in one recent case (Yugoslavia). There are close similarities between the bank monitoring arrangements in these four cases. (Ecuador and Yugoslavia have monitoring arrangements with official creditors also.) All of these bank monitoring arrangements foresee the provision of annual reports by the Fund on the countries’ quantified financial program and semiannual reviews of the implementation of that program. These semiannual staff reports would be released by the member countries to their creditor banks. There are some variations of detail in the monitoring procedures but, in each case, key policy areas such as monetary and fiscal policy are to be reviewed.

The effectiveness of enhanced surveillance depends crucially on creditors’ performing their own assessments of the debtor country’s policies and, if necessary, seeking to influence these policies in a timely fashion. The move toward serial MYRAs, rather than block restructuring, has reflected banks’ desire to apply conditions separately to each proposed segment of a restructuring. In a serial MYRA, banks have successive opportunities to halt the restructuring if they are not satisfied that adequate policy implementation has taken place, whereas in a block restructuring, the conversion of maturities falling due in the future could be halted only by an event of default.

In a serial MYRA, banks have made the approval of each annual restructuring conditional on some form of satisfactory Fund involvement, which may be defined as a Fund arrangement or enhanced surveillance or—more broadly—as adequate monitoring procedures. In such cases banks may reconsider at the time of each tranche of the restructuring the specific form of monitoring procedures and policy programs that are satisfactory. The following background information focuses on the overall duration of enhanced surveillance and linkage between the monitoring arrangements and the continuing implementation of restructuring.

Venezuela’s request for enhanced surveillance was approved by the Fund’s Executive Board in May 1985 to support a bank MYRA that covers 1983–88 maturities. Enhanced surveillance is scheduled to continue through 1997, until the final amortization of the restructured debt. Because the Venezuelan MYRA restructures a complete block of maturities, the agreement does not include a specific provision for a date on which banks may vote to discontinue the restructuring. At any time, however, two thirds of the banks may call an event of default if they consider Venezuela’s economic program unsustainable. Also, a decrease in the operating reserves of the Central Bank below $2 billion would trigger an event of default. The Fund’s Executive Board has discussed three staff reports on Venezuela under enhanced surveillance. However, because the bank MYRA with Venezuela has not been implemented, banks have not as yet received any of these reports.

Enhanced surveillance for Mexico was due to begin in January 1986, after the extended Fund facility had expired, and will continue through 1990 or 1994, depending on the date of final repayment of the 1983 new money package. The MYRA restructures maturities falling due during 1985–90 in two blocks of three years each. In addition, in the event of Mexico’s economic situation or prospects deteriorating to the point that it would be unable to meet its financing requirements through normal market channels, it agreed to seek financing from other sources, which could include a request to use Fund resources. The Mexican authorities’ request to use Fund resources was approved in principle by the Fund’s Executive Board on September 8, 1986.

Ecuador agreed with its bank creditors to request enhanced surveillance in support of a MYRA that restructures maturities due in 1985–89, with final repayment in 1996. For the restructuring of the 1985–86 maturities, Ecuador was required to be under a standby arrangement through both years. After 1986, the restructuring is “serial,” insofar as there is explicit provision for a majority of banks to halt the restructuring in any year, if Ecuador’s financial program is judged inadequate by the banks, or if its external situation or prospects deteriorate. The bank MYRA does not envisage enhanced surveillance beginning until after the existing Fund arrangement expires in mid-1987. Enhanced surveillance would continue for ten years to the final amortization payment of the restructured debt.

The bank MYRA for Yugoslavia covers restructuring of maturities falling due in 1985–88, with the final amortization payments in 1996. Enhanced surveillance for Yugoslavia began on the expiration of the Fund arrangement on May 16, 1986 and is to continue through 1991. A novel aspect of the bank MYRA with Yugoslavia was the inclusion of a trigger mechanism to indicate likely difficulties in meeting future repayments. The major purpose of this mechanism is to shorten the time lag between implementation of remedial action, if needed, and the diagnosis of such a need, and to facilitate the assessment of the situation by creditor banks. Remedial action could potentially lead to an approach to the Fund by the authorities. The trigger clauses agreed between the creditor banks and the authorities apply to external reserves and export earnings. The experience with these trigger mechanisms is yet to be tested.

Developments in Restructuring Packages

The previous study on international capital markets discussed the introduction of new forms of financing techniques in new money packages. Arrangements such as currency redenomination, relending and on-lending, loan sales and swaps, and conversion of debt to equity have been introduced as ways of diversifying modalities in some financing packages and allowing banks to match their claims more closely with their long-term business interests and customer relations. New money packages have also included trade facilities or commitments to provide trade financing, partly reflecting banks’ preference to provide financing that is linked to trade, which may also support the operations of the domestic business clients. Some points of interest in recent restructuring packages are reviewed below.

Currency Redenomination

Under several recent restructuring agreements, banks have been permitted, at their option, to redenominate existing loans in their domestic currencies or the ECU. For non-U.S.-dollar-based banks, such redenominations reduce the funding risk and the effect of future exchange rate movements on the banks’ claims relative to their domestic currency capital. Banks may elect to redenominate their loans before the restructuring agreement is signed, or for a specified period thereafter. The period for such conversions has varied, extending up to four years. Between 50 and 100 percent of existing loans denominated in currencies other than the banks’ home currency have been eligible for redenominations. In the case of Venezuela, an additional effective ceiling limits conversions to about one seventh of the total debt restructured.

Some banks consider a currency redenomination option as an incentive to participate in financing packages. For debtor countries, the benefits from currency diversification are difficult to predict. Possible savings on interest payments have to be weighed against the potential increase in debt and debt service payments from a further decline of the U.S. dollar. During 1985, the timing of a currency redenomination from U.S. dollars into one of the other major currencies would have been crucial in determining the overall gain or loss for a debtor.

Currency redenomination options exist in the restructuring agreements with Argentina, Mexico, the Philippines, Uruguay, Venezuela, and Yugoslavia. Discussions with banks suggest that where options have already been exercised, banks in Japan and Switzerland have taken advantage of such schemes to redenominate their claims; U.K. banks have redenominated about half of eligible claims, and Italian banks have in some cases switched loans into ECUs.

The agreement between Argentina and the banks envisages that almost a quarter of the debt restructured is eligible for redenomination. Significant amounts of restructured debt are believed to have been redenominated from U.S. dollars into Japanese yen and deutsche mark, although the exact amounts are not known.

In the case of Mexico, about half of the restructured debt is covered by a currency redenomination option; of this portion, a maximum of 50 percent is eligible for redenomination. Mexico’s creditor banks seem to have maintained the U.S. dollar denomination of their assets to a large extent, although banks estimate that up to $5 billion, or 7 percent, of Mexico’s debt may have been redenominated largely into sterling, yen, and deutsche mark claims.

In the case of the Philippines, the currency composition of the first tranche of new bank money indicated a continuing strong preference for the U.S. dollar, with a share of 70 percent, but also a considerable interest in yen (18 percent) and ECU (5½ percent) denomination. The MYRA agreement for Uruguay permits banks to redenominate the principal restructured at each annual advance. This selection can be changed prior to each annual advance date during the consolidation period.

The restructuring agreement with Venezuela, which includes a redenomination option, had not been implemented by mid-1986. The Venezuelan authorities have indicated that they expected that many Japanese banks and German banks could eventually elect to redenominate their loans into their home currencies, and that the overall ceiling of $3 billion could be reached. The MYRA agreement for Yugoslavia permits banks to redenominate the restructured debt at each refunding date into one of a list of “preferred” currencies.

A different form of currency redenomination has taken place with Sudan, where, at the option of the debtor, all of the restructured bank debt (almost $1 billion) was converted into Swiss francs from U.S. dollars. This transaction was part of a modification of the 1981 restructuring agreement and was signed in October 1985. The main reason for this transaction was to reduce the interest obligation on the restructured debt. The conversion was undertaken at a Swiss franc/ U.S. dollar exchange rate of Sw F 2.17 per US$1; at the end of September 1986, the Swiss franc/U.S. dollar exchange rate was Sw F 1.64 per US$1.

On-Lending and Relending

On-lending and relending both involve the reallocation of credit to a different debtor within the same country. On-lending occurs when the lender and the original borrower agree that the proceeds of a new money loan will be transferred to a new obligor who takes over the obligation to repay from the original borrower. The latter, however, often guarantees repayment of the loan. Relending involves the repayment of an existing debt by the original borrower to the lender, which lends the proceeds to other borrowers in the country. Relending and on-lending enable banks to maintain or develop business relationships with clients in developing countries, to support the export activities of their customers, and, more generally, to reallocate the credit risks among different borrowers within a certain country. However, switching in the direction of lending under these arrangements can only be accommodated to a modest extent within the framework of a financial program.

Provisions for on-lending and relending in restructuring or new money packages in Argentina, Brazil, Chile, the Philippines, and Venezuela were described in the previous study on international capital markets. In Brazil, relending under the 1983 and 1984 restructuring and the 1985 interim arrangement and on-lending under the 1983 and 1984 new money packages amounted to $8.8 billion in 1984 and to $5.5 billion in 1985. Since September 1985, the Central Bank of Brazil has been reluctant to approve new applications for relending and on-lending, and no more than $1 billion is expected for relending and on-lending operations in 1986. Under Argentina’s $3.7 billion new money agreement signed in August 1985, $510 million was estimated to have been on-lent up to July 1986. In Mexico, a modest amount of on-lending has been permitted, while in the cases of Chile and the Philippines no significant relending or on-lending has taken place (despite the provisions in the respective restructuring and new money agreements). The Venezuelan MYRA allows relending, but the process may only start in 1987.

Loan Swaps and Sales

Banks in various countries have engaged in sales and swaps of loan claims to eliminate their claims on a certain country and to concentrate claims on a country where prospects and future relations were viewed more favorably. In addition, some major banks have assembled packages of loans from banks with small exposures to provide finance for industrial clients to make investments in certain developing countries through debt-to-equity swaps.

No standard practice exists in the banking community on the treatment of new money obligations associated with sold or swapped loans. For some banks, the major reason for selling such loans appears to be reluctance to participate in future concerted lending packages. In certain instances, however, banks discovered that, even though they had sold claims on particular developing countries, they had not been released from obligations to contribute to new money packages. Banks observed that it would be crucial that legal arrangements clarify that the potential new money obligation of a “selling” bank had been extinguished in a manner definitive and satisfactory to all partners.

One factor that has kept the market for loan sales small and has inhibited larger banks from participating is the potential impact of discounted sales on the valuation of existing assets by auditors and bank supervisors. Banks are concerned that loan sales or swaps on a large scale could require, at a future stage, that the value of comparable assets be marked down. Nevertheless, in discussions with banks it appeared that the loan sale and swap market is expanding. Market-makers have estimated that $10 billion to $15 billion (counting both sides of the transactions) have changed hands in this market, compared with over $300 billion of bank debt outstanding to countries that have restructured since 1982.

Discounts associated with loan sales have ranged from 10 percent to 90 percent; variations among countries, and over time for some countries, have reflected the perceived creditworthiness of the debtor countries and the levels of provisioning and write-offs. Debt of the major debtor countries in the Western Hemisphere has traded at 20–40 percent discounts, according to market participants, while discounts between 70 percent and 90 percent have been observed for those countries in the Western Hemisphere with severe debt problems, such as Bolivia and Peru, and for some sub-Saharan African countries.

Although debt of all major developing countries that have recently restructured their bank debt seems to have been traded at times, the quoted discounts may not be representative insofar as relatively few transactions have taken place. Also, information on the amounts of traded debt for individual countries is limited, except in cases such as Chile, where certain debt has been subsequently converted to local currency through formal arrangements requiring either Central Bank approval or the intermediation of Chilean financial institutions (see below). Thus, much of the $1 billion of debt conversions in Chile between June 1985 and September 1986, equivalent to about 7 percent of total bank debt, involved debt bought at a discount. Discounts on this debt appear to have been on the order of 30 percent.

To facilitate the trading of bank claims on certain developing countries that have restructured their bank debt, some banks have studied the possibility of securitizing trade and interbank lines that have been effectively frozen as part of financing agreements. A few Mexican banks have issued bonds to replace their interbank lines; some Brazilian banks were also in the process of arranging a note issuance facility for this purpose.

Debt-to-Equity Conversions

Country Experience

In recent years, countries have adopted various arrangements to allow the conversion of loans into domestic currency to meet local currency obligations and make local investments. The most fully developed scheme is that of Chile, where the conversion of selected categories of external debt into equity is permitted under Central Bank regulations whether the holder of the loan is the original creditor or whether the loan claim has been purchased by its present holder, such purchases being mainly at a discount. In addition to the possible use of discounted claims by nonresidents for foreign direct investment, both residents and nonresidents are allowed to repatriate such external loan claims, on a limited scale, for certain other specified and restricted uses.

As regards the exchange of external debt claims for equity, the Chilean law distinguishes between the direct exchange of debt for equity in the debtor company, and other types of equity conversion. The Foreign Investment Law (Decree No. 600), in effect since 1977, allows external creditors to capitalize (i.e., convert into equity) debt claims held on Chilean companies, subject to Central Bank approval, which is granted on a case-by-case basis. Under this law, dividends and profits may be remitted immediately and capital may be repatriated after three years. It is understood that some banks have recently made conversions of debt to equity under these provisions.

In addition, in the context of the 1985 financing package, the Central Bank has introduced a broader mechanism that allows nonresidents to convert certain external debt claims—that may have been bought at a discount—into equity in a separate entity. Under the Central Bank regulations, the holders of selected long-term public and financial system debt may convert their claims into local currency at full face value, even if such claims have been bought at a discount. Such conversions are subject to the agreement of the debtor, which has to renounce its future claims on foreign exchange for servicing the debt. The peso-denominated loans may then be sold to Chilean financial institutions or to the government-owned development bank or—in some cases—prepaid by the debtor. The local currency proceeds obtained from the sale, prepayment, or continued servicing of the loan may be used for equity investment, subject to approval from the Central Bank. The Central Bank may require as a condition of its approval of the investment that a specified percentage be made in foreign exchange.

The conditions attached to foreign direct investment to be made through this mechanism are much more stringent than those applying to foreign investment under Decree-Law No. 600. In particular, capital repatriation is not permitted for ten years following the investment. Profits and dividends cannot be remitted in the first four years, and no more than 25 percent of any profits earned during this four-year period may be remitted in a single subsequent year. However, profits and dividends earned after the fourth year may be freely remitted.

Under a separate regulation, the Central Bank allows both residents and nonresidents to repatriate discounted loan claims for certain other specified uses, provided that the foreign exchange used to buy the claims was not obtained in the official market. The local currency proceeds from the disposal of claims under this regulation must be used to repay domestic debt owed to domestic financial institutions, or to acquire goods and financial assets owned by these institutions.

Since discounted external loan claims may be converted into pesos at the official exchange rate, it is worthwhile for residents to purchase foreign exchange in the parallel market, in order to buy such claims for as long as the premium in the parallel market for foreign exchange is less than the discount on the loan claims. In order to restrict the amount of such transactions, the Central Bank sets monthly limits on the authorizations that it grants for the financial system to participate in these operations. As of September 1986, financial institutions were paying a premium to the Central Bank for the authorizations, which are allocated through an auction system.

Consistent with the Central Bank regulations, the bank debt restructuring agreements of 1985 allow for the conversion of debt both for direct investment and for the repayment of domestic debt to Chilean financial institutions or for the purchase of selected assets from such institutions. In the first nine months of operation of the two schemes in Chile, transactions under both amounted to $1,038 million. Of this, some $325 million involved direct investment conversions with remittance rights.

While provisions for the conversion of loans to equity have not been a formal part of restructuring agreements between Brazil and commercial banks, this type of transaction has been permitted under the Foreign Investment Law (Law No. 4131), in effect since the mid-1960s. Fiscal incentives to promote such conversions (tax credits amounting to 5–10 percent of the amount of capital) were introduced in December 1982, and eliminated in June 1984. Conversion of debt to equity in Brazil totaled $731 million in 1984, but declined to $530 million in 1985. These conversions may have involved the debt of a parastatal enterprise. Although there has not been a new formal scheme introduced in Brazil, market sources have stated that a large European bank recently arranged for the pooling of loan claims sold at a discount by other banks and the subsequent sale of these claims to a large corporate client which wished to expand its investments in Brazil. Conversions of debt to equity in Brazil are estimated by market sources at $600 million for 1986.

The 1985 agreements related to the restructuring of Mexico’s public sector debt to commercial banks provide for the possibility of capitalizing the debt of certain public sector obligors. Subject to written agreement between creditor and debtor, and approval by the Mexican authorities, external loan claims may be exchanged for certain qualified investments. Such investments cover equity in nonpriority, nonstrategic public sector companies, whose sale is being promoted by the Mexican authorities, and equity in private sector companies. The qualified investments must not benefit from guaranteed dividends payable irrespective of earnings and profits and must not have more favorable redemption terms than the loans for which they are being exchanged. In addition, such holdings are not transferable to any public or private sector Mexican entity before January 1, 1998.

Considerable interest was expressed by banks in the debt-to-equity conversion clause in Mexico’s restructuring agreement. This clause has been applied on an ad hoc basis, with a number of separate negotiations between claim holders, potential investors, and the Mexican authorities. As in Chile, the Mexican authorities seek to ensure that investors provide additional foreign exchange inflows.

A debt-to-equity swap was completed in May 1986 with Nissan Motor Company of Japan. Nissan bought approximately $50 million of loan claims from a number of banks and negotiated the exchange of these claims for pesos, which were then invested in Nissan’s Mexican subsidiary. As provided for in the bank agreements relating to Mexico’s debt restructuring, the new shares may not be sold until 1999. Market participants have indicated in discussions with the authors that a number of other banks and brokers were attempting to arrange similar conversions.

Information has recently become available on a debt-to-equity conversion scheme for Nigeria. Nigeria restructured outstanding uninsured arrears by issuing promissory notes. Holders of such notes have the option of redeeming the notes in Nigerian naira at any time on such terms as may be mutually agreeable between the holder and the Nigerian Central Bank. If an approved long-term investment in Nigeria is made with those proceeds, the investment will be treated as having been made in foreign currency and be accorded approved status for purposes of taxation and repatriation of dividends and capital in accordance with relevant laws in Nigeria in force at the time. However, there are no indications that this option has been used to a significant extent.

The bank debt restructuring agreement for the Philippines allowed for the early repayment in pesos of public sector obligations, subject to the approval of the Central Bank. The Central Bank has allowed the conversion of debt into equity to meet local currency obligations and to make local investments. Recently the Government has also introduced a new scheme of debt-equity conversion modeled after the Chilean experience. Under this plan, loans that are acquired at a discount can be converted at face value into long-term peso investments; by mid-October 1986, a total of $18 million of conversions had been approved. The plan makes a distinction between preferential sectors where the Government wants to encourage foreign investments and other nonpreferential sectors. More favorable conditions with regard to repatriation of dividends and conversion fees will apply to investment in preferential sectors.

Other Aspects

In addition to equity conversions initiated and supported by the authorities in debtor countries, individual bank creditors have swapped loan claims into equity on a private basis. A further possible mechanism for converting debt to equity is under study in the International Finance Corporation. This would involve the use of a trust fund that would pool loan claims from a number of banks in order to purchase investments, or shares, in private companies in debtor countries. Such schemes may be particularly appropriate where countries have a developed private sector and stock market.

Banks have indicated that conversions could in the future allow countries to extinguish some of their external indebtedness and banks to eliminate unwanted loan claims. They noted, however, that there was still considerable uncertainty about the scope for possible debt-to-equity conversion. Commercial banks’ attitudes toward debt-to-equity conversion differ according to banks’ long-term interests in developing countries. In addition, they may be influenced by national banking regulations (as described in the section on “Banking Supervision,” below). Some banks indicated their interest in debt-to-equity conversion in order to develop relations with financial institutions in debtor countries, to expand existing operations and obtain new nonbank customers, or simply to diversify existing claims. Other banks noted that while they were interested in arranging for the brokering of loan claims between banks that wished to dispose of their debt and corporate clients who wished to invest in a debtor country, they did not contemplate using their own loan claims for such transactions. Concern over the valuation of remaining claims was usually cited in this context.

The success of such schemes depends crucially on available investment opportunities in debtor countries and on banks’ willingness either to dispose of loan claims at a loss or to accept a longer-term commitment to a debtor country in the form of equity participation. Typically, conversion schemes approved by debtor countries will involve restrictions on profit remittances and on the repatriation of investment capital for a period of time. In Chile, for example, such restrictions are designed to have similar impact to that of the terms of the debt restructuring.

Some debtor countries have developed mechanisms to capture part of the discount at which their claims were being sold in the secondary market and to avoid subsidizing foreign investments that would have taken place anyway. Countries have also placed ceilings on the total amount of conversions, so as to limit excessive monetary expansion through the provision of local currency counterparts to redeem loans and to avoid squeezing economic activity on account of this early repayment of debt.

Transition to Spontaneous Lending

A few countries that have experienced debt difficulties since 1982 have been able to move toward loans from a limited group of banks with longer-term business interests to secure spontaneous lending. In Côte d’Ivoire, Ecuador, and Uruguay, such lending has been a step in the direction of a return to normal market relations between debtors and creditors. Banks have noted that for such spontaneous lending to occur, banks must have confidence in the economic policies and prospects of the debtor country and that the country’s financing needs are relatively small. Some banks expressed concern that participation in these lending packages might not lessen their contribution to a future concerted lending package if such lending would again be required.

A transition to spontaneous bank financing has been possible where it has been confined to banks wishing to develop their business interests in the country concerned. In many cases, banks expressed a preference for linking their lending to particular projects of trade operations in countries where they have close business ties, rather than to advance general medium-term balance of payments financing. Banks have also noted that, in some cases, such spontaneous lending could be facilitated by cofinancing with the World Bank. In both Uruguay and Côte d’Ivoire, spontaneous lending was arranged in the context of a cofinancing with the World Bank. In the case of Uruguay, commercial banks are to provide half of a $90 million energy loan. The loan had been oversubscribed by banks and the agreement was signed in October 1986. The banks’ approval of the cofinancing was dependent on agreement on the MYRA, which was signed in early July.

In Côte d’Ivoire, regional ties of some banks, together with a World Bank cofinancing, facilitated a loan involving fewer than a dozen banks. These banks disbursed a $32 million loan in April 1986. The loan, cofinanced with the World Bank, was in support of a large highway project, which is also being supported by a regular World Bank “A” loan, by a loan from the African Development Bank, and by some bilateral official creditors.

A somewhat different arrangement has been pursued in Ecuador. In this case, Ecuador negotiated a revolving trade facility with a limited group of banks to prefinance oil exports. About 50 banks participated in a syndicated financing arrangement that will prefinance Ecuador’s foreign oil sales. The facility was oversubscribed, and the maximum amount was raised from $200 million to $220 million (the precise amount will depend on the price of oil). In addition, Ecuador plans to increase its use of existing trade lines to provide a further $100 million balance of payments financing this year.

In the context of a transition to spontaneous lending, banks mentioned innovative techniques that could facilitate re-entry to international capital markets for countries with debt-servicing difficulties. Banks pointed to the recent borrowing experience of Turkey and Hungary (although Hungary has not needed to restructure its debt), as illustrating how new techniques have been used by developing countries. For Turkey, short-term bond issues—placed mainly with banks—were a part of its return to spontaneous financing. Turkey has now entered the Eurocommercial paper market, following its use of note issuance facilities in early 1986. In June 1986, Turkey mandated Bankers Trust to arrange a $125 million loan facility to back up issues of commercial paper to be guaranteed by Fuji Bank of Japan.

Hungary recently borrowed in the Eurobond market through the issuance of $250 million in the form of 20-year floating rate notes which were “collateralized” by a zero coupon U.S. Treasury bond of 20-year maturity and a cash reserve fund that will be invested in short-term U.S. dollar securities. This collateral is intended to secure both the principal of the notes—through the zero coupon bond—and the interest payments expected to fall due after an estimated 12–15 years. In those later years the combined value of the cash reserve fund and the zero coupon bond would cover the principal due on the Eurobond. In addition, earnings on the cash reserve fund in those later years would be used to pay the interest payments on the floating rate notes. Market participants believed that the arrangement had enabled Hungary to obtain longer-term bond finance, but at a higher effective spread than on more traditional medium-term issues.

Some banks also noted that “transferable loan facilities,” which allow banks to increase the tradability of their assets, have been used by several countries in recent years, including countries that have refinanced their debt and other developing countries that have retained access to spontaneous flows. These instruments are a hybrid between a loan and a bond, and have generally been held by banks in anticipation of a future increase in countries’ creditworthiness, rather than being transferred. About $11 billion of transferable loan facilities (including both transferable loan certificates and transferable loan instruments) have been established.

Broadly, banks have indicated that they place great importance on developing the financing techniques described above so that the diversity of banks’ interests in developing country business may be recognized as debtors regain access to spontaneous finance. They cautioned, however, that a premature move away from concerted financing could prove costly for debtors and make it difficult for them to generate sufficient private finance.

Banking Supervision

Since the onset of the bank debt crisis in 1982, bank supervisors have sought to strengthen banks’ balance sheets. In strengthening balance sheets, the quality and diversification of earnings are the first defense against potential future losses. Increased provisioning against country risk exposure and, more generally, a buildup in banks’ capital have also taken place.


Provisioning practices and the role of the supervisory authorities differ across countries according to their regulatory and accounting framework and the tax treatment of loan-loss reserves. Supervisors in the Group of Ten countries and Switzerland review and compare provisioning levels on a regular basis in the Basle Committee. They have noted that considerable variations exist in national provisioning practices and levels, notwithstanding efforts to coordinate the strengthening of banks’ balance sheets and a generalized move toward higher provisioning.

In some countries, such as the Netherlands and the United States, banks are required to make specific provisions against their claims on each debtor, reflecting the risk associated with a particular type of claim on an individual sovereign borrower. Specific provisioning also allows for different categories of loans to the same country to be treated differently, according to the debt-servicing record. In both the Netherlands and the United States, for example, trade finance may be excluded from provisioning requirements where it has been regularly serviced. Under such a regime, loans may also be upgraded as a result of an improved assessment of countries’ creditworthiness and of the likelihood that the loans will be repaid.

Some other countries (for example, Canada and Japan) do not estimate directly the risk attached to claims on individual countries, but instead require banks to make “basket” provisions against a number of countries. Typically, countries are included in the basket for a number of years after a debt restructuring or moratorium. If this approach is adopted rigidly, loans to the group of countries in the basket are treated uniformly regardless of the type of loan and the current prospects of the country. If such a basket rule is in operation, there may be a particular advantage to MYRAs, as these allow for an extended period of restructuring but without necessarily prolonging the time of inclusion in the provisioning basket.

For other industrial countries, the provisioning regime falls somewhere between these two approaches. Typically, a judgment on the adequacy of an individual bank’s overall level of loan-loss reserves is made initially by the bank management and then by the external auditors. Supervisory authorities then review bank practices and may recommend, or (where provisioning is mandatory) require, increases in provisioning in individual cases, taking into account banks’ experience and the general judgment within the banking community of the riskiness of particular overseas exposure. In principle, this approach allows for trade financing to be treated more flexibly and for banks’ claims on individual debtor countries to be upgraded if commercial banks and their auditors recognize an improvement over time in the quality of such assets.

From a supervisory perspective, the advantage of basket provisioning is that it limits the need for judgments about individual countries. In discussions with the authors, banks noted, however, that mandatory basket provisioning, if implemented rigidly for a fixed number of years, could inhibit new lending—including spontaneous trade credits—to countries in the basket and might provide inappropriate signals to countries that are implementing adjustment measures. In Japan, however, there is some evidence of resumed lending to at least one country still in the provisioning basket. The disadvantages attributed to the basket approach are also possible where any provisioning practices are applied by the authorities without regard to the borrowers’ economic performance and payments record.

There are marked differences within many countries in the provisioning levels of different domestic banks, as supervisors have allowed time for individual institutions to build up their reserves. Some banks indicated that they intend to continue to build up reserves against their exposure to those countries that have experienced debt-servicing difficulties, even without pressure from supervisory authorities. In other cases, banks noted that the need to make an immediate set-aside out of current income for a proportion of any new loan is an important disincentive to lending.

For an individual bank, decisions concerning provisioning reflect a number of factors in addition to the assessment of risk involved in loan exposure. Such factors include the impact of an increase in reserves on cash flow and on the ability to distribute dividends, as well as the expected attitude of the supervisory authorities. Tax treatment has been an important factor influencing loan-loss reserves. In some countries, banks are able to claim large tax deductions on the basis of their provisioning against country risk. For profitable banks in these countries, it has been worthwhile to make substantial provisions in order to benefit from tax deductibility. For countries outside the United States, the recent sharp decline in the value of the U.S. dollar has tended to raise the value of existing provisions (which are generally made in local currency) when measured in relation to U.S. dollar-denominated cross-border loans.

Average provisioning levels against a group of countries that have experienced debt-servicing difficulties are now in excess of 20 percent or more of banks’ loan claims for some continental European countries. At the other end of the spectrum, the supervisory regime in the United States—which places emphasis on capital and general reserves rather than specific provisions—has led to a low average level of provisioning against country risk. However, the differences in provisioning levels among banks within a country—which in some cases are substantial—may perhaps be of greater significance for the robustness of financial markets than differences in national averages.

Cohesion among commercial banks may be more difficult to achieve because of disparities in loan-loss reserves. Banks that are generally better provided against their exposure to debtor countries, and thus presumably better able to withstand losses on such claims, may have less incentive to participate in new money packages than banks with smaller provisions. However, the interlinking of the banking system means that all banks have an interest in the continued strength of the financial system as a whole. Moreover, although there are some banks with high provisioning levels, authorities and bankers generally have indicated that the average level of loan-loss reserves in most national banking systems is not so high as to make banks indifferent to the success of the debt strategy. Finally, even in cases where loans have been written down, banks will benefit from improvements in the creditworthiness of debtor countries.

Capital Adequacy

There has been a general strengthening of banks’ balance sheets relative to their exposure to developing countries since 1982. All size categories of U.S. banks continued to increase their capital in 1985 and during the first half of 1986 (Table 51). As a result of this increase and a small decline in U.S. banks’ claims on developing countries, there was a further drop in the ratio of U.S. banks’ exposure to developing countries to their capital to below the level in 1977. The ratio for the nine money center banks declined to 186 percent in the first half of 1986 from 308 percent in 1982; the ratio for the next 15 largest U.S. banks decreased to 103 percent from 202 percent in 1982; and the ratio for regional banks declined to 45 percent from 82 percent in 1982. For banks outside the United States, the depreciation of the U.S. dollar during 1985–86 and a continued buildup of their capital and provisions have reduced considerably these banks’ exposure to developing countries relative to capital and provisions. Consequently, the unprovisioned exposure of banks in industrial countries to developing countries relative to capital has diminished sharply since 1982.

The ultimate safeguard against banks’ potential losses on international lending, as on any other credit risk, is capital. There is growing agreement among supervisors that solvency ratios need to recognize that exposure to some sovereign borrowers may be more risky than to others. Risk-asset ratios that require different capital backing for different assets provide regulatory authorities with such a tool. While these ratios typically treat all commercial and industrial loans in a similar manner regardless of the country of residence of the borrower, they do sometimes give a higher weight to, and thus require higher capital backing for, loans to sovereign borrowers in developing countries. This approach—termed “zoning”—provides a buffer against the transfer risk believed to be inherent in claims on this broad class of borrower.

Supervisors recognize that the boundary between groups of countries would inevitably be arbitrary and, at least for the time being, not based on a country-by-country risk assessment. They noted that risk-asset ratios involve only the broad classification of assets into a relatively limited number of simply defined categories. A more detailed measurement system could effectively amount to credit allocation. For domestic commercial loans, the designation of the same risk weighting indicated that banks were expected to make their own judgments of credit risk and lending opportunities between individual borrowers. While the impact of zoning on borrowing costs for individual developing countries is hard to gauge, some supervisors have commented that it is likely to be slight.

Risk-asset ratios are already used in some European countries. These ratios vary in their treatment of sovereign risk and in the relative capital weighting that they accord to different borrowers. The risk-asset ratio proposed by the United States and circulated for comment in early 1986 would assign a higher risk weight to loans to sovereign borrowers from developing countries than to industrial countries. The U.S. regulators proposed initially to use the Fund’s country, classification. In placing on record the basis of the Fund’s country classification system, the Fund’s Deputy Managing Director, in a letter to the U.S. federal regulatory authorities, stressed that “in particular, the Fund’s classification of industrial and developing countries was devised to facilitate economic analysis and not with a view to prudential concerns. The proposal issued by the Federal regulatory agencies notes that the country classification adopted by the U.S. supervisors for capital adequacy purposes may be modified, and might not at all times in the future coincide with the classification adopted by the Fund for economic analysis.” The Deputy Managing Director noted that the Fund “would understand this acknowledgement to mean that the judgements of prudential classification are in the final analysis based on assessments made by supervisory authorities.”

A limit on the concentration of assets to a particular borrower, type of borrower, or group of economically related borrowers is another common method of regulating capital cover for particular assets and avoiding concentration of risk. Such limits on large exposures do not generally appear to have inhibited lending to developing countries, partly because typically more than one borrower is involved in banks’ sovereign lending.

Other Issues

Banks’ interest in financing techniques such as loan sales and swaps and the conversion of debt to equity, and their attitudes to alternatives to concerted lending such as interest capitalization, can be influenced by supervisory and accounting practices.

Exchanges between creditors of existing loan claims, or conversion of loan claims into a different kind of asset, may change the valuation of these claims and, potentially, have an impact on the valuation of similar claims. Banks have indicated some concern that if a secondary market in loans became well developed, the discounts applied in such transactions might call into question the value of banks’ remaining claims on the same borrower. For this reason, few institutions with large exposures to developing countries have engaged directly in discounted disposals of loan claims, although a number have been involved in brokering sales by other banks. The banks participating in the secondary market for loan sales and swaps have tended to be banks with smaller exposures that have sought to eliminate completely their exposure to a particular country, or to concentrate their holdings in a few countries.

Supervisory treatment of interest capitalization and concessional interest rates also varies across countries, with experience so far mainly limited to domestic loans. In continental Europe, capitalized or deferred interest would not generally be accrued or—if accrued—would be provided against. Under the U.S. regulatory system, banks may be able to continue to accrue such income if loans are well secured and in due process of collection. A proposed change in the U.S. treatment of renegotiated farm and energy loans which carry below-market interest rates suggests that these loans should not be treated as “substandard” as long as the principal is deemed recoverable over the life of the loan.

The conversion of debt to equity involves further issues, including the valuation of the equity claims, if banks hold these themselves; the changes that may be generated in banks’ capital requirements by a switch from debt to equity; and the legal or supervisory restrictions on banks’ holdings of equity participation in nonbanks. Treatment of equity holdings in general varies considerably across countries, from those where there is a long tradition of banks or bank holding companies having equity in nonfinancial companies (e.g., the Federal Republic of Germany and France) to others where such equity participation is much less usual, or even prohibited under existing regulations. In a number of countries, there are restrictions on the proportion of a nonbank company that may be held by a bank.

In cases where equity holdings are allowed, the regulatory treatment for the purpose of assessing capital adequacy may vary. In some cases, it may be treated as more risky than a loan, and in other cases as less risky. In Germany, for example, no capital cover is required—at least in principle—for holdings of listed securities. In the United Kingdom, investments in subsidiaries and associated companies and trade investments are treated like a fixed asset and must be deducted from capital.

In general, most supervisory regimes allow banks to take a noncontrolling interest in a foreign nonfinancial company under at least some conditions. Such interests would typically require a higher capital backing than would a loan, unless those interests involved listed securities that may be readily sold on a stock exchange, which is not usually the case with the type of equity that may be obtained by banks in exchange for their loan claims on developing countries. The more usual form of equity participation by a lending bank in a developing country would be in a local bank or other domestic financial company.

Outflows of Private Capital

Outflows of private capital have often been associated with the concept of “capital flight.” However, capital flight is not easily defined or measured. Some observers have defined capital flight broadly, including all acquisitions of foreign assets by the private sector. Such a definition, however, measures all outflows of private capital, including a number of transactions that would not normally be termed flight capital, notably the extension of trade credits, transactions balances in overseas banks held by local firms engaged in international trade, direct foreign investment, and certain portfolio diversification. This section discusses the approaches taken in some recent studies to measuring outflows of private capital and estimating the impact of various factors on such outflows.

There have been a number of efforts by various observers to measure outflows of private capital, with a particular view to identifying capital flight. In practice, such outflows are often unrecorded in balance of payments statistics, and comprehensive information is not available on the foreign asset holdings by residents of developing countries. Estimates of outflows of private capital in developing countries have been derived in various studies from banking statistics, balance of payments data, and external debt statistics, based on three basic methodological techniques.

Outflows of private capital have been defined as the change in deposits with foreign banks by nonbank residents. During 1980–86, international bank deposits held by nonbanks in developing countries rose by almost $80 billion to $185 billion at mid-1986, amounting to 31 percent of these countries’ bank debt or 21 percent of their total debt. Depositing by nonbanks from developing countries has shown an uneven pattern during 1982–86; such depositing slowed from $17 billion in 1982 to $1 billion in 1984 before increasing to $21 billion in 1985; during the first half of 1986, nonbanks reduced their deposits by $3 billion.

These deposits include foreign currency working balances of firms engaged in trade, tourism, or transportation (such as airlines or shipping), which would not generally be considered capital flight. At the same time, this approach does not capture all forms of capital flight, for example, acquisitions of securities or real estate. The principal advantage of depositing data is that they are available on a quarterly basis and with relatively short delays.

Short-term capital outflows and net errors and omissions have been summed to obtain an estimate of total private capital outflows based on balance, of payments data. Cumulative private capital outflows from capital importing developing countries during 1980–85 totaled $135 billion, equivalent to 25 percent of their bank debt and 15 percent of their total external debt. Annual private capital outflows declined by this measure from a peak of $37 billion in 1982 to $10 billion in 1985.

This methodology incorporates a broader definition of private capital outflows, but it may not fully measure such outflows due to deficiencies in the data on balance of payments, especially for the capital account. For example, if any balance of payments receipt was overstated or payment understated, then errors and omissions would increase correspondingly, as would the defined measure of private capital flows. These errors and omissions could include discrepancies unrelated to private capital movements. On the other hand, some private capital outflows may not be captured in the balance of payments statistics at all due to overinvoicing of imports or underinvoicing of exports and to offsetting unrecorded capital flows. Balance of payments data in many countries are available only on an annual basis and with substantially longer lags than for banking statistics.

Private capital flows have also been defined as a residual by subtracting from the increase in external debt the accumulation of foreign assets by the domestic banking system and recorded current account deficits, and adding to it net nondebt-creating flows such as direct investment. Total external debt of capital importing developing countries grew by $400 billion during 1980–85, while the recorded current account deficits, less nondebt-creating flows, plus the increase in foreign exchange reserves amounted to only $250 billion during 1980–85. Based on this method, the acquisition of foreign assets by private residents of developing countries in 1980–85 may have amounted to $150 billion. This estimated acquisition of foreign assets was equivalent to 27 percent of the debt owed to commercial banks at the end of 1985 and 17 percent of total debt. According to this definition, private capital outflows declined from $50 billion in 1982 to $19 billion in 1985.

This residual approach is likely to produce an inaccurate estimate of private capital movements for several reasons. Private capital flows would be understated to the extent that there is overinvoicing of imports and underinvoicing of exports. An appreciation in the value of the U.S. dollar relative to other major currencies would decrease the U.S. dollar equivalent of debt denominated in other currencies, with no corresponding balance of payments flow, and would lead to a downward bias in the estimate of capital outflows. This bias may be partly offset by a countervailing movement in the foreign assets, owing to similar valuation effects. Because developing countries are generally net debtors, a stronger U.S. dollar would tend to lower the estimate of private capital outflows, while a weakening in the U.S. dollar would have the opposite effect.

These estimates of private capital flows are not accurate indications of the magnitude of capital flight. Any underreporting of current account deficits by developing countries would automatically increase private capital outflows. This approach also defines as flight capital certain acquisitions of foreign assets by domestic residents (e.g., extensions of trade credits and increases in working balances held in foreign exchange) that are not normally viewed as capital flight. Any transaction not explicitly subtracted from the increase in external debt is considered capital flight.

The three methodological approaches discussed above result in substantial differences in annual and cumulative data for private capital outflows from developing countries. Notwithstanding these differences, two common features emerge. First, cumulative private capital outflows for developing countries on these measures amounted to 15 to 21 percent of their total external debt. Second, these capital outflows declined between 1982–83 and 1985, albeit to different degrees depending on the methodology employed.

In addition to the approaches outlined above, one recent study attempts to distinguish between flight capital and other foreign assets held by domestic residents by excluding from the measure of flight capital foreign assets that result in interest payments recorded in the balance of payments.

Various estimates of capital flight for five developing countries for the period 1976–85 are provided in Table 52, based on several recent studies and the Fund’s international banking statistics. The sources used, and the main characteristics of the authors’ methodology, are identified in the table. While private capital flows in other developing countries were also studied by these authors, this list has been confined to those developing countries reviewed in all studies.

Comparisons of these studies are hampered by differences in definition, methodology, countries covered, and time periods reviewed. Even when these differences are adjusted for, however, the different studies show substantial variations in the magnitude and even in the direction of private capital movements.

Two of the studies cited in Table 52 (Cuddington and Dooley) have attempted empirical analysis of the factors influencing private capital flows in selected developing countries. These studies suggest that private capital movements were affected by factors that shape residents’ attitudes toward domestic financial assets. These factors included exchange rate movements, domestic inflation, interest rate differentials, and risk premiums, although the importance of these factors differed among the developing countries studied.

Cuddington10 ran separate regressions of the real exchange rate, domestic inflation, and the differential between domestic and U.S. dollar interest rates on capital flight for eight developing countries. He concluded that the most important contributor to capital flight seemed to be the extent of disequilibrium of the real exchange rate. The exchange rate had a statistically significant impact on private capital outflows in the developing countries that had substantial private capital outflows, but did not appear to have been a major factor in the remaining developing countries, which had relatively smaller amounts of private capital outflows. High inflation also was shown to encourage private capital outflows. According to Cuddington, the incentive for private capital outflows was also “greatly exacerbated in economies suffering from repressive financial policies that keep real rates of interest in the domestic economy considerably below those prevailing abroad.” He concluded that capital controls “significantly reduced” private capital outflows in some of the cases studied. He added, however, that although capital controls could reduce private capital outflows, this did not mean that such controls should be imposed. He suggested that a durable solution to private capital outflows needed to address the underlying causes of such capital movements, such as expansionary fiscal and monetary policies and exchange rate overvaluation.

Dooley also empirically tested the determinants of private capital outflows, using a group of seven developing countries over the period 1976–85. His results suggest that net private capital outflows were most likely “in circumstances in which residents perceive risks to income derived from domestic claims but nonresidents perceived relatively smaller risks on credits to the residents of the country studied.”11 Risks to residents as measured by domestic inflation and domestic financial repression are found to have a positive influence on the level of private capital outflows, while the reluctance of nonresidents to acquire claims on the country, as measured by the risk premium on such claims, is found to have the expected negative effect on net private capital outflows.

Direct Investment

During 1984–85, foreign direct investment averaged $11 billion, only somewhat less than the sharply reduced level of bank lending during those years. These investment flows were slightly below the average level recorded before the onset of widespread debt-servicing difficulties and significantly below the average level of bank lending ($80 billion) during 1979–82.

Discussions by the authors with multinational corporations and comments by banks concerning attitudes of their corporate customers have indicated that multinational corporations and banks consider direct investment an appropriate form of finance for developing countries. Also, a number of developing countries consider that there may be broad advantages in greater reliance on foreign direct investment, and thus are studying or undertaking measures to encourage such flows. The Multilateral Investment Guarantee Agency (MIGA), established under the auspices of the World Bank, and the International Finance Corporation (IFC) are expected to assist in attracting nondebt-creating flows to developing countries.

Developing Countries’ Direct Investment Policies

Foreign direct investment in developing countries has been concentrated in a few countries, predominantly the higher-income countries of Asia and Latin America. A review of recent changes in foreign direct investment regulations in some of these countries, as well as in a few selected countries where significant changes have occurred, is given below.

In Argentina, the authorities have recently liberalized investment policies in the petroleum sector, increasing the percentage of foreign equity participation. In Mexico, the authorities have indicated that over the last four years they have conducted a selective policy of promoting foreign investment, with special emphasis in areas related to non-oil exports and the transfer of technology, within the current legal framework. During the period 1983–85, more than 150 projects with 100 percent foreign-owned capital have been approved. Steps have been taken recently, and will be strengthened in the future, to streamline the administrative procedures for initiating and approving foreign investment projects, so as to increase the flow of foreign capital in appropriately selected sectors. In particular, approval of investment projects fundamentally oriented to export markets will be granted automatically if after a period of 30 days no contrary decision has been reached. Special efforts also will be made to facilitate an enlarged access to the Mexican market by small- and medium-sized foreign companies.

Five Latin American countries—Bolivia, Colombia, Ecuador, Peru, and Venezuela—are members of the Andean Pact. Decision No. 24, adopted December 31, 1970 by the Commission of the Cartagena Agreement, established rules for the common treatment of foreign direct investment, which act to discourage these flows. Some countries interpret the provisions liberally, or have taken steps to liberalize this decision. Thus, in Bolivia, foreign investment has recently been permitted in mineral smelting, transport, retail sale of refined oil products, and petrochemicals. Colombia has taken steps within the past two years to liberalize its investment regime. Ecuador has taken some measures to improve the regulatory climate, including a decision to join the International Center for Settlement of Investment Disputes (ICSID). It has also opened up its petroleum and mining sectors to foreign direct investment and has been loosening profit remittance and ownership transformation rules. In a decree issued in June 1985, Venezuela introduced a greater degree of flexibility in relevant laws. Specifically, the decree liberalized foreign investment in four sectors of the domestic economy (agroindustry, agriculture, construction, and tourism).

As regards Asian countries, the authorities in Korea have taken a number of steps aimed at broadening the scope of foreign direct investment. The Amended Foreign Capital Inducement Act (1983), which came into effect on July 1, 1984, provides that foreign investment will be allowed in all industries except those specified in a “negative” list; previously, investment was only allowed in those areas detailed in a “positive” list. In addition, the scope for foreign direct investment was expanded; foreign participation became freely permitted in 660 industrial sectors, or 66 percent of the total, compared with about 61 percent previously. In October 1985, an additional 102 industries were removed from the negative list, raising the liberalization ratio to 76 percent. Partnership with local firms is no longer required of companies with foreign investment. Since July 1984, the repatriation of foreign capital has also been freely permitted.

In Malaysia, foreign investors, in general, could own up to 70 percent of equity in a Malaysian company, provided that such investment was export oriented. In January 1986, authorization was granted for Singapore investors to retain up to 70 percent of the shares in their companies in Malaysia, even if such investment was not export oriented. For companies serving the domestic market, foreign shareholdings are required not to exceed 30 percent after 1990.

In June 1986, Indonesia almost doubled the number of sectors open to foreign direct investment. Under the Government’s new investment priority list, projects in agricultural, industrial, mining, public housing, and road construction sectors, which previously had been reserved for Indonesian companies, are now open to foreign investors. To encourage foreign investment, the Government of Thailand has waived the growth restrictions specified in the Alien Business Law for most activities until the end of 1986.

In India, foreign collaboration approval procedures were simplified in August 1985. The Government has also encouraged investment of private capital, including foreign capital, in infrastructural projects that had hitherto been reserved almost exclusively for the public sector. Since September 1985, foreign firms offering advanced technology are allowed to invest up to 25 percent in the equity capital of existing industrial units subject to certain conditions, in addition to the 40 percent foreign equity participation already allowed in new ventures.

In China, the State Council announced in 1984 that foreign direct investment would be permitted in 14 selected coastal cities. In early 1985, new plans to open four large industrial regions to foreign investment and trade were announced. It was also announced in 1985 that foreign oil companies would be allowed to participate in exploration and development of oil and gas reserves in nine provinces and one autonomous region. Also, in August 1985 the establishment of the first foreign bank branch office in the country since 1949 was approved, and in December 1985, a joint venture bank, the first with foreign capital participation, was opened.

In the last few years, important changes have occurred in the policies on foreign direct investment in Turkey and Yugoslavia. The authorities in Turkey have actively encouraged foreign direct investment since late 1983, when a decree granting foreigners permission to invest in certain activities not categorized as industrial was introduced. The free transfer of profit and capital repatriation for all approved investment was extended to such investments. In Yugoslavia, a new law on joint ventures was approved in November 1984 which significantly liberalized previous regulations. A number of important restrictions were lifted, and administrative procedures were simplified. Further, as part of the policy to stimulate foreign investment in Yugoslavia, the limit on profit remittances contained in the foreign exchange law was liberalized somewhat in December 1984.

There have not been major changes in recent years in policies on foreign direct investment in Middle Eastern and African countries, where foreign direct investment in general plays a smaller role than in other regions. On the occasion of the 1986 budget speech, the Minister of Finance of Nigeria waived the requirement which had been in effect since January 1, 1985 that dividends due to nonresidents could be reinvested in new companies only if an additional amount equal to half of the remittable amount was imported. Such amounts can now be reinvested without this import requirement.

Foreign Portfolio Investment

Portfolio investment has not yet become an important source of finance for developing countries, although its contribution has grown for some countries. In discussions with the authors, investment banks in major financial centers have pointed to an interest in portfolio investment in selected developing countries by some clients as part of a general diversification strategy. In some cases, private sector firms in developing countries have issued bonds convertible into equity. More generally, in some recent bank debt-restructuring agreements, provision has been made to permit the conversion of debt into equity (see above). The limited size of national stock markets and small turnover in these markets in most developing countries have made portfolio investments in such countries less attractive to foreign investors. The size of the equity markets, in terms of number of listings, market capitalization, and trading volume, in selected developing countries is shown in Table 14.

Table 14.Equity Markets in Selected Developing Countries, 1980 and 1985
Number of ListingsMarket CapitalizationTrading Volumes
(In billions of U.S. dollars)(In billions of U.S. dollars)
Western Hemisphere
Source: International Finance Corporation.

Foreign exchange restrictions in many developing countries and territories prohibit listings of foreign companies. The exceptions include, inter alia, Singapore, where many Malaysian stocks are listed, together with some other foreign stocks, and Hong Kong, where many international stocks are listed and actively traded. In several countries (e.g., Mexico and Nigeria), the local subsidiaries of foreign companies were among the earliest listings. International investors are of major significance only in Hong Kong, Malaysia, and Singapore, whose markets have been used for many years by internationally oriented mutual funds and other institutional investors. Foreign individual investors have also invested in a number of locally incorporated funds in these countries. Korea and India have introduced country funds to international investors in recent years.

In January 1981, the Korean Government announced a four-stage capital market liberalization program, with the intent of completing the liberalization process by the early 1990s. The first stage allowed foreigners to acquire Korean securities indirectly through investment trusts. Five open-ended investment trusts had been established by mid-1986, with a total capitalization of $140 million, and one closed-end fund, the Korea Fund, had been established. The second phase was inaugurated in November 1985, when the Government authorized qualified Korean companies to issue convertible bonds and depository receipts abroad. Although 14 companies now meet the Government’s eligibility requirements, the only offering to date has been for $30 million of convertible bonds by Samsung Electronics Company.

Investment funds have also been established in other countries. The Mexico Fund, which can invest in a broadly diversified portfolio of Mexican shares, was introduced in 1981 for an amount of $120 million. International investors can buy into this fund without the usual restrictions on portfolio investment. It is listed on the New York and London stock exchanges. In Thailand, foreign investors can, as an alternative to investment in individual firms, invest in the Bangkok Fund—a mutual investment fund. Investments are limited to listed securities or to securities of companies intending to list their shares at a later date. The size of this fund is about $10 million. Plans are under way for the launching of a separate “on shore” Thai Fund, for $30 million, to be sponsored by the IFC. An equity mutual fund for India was launched in July 1986 for about $100 million. The Indian Fund is a closed-end unit trust that enables Indian nonresidents to invest in listed shares on the stock exchanges in India. It is expected that the portfolio would initially emphasize the fertilizer, chemical, pharmaceutical, electronics, telecommunications, and computer industries.

The Role of the World Bank and the IFC

The IFC often assists developing countries in drafting or revising investment codes, laws, and regulations that govern private direct or foreign investment. Further, acting directly as an investor, the IFC promotes the flow of foreign investment to developing countries and seeks to stimulate the domestic private sector, and it provides services that help to bring domestic and foreign investors together. Its presence often serves to raise the confidence of foreign investors, and as a neutral partner it helps to structure projects so that the benefits are shared equitably among local public and private investors and foreign interests.

The IFC has also stimulated foreign portfolio investment in developing countries, for example, through the development of local markets by establishing specialized equity funds for individual countries. The Korea Fund is one example. This fund is a closed-end investment company registered with the U.S. Securities and Exchange Commission and listed on the New York Stock Exchange. It is expected that normally at least 80 percent of the fund’s assets will be invested in Korean listed stocks. The IFC acted as one of the co-lead managers of the underwriting. The capitalization of the Korea Fund has risen to $100 million. In addition to country-specific funds, earlier this year the IFC set up a $50 million mutual fund—the Emerging Markets Growth Fund—to invest in publicly listed shares in a number of developing countries. This closed-end fund, which is to be capitalized from 7 capital exporting countries, is expected to select about 25 companies initially, in 9 emerging stock markets in Asia and Latin America (Argentina, Brazil, Chile, India, Korea, Malaysia, Mexico, the Philippines, and Thailand). The IFC envisages that, over time, the fund’s capital could be increased to over $500 million, with investments in some 20–25 developing countries.

The World Bank has also taken some international initiatives on foreign investment. The establishment of the ICSID in 1965 has helped to improve the framework for direct investment by providing acceptable procedures for the settlement of disputes between foreign investors and their host countries. The membership now totals 78 countries, with 4 other signatories expected to become members soon.

As mentioned earlier, the MIGA, established under the auspices of the World Bank, is designed to improve the investment climate in developing countries through issuing guarantees for foreign investment against noncommercial risks and supplementing the activities of the Bank and IFC in promoting such investments by carrying out research, providing information, rendering technical assistance, and encouraging policy cooperation. The MIGA will finance itself from its own revenues, notably from premiums charged for its guarantees, and it will have its own share capital. With the signing of the convention by the United States on June 18, 1986, enough signatures had been collected to permit establishment of a preparatory committee for the agency. The World Bank convened a preparatory conference in September that formulated MIGA regulations and policies. Signatories of the convention totaled 50 in October 1986, of which 39 were developing countries. Signature of the convention indicates an intention to join the agency, but only ratification, which normally entails a legislative process, constitutes a binding commitment. The MIGA convention will enter into force upon ratification by 5 capital exporting and 15 capital importing countries representing at least one third of the authorized capital of $1.1 billion. In October 1986, four countries had ratified the MIGA convention.

In September 1986, the IFC announced a new arrangement—the guaranteed recovery of investment principal (GRIP)—for encouraging equity investment in developing countries. Under the GRIP, instead of making a direct equity investment, the investor deposits the funds with the IFC. The IFC makes the equity investment in its name and assumes the full risk of principal loss and is fully obligated to repay the deposit when it matures on a specified future date, say, in 10–20 years. Dividends and capital gains would be shared by the IFC and the investor in agreed proportions. On maturity of the deposit, the investor has three options: (1) cash in the deposit with the IFC, retaining the equity shares; (2) buy the shares from the IFC for the deposit plus a premium based on a pre-agreed formula; and (3) pay a pre-agreed fee to extend the arrangement for a specific period. In order for an investment project to qualify for GRIP participation, it must be commercially viable, it must be a private sector capital project with newly issued shares, and it must benefit the host developing country.

Recent Activities of Multilateral Development Banks

Together, the activities of the World Bank (International Bank for Reconstruction and Development—IBRD and International Development Association—IDA), the African Development Bank Group (AfDB), the Asian Development Bank (AsDB), and the Inter-American Development Bank (IDB) provide an important part of capital flows to developing countries. The World Bank is by far the largest of the multilateral development banks, accounting for about three fourths of the aggregate lending. Recent developments in the general lending activities of the four multilateral development banks are discussed, with particular emphasis on their efforts to catalyze private resources through cofinancing and other means.

Overview of Lending Activities

Loan disbursements of the multilateral development banks increased steadily during 1980–84, but declined somewhat in 1985 (Table 15). Their net disbursements in 1985 amounted to about $11 billion (Table 53), compared with $9 billion of commercial bank net flows, $8 billion of net bond issues, and $22 billion of net disbursements of bilateral development assistance.

Table 15.Overview of Lending Activities of Multilateral Development Banks, 1980–851(In billions of U.S. dollars)
Own lending activities
To all members
Gross disbursements8.49.711.813.114.814.7
Net disbursements6.37.69.410.211.310.6
To 15 heavily indebted countries
Gross disbursements3.
Net disbursements2.
All lenders29.911.114.712.213.516.1
Multilateral development banks’ own resources5.
Commercial lenders31.
Sources: Data provided by the African Development Bank, the Asian Development Bank, the Inter-American Development Bank, and the World Bank.

The decline in net disbursements by multilateral development banks (except the AfDB) in 1985 was concentrated in the heavily indebted middle-income countries, and especially in World Bank lending to Brazil, while disbursements to other groups of countries declined only marginally. The decline reflected partly a slowdown in the implementation of projects that had been approved in previous years. The slowdown was brought about largely by constraints on the availability of local funds of the implementing agencies, which reflected, in turn, the general economic difficulties faced by a number of borrowing countries. Among heavily indebted countries, disbursements to Chile, Colombia, Mexico, Morocco, and Nigeria continued to rise throughout 1980–85.

While disbursements are relevant in gauging the contribution to external financing in any given year, they tend to reflect poorly shifts in lending activities of the development banks in that year because of lags between commitments and disbursements. Commitments better reflect recent activity, although a period of time is also required for the preparation and negotiation of loans before commitments can be made. The timing of commitments is also influenced by the external environment, such as general investment activities, and by domestic resource availability in the borrowing country.

Commitments of the multilateral development banks in 1985 significantly exceeded the high level achieved in 1983, primarily because those of the World Bank rose strongly. The increase in commitments to the heavily indebted countries (particularly Brazil, Chile, Mexico, and Morocco) accounted for the bulk of the increase in total commitments in 1985, and the World Bank’s policy-based lending contributed about one fifth of the increase in commitments to the heavily indebted countries. The strong recovery in commitments of quick-disbursing policy-based lending of the World Bank in 1985/86 and actual disbursements in the first half of 1986 imply that development bank disbursements are likely to increase in the future, in particular for the heavily indebted countries.

The bulk of the lending of multilateral development banks continues to be geared to project assistance. Such lending generally aims at financing specific investments to create or maintain productive assets and economic and social infrastructure and is normally disbursed over a period of years in line with the implementation period of the project. In the context of determining the economic and financial feasibility of the project, multilateral development banks generally consider and assess the economic setting in which the project is to be made and the policies pursued by the government that may affect the chances of success of the project. Other traditional types of loans extended by multilateral development banks include technical assistance, financial intermediary loans, and reconstruction and emergency assistance.

Policy-Based Lending

Acceptance of the concept of policy-based lending to promote structural and sectoral reform has varied among multilateral development banks, with the World Bank engaging in such lending much more than other development banks. Over the past few years, the World Bank has increasingly extended fast-disbursing structural and sector adjustment loans in support of programs of policy reforms. In March 1986, the World Bank’s Executive Directors expressed general support for an increase in policy-based lending in the next three years to 15–20 percent of total lending, and the share of policy-based lending in total commitments rose close to the upper end of the range in fiscal year 1986.

With the aim of helping revive growth in heavily indebted countries, the World Bank is developing economic programs spanning three to five years that cover policy issues at both macroeconomic and sector levels and identify financing requirements of the borrowing countries. On the basis of these programs, the Bank envisages reaching broad agreements with individual governments on a medium-term integrated framework, within which agreement is typically expected on the size and the broad composition of the public investment program and on a program of policy actions for the following 12–18 months. The Bank staff collaborates closely with the Fund staff in developing these programs.

Although they do not engage in policy-based lending in the same form and to the same extent as the World Bank, the regional development banks in varying degrees do engage in some form of policy lending. This may take the form of nonproject assistance (as in the case of the AfDB) extended as cofinancing with the World Bank to support specific policy reforms. Or it may take the form of general purpose loans partly for balance of payments assistance (as in the AsDB). While the IDB has not engaged in structural adjustment policy-based lending thus far, in the evaluation of projects it assesses policy measures at the project and sector level that are necessary for the success of the project. At present, in the context of the IDB’s Seventh General Replenishment of Resources, discussions are taking place on what types of policy-based lending the Bank should engage in. In the AfDB, a broad consensus has been reached that nonproject lending in support of comprehensive strategies aimed at promoting growth and structural reform should be increased, although such lending should account for a limited share of total lending in the initial stage.

Lending Activities of Individual Multilateral Development Banks

World Bank

Over the past six years loan commitments of the World Bank (IBRD plus IDA) increased by 42 percent, rising from $11.5 billion in 1979/80 (fiscal year July 1-June 30) to $16.3 billion in 1985/86 (Table 54). While gross disbursements increased from $5.8 billion in 1979/80 to $11.4 billion in 1985/86, net disbursements over the same period expanded by $3 billion to $7.5 billion in 1985/86.

In 1985/86, loans and credits were 13 percent higher than in 1984/85 and comprised IBRD loan commitments of $13.2 billion and IDA credits of $3.1 billion. In addition, commitments from the Special Facility for Sub-Saharan Africa totaled $0.8 billion in 1985/86. Gross disbursements in 1985/86 increased by 3 percent, compared with the previous fiscal year, and comprised disbursements of $8.2 billion from IBRD and $3.2 billion from IDA. Net disbursements declined by 8 percent in 1985/86 compared with 1984/85. In the first half of 1986, the World Bank continued to expand commitments, particularly of fast-disbursing, policy-based lending.

The World Bank has provided both traditional project loans and policy-based structural and sector adjustment loans to the heavily indebted countries. Project loans continue to constitute the bulk of World Bank loans to these countries; in 1985/86, World Bank commitments of such loans rose by 4 percent to $3.9 billion. Policy-based lending commitments to this group of countries rose sharply to $2.1 billion in 1985/86, equivalent to 35 percent of total lending to these countries, which was a higher proportion than in previous years and was also well above the ratio for all countries as a whole. In many of these cases, the concerned member had a stand-by, extended, or monitoring arrangement with the Fund, or such arrangement was under negotiation, when the World Bank extended a structural or sector adjustment loan.

The range of World Bank lending instruments includes traditional project loans geared to financing specific investments to create and maintain productive assets and economic and social infrastructure, financial intermediary loans to provide funds for enterprises and small and medium-size farmers, technical assistance loans, sector investment and maintenance loans, and emergency reconstruction loans to support rebuilding activities after disasters; it also includes the nontraditional structural and sectoral adjustment loans.

In 1980 the World Bank introduced its structural adjustment lending program. Structural adjustment loans (SALs) are made in support of programs of economic and institutional policy reforms and carry relatively short disbursement periods (averaging about two years). Progress in implementation of program policies is monitored by the World Bank, forming the basis for release of loan tranches or the processing of further SALs. The reforms undertaken with SAL support have generally aimed to increase efficiency of the economy through changes in pricing and trade policies, in the size and structure of the public investment program, and in the extent of the government’s controls on private activity. While SALs support economy-wide reforms, sector adjustment loans support programs of sectoral policy and institutional change, including restructuring of capacity, to improve resource allocation within an acceptable macroeconomic framework. Sector adjustment loans generally provide fast-disbursing funds with an average disbursement period of two years; in some cases, they may be combined with support of sectoral investment programs, and disbursement may extend to five to six years.

The relationship between adjustment loans, structural or sector, and other lending instruments and the role of these instruments in the World Bank’s country assistance strategy, depends on the situation of the particular country. On an overall basis and especially since 1982, the balance of World Bank lending has shifted to emphasize more policy-based lending, which increased from 3.7 percent of total World Bank commitments in 1979/80 to 11 percent in 1984/85. Policy-based lending increased sharply in 1985/86 to over $3 billion, accounting for nearly 19 percent of total World Bank commitments.

In early 1986, the World Bank defined several dimensions to an expanded role of the Bank in efforts to revive growth in the heavily indebted middle-income countries. These included (1) assistance in the development, implementation, and monitoring of medium-term adjustment programs in pursuit of the objectives of member countries committed to policy reform; (2) expansion of the World Bank’s own lending in support of such programs; and (3) extension of the World Bank’s catalytic role and, consistent with its role as “preferred creditor,” help in establishing a process for coordinated mobilization of private and official support of developing countries’ efforts. In the implementation of this strategy, the World Bank has emphasized the importance of close collaboration with the Fund; such collaboration helps promote, inter alia, consistency of policy advice by the two institutions and coordination of efforts to harness financial resources for members, while avoiding cross-conditionality.

Inter-American Development Bank

Loan commitments of the IDB rose from $1.9 billion in convertible currencies in 1980 to $3.3 billion in 1984, before declining to $2.9 billion in 1985 (Table 55). Gross disbursements increased from $1.3 billion in 1980 to $2.2 billion in 1984 but declined to $2.1 billion in 1985. The decline in new commitments and disbursements in 1985 reflected primarily the economic difficulties facing a number of Latin American countries, which in some cases experienced a contraction in investment activities and reductions in the availability of funds for domestic cost financing. Amortization payments to the IDB increased from $0.3 billion in 1980 to $0.4 billion in 1985.

According to the IDB’s charter, its loans have to contain project objectives; these include multisector credit programs with a single objective, which are divided among various projects within a number of sectors (e.g., increasing export capacity of the economy, or rehabilitation of the economy following a natural disaster). During 1984/85, the IDB had a special program for industrial rehabilitation through which $1.5 billion was committed for the financing of imports of industrial inputs by efficient producers, at a time when normal commercial financing for such activities decreased substantially. The IDB’s loans entail policy conditionality at the project and/or sector level, aimed at ensuring success of the project. The IDB does not, however, engage in policy-based structural adjustment lending of the type undertaken by the World Bank. IDB’s loans are generally disbursed in the medium to long term, depending on the execution period of the project; most loans have a disbursement period averaging about four years.

African Development Bank

New lending commitments of the African Development Bank Group, which comprises the African Development Bank (AfDB), the African Development Fund, and the Nigeria Trust Fund, doubled from $571 million in 1980 to $1,154 million in 1985 (Table 56). At the same time, disbursements more than doubled from $220 million in 1980 to $531 million in 1985. Commitments and disbursements increased gradually until 1983, but decreased significantly in 1984. The decline was due in part to a revision of priorities and investment programs of member countries in the context of adjustment programs undertaken in response to external and internal imbalances. There was also some withdrawal of projects for technical reasons and of some projects whose beneficiaries were under sanction for arrears vis-à-vis the African Development Bank. In 1985, however, both loan commitments and disbursements rebounded strongly, by 31 percent and 84 percent, respectively. This was due to a large extent to increases in nonproject lending and cofinancing.

Lending has mainly been directed to investments in agriculture and basic infrastructure. As of December 31, 1985, 26 percent of accumulated loan approvals of the AfDB Group have been for projects in the agricultural sector, 24 percent in the transport sector, 11 percent in the water supply and sewerage sector, and 9 percent in the power supply sector; 10 percent was for participation in regional development banks. The AfDB Group has been placing greater emphasis on the agricultural sector and on environmental issues. The Group initiated a shift away from purely project financing in 1983. Nonproject lending has thus far been undertaken on a small scale, and occasionally by way of cofinancing with other multilateral development banks.

Asian Development Bank

Loan commitments of the AsDB rose from $1.4 billion in 1980 to $2.2 billion in 1984 before declining to $1.9 billion in 1985 (Table 57). Gross disbursements increased from $0.6 billion in 1980 to $1.0 billion in both 1984 and 1985, while net disbursements amounted to $0.8 billion in each of the last two years. The decline in the level of lending commitments in 1985 was attributable to the economic difficulties faced by a number of developing member countries, which affected both the implementation of ongoing projects and initiation of new projects.

The bulk of the AsDB’s financing has been designed to support specific investment projects. In certain cases, the AsDB has also provided program, sector, and multiproject loans. As the AsDB mainly provides financing for projects with a long period of gestation, its loans generally have a relatively long disbursement period. The covenants applying to project loans usually cover wide areas and are difficult to generalize. The primary intention is to ensure successful implementation of projects, including realization of intended benefits. The covenants are, therefore, project or sector specific. Loans of the AsDB are rarely staged in tranches. Therefore, the release of loan proceeds is mostly not conditional on compliance of covenants, although breach of covenants entitles the AsDB to suspend disbursements.

The AsDB’s charter does not allow for the provision of structural adjustment loans in support of general policy reforms, and the AsDB does not engage in the development of macroeconomic and structural policy programs. However, in order to improve the general economic performance in its member countries, the AsDB has in recent years become more deliberate in programming its assistance, and has increasingly engaged in policy dialogues, particularly on sectoral issues that affect project performance, such as the level of charges for irrigation and energy and interest charges in the banking sector. It is now a formal requirement of AsDB’s program lending that policy dialogue is undertaken on institutional and policy issues relating to the longer-term operations and management of the sector concerned; key policy issues are identified and discussed at the time of appraisal of program loans.

The AsDB provides nonproject financing in the form of program loans under which importation of essential inputs, such as fertilizers and gasoline, is financed. The AsDB has placed a greater emphasis on nonproject financing since 1983. Commitments of nonproject loans increased from 1.4 percent of total commitments in 1980 to 5.8 percent in 1984, when they amounted to $130 million, but fell to 2.0 percent of total commitments in 1985. Disbursements of nonproject loans totaled $165 million in 1985 (16.3 percent of total disbursements), compared with a total of $12 million during 1980–83 (0.4 percent of total disbursements). The AsDB has recently gained some experience with general purpose loans. In 1986, a loan of $100 million to the Philippines was arranged to provide the Government with local cost financing with which to help implement ongoing AsDB projects, while at the same time serving as a balance of payments support loan for imported goods essential to the structural adjustment of the Philippine economy.

The AsDB recently has laid increased emphasis on assistance to the private sector, reflecting the growing interest of developing member countries in stimulating private enterprise. The principal objectives of the AsDB are (1) to increase financial and technical assistance to private enterprises, both from AsDB’s own resources and through catalyzing other external resources; (2) to hold policy dialogue with developing member countries on private sector development; (3) to strengthen financial institutions and capital markets in developing member countries; and (4) to assist developing member countries in privatizing public sector enterprises where appropriate. In late 1985, the AsDB introduced loans to private enterprises, which do not require government guarantee. This new policy will be limited initially to $100 million in the period ending 1987. To strengthen the development of financial systems in developing member countries, the AsDB will make resources available to suitable institutions through lines of credit and equity investment facilities, and provide technical assistance to securities market institutions and government regulatory agencies. Furthermore, the AsDB will assist developing member countries in their efforts to privatize enterprises, including publicly owned and controlled financial institutions.


Apart from the direct lending operations, multilateral development banks play an important catalytic role in mobilizing sources of financing through various techniques of cofinancing with official and commercial sources. Lenders participating in cofinancing benefit from these banks’ input in the analysis of projects, supervision of their implementation, and, in some cases, the administration of loans until their full repayment.

External financing committed by all lenders, including the multilateral development banks, through cofinancing arrangements increased from $10 billion in 1980 to $16 billion12 in 1985. While the multilateral development banks’ own resources committed in these arrangements nearly doubled to $9 billion in that period, the contribution of commercial lenders remained broadly unchanged.

The multilateral development banks have developed various new instruments for cofinancing with commercial lenders in recent years. The new instruments incorporate different techniques, but all cofinancing instruments involve a multilateral development bank directly in a commercial bank syndicated loan, although in different manners. The four principal techniques developed by multilateral development banks for cofinancing have been direct financial participation, guarantees, contingent participation in later maturities, and sale of participations or complementary loan contracts. These techniques provide commercial banks with varying degrees of financial protection.

World Bank

The World Bank is playing an increasing catalytic role in mobilizing additional capital flows from private and official sources in support of developing countries’ adjustment efforts without formal cofinancing being involved. Such a role involves increased collaboration with both commercial banks and official creditors. This broader catalytic role of the World Bank is not reflected in the statistics on cofinancing, which refer only to transactions where formal links are involved and do not measure the total packages mobilized. On this narrower measure, the World Bank Group’s contribution to cofinanced operations totaled $5 billion in fiscal year 1986 (22 percent of the total costs of the cofinanced projects), compared with $6.5 billion in fiscal year 1985 (Table 58).

The decline in World Bank cofinancing in 1985/86 reflected the difficult economic situation facing developing countries which affected their investment activities, including initiation of new projects. The World Bank contribution was associated with other cofinanciers’ contributions of $5 billion in 1984/85 and $3.5 billion in 1985/86. Cofinancing with official development agencies remains by far the largest source, amounting to $2.5-2.6 billion in the past two years and covering 89–102 projects. This form of cofinancing provides development-oriented lending on highly concessional terms and is particularly critical for the poorest member countries that are affected by the slow growth in concessional flows and by the limits on IDA resources. For those developing countries that can borrow on market terms, the focus of effort remains on cofinancing with export credit agencies and commercial banks. The bulk of cofinancing activities with private creditors is with developing countries that have some, though still marginal, access to commercial funds.

In January 1983 the World Bank’s Executive Directors approved the introduction of new instruments for cofinancing between the World Bank and commercial banks on a pilot basis. The new instruments reflect various approaches, all of which involve the World Bank directly in a commercial syndicated loan either by taking a direct participation in it or accepting a contingent commitment in respect of a part of the loan. The four options whose use was authorized during the trial period were as follows: (1) direct financial participation: the World Bank participates in the later maturity of a commercial loan to encourage the banks to extend their own maturities and to achieve a lengthening of maturities beyond a point to which the commercial banks would commit themselves; (2) guarantee: the World Bank guarantees the later maturities of a loan made by commercial banks and thus provides an incentive for the co-lenders to fund the later maturities with short-term funds for a significantly longer period than would otherwise be the case; (3) contingent obligations after level payments: the World Bank takes a contingent participation in the final maturity of a commercial loan designed with fixed repayment installments that combine floating interest and variable principal components; and (4) sale of participations: the World Bank arranges the commercial loan but this would be sold by previous agreement to participating commercial banks, with the sale on a “nonrecourse” basis with the World Bank remaining the lender of record. In the meantime, the pilot program was completed, and the various options, with the exception of the last option described above, have been approved for regular use.

While the new cofinancing instruments are available for use in the heavily indebted countries, in the context of the World Bank’s strategy for assistance to these countries, primary emphasis is placed by the World Bank on developing and supporting growth-oriented adjustment programs. This is considered by the World Bank to be the most effective means through which the quality of the existing and the incremental exposure of external lenders can be increased in these countries. The World Bank recognizes, however, that formal cofinancing links could have a role, if necessary, to complete a financing package in support of the adjustment programs for heavily indebted countries. While normally the World Bank would expect such cofinancing to involve direct financing, it is prepared to consider the selective use of guarantees and other cofinancing instruments in heavily indebted countries on a case-by-case basis.

Inter-American Development Bank

The IDB initiated its complementary financing program in 1976 in an effort to secure additional private financing for Latin America. Under the program, resources from private commercial banks and other financial institutions are channeled to projects in Latin America. Using this financing technique, in 1985 the IDB secured a loan of $52 million in the private capital market for a project in Colombia for which the IDB provided $28 million from its own resources. This was the first time since 1983 that the IDB employed this technique. The total amount of commercial financing raised through the complementary financing program amounted to $260 million in 1980–85, while “matching” IDB financing from its own resources for the same projects amounted to $393 million.

Under the complementary financing program, the IDB signs two loan agreements with the borrower. The loan made with IDB resources is retained in the IDB’s own portfolio, while the other is sold to commercial banks at prenegotiated terms. Since the latter remains from beginning to end an IDB loan, it is subject to the same policies as regards default and rescheduling as are ordinary IDB loans. There is no legal nexus between the borrower and the participant commercial bank; neither is there any recourse by the participant bank against IDB, so long as the latter acts in good faith and exercises the same care in administering the loan as it does in relation to loans funded from its own resources. The IDB administers all aspects of the loan, ensures that the proceeds are devoted entirely to the project in question, and collects the repayments and interest from the borrower and passes them on without charge to the participants in proportion to their shares.

Since the complementary loans are IDB loans, the terms and conditions—while less favorable than those made with IDB’s capital—are generally more favorable to the borrower than those that could have been obtained by the borrower independently. Participating financial institutions benefit from IDB’s input in the analysis of the project, supervision of its implementation, and administration of the loan until its full repayment, as well as from the lower risk of default implied by the consequences such a default will entail on the borrower’s relations with the IDB.

The IDB also employs “parallel” and “cofinancing” arrangements in its financing activities. Since its inception, the IDB has engaged in parallel financing with the World Bank. During 1980–85, the IDB committed $1,257 million in parallel financing for projects for which the World Bank approved $1,337 million. The IDB’s cofinancing operations are carried out with various international economic and development institutions other than the World Bank. During 1980–85, the IDB committed $2,105 million for projects that were cofinanced with other institutions, suppliers, and commercial banks. International institutions committed $191 million for those projects, while other sources furnished $1,153 million (Table 59). In 1985, the IDB cofinanced projects in the Dominican Republic, Guyana, and Panama, with the Special Fund of the Organization of Petroleum Exporting Countries (OPEC), the European Community, and the International Fund for Agricultural Development, respectively.

African Development Bank

Cofinancing has generally accounted for a significant share of the AfDB Group lending. Projects are cofinanced with other regional or multilateral development banks and with bilateral donors. The AfDB’s commitments for cofinanced projects declined from $301 million in 1982 (39 percent of total AfDB Group commitments) to $197 million in 1984 (22 percent) (Table 60). Total commitments from all sources for the cofinanced projects (including those of cofinanciers) also declined substantially during this period, reflecting largely a decline in projects cofinanced with Arab development banks and funds. The AfDB Group’s commitments for cofinanced projects rebounded sharply in 1985 to $452 million (39 percent of total AfDB Group commitments). The increase reflected mainly larger cofinancing with the World Bank.

The AfDB has established a special cofinancing unit in order to ensure efficiency and effective use of cofinancing resources. It has also intensified aid coordination by establishing continuous contacts with other donors and by involvement in consultation activities and round tables. In 1985, the AfDB Group held six coordinating meetings with a number of financial institutions with which it has cofinanced projects; these included the Arab Bank for Economic Development in Africa, the European Investment Bank, the European Development Fund, the Caisse Centrale de Coopération Economique, the World Bank, and the OPEC Fund for International Development. The AfDB Group cofinanced eleven projects with the World Bank, which has emerged as the former’s largest cofinancier.

Asian Development Bank

The AsDB actively pursues cofinancing operations with official and commercial sources. Total commitments of participating creditors for cofinancing operations declined from $2.5 billion in 1984 to $1.8 billion in 1985; within these totals, the contributions of the AsDB fell from $1.4 billion in 1984 to $1.2 billion in 1985 (Table 61). This decline reflected the slowdown in initiation of new projects by a number of developing member countries in 1985.

The AsDB has emphasized greater participation of official sources in its lending programs. Such participation has been viewed by the AsDB as improving the terms and quality of the cofinanced resources. Cofinancing was arranged mainly with the World Bank, as well as with other multilateral agencies such as the United Nations Development Program, the OPEC Fund for International Development, and the International Fund for Agricultural Development. These multilateral agencies contributed $136 million of cofinanced resources in 1985, compared with $383 million in 1984. Bilateral donors and export credit agencies together provided over $400 million in 1985.

Cofinancing with commercial banks decreased sharply in 1985, partly reflecting the cautious attitude of a number of traditional middle-income country borrowers in respect of commercial borrowing. While the AsDB has available a variety of cofinancing techniques, most cofinancings with commercial banks have been of two kinds—parallel loans and participation financing. Parallel loans allow the AsDB and the commercial banks to administer their loans independently. Participation financing involves purchase by commercial banks of all or part of the loan entered into by the AsDB, albeit without the legal guarantee of the AsDB in the eventuality of default by the borrower.

In order to expand arrangements with commercial credit sources, the AsDB is currently promoting a somewhat different technique called the Complementary Financing Scheme. Under this procedure, the AsDB signs two loans, of which the first is held in the AsDB’s own portfolio, while the second is simultaneously sold to a group of commercial banks on prenegotiated terms. The commercial banks do not receive a formal guarantee from the AsDB, but the AsDB remains the lender of record. A default in the complementary loan purchased by commercial banks would involve a direct default on an agreement between the borrower and the AsDB. However, in the view of the AsDB, a default by the borrower on such an agreement is unlikely, because of the consequences this would entail in the borrower’s relations with the AsDB. The stronger security provided to commercial banks by complementary financing is viewed by the AsDB as important in its effort to attract external financial resources.

In interpreting these flows, it is necessary to bear in mind that the Fund’s international banking statistics series are based on a balance of payments approach to recording credit flows. Data are obtained from direct reports by a member country’s banks and “derived” reports based on the geographical positions of banks located in major banking centers. Overcounting of bank claims in certain developing countries may occur when loan claims on non-banks are transferred to banks (i.e., the central bank). Such transfers should result in an increase in interbank claims offset by a decline in claims on nonbanks. However, international banks that report their claims on banks and nonbanks may not properly reclassify their claims on nonbanks, which would result in overcounting of those claims and an overestimation of lending.

As indicated in Table 2, data published by the BIS estimate bank lending to developing countries (adjusted to the Fund’s scheme of country classification) at $14 billion in 1985 and net repayments of $10 billion during the first half of 1986. The difference between the IBS and BIS estimates is distributed over a wide range of developing countries. For a group of countries with significant bank debt, an attempt has been made in compiling the IBS data to identify reporting problems such as double counting or misclassification of interest arrears.

Bank for International Settlements and Organization for Economic Cooperation and Development, Statistics on External Indebtedness: Bank and Trade-Related Non-Bank External Claims on Individual Borrowing Countries and Territories at End-December 1985 (Basle, July 1986).

Various assumptions can be made about the currency composition of officially guaranteed claims. It is assumed in the text that the currency composition of guaranteed bank lending was identical to the currency composition of bank claims. The currency composition of export guarantees may, however, include a higher share of domestic currency claims than in total bank claims, which would reduce the guaranteed bank lending flow.

Data on borrowing through international bond markets are based on gross issues.

Data on consolidated bank claims by nationality of banks are regularly published by the United States and the United Kingdom. The Deutsche Bundesbank publishes some data on the geographical claims of domestic banks and their foreign branches and subsidiaries; consolidated information has been published once for a limited number of countries, but is not published on a regular basis. Data on the currency composition of claims, which can be used to correct the data for the movements in exchange rates in deriving lending flows, are not available for the Federal Republic of Germany, the United Kingdom, or the United States, and thus the changes in claims derived from these series have to be interpreted with caution.

Midland Bank, Annual Report 1985.

Crocker National Bank, Annual Report 1985. Other falls included $0.5 billion in Asia and $0.2 billion in the Middle East and Africa.

For the complete text of this summing-up, see Selected Decisions of the International Monetary Fund and Selected Documents, Twelfth Issue (Washington, 1986), pp. 19–22.

John T. Cuddington, Capital Flight: Issues, Estimates, and Explanations, Princeton Studies in International Finance No. 58 (Princeton, New Jersey: Princeton University Press, forthcoming, 1986).

Michael P. Dooley, “Capital Flight: A Response to Differences in Financial Risks,” Staff Papers, International Monetary Fund (Washington), Vol. 33 (December 1986).

There is a certain degree of double counting, as these amounts include commitments made by multilateral development banks for cofinancing arranged by other multilateral development banks.

    Other Resources Citing This Publication