Regional Pattern of Capital Flows

Except for countries in the Middle East, the decline in bank claims on developing countries in the first three quarters of 1988 was broadly distributed across regions, while bond financing continued to be primarily available to developing countries in Europe and Asia. Bank claims on developing countries in Africa declined by $1.5 billion since the end of 1986, mainly reflecting a reduction of $0.8 billion in claims on Nigeria.¹⁸ Except for bond issues by Algeria, African countries have not raised financing in the international bond markets in recent years. Bank claims on countries in Asia also declined, by $3.8 billion, in the first three quarters of 1988 compared with bank lending of $5.1 billion registered during the same period in 1987 and $14.7 billion for the year as a whole. These developments predominantly reflected reductions in claims on residents of Taiwan Province of China during the first three quarters of 1988, after substantial increases recorded in 1987. Observers consider that this borrowing was primarily a hedge against possible changes in foreign exchange rates. Moreover, reflecting strong balance of payments positions, Korea and Malaysia have significantly reduced their liabilities to commercial banks in recent years and have also reduced their issuance of new bonds. China continued to rely on bank financing ($4.8 billion in 1987 and $5.6 billion during the first three quarters of 1988), as well as on the bond market.

For developing countries in Europe, bank claims declined in both 1987 and the first three quarters of 1988, although this aggregate trend concealed quite diverse developments among these countries. In particular, bank claims on countries in Eastern Europe that are experiencing balance of payments problems and that do not have spontaneous access to international capital markets have fallen (Poland, Romania, and Yugoslavia); whereas claims on a number of countries that continue to have access to capital markets have been increasing (in particular Hungary and Turkey, but recently also Greece and Portugal). The latter group of countries have also stepped up the issuance of international bonds from $2.0 billion in 1987 to $4.1 billion in 1988. In the Middle East, the most significant changes in recent years were Saudi Arabia’s return to the bank market for $2.3 billion in 1987, compared with net repayments of $0.7 billion in 1986, and the renewed increase in borrowing from banks by the major Middle Eastern oil producers in 1988. Except for a minor bond issue by Israel in 1987, countries in the Middle East have not been raising funds in the international bond market. Changes in bank claims on developing countries in the Western Hemisphere have recently declined sharply, from an increase of $3.4 billion in the first three quarters of 1987 to a net reduction of $8.7 billion during the same period of 1988. This reversal was largely accounted for by Mexico and Brazil, although in general it can be attributed to an increase in debt conversions and a slowdown in disbursements under concerted lending arrangements.

Terms

Average spreads on bank lending declined very substantially between 1983 and 1986 but have remained relatively stable since then (Chart 9). The reduction was particularly pronounced for lending to countries with debt-servicing problems, both in the form of concerted new money packages and restructurings of existing debt. While spreads over LIBOR on bank lending under concerted commitments fell from 225 basis points in 1983 to 83 basis points in 1988 and spreads on restructured debt fell from 193 basis points to 83 basis points, spreads on spontaneous lending to developing countries as a group only fell from 72 basis points in 1984 to 57 points in 1988 (Table A28).

View Full Size

Terms on International Bank Lending Commitments, 1976–88

Sources: Organization for Economic Cooperation and Development, Financial Market Trends; International Monetary Fund, International Financial Statistics; and Fund staff estimates.1 New, publicized long-term international bank credit commitments.

• Download Figure
• Download figure as PowerPoint slide

The reduction was more pronounced for the three largest debtor countries (Argentina, Brazil, and Mexico) than for the remaining countries involved in restructuring and concerted lending packages, although the difference has narrowed recently. The latter development reflects renegotiation by a number of smaller creditors of previously agreed multiyear restructuring agreements (MYRAs). In addition, average maturities under restructuring agreements have lengthened from 7½ years in 1983 to about 18 years in 1988. Maturities for concerted lending arrangements have also lengthened considerably, although much less than for restructuring arrangements.

Trends in Bank Exposure

In the normal conduct of business, banks seek to maximize post-tax returns on equity over some period, perhaps guided by some dynamic considerations about their strategic position in the market. To achieve this objective, they undertake the least costly new activities, seek to improve the quality of their portfolios, and restructure their balance sheets so as to minimize the cost of raising capital. As the sense of crisis faded and balance sheets strengthened, many banks became increasingly reluctant to increase their exposure to developing countries and subsequently took steps to swap or liquidate some of their claims. By mid-1988, the share of BIS reporting banks’ claims on capital importing developing countries in total external assets had declined by about 6 percentage points since its peak in 1985, mostly because of a decline in claims on the Western Hemisphere (Table A29).

The pace at which commercial banks trimmed their claims on developing countries apparently accelerated during the first part of 1988. In particular, U.S. banks’ claims declined by $18.8 billion in the first three quarters of 1988, or at an annualized rate of 15.5 percent, the highest recorded in recent times (Table A30). Their claims on Western Hemisphere developing countries declined the most in absolute terms, although in relative terms the reduction was comparable with that of other regions. Although U.S. banks of all sizes shed their claims, smaller banks showed the most dramatic reduction in terms of their total assets. As will be discussed, this was reflected in falling prices for such claims in the secondary market. U.K. banks’ claims on all regions, except the Middle East, also declined during the first half of 1988, signaling a significant departure from a recent stable trend. The observed decline was greater for claims on African and European countries (Table A31).

Data on claims of banks in other major industrial countries are either available with some delay, very limited, or presented on an unconsolidated basis. Despite these shortcomings, the decline in claims on developing countries was also apparent for banks in France and Italy. In the case of French banks, for which data is only available up to the end of 1987, the declining trend is dominated by a rapid reduction in claims on countries in the Western Hemisphere. Based on information available up to the second quarter of 1988, Italian banks’ claims also showed a downward trend.

By contrast, German and Japanese banks’ claims on developing countries appear to have grown somewhat, although with marked regional differences. German banks’ claims, adjusted approximately for exchange rate changes, appear to have increased somewhat during the first three quarters of 1988, especially those on Europe and the Western Hemisphere (Table A32). While no information is available on Japanese banks’ claims by region, there are clear indications that they have continued providing financing to some developing countries, particularly to those in Asia. Simultaneously, Japanese banks have continued transferring some existing claims to a factoring company set up by them in the Cayman Islands and at times selling small amounts outright.

Specific Agreements

During 1988, Chile and Uruguay signed agreements with bank creditors on the amendment of earlier restructuring packages. Malawi signed a new rescheduling, Nigeria reached agreement in principle on a restructuring, and earlier agreements were finalized for three countries (Gabon, The Gambia, and Morocco). Final agreement on a package involving new money was reached by Brazil and Yugoslavia.

Chile reached a new agreement with its commercial bank advisory committee in 1988 to amend previous debt agreements, improve the terms on its MYRA and provide for greater flexibility in debt management policies. This agreement appeared to reflect banks’ response to Chile’s good payments record, the substantial adjustment that has occurred since 1982, and the improvement in debt-servicing capacity, as well as the expected absence of a new money request in the period immediately ahead. It included a reduction in spreads on most bank debt (from 100 basis points over LIBOR to 81 basis points on rescheduled debt, except for the new money packages of 1983–85, for which the reduction was from 113 to 88 basis points) that put Chile’s terms in line with those of major debtors. The fee paid by banks on the World Bank guarantee under the 1985 cofinancing was reduced by ¼ of 1 percent to facilitate the reduction in spreads. Under the agreement, Chile committed itself not to ask for new money on a concerted basis before the end of 1989. Should it do so, spreads and guarantee fees would revert to their previous level.

Nigeria reached an agreement in principle in September 1988 on a new rescheduling that covered medium-term bank loans (including those covered in an earlier rescheduling of November 1987), trade credits covered by the November 1987 refinancing agreement, and interest arrears on letters of credit. The latter were to be repaid monthly over three years, while the medium-term loans were restructured over 20 years and the trade credits over 15 years. The interest rate spread on the trade credit was 81 basis points over LIBOR, while that on the medium-term debt was 88 basis points. The agreement in principle included some debt conversion options, including exit bonds with 25 years’ maturity and 15 years’ grace and a fixed 6 percent interest rate.

In March 1988, Uruguay reached agreement on significant changes in its July 1986 MYRA with commercial banks that restructured the bulk of outstanding bank debt (about $1.7 billion). The latest agreement changed the MYRA from a serial MYRA, with agreement to be reached on successive annual reschedulings, to a block MYRA where all maturities were rescheduled at the outset. It extended the amortization schedule of both past and future maturities and covered, in addition, maturities falling due in 1990 and 1991, the only portion of Uruguay’s commercial bank debt not previously rescheduled ($100 million). The spread was reduced to a uniform 88 basis points over LIBOR. As a result of these changes, interest obligations were reduced by about $10 million a year and the first amortization payments under the MYRA were deferred until 1991. Uruguay has not sought concerted new money from commercial banks since 1983; it obtained a $45 million nonconcerted loan under a cofinancing arrangement with the World Bank in conjunction with the original MYRA in 1986.

In November 1988, Brazil finalized an agreement in principle with its major commercial bank creditors on a package involving new finance of $5.2 billion, a reduction in spreads to 81 basis points over LIBOR, comparable with spreads agreed with other major debtors; the refinancing of $61 billion of commercial bank debt falling due until 1993, with a grace period of 7 years and a maturity of 19 years and a retiming of interest payments on public sector debt from a quarterly to a six monthly basis. The new financing is comprised of a parallel financing facility with the World Bank of $3.31 billion, cofinancing with the World Bank of up to $0.62 billion, $0.68 billion in new money bonds, and up to $0.6 billion in a medium-term trade deposit facility. An early participation fee of ⅜ of 1 percent was paid to banks confirming participation on or before August 5, 1988 and a fee of ⅛ of 1 percent if confirmation was received after that date but before September 1, 1988. In essence, the new financing is of a traditional medium-term nature, although the form of the new money facilities—including the links to World Bank loans and the new money bonds—is more differentiated than previously.

The agreement also provides for the use of some new financing instruments—including an expanded debt-equity conversion scheme that favors banks that participate in the new concerted lending, debt-to-debt refinancing operations, and exit bonds. Exit bonds (Brazil Investment Bonds) were issued, to a maximum of $15 million a bank, for a total of $1.1 billion, with a fixed interest rate of 6 percent, 26 years’ maturity, and 11 years’ grace. These bonds are eligible for debt-equity conversion and can be exchanged at par for indexed cruzado-denominated Brazilian Treasury obligations. Another new element in the Brazil agreement was a shift in the base date determining banks’ contribution to the new money facilities. Erosion of the original 1982–83 base date exposure has in general added to the difficulty in obtaining full participation in new financing packages. In this case, the exposure date was moved to March 31, 1987, in effect legitimizing the exit of banks that had reduced exposure before that date. All debt converted into equity with the approval of the Central Bank has been excluded from the new money base, as is becoming a general practice. Banks that have contributed to concerted new money packages and trade facilities will not be additionally penalized as the base exposure is defined to exclude any such lending since January 1, 1983.

A major issue in the new financing package was the nature and timing of the linkage to the Fund and the World Bank. The communiqué on the agreement in principle reached between Brazil and the Bank Advisory Committee noted the expectation that a Fund-supported program would be in place in late July 1988, and that bank disbursements would be parallel to Brazil’s Fund-supported program. The arrangement was approved in principle by the Executive Board of the Fund in July 1988 and entered into effect in August, after a “critical mass” of participating banks had agreed to the financing package. The bank agreement does not contain direct linkage of each disbursement to each drawing under a Fund arrangement as has characterized most concerted bank lending. The first tranche of new money ($4 billion) was made contingent on World Bank approval and disbursement of selected loans and the termination of the interest moratorium to banks. The signing of a Paris Club Agreed Minute and further World Bank disbursements were included as conditions for disbursement of the second tranche of new money, as was the requirement of a status report from the Managing Director to the Bank Advisory Committee confirming that Brazil is proceeding with its economic program and is making progress toward achieving the goals established in the program. The third and final tranche was made conditional upon approval and disbursements under selected World Bank loans and progress being made on the negotiation of bilateral agreements under the Paris Club. In addition, the bank disbursement was made conditional on Brazil having made the second purchase under the stand-by arrangement and either having made, or being authorized to make, the third purchase. A majority of 85 percent of bank creditors may waive this condition.

Ecuador was another case where new money ($350 million) under a recent financing agreement was essentially to refinance interest arrears. Since a critical mass of bank commitments was obtained in January 1988, however, commercial banks made little progress in finalizing the new money package. Interest payments were later suspended, and the authorities began seeking additional bank financing to regularize arrears and meet requirements for 1989. Difficulties in obtaining financing also arose in the case of Côte d’Ivoire, which was seeking a concerted loan to refinance interest arrears. In this case, it was intended to tackle the expected refusal of some banks to subscribe to a new money loan ($150 million) with a new legal approach aimed at overcoming the “free rider” problem, through essentially lifting out some banks from the original agreements; interest payments were again suspended in mid-1988, however, and the process of obtaining agreement to the financing package stalled.

In another case of interest arrears, Costa Rica has been involved since mid-1987 in negotiations with commercial banks on a financing package that would include a rescheduling of principal, as well as financing of interest arrears. The authorities were seeking financing features that would permit a reduction of the debt-servicing burden, for example, through a debt buy-back. Interest payments to commercial banks were suspended in early 1988, except for some partial payments at the end of the year, but a tentative agreement was reached with banks in December 1988 on certain details of a debt buy-back that would be part of a two-stage process of debt reduction.

In the case of Yugoslavia, an agreement was reached in September 1988 with commercial banks on a new concerted loan without a prior accumulation of interest arrears; the agreement, however, does not include a traditional medium-term loan but rather the provision of new finance through a short-term trade facility. The new $300 million revolving facility is based on prorata contributions from creditor banks and funds will be freely available to the National Bank of Yugoslavia until June 30, 1990. Thereafter, deposits in the facility will be available to creditors as collateral against short-term trade credits of up to 180 days. The facility has a five-year maturity, with repayment generally due by the end of 1993. The spread of ⅞ of 1 percent over LIBOR is higher than that for Brazil and Mexico, but the same as the spread on Argentina’s 1987 new medium-term loan. The financing package included an early participation fee of ¼ of 1 percent for amounts committed before the end of June 1988. It provides for a variety of debt exchange options, including debt-equity swaps, debt-bond swaps, and debt-commodity swaps. In addition, the package includes long-term exit bonds with a fixed interest rate of 5¾ percent and maturity and grace periods of 20 years and 10 years, respectively.

Morocco experienced difficulty in obtaining all signatures for its rescheduling agreement. In this case, a new rescheduling agreement was drawn up between Morocco and participating banks, effective in January 1988, without the participation of one recalcitrant bank. The nonparticipating bank will, nonetheless, receive payments as if it had participated in the new agreement, which itself is conditional upon the nonparticipating bank receiving no more favorable treatment than the participating banks. Negotiations later began on a new financing package for 1989.

In addition to the concerted packages described above, financing agreements for two countries were reached in 1988 that included elements of spontaneous financing. A $1 billion quasi-spontaneous bank loan for Colombia was finalized in early January 1988, with a first disbursement of $685 million in May 1988. Although organized by an advisory committee of banks, the loan was not based on proportional participation by all creditor banks. This loan was preceded by a concerted loan for $1 billion in 1985. Also in 1988, Venezuela raised about $250 million from commercial banks through new bonds, including a $100 million five-year fixed rate issue of Eurobonds, yielding 3½ percent above the U.S. Treasury note rate. In December 1988, the authorities issued $500 million in new bonds, with a spread of 1.25 percent over LIBOR, and exchanged them at par for $400 million of restructured debt and $100 million of new money. In January 1989 commercial banks supported Venezuela’s request to defer principal payments on public debt, pending negotiation of a new comprehensive financing package.

Secondary Market

The secondary market for developing country debt that emerged since 1982 expanded notably in 1987–88, at the same time as the prices for most countries’ debt moved considerably. Nevertheless, this market remains rather illiquid, with high turnover reflecting the difficulties associated with matching supply and demand in the heterogeneous market, rather than high liquidity. Frequently, a chain of transactions is required in order to match final users of the loan claims with the type of claim that they require and in the amount that they need. Although market participants have made progress in designing standardized documents during the period under review, each transaction was still very time consuming and required complex documentation. Market activity during the period was concentrated in somewhat small transactions for only a few major debtor countries. Larger transactions, or even small transactions for less actively traded paper, were difficult to arrange and took several months. Initially, secondary market transactions were mainly engaged in by banks that wished to rearrange their portfolios. Banks in industrial countries used the market to swap out of countries where they felt less prepared to manage the risks and to increase exposure in countries where they had more business interests. Banks in debtor countries in Latin America also used this market to reduce cross-border exposure by swapping loans to other countries for loans to borrowers in their home country.

The cash segment of the market has grown as debt conversion and debt settlement arrangements promoted or sanctioned by debtor countries have become more important. A number of debtors have established debt conversion schemes under which debt claims that may have been purchased at a discount in the secondary market could be exchanged for equity investments or for other debt instruments. Cash demand has arisen from these schemes, as nonresident corporations or investors that may wish to finance their direct investment through debt-equity swaps must acquire loans in order to carry out debt conversion transactions. Cash sales enable banks to exit, albeit at a loss, from involvement in a particular country. Private sector nonbank residents of debtor countries have also employed debt conversion schemes as a profitable means to repatriate capital, and subsidiaries of banks from debtor countries are reported to have made purchases in the market. It also appears, according to market sources, that some private and public sector enterprises have carried out transactions to cancel or retire debts at less than face value. This may happen through debt settlements arranged by borrowers that use foreign exchange to buy back debt at a discount, either directly or indirectly through loan swaps.

Many observers wonder whether the secondary market for bank claims may also in time broaden so as to attract investors from outside the existing pool of private creditors of developing countries. So far, there has been virtually no private nonbank portfolio demand for discounted sovereign debt. Potential institutional investors may, in some cases, face legal or policy obstacles to making such investments. In addition, the difficulty of carrying out technical credit and financial analysis for sovereign borrowers has probably inhibited outside interest in this market.

The total volume of claims sold in the secondary market is difficult to gauge. As noted earlier, the lengthy chain of transactions means that estimates of gross volume considerably overstate, probably by several factors, debt sales by original holders and the extent to which such claims have been retired by the debtor. Nonetheless, the trend in the volume of gross transactions in the secondary market may give a guide to changes in underlying final sales. Estimated gross transactions climbed from less than $5 billion a year in 1985–86 to a range of $15 billion to $20 billion in 1987, and considerably more during 1988—perhaps as much as $40–50 billion, according to some market participants’ estimates.

Published discounts for bank claims reflect a mixture of actual transactions, bid or offer prices by dealers, and judgment. For small amounts of actively traded debt, it may be possible to transact at published prices. For large amounts of such debt, however, or for less actively traded debt, transactions may need to take place at a discount from published prices in order to find a buyer; conversely, moves to acquire a large amount of claims (for example, to carry out a major debt-equity swap) could push prices up. Market intermediaries have indicated that the prices published by Salomon Brothers and others for six actively or regularly traded countries (Argentina, Brazil, Chile, Mexico, the Philippines, and Venezuela) were representative of the price levels at which most transactions in those claims took place. Listed prices for other countries’ bank debt, however, are less representative.

Prices in the secondary market appear to be influenced both by general factors that affect the entire market and by country-specific developments. In 1987, prices of claims on most major debtors dropped after the round of increased provisions against secondary market loan claims by banks from the United States, United Kingdom, and Canada, and the indications that larger banks were possibly willing to dispose of claims in the market. The average discount on claims on the 15 heavily indebted countries increased from just over 30 percent at the end of 1986 to almost 55 percent in October 1987. Prices recovered somewhat and stabilized in the first half of 1988, with discounts of a little over 50 percent on average, but there were further steep price declines for debt of most major countries in the second half of the year and the average discount rose to 60 percent by December. The discounted prices on typical bank claims of the heavily indebted countries ranged from less than 15 percent of face value (Bolivia, Ecuador, and Peru) to just under 60 percent of face value (Chile, Colombia, and Uruguay) at the end of 1988. For the three largest debtors, prices have diverged somewhat, with Mexican and Brazilian claims recently trading at about 35–40 percent of face value and claims on Argentina trading at about 20 percent.

Secondary market prices and the implicit yields on traded bank claims do not necessarily reflect considered assessments of debtors’ creditworthiness; short-term supply and demand conditions, driven by portfolio requirements and equity-swap opportunities, appear to be the major market determinants. Reductions in the volume transacted under official debt conversion schemes, such as the suspension of the Mexican scheme in late 1987, have had an immediate negative impact on prices, while Bolivia’s direct buy-back scheme had an immediate positive effect on prices. There was a major break in the market in the second quarter of 1987, following the sharp increase in banks’ provisioning levels. In addition to technical factors (notably debt conversion schemes), other factors, such as the payment or nonpayment of interest, the level of debt in relation to exports, and whether or not a debtor has rescheduled, had a significant effect on prices.

For potential investors in bank claims of developing countries, and for debtor countries themselves when evaluating debt buy-backs and exchanges, secondary market prices may give an indication of yields that can be compared with alternative uses of funds. It is generally assumed that, since sovereign loans are usually rescheduled when the borrower is unwilling or unable to make amortization payments, perpetual yields may be a better measure of perceived credit quality than yields to contractual maturity. Perpetual yields were about 25 percent each for Brazil and Mexico and about twice that for Argentina at the end of 1988. If the secondary market were well developed, the extent to which the yield on a sovereign loan exceeded the yield on an equivalent risk-free security would reflect the market’s perception of the credit risk associated with the sovereign loan. Few market observers view this yet to be the case. Especially since mid-1987, yields on the loan claims on developing countries have climbed substantially. The lack of liquidity in the secondary market, however, means that such spreads should be treated with caution; they do not merely reflect credit risk but also a variety of other factors, including the liquidity preference of claim holders.

Banks’ willingness to supply claims to the secondary market is influenced by several factors, including the discount on the claim, provisioning levels, tax treatment, accounting practices, and capital adequacy considerations. The importance of these factors has varied over time and according to banks’ nationality, exposure, and profitability. Before the mid-1987 round of substantial bank provisioning, the chief source of supply for the market was smaller banks in continental Europe and in the Middle East that had no continuing business interests in the debtor country and wished to reduce exposure even at a loss. More recently, U.S. banks have accelerated their disposal of loan claims on heavily indebted countries. Citicorp, for example, announced that since mid-1987 it had reduced loans to developing countries by $2.4 billion. This was reportedly accomplished at an aggregate discount of 33–35 percent. During the past six years, the claims on developing countries of U.S. banks as a group have declined and capital bases have increased in both absolute and relative terms (Table 7). Total external claims of U.S. banks on developing countries peaked at $150 billion in 1983 and declined to $98.8 billion in 1988, even though the total assets of those banks continued to rise. In relative terms, the exposure of U.S. banks to developing countries peaked in 1982 at 11.7 percent of total assets and fell to 5.9 percent by 1988. During the same period the total capital of the banks increased steadily. The result for the U.S. banking system as a whole was a dramatic change in the ratio of capital to developing country exposure (from 47.8 percent in 1982 to 137.5 percent in 1988).

Table 7.

Assets and Capital of U.S. Banks, 1978–88

Sources: Federal Financial Institutions Examination Council, Country Exposure Lending Survey; and International Monetary Fund, International Financial Statistics.

¹The data presented in this table are based on exposure; that is, they are adjusted for guarantees and other risk transfers. The data for 1988 are as at the end of September, the latest available at time of publication.

Table 7.

Assets and Capital of U.S. Banks, 1978–88

		1978	1979	1980	1981	1982	1983	1984	1985	1986	1987	1988
		(In billions of U.S. dollars)
External claims on developing countries1		81.5	91.9	110.9	132.6	147.7	150.0	145.6	133.1	121.8	111.5	98.8
Total assets		823.6	941.3	1,066.3	1,164.5	1,261.0	1,336.0	1,413.0	1,529.0	1,613.0	1,633.0	1,671.0
Capital		45.5	49.7	56.9	62.7	70.6	79.3	92.2	105.4	116.1	129.2	135.9
		(In percent)
Memorandum items
	Capital to total assets	5.5	5.3	5.3	5.4	5.6	5.9	6.5	6.9	7.2	7.9	8.1
	External claims on developing countries to total assets	9.9	9.8	10.4	11.4	11.7	11.2	10.3	8.7	7.6	6.8	5.9
	Capital to external claims on developing countries	55.8	54.1	51.3	47.3	47.8	52.9	63.3	79.2	95.3	115.9	137.6

Sources: Federal Financial Institutions Examination Council, Country Exposure Lending Survey; and International Monetary Fund, International Financial Statistics.

¹The data presented in this table are based on exposure; that is, they are adjusted for guarantees and other risk transfers. The data for 1988 are as at the end of September, the latest available at time of publication.

View Table

Table 7.

Assets and Capital of U.S. Banks, 1978–88

		1978	1979	1980	1981	1982	1983	1984	1985	1986	1987	1988
		(In billions of U.S. dollars)
External claims on developing countries1		81.5	91.9	110.9	132.6	147.7	150.0	145.6	133.1	121.8	111.5	98.8
Total assets		823.6	941.3	1,066.3	1,164.5	1,261.0	1,336.0	1,413.0	1,529.0	1,613.0	1,633.0	1,671.0
Capital		45.5	49.7	56.9	62.7	70.6	79.3	92.2	105.4	116.1	129.2	135.9
		(In percent)
Memorandum items
	Capital to total assets	5.5	5.3	5.3	5.4	5.6	5.9	6.5	6.9	7.2	7.9	8.1
	External claims on developing countries to total assets	9.9	9.8	10.4	11.4	11.7	11.2	10.3	8.7	7.6	6.8	5.9
	Capital to external claims on developing countries	55.8	54.1	51.3	47.3	47.8	52.9	63.3	79.2	95.3	115.9	137.6

Sources: Federal Financial Institutions Examination Council, Country Exposure Lending Survey; and International Monetary Fund, International Financial Statistics.

¹The data presented in this table are based on exposure; that is, they are adjusted for guarantees and other risk transfers. The data for 1988 are as at the end of September, the latest available at time of publication.

Debt Conversion Schemes

Debt conversion schemes have been established in several debtor countries (e.g., Argentina, Brazil, Chile, Costa Rica, Mexico, the Philippines, Uruguay, Venezuela, and Yugoslavia), in many cases in the context of bank restructuring agreements (Table 8). During the period 1984–88, an estimated $16.5 billion in bank debt was converted under officially recognized schemes. This amount represented less than 5 percent of outstanding bank debt of those countries with active conversion schemes, although in some cases (notably Chile), a substantially larger share of bank debt has been retired. The price at which debt is converted varies under these schemes. In some cases, it may depend upon the sectoral distribution of the resulting investment, the type of domestic claim exchanged for the external claim, or the nature of the external claim being exchanged.

Table 8.

Debt Conversions, 1984–88¹

(In millions of U.S. dollars)

Sources: Central Bank of Argentina; Central Bank of Brazil; Central Bank of Chile; Mexico, Ministry of Finance; Central Bank of Philippines; Bank of Jamaica; and Fund staff estimates.

¹Face value of debt converted under officially recognized operational schemes.

²Figures do not include the exchange of $64.4 million of Bolivian debt to banks for $7.1 million in 25-year collateralized bonds as part of a buy-back scheme.

³The annual breakdown of conversions is estimated.

⁴January–October 1988.

⁵Excludes direct buy-back of $299 million in November 1988.

⁶January–June 1988.

⁷Does not include the exchange of $3,671 million of medium-term bank debt for $2,556 million in 20-year collateralized new Mexican bonds. Also does not include prepayments at a discount of private sector debt, estimated at $6–8 billion since August 1987 signing of agreement to restructure FICORCA debt.

Table 8.

Debt Conversions, 1984–88¹

(In millions of U.S. dollars)

		1984	1985	1986	1987	19882
Argentina		313	4693	—	—	1,362
Brazil		731	537	176	290	1,9644
Chile		—	324	987	1,983	2,6065
Costa Rica		—	—	7	104	29
Ecuador		—	—	—	127	258
Honduras		—	—	—	6	46
Jamaica		—	—	—	—	4
Mexico		—	—	413	1,741	1,3567
Philippines		—	—	15	266	457
Uruguay		—	—	—	—	128
Venezuela		—	—	—	45	49
	Total	762	1,330	1,598	4,562	8,217

Sources: Central Bank of Argentina; Central Bank of Brazil; Central Bank of Chile; Mexico, Ministry of Finance; Central Bank of Philippines; Bank of Jamaica; and Fund staff estimates.

¹Face value of debt converted under officially recognized operational schemes.

²Figures do not include the exchange of $64.4 million of Bolivian debt to banks for $7.1 million in 25-year collateralized bonds as part of a buy-back scheme.

³The annual breakdown of conversions is estimated.

⁴January–October 1988.

⁵Excludes direct buy-back of $299 million in November 1988.

⁶January–June 1988.

⁷Does not include the exchange of $3,671 million of medium-term bank debt for $2,556 million in 20-year collateralized new Mexican bonds. Also does not include prepayments at a discount of private sector debt, estimated at $6–8 billion since August 1987 signing of agreement to restructure FICORCA debt.

View Table

Table 8.

Debt Conversions, 1984–88¹

(In millions of U.S. dollars)

		1984	1985	1986	1987	19882
Argentina		313	4693	—	—	1,362
Brazil		731	537	176	290	1,9644
Chile		—	324	987	1,983	2,6065
Costa Rica		—	—	7	104	29
Ecuador		—	—	—	127	258
Honduras		—	—	—	6	46
Jamaica		—	—	—	—	4
Mexico		—	—	413	1,741	1,3567
Philippines		—	—	15	266	457
Uruguay		—	—	—	—	128
Venezuela		—	—	—	45	49
	Total	762	1,330	1,598	4,562	8,217

Sources: Central Bank of Argentina; Central Bank of Brazil; Central Bank of Chile; Mexico, Ministry of Finance; Central Bank of Philippines; Bank of Jamaica; and Fund staff estimates.

¹Face value of debt converted under officially recognized operational schemes.

²Figures do not include the exchange of $64.4 million of Bolivian debt to banks for $7.1 million in 25-year collateralized bonds as part of a buy-back scheme.

³The annual breakdown of conversions is estimated.

⁴January–October 1988.

⁵Excludes direct buy-back of $299 million in November 1988.

⁶January–June 1988.

⁷Does not include the exchange of $3,671 million of medium-term bank debt for $2,556 million in 20-year collateralized new Mexican bonds. Also does not include prepayments at a discount of private sector debt, estimated at $6–8 billion since August 1987 signing of agreement to restructure FICORCA debt.

In Chile, where the rules under which private foreign investment operates are relatively liberal, the sectoral allocation of debt-equity conversion is largely determined by the private investor. However, the Central Bank reviews each application and limits debt conversion to 10 percent of the total equity of the project for large natural resource-based projects to try to increase the amount of additional investment. Large-scale forestry projects have received much of the investment. In Costa Rica, export-promoting and import-substituting sectors have been given priority; this is the same in Venezuela (in addition, 11 designated priority sectors have been specified). In the Philippines, priority sectors (to which a more favorable conversion rate is applied) include export-oriented production, agriculture, health, education, middle-income housing, and the assets held by the Asset Privatization Trust (APT). In Mexico, the emphasis was on encouraging industrial investment. The Uruguay scheme does not give any particular sector priority, but the Government will carry out an evaluation of the social impact of proposed projects. In Argentina, investments are required to be directed toward increasing the productive capacity of the economy; in Brazil, investment in the Northeast and Amazon regions is particularly favored.

The discount captured by the government has varied from as little as 3 percent of the face value of converted claims to as much as 50 percent to 60 percent. In Brazil, the discount in auctions in 1988 averaged 20 percent; however, new loans under the 1988 financing package may be converted at face value up to a certain limit, as described below. In Chile, the conversion of external for domestic debt has taken place through an auction system with the Government’s share of the discount varying from 15 to 30 percent of face value, in accordance with changes in the secondary market price. Debt conversions for mortgages under a temporary scheme are taking place at face value, with the benefit thus shared directly between the domestic mortgage holder, who purchases the debt at a discount, and the financial intermediary. In the case of debt-equity conversions, the exchange price is determined on a case-by-case basis. In Costa Rica, the Central Bank in exchange for debt issues bonds that may be used to acquire cash for approved investments. The Central Bank has shared substantially in the discount available in secondary markets. In Uruguay, the Government share is required to be at least 12 percent of the face value of the debt for debt-equity swaps and is determined as two fifths of the discount on the secondary market (which must be at least 30 percent) for the debt-for-debt swaps. In Mexico, the Government share was negotiated on a case-by-case basis.

In the course of 1988, Argentina instituted a number of debt conversion programs, including a public debt-equity conversion program, a private debt-equity conversion program, and a debt-rediscount cancellation program. Debt with a total face value of $1.4 billion was retired during the year under these schemes. In accordance with the 1987 new bank financing agreement, the Central Bank of Argentina held the first auction under the new public debt-equity conversion scheme in January 1988. During 1988, debt with a face value of $785 million was auctioned under this scheme at an average discount of 62 percent. An overall limit of $1.9 billion was set on the dollar equivalent of the discounted value of debt converted over the next five years. In addition, $89 million was auctioned in 1988 under a separate scheme applying to private sector debt. Under a further scheme, using debt conversions to cancel rediscounts, the Central Bank would, under certain conditions, accept public and publicly guaranteed external debt as prepayment of rediscounts.

Brazil recently revived its debt conversion scheme, with auctions of $0.3 billion of debt in January 1988; a further $0.6 billion was also converted under the old conversion scheme, which is being phased out. The November 1988 MYRA included provision for the conversion of restructured debt into equity, new money facilities, and exit bonds. In addition, loans under the parallel new money facility may be invested in equity in Brazil at face value beginning one year after November 1988; these investments are subject to a limit of $50 million a month for a three-year period, with an overall limit of $1.8 billion.

In Chile, debt conversions under two main official schemes extinguished $6 billion of bank debt during the period from mid-1985 to 1988 (more than one third of Chile’s long-term bank debt). Recently, operations under Chapter 18 (which allow conversion into debt instruments with no remittance rights and are intended to stimulate the return of flight capital) expanded rapidly and were thought by the authorities to have contributed to pressures in the parallel exchange market. The authorities have now limited these operations, other than transactions under a special temporary scheme for small mortgage holders introduced in March 1988, which allows them to participate in debt conversions and to use their profits to reduce outstanding mortgages. Mortgage-related operations led to over $100 million of conversions in 1988. The possible total stock of foreign debt to be converted under this mechanism is about $300 million.

In December 1987, Uruguay established a debt-equity swap scheme to begin in April 1988. Under this scheme, an overall limit of $55 million could be converted during 1988 and 1989. This limit was intended to contain the impact of the scheme on the management of monetary policy. An auction procedure is to be followed to determine the projects to be accepted under the scheme, and a minimum discount, to be appropriated to the Central Bank has been set at 12 percentage points of the face value of the debt. At the time that the scheme went into operation, the secondary market price for Uruguay’s debt was trading at a discount of about 40 percent. A separate scheme, established in August 1987, allows conversion of external debt to domestic debt by selected domestic private debtors, who can cancel their liabilities to the Central Bank with public external debt purchased in the secondary market. The purpose of the scheme is to allow the Central Bank to recover at least part of its nonperforming private sector loan portfolio and to share in the discount in the secondary market, while providing domestic debt relief. The authorities expect that such external debt purchases would be financed by capital repatriation and not by domestic credit that might endanger the monetary program. Transactions under this scheme amounted to $128 million in 1988.

Debt Buy-Backs and Exchanges

Other techniques that allow debtor countries to benefit from discounts on their debt in secondary markets are cash buy-backs, or the exchange of existing debt for new debt with different terms; however, direct buy-backs and many debt exchanges are usually not technically possible without a waiver of the prepayment sharing provisions in loan agreements. Thus far, creditor banks have been reluctant to agree to buy-backs or debt exchanges in which existing foreign exchange resources of the debtor government would be used; banks have argued that such resources should be used for servicing existing debt. They have also been concerned about moral hazard implications and about stockholder reactions to any increases in exposure, for example, through new lending, that may be seen as providing debtors with cash resources for a buy-back that involves a loss for participating banks. For middle-income debtors in particular, banks have so far strictly limited the amounts of debt that may be exchanged or bought back. Nevertheless, banks have approved the use of this technique by governments in three recent cases (Bolivia, Chile, and Mexico). More limited schemes were adopted in two other instances (Mexico and Peru), and banks are engaged in negotiations with a number of countries on possible debt-reduction operations.

A Bolivian buy-back scheme went into operation in 1988. Almost half of outstanding principal has now been retired or converted into collateralized bonds under this scheme. In November 1987, a voluntary contribution account to administer funds donated for the buy-back was set up in the Fund, commercial banks completed signing an authorizing amendment to the 1981 financing agreement to allow a buy-back, and the Bolivian authorities issued a decree establishing the domestic legal framework for the transaction. In early January 1988, a letter from Bolivia was distributed to all creditor banks in which Bolivia offered to buy its debt back at 11 cents on the dollar. This was in line with the secondary market bid price at that time; however, the price had moved up in November 1987 from 8 cents, in anticipation of the buy-back. Bolivia also offered an exchange of debt for fully collateralized bonds at 11 cents on the dollar. The four-month period for the buy-back initially ran from November 12, 1987 to March 12, 1988; during this period, the sharing provisions under the bank legal agreements were suspended. The offer of collateralized bonds was extended, however, beyond March 12, 1988.

Under the buy-back itself, a total of $253 million of principal and interest arrears was repurchased at face value at a cost of $27.8 million. A further $15.8 million was donated to Bolivia, including $0.6 million as a result of a debt-for-nature swap under the auspices of World Conservation International, and $64.4 million was converted at a discount into bonds. Fifty-five banks participated. However, because not all of them tendered all of their claims on Bolivia, the universe of bank creditors shrank somewhat less, from an estimated 130 to some 80 to 85 banks. The buy-back appeared to have a significant impact on the secondary market price for Bolivian debt. The average price in 1986 had been 6 cents, and the price rose when discussions of a buy-back began late in 1986. After falling back in mid-1987, it rose further later in the year to an offer price of as high as 13 cents during the buy-back period and once the trust fund in the Fund had been established.

After Mexico’s public sector debt to commercial banks had been restructured, banks agreed in August 1987 that private sector debt covered by the Foreign Exchange Risk Coverage Trust Fund (FICORCA) would be rescheduled on comparable terms. During 1987, Mexican private borrowers paid $2.7 billion for the prepayment of debt to commercial banks with a face value of $3.7 billion—a discount of 27 percent of face value. Debt with a face value of a further $6–8 billion was believed to be prepaid in 1988. As a result, the total amount of private sector debt to be rescheduled may have been reduced from $9 billion to about $4 billion.

The exchange of Mexican public sector debt under which certain existing medium-term bank debt was voluntarily exchanged for newly issued collateralized 20-year Mexican bonds was announced in late December 1987. Mexico offered a 20-year marketable international bond to its creditors at a spread over LIBOR of 163 basis points, or double the spread on rescheduled Mexican bank debt. The newly issued bonds will be amortized in a single payment, that is, a bullet, but will have call provisions that permit earlier amortization at the discretion of the Mexican authorities. The principal value of the bonds was collateralized with a nonmarketable zero-coupon U.S. Treasury bond maturing in 20 years.

The Mexican exchange was completed at an auction on February 26, 1988, with 139 banks making bids covering $6.7 billion in old debt. Mexico accepted bids from 95 banks for $3.7 billion in claims. These claims were exchanged for $2.6 billion in new bonds, with a resulting drop in Mexico’s debt of $1.1 billion, much less than some early estimates of the possible impact of the exchange. Among successful tenders, Japanese banks offered the largest amount, followed by U.S. and Canadian banks. Only two or three U.S. money-center banks reportedly participated out of about 30 U.S. banks that tendered bids. The average exchange ratio of 1.43 was equivalent to a price of about 70 cents per dollar. Mexico accepted bank tenders up to a price of 75 cents on the dollar. About three fourth of the bids were concentrated in the range of 60 cents to 75 cents on the dollar and within this band, most exchange offers were between 65 cents and 71 cents on the dollar. These bid prices were in line with the market’s financial analysis. According to this analysis, Mexico’s ability to remain current on its interest payments would not be materially affected by the exchange; although the bond was enhanced by the collateralization of principal, much of its present value was in the stream of interest payments.²⁰

The 1988 agreement between Chile and its commercial bank creditors built on the Mexican experience by amending existing rescheduling agreements to allow for direct debt exchanges and buy-backs by the Government, by permitting the issuance of future collateralized debt, and by allowing the use of collateral to undertake hedging operations. The agreement allowed Chile to use up to $500 million for either cash buy-backs or the exchange of collateralized new debt for old. Not more than $2 billion of existing debt, however, was to be extinguished or converted in this way. The financing of the new transactions was to come from excess reserves, resulting from higher-than-expected copper prices in 1988, that were accumulated in a special Copper Fund set up under a World Bank Structural Adjustment Loan. In November 1988, the Central Bank used $168 million of reserves to buy back $299 million of its external debt, implying an average discount of 44 percent. The Central Bank received bids from 129 banks to sell $822 million of debt, and it decided to accept all bids offering a price of $0.575 per dollar or less. In addition to the provisions allowing for collateralized debt conversions and exchanges, the proposed amendments would allow for the limited collateralization of future bank financing, thus incorporating the idea of the subordination of old debt to new. Chile would be allowed to borrow on a collateralized basis up to $100 million in 1988, $200 million in 1989, and $200 million a year thereafter with an overall limit of $500 million at any time after 1989. Up to $200 million of such future new money may be secured by exportable assets and related receivables.

The treatment of debt buy-back and exchange operations and of the necessary waivers of some clauses in bank documents has varied. Under the terms of Mexico’s 1985 restructuring agreement, which introduced amendments to allow debt-equity operations, banks holding a simple majority of total exposure (over 50 percent) could waive clauses related to the priority rank for payments and negative pledge clauses; 50 banks, compared with a total of about 500 creditor banks that originally signed these documents, account for such a majority. Waiver of payment-sharing provisions, conversely, required approval of all banks. Before the debt exchange announced in late 1987 could take place, a waiver of these clauses was obtained. In the Chile case, the 1988 rescheduling proposals gave approval in advance to both debt exchanges and buy-backs within predefined limits, subject to approval by a reduced majority (two thirds), rather than by all of the banks as in the original loan agreements. The amendments allow the qualified majority to waive the provisions governing payments sharing, pari passu, and negative pledge clauses. For amounts larger than those specified in the amendments, approval of all banks would be required. The new agreement also included a modification to allow the payment or prepayment of external debts in local currency on a voluntary, negotiated basis. Such payments would not need to be offered on a pro rata basis and would not trigger sharing or mandatory prepayment clauses.

Risk-Management Techniques

Shifts in interest rates and exchange rates have had a serious impact on countries’ debt-servicing obligations. Over the past few years, there has been increased diversification of currency composition of bank debt of developing countries and a general shift to longer interest payment periods. However, in most rescheduling agreements, the option to switch from U.S. dollars into other currencies is at the discretion of the lender and not the borrower. The option of paying interest on a six-month, rather than three-month, basis was included in most of the recent rescheduling agreements and is at the discretion of the borrower. In one case, Chile, the debtor country has been able to secure interest rates for a year ahead by switching to a one-year interest payment period. Six-month interest rates have generally been higher than those for three months in recent years; however, the country that opts for a longer payment period has the advantage of securing its payments obligations for a period ahead, and may deploy reserves elsewhere in the meantime.

As countries have rescheduled their debt, this has often led to a bunching of interest payments around a particular date. In the case of Chile, the retiming of interest payments to a one-year basis would have led to an extreme concentration of a full year’s interest obligations at one time in the year; this bunching was partially avoided by dispersing payments dates to some extent as part of the 1987 rescheduling agreement. In other cases, such as Argentina, there is some bunching in two months of the year; however, in this case the number of different facilities and rescheduling agreements with banks means that, while interest payments in one month may be twice as large as those in surrounding months, the payments due in the two heavy months still represent only about 30 percent of annual interest payments to banks.

Market techniques to reduce exposure to interest rate or currency fluctuations are being examined by some heavily indebted developing countries as they generally do not have the option of issuing medium-term debt with fixed interest payments. Hedging instruments available in financial markets can, however, be used to provide these countries with the equivalent of such an option. They may use these instruments effectively to transform their floating rate medium-term bank debt into debt with partially fixed interest payments, at costs not higher than those for the most creditworthy borrowers. Instruments that would permit developing countries to fix interest payments by effectively lengthening the coupon period of their debt fall into two basic categories: options, such as interest rate caps, and futures. (The swap market, which would allow a similar transformation but for much longer maturities, is not yet open to countries with debt-servicing difficulties because of the credit risk involved.)

Options instruments are one sided in that they allow debtors to insure against the risk of an increase in interest rates, while still allowing them to benefit from a decline in rates. This aspect also removes the question of debtors’ credit risk; once the debtor has purchased the option, only the bank counterparty will be called upon to deliver in case of a rise in rates. The asymmetry means, however, that there is a significant initial cost in purchasing options. The cost, for example, of purchasing a 10 percent cap on rates for three years was about 1.25 percent of principal when LIBOR was 8.5 percent. This cost can be limited by buying a tailored cap that insures selectively against interest rate increases, for example, for rate increases to between 10 and 12 percent. The cost of such a tailored cap would have been reduced to 80 basis points of principal, given the same 8.5 percent LIBOR.

The market in LIBOR caps is sufficiently liquid that market participants estimate it would be possible to buy caps to cover as much as $5 billion of debt in one day without moving the price. The market is particularly deep for caps covering the next five years, but it is possible to go as far as eight years. One advantage of LIBOR caps is their simplicity: a debtor country with limited technical and managerial expertise could more easily use this market than, for example, the futures market.

Futures instruments are symmetrical in that an indebted country that hedges in this market against a rise in interest rates will be locked into a particular rate for the period of the contract. The country would “gain” if rates rise and “lose” if rates fall during that time. Of course, the loss is only relative to the position without a hedge, rather than to the position at the beginning of the period. While there is no initial cost (and the transactions costs are relatively small), the issue of credit risk is dealt with through margin calls. These eliminate incentives to default on futures contracts and are operated by the futures exchanges, which constitute the counterparty for all contracts in futures markets. An initial margin must be deposited at the outset of a contract, and daily margin calls require further deposits to match the likely losses on each contract from the day’s movement in interest rates. Indebted countries operating in this market would have to be prepared to meet such margin calls, perhaps through previously arranged lines of credit. Waivers to existing bank loan agreements would be required if such credit lines were to be collateralized.

Futures markets are extremely deep, but only for three-month Eurodollar rates, and for a limited period ahead. Nevertheless, some techniques exist for “imperfect” hedging of longer interest payment periods, and for perhaps as long as 18 months ahead. Moreover, one major debtor—Chile—has already used these markets to hedge against interest rate increases on a portion of its sovereign debt. Chile is now in the process of expanding its hedging operations; the 1988 amendments to its rescheduling agreements allow it to undertake collateralized operations in order to hedge against risk. A potential disadvantage of futures for some debtor countries is that considerable technical expertise, and tight management control, is needed for best use of these markets. Although it is possible to set simple hedging rules, a system is needed to ensure that such rules are followed by all market operators. Typically, companies and other borrowers have chosen to have their hedging operations in-house, although it would be possible to hire outside consultants. In some indebted countries, there may be some domestic expertise in commodities futures markets that could be tapped by the government.

For debtor countries, a number of issues are being raised by the different risk-management techniques. First, the absolute cost of cover must be assessed against the protection it affords over and above the existing protection inherent in 6-month or 12-month interest payment periods; second is how much liquidity may be tied up in margin requirements, or needs to be set aside for possible margin increases; third, the market beyond the near term may not at present be sufficiently deep to make hedging of large amounts of sovereign debt feasible; and finally, the scarcity of technical and management expertise in debtor countries may severely limit the usefulness of some of the currently available hedging instruments. Market developments may, in time, be able to counteract some of these difficulties.²¹

Bank Provisioning

The level of provisioning against sovereign risk claims and the tax and accounting treatment of that provisioning have been key influences on banks’ attitudes toward financing for developing countries that have restructured, and in particular, on banks’ preferences for certain financing modalities. For banks in some countries, all provisions against sovereign exposure are specific; that is, they apply against particular loans to particular countries. In others, basket provisioning is used, involving loan-loss reserves equivalent to a proportion of all claims against a group of countries. In the United States, there are both general reserves earmarked against sovereign risk and specific provisions against claims on individual countries. These different types of provisioning have different implications for banks’ balance sheets and tax treatment. In countries such as the United Kingdom and the United States, where levels of banks’ loan-loss reserves had previously been low, developments took on special significance in the past year. Increases in provisioning above mandated levels were generally initiated by banks themselves, but supervisory authorities in a number of countries reinforced this movement. In Canada, the band of mandatory provisions on a basket of countries that have experienced debt-servicing difficulties was increased in 1988 from a range of 30 percent to 40 percent of the face value of claims to a range of 35 percent to 45 percent, and four countries have been added to the basket, bringing the total debtor countries to 38. It is expected that tax deductibility will be granted covering substantially all relevant provisions.

In Japan, the maximum level of provisioning against claims on 36 countries that have experienced debt-servicing difficulties was increased in early 1988 from 5 percent to 10 percent, and was raised further at the start of 1989 to 15 percent. The general limit on the tax deductibility of provisions has, however, remained at 1 percent of the face value of rescheduled claims and of exposure increases since the base date. Swiss banks were required to have provisions equivalent to at least 35 percent of claims on some 60 to 70 developing countries by the end of 1988, compared with 30 percent at the end of 1987. The mandatory provisioning target is expected to be raised further. Most major Swiss banks had, however, already exceeded the mandatory provisioning level at the end of 1987.

Tax deductibility of loan-loss provisions against claims on developing countries became more widespread during 1987, as tax arrangements in countries with more limited deductibility were modified and banks in other countries moved to take greater advantage of existing possibilities for tax deductions. In the United Kingdom, the establishment of a more quantified framework for banks’ provisioning decisions (the Bank of England matrix) appears to have facilitated the agreement with the tax authorities. In Japan, while explicit tax concessions were not granted, banks were allowed, in the context of certain new money packages, to sell loans at a discount to an offshore company in the Cayman Islands set up jointly by major banks and to claim tax deductions on the loss incurred by this transaction.

Generally, the United States tax authorities will not allow banks a tax deduction for reserves against developing country exposure until a loss is realized, unless a specific provision is mandated by the supervisory authorities; however, banks intending to dispose of their loans at a loss, for example, through sales in the secondary market, may show a deferred tax credit in their profit-and-loss accounts when they add to loan-loss reserves. On the other hand, the deferred tax credits recorded by some banks when loan-loss reserves were increased sharply in 1987 may be disallowed unless a loss is realized soon. In contrast to these moves toward tax deductibility, it has been suggested that banks in some countries, such as Switzerland and the Federal Republic of Germany, where reserves are tax deductible may begin to encounter some resistance to extremely high provisioning levels (these approach 70 percent to 80 percent in some cases). It should therefore be noted that, in general, countries providing tax deductibility for such reserves do not also allow banks to include the reserves in their capital bases for regulatory purposes, a matter which the next section addresses.

Observers have noted that regulations and accounting treatment in some countries have inhibited banks’ use of certain menu options. It has become apparent over the recent past, however, that authorities in a number of countries are willing to show flexibility in order to encourage the pursuit of the debt strategy through the development of the new financing instruments in the menu. At the same time, banks’ reluctance to engage in general purpose medium-term lending has increased. Some banks have favored cofinancing with the World Bank as adding to the security of loans. Cofinancing loans with interest-sharing provisions (so-called B-loans), which are most attractive to banks, are not, however, generally available for countries whose debt-servicing difficulties have limited their market access. Other cofinancing loans, and more loosely linked parallel financing, have been an important menu item in packages (Argentina 1987 and Brazil 1988). As described below, guarantees, including those from the World Bank, will be particularly attractive to banks with high exposure in the context of the new Basle capital adequacy standards.

In a number of cases, trade credits are viewed more favorably by regulators than new medium-term money. In Japan, bill finance for trade purposes is excluded from provisioning. In the United Kingdom, a high proportion of trade-related debt that may be serviced more regularly could warrant lower provisioning levels within the bands indicated by the provisioning matrix. In the United States, mandated specific provisions distinguish between different types of debt, depending on the payments record and outlook, and banks have also generally excluded trade-related debt from their exposure when determining the level of general loan-loss reserves to be established in relation to developing country risk.

There has been an increasing move toward the use of securities in debt restructurings. Subscription to the original exit bonds or alternative participation instruments in the 1987 Argentina financing package was very poor. Most banks found the interest rate unattractive and preferred either to dispose of their claims entirely through outright sales or swaps or to hold on to their claims in the hope of collecting interest financed by the contribution of other banks that participated in the new money package. Exit bonds in subsequent financing packages have been priced more attractively. New money bonds have also been included in recent financing packages. Banks holding securities often need to take a decision whether to hold them until maturity, and therefore to account for them at book value, or whether to use them as investment instruments, in which case it would be appropriate to account for them at their market value. When securities are listed, there is a greater presumption that they may be traded and thus marked to market.

Future Financing

With the move toward greater provisioning, new money in the form of traditional medium-term credits has become particularly unattractive when it requires new provisioning, out of profits, against the new claims. However, since loan-loss provisions are tax deductible in most countries, with the notable exception of Japan and the United States (except in the limited cases where country-specific provisions are required), the cost to banks of setting aside these reserves is partially offset. Banks’ degree of exposure is a particularly important influence on their attitude to new money. Within the United States, for example, a smaller bank that has reserved heavily against its developing country exposure, and written off some of its claims, may be unwilling to join a new money package and take the risk that its actions will provoke interest arrears. On the other hand, the risk of interest arrears, a subsequent shift of loans to nonaccrual status and, possibly, a mandated write-down may be judged quite differently by a major bank with high exposure and relatively weak capital cover. Japanese banks increasingly have traded their willingness to participate in traditional concerted new lending for acceptance that a tax deductible loss is involved in their holdings of similar claims. Thus, Japanese participation in the 1987–88 new money packages for Mexico, Argentina, and Brazil took place after agreement with the Ministry of Finance on the treatment of sales to the Cayman Islands holding company.

The recent move by a number of debtors toward obtaining de facto financing through the accumulation of interest arrears also has regulatory implications. In countries where rescheduling by itself is sufficient to cause supervisors to mandate or require provisions (Canada, France, Japan, and Switzerland) and, in particular, in countries where provisioning applies to a basket of countries rather than depending on the particular circumstances of each country, temporary interest arrears do not have an automatic repercussion on banks’ balance sheets. In other countries, notably the United States and the United Kingdom, the existence of interest arrears beyond a certain period is a major factor in determining provisioning levels.

For a highly exposed bank in the United States with a significant proportion of profits from developing country lending and high exposure relative to capital, the use of interest arrears as a financing technique could be particularly unwelcome. First, interest arrears that persist for more than 90 days mean that interest may no longer be accrued, with an immediate impact on profits. Once a loan is placed on a nonperforming basis, it may be difficult to change its status. A further step would be the designation of these claims by the supervisory authorities as value-impaired, requiring specific provisions (allocated transfer risk reserves), which are then taken out of capital.

For an exposed bank in the United Kingdom, interest arrears would probably add to required provisioning levels, as they are one of the factors taken into account when deciding on the level of provisioning against sovereign risk. Topping up earmarked reserves would reduce both book profits and capital, as reserves are not treated as part of capital in the United Kingdom. For a Japanese bank, interest arrears do not have an immediate impact. They do not trigger provisioning, and income may be accrued for up to 180 days. In the Federal Republic of Germany, by contrast, although the rules for income accrual are very flexible, allowing interest to be accrued for 365 days, most banks behave more conservatively and are likely to respond to an accumulation of interest arrears by placing loans on a nonperforming basis, increasing provisions—unless they are already very high—and, if interest payments on the claims are resumed, using these cash receipts to apply against principal. In Switzerland, there is no fixed time after which accrual is disallowed, but once the eventual collection of interest becomes in doubt, banks must move to a cash accounting basis.

Interest capitalization on a negotiated basis has been favored by some banks, particularly in continental Europe, partly on the grounds that it is easier for bank managements to accept this than to take an active decision to increase exposure through new money. Some banks also believe that the free rider problem would be lessened if this financing modality were used. In many countries, capitalized interest is treated just as new money for supervisory purposes, in that it is shown as an increase in exposure that may have to be provided against.

Banks in some countries face considerable uncertainty about the regulatory and accounting treatment of negotiated interest concessions. In Japan, there are no special supervisory rules for negotiated interest concessions. In the United States, in the case of domestic loans, a special accounting rule (F.A.S. 15) for troubled debt restructurings allows a restructured loan to be maintained on a bank’s books at face value even if concessions are made on interest, provided that the total expected value of cash receipts on interest and principal exceeds the original contractual principal value of the loan. Initial indications are that this may be extended to concessions made in connection with the restructuring of sovereign loans, for example, in the case of exit bonds involved in the Brazilian restructuring.

A well-provisioned bank in continental Europe may favor a negotiated concession on interest that makes ultimate loan collection more likely and improves the creditworthiness and prospects for debt servicing of an individual debtor. Any reduction in profit from the decline in interest receipts may not be an obstacle to such a transaction, as the bank may already be using a part or all of the interest receipts on those claims to add to provisions; moreover such a bank may not, for tax reasons, be trying to maximize short-term profits, while it may also view the lower income as more secure. In the Federal Republic of Germany, interest concessions by themselves would not trigger a write-down of claims, although banks are likely to have already written down claims on those debtors for whom concessions might be considered. Since provisions in the Federal Republic of Germany are specific and reduce banks’ exposure, banks only need to fund the nonprovisioned proportion of their claims on troubled debtors. Interest rate concessions up to that extent would not therefore produce a running income loss.

The Expanded Menu

A number of menu options and financing instruments have been designed to complement the provision of new money, such as debt-swapping schemes, which involve banks in selling loans or converting these into equity, and, more recently, limited debt exchanges involving the securitization of loan claims. One issue raised by these instruments is that of portfolio contagion or contamination. For banks with larger exposures, disposing of a part of their claims on troubled debtors may be unattractive because of the implications for the valuation of the rest of their portfolio. Special treatment by some supervisors has, in practice, overcome this obstacle for particular new financing instruments, notably the Mexican debt exchange. In Japan, the authorities indicated that losses realized through purchasing the bond would be tax deductible, but the remaining portfolio would not be contaminated. In the United States, it was made clear, in a joint letter from the U.S. federal bank supervisors to Mexico’s legal counsel, that participating in the exchange through tendering a portion of eligible claims would not contaminate the rest of a bank’s portfolio and the tender would not have a wider impact on banks’ balance sheets unless existing loan-loss reserves were judged to be insufficient in light of the tender price. In other countries, such as the Federal Republic of Germany, Switzerland, and the United Kingdom, only those claims that were actually exchanged affected the banks’ balance sheets.

Debt sales and exchanges at a discount from face value would generally involve a loss on the exchanged claims, except for some banks in continental Europe where provisioning has exceeded the secondary market discount, and where a loan sale or exchange could involve realizing a profit. In such cases, banks could, in some countries, choose to shift existing provisions to another debtor rather than take the profit. Since existing loan-loss reserves are tax deductible, there is in general no tax advantage for such banks in realizing a loss; taxable income would obviously increase if the loss were less than the reserves. For most major banks in the United States, particularly the larger money-center banks with high exposure, the level of provisions, by contrast, remains well below what would be needed to cover losses if loans were exchanged at current secondary market prices for most debtors. Reserves that were used to offset the loss would also be removed from capital, thus worsening banks’ capital-asset ratios. Under certain circumstances, banks may, however, wish to realize a loss in order to receive a tax credit. In Japan, loan sales generally are deemed to contaminate the rest of banks’ portfolios, although with prior approval from the Ministry of Finance specific exceptions have been made to facilitate both the Mexican debt exchange and the Bolivian buyback, as these have been viewed as consistent with the objectives of the debt strategy. The Cayman Islands scheme has also explicitly used the mechanism of loan transfers to realize a tax deductible loss on the transferred claims without contaminating the rest of banks’ portfolio of similar claims.

Another issue affecting banks’ willingness to engage in debt exchanges is the regulatory treatment of the converted debt itself. In the United Kingdom, while the new Mexican bonds have, in practice, been deemed not to be subject to provisioning for supervisory purposes, they are included in the provisioning matrix as Mexican risk. The collateral provided by the zero coupon bond could, however, be used to justify lower provisioning levels when banks hold discussions with the supervisory authorities. Under the normal accounting conventions in Canada, France, the Federal Republic of Germany, Switzerland, and the United Kingdom, if debt is converted into securities the latter must be valued at market prices. Thus, banks could show a loss on the exchange of loan claims for bonds at an implicit discount, and a further loss if the bonds themselves declined in value. In the United States, banks have been given the flexibility to choose whether to disclose debt converted under the 1988 Mexican scheme as loans or as investment securities. For some banks, this flexibility may help overcome restrictions on lending limits. The initial valuation of the bonds would represent a fair, or market, value; the subsequent valuation would depend on how they were categorized—if booked as loans their value would remain unchanged; if booked as investments their value would fluctuate with the market. In Japan, banks will probably have a choice once trading begins between valuing the bonds at face value or at market value.

For U.S. banks, loan swaps had been favored over loan sales by banks with large exposures both because of tax benefits and because large-scale cash sales threatened to swamp the secondary market and depress prices. Until 1985, the accounting treatment of such swaps appeared to allow more discretion in the valuation of remaining claims. After a 1985 ruling, this discretion was narrowed; moreover, losses from sales and swaps were supposed to be taken directly out of earnings and not from loan-loss provisions, as they were viewed as the result of a business decision to engage in the market rather than an unavoidable loss for credit risk reasons. Since the 1987 provisioning changes, banks have in practice been able to use their loan-loss reserves to offset losses from secondary market operations. For banks with smaller exposures, loan sales can be more attractive since they potentially liquidate an entire portfolio, thus avoiding the possibility of contamination as well as the obligation to participate in future new money packages. In general, debt conversions through more traditional schemes have provided some banks with the opportunity to earn brokerage fees. Some larger banks have engaged in direct debt-equity swaps, although there are particular regulatory and tax obstacles in individual countries.

A number of policy changes were made in the United States in the period under review to facilitate debt-equity conversions by banks. Prior to August 1987, U.S. banks’ equity participation in nonfinancial firms was limited to 20 percent of that firm’s equity. After that, the Federal Reserve Board allowed U.S. banks under its authority to make equity investments up to 100 percent in nonfinancial companies in developing countries that had restructured their external debt since 1980, provided that the nonfinancial corporations were state-owned companies in the process of being privatized; divestment, however, was required within five years (or ten years with special approval).

In February 1988, these rules were further relaxed by allowing U.S. banks to swap debt for up to 40 percent of the shares in any private sector nonfinancial company from a heavily indebted developing country so long as the bank does not hold the largest block of shares. The bank would be permitted to hold the investment for up to two years after the end of the period during which the debtor country restricts full repatriation of the investment, but not more than 15 years altogether. In addition, the Comptroller of the Currency has taken an even more liberal view, issuing two letters in 1988 that allow debt to be exchanged for equity under the banks’ statutory authority to take action to improve the value of previously contracted debt. The underlying policy is that banks should be allowed to take actions that may improve their balance sheets. For some major international banks, debt-equity swaps are an attractive option for portfolio diversification, and the removal of regulatory obstacles has encouraged the development of these schemes.