Information about Sub-Saharan Africa África subsahariana

Official Debt Reduction: Toronto Terms, Trinidad Terms, and Netherlands Initiative

I. Patel
Published Date:
December 1992
  • ShareShare
Information about Sub-Saharan Africa África subsahariana
Show Summary Details

Francesco Abbate and Anh-Nga Tran-Nguyen*

The Setting

In 1988, Paris Club creditors introduced the “Toronto terms,” a major policy advance in the rescheduling of the official bilateral debt owed by low-income countries. The implementation of the Toronto terms has not, however, resulted in debt relief commensurate with the weak debt-servicing capacity of most low-income African countries.

Recognizing the inadequacy of the Toronto measures, the governments of the United Kingdom and the Netherlands have recently put forward proposals for massive debt reduction. The U.K. proposal, known as the “Trinidad terms,” consists of reducing the stock of Paris Club debt by two thirds and rescheduling the remaining one third over 25 years, with interest payments capitalized during a five-year grace period; debt service would then grow as a function of the debtor’ s export capacity.1 The Dutch proposal calls for full forgiveness of official bilateral debt owed by the least developed and other low-income countries facing severe debt problems, provided they are implementing sound economic policies.

The debt-service profiles resulting from the implementation of the Toronto and the Trinidad terms (the latter with two different rates of growth of debt-service payments, 5 percent and 8 percent) demonstrate that the debt-service reduction under the Trinidad terms is substantial (Chart 1).2 In fact, the Trinidad terms are concessional; the resulting grant element would amount to about 67 percent, while the combined grant element of the three Toronto options is only 20 percent. Moreover, under the Toronto terms, debt-service obligations generally must be rescheduled repeatedly, sometimes every year (the assumption used in the chart). The resulting debt service would increase sharply from Year 9—reaching a level slightly below the debt service due in the absence of debt relief—and would peak in Year 14 at a level almost four times higher than the debt service under the Trinidad terms.

Chart 1.Debt-Service Profiles: Toronto Options and Trinidad Terms 1

Source: United Nations Conference on Trade and Development (UNCTAD), Trade and Development Report, 1991 (Geneva, 19911, p. 54.

1Debt-service payments are derived from a hypothetical loan of $12 million, with a maturity of ten years and an interest rale of 9 percent

This paper examines the adequacy of the three debt-relief measures in relation to the debt-servicing capacity of beneficiary countries. These are the low-income, debt-distressed countries in sub-Saharan Africa, which are recipients of the World Bank’ s Special Program of Assistance (SPA).3 The schemes are assessed here according to their ability to remove the debt overhang. The analysis is first carried out for the SPA countries as a whole and then for each of these countries.

The Impact of Alternative Debt Relief Measures on SPA Countries as a Whole

To evaluate the impact of the three measures on SPA countries as a group, their overall export earnings have been projected through 1997 on the basis of two different annual average rates of growth, 5 percent and 8 percent both in nominal terms (Table 1). The rate of 5 percent might look quite modest but, in reality, is somewhat optimistic when compared with the negative growth rate of –3 percent registered during the past decade.

Table 1.Alternative Debt-Service Obligations of 22 SPA Countries1(In millions of U.S. dollars)
Year 8 (1997)
Year 1 (1990)Export growth of 5 percentExport growth of 8 percent
Export of goods and services12,01116,90120,585
Total debt service
No scheduling6,3834,5904,590
Toronto terms34,5664,4424,442
Trinidad terms34,2843,7903,734
Dutch proposal34,2843,4523,452
Debt service on eligible Paris Club debt4
No rescheduling1,8551,1361,136
Toronto terms5282990990
Trinidad terms0338282
Dutch proposal000
Source: Author’ s calculations based on World Bank, Special Program of Assistance: Proposal for the Second Phase (Washington: World Bank, 1990), Annex D.4.

Data exclude Zambia.

20 percent of exports.

Net of expected rescheduling of non-Paris Club debt on current rules.

Excludes debt contracted after the cutoff date. The stock of eligible Paris Club debt amounted to about $10 billion in 1989.

On the assumption that the debt service is rescheduled each year.

Source: Author’ s calculations based on World Bank, Special Program of Assistance: Proposal for the Second Phase (Washington: World Bank, 1990), Annex D.4.

Data exclude Zambia.

20 percent of exports.

Net of expected rescheduling of non-Paris Club debt on current rules.

Excludes debt contracted after the cutoff date. The stock of eligible Paris Club debt amounted to about $10 billion in 1989.

On the assumption that the debt service is rescheduled each year.

To determine the amount of debt service these countries would be able to pay, a “normative” debt service equivalent to 20 percent of projected export has been chosen in light of a number of considerations. During the past decade, actual debt-service payments of severely indebted low-income countries averaged about 25 percent of their exports of goods and services (after reschedulings and accumulation of large arrears). However, despite significant capital inflows, these countries as a group (and especially the SPA countries) experienced negative growth rates of per capita gross domestic product (GDP). A debt-service ratio of 20 percent would release additional resources to finance some per capita GDP growth, especially if the export growth rate is weak and if external capital inflows are not buoyant, as is expected. It should be noted that a debt-service ratio of 20 percent is still quite high relative to the actual debt-service ratio of 12 percent, which the low-income African countries registered in 1980–81.

The normative debt service may be compared with the debt service that debtor countries would have to pay under the Toronto terms, the Trinidad terms, or the Dutch proposal. This “post-relief debt service is obtained by applying the three measures to scheduled debt service, as projected by the World Bank on the basis of the existing (end of 1988) stock of debt, disbursements from existing commitments, and new commitments foreseen as of 1990. The World Bank projections might therefore underestimate the debt service due in 1997.

If the post-relief debt service exceeds the normative debt service, then additional debt-relief measures are needed to bring the debt service into line with the servicing capacity of debtor countries. The additional debt-service reduction that would be necessary under the different schemes is show in Table 2.

Table 2.Debt-Service Ratios and Additional Debt-Service Reduction Requirements in 1997 for 22 SPA Countries(Values in millions of U.S. dollars; ratios in percent)
Export Growth
5 percent8 percent
Debt-service ratioAdditional debt-service reduction2Debt-service ratioAdditional debt-service reduction2
Toronto terms26.31,06221.6325
Trinidad terms22.441018.10
Dutch proposal20.47216.80
Source: Table 1.

Data exclude Zambia.

To attain a 20 percent debt-service ratio.

Source: Table 1.

Data exclude Zambia.

To attain a 20 percent debt-service ratio.

The Toronto terms appear to be inadequate under both export growth assumptions, because large amounts of additional debt-service reduction are required. The additional reduction would equal about 23 percent of total scheduled debt service (before debt relief of any kind) when the export growth rate is 5 percent, and approximately 7 percent when the export growth rate is 8 percent. When the export growth rate is high, both the Trinidad terms and the Dutch proposal would release enough resources to service the remaining debt; however, when the rate is low, both proposals would require additional debt-service reductions equivalent to about 9 percent of scheduled debt service under the Trinidad terms and some 2 percent under the Dutch proposal.

It should be noted that the three schemes do not cover debt contracted after the cutoff date.4 For SPA countries as a whole, such debt represents about 11 percent of total nonconcessional bilateral debt owed to Paris Club creditors; but for some individual countries the share is much higher. In the case of Uganda, for example, it is about 50 percent. Moreover, debt owed to non-Paris Club creditors—accounting for 25 percent of total bilateral nonconcessional debt—is not subject to the same debt-relief measures. The amount of required additional reduction under the Trinidad terms could be met—at an aggregate level—by advancing the cutoff date and extending the scheme to debt owed to non-Paris Club creditors.

The maintenance of the cutoff date is central to the Paris Club policy of protecting new export credits from rescheduling. However, the increase in concessional debt relief obtained from bringing forward the cutoff date would far outweigh any possible adverse impact on new export credits. In any case, most low-income African countries will be able to attract and afford only highly concessional flows for many years to come. Alternatively, if the cutoff date policy is not changed, even greater aid flows will be required in some cases to compensate for the servicing of post cutoff-date debt.

Impact of the Trinidad Terms on Individual SPA Countries

The results described above do not distinguish between countries; yet, individual countries’ debt burden and need for debt relief differ significantly. Estimates of the impact of the Trinidad terms on debt-service ratios of each of the 18 SPA countries highlight the differences (see the appendix).5 These estimates have been made assuming that the Trinidad terms would have been applied in 1990 on the overall debt (including post cutoff-date debt) owed to Paris Club creditors, after taking into account cancellations of official development assistance (ODA) debt that were granted by donor countries in 1989 and 1990. Exports are projected to grow by 5 percent annually. After five years of grace and interest capitalization, debt-service payments would also increase by 5 percent a year, in line with following the suggestion contained in the Trinidad terms’ formula linking payments growth to export growth.

It is noteworthy that the debt-service ratios have been calculated on the basis of World Bank projections of debt-service payments on the existing (end of 1989) stock of long-term debt. As is the case for SPA countries as a group, these projections tend to underestimate future debt-service obligations, since new commitments, short-term debt, and IMF credit, are not considered.

In examining the impact of the Trinidad terms on debt-service ratios, one can divide the SPA countries into four groups: Group I. For Benin, Central African Republic, Chad, Malawi, and Togo, whose debt-service ratios originally did not exceed 20 percent, the Trinidad terms would result in a reduction in their ratios of a few percentage points. Most of the debt of these countries is on concessional terms. Group 2. Guinea, Niger, Senegal, and Zaïre would benefit significantly from the Trinidad terms. Their debt-service ratios would be lowered from levels originally ranging from 25 to 35 percent to less than 20 percent by the Year 2. Group 3. For Madagascar, Mali, Mauritania, Tanzania, and Uganda, debt-service ratios would decline below 20 percent from Year 6 to Year 8 and beyond. But for the first five to seven years, they would continue to be afflicted by the debt overhang, although for Madagascar and Tanzania, the immediate debt-service reduction would be considerable. In this group, about 40 percent of total debt is owed to Paris Club creditors.

Group 4. Guinea-Bissau, Mozambique, Sao Tome and Principe and Somalia would maintain high, sometimes extremely high, debt-service ratios. While the Trinidad terms would have a significant impact on their ratios, they would remain far above 20 percent. For instance, in Year 8, debt- service ratios would be about 60 percent in Guinea-Bissau and some 30 percent in Mozambique. In Sao Tome and Principe and Somalia, they would be in the 30–40 percent range. All countries in the group owe a large part of their debt to either non-Paris Club bilateral creditors, or multilateral institutions. Guinea-Bissau, for example, owes about three fourths of its official bilateral debt to non-Paris Club creditors and approximately half of its total debt to multilateral institutions.


If the Trinidad terms are applied, almost half of the SPA countries (i.e., the nine countries in Groups 1 and 2) would register debt-service ratios lower than the normative level of 20 percent within two years. But for the remaining countries, saddled with an extremely heavy debt burden, the initiative would not, by itself, remove the debt overhang. This conclusion is reinforced by the fact that debt-service projections do not include debt contracted after 1989 and assume that debt reduction would also apply to post cutoff-date debt, a scenario more favorable to debtors than that which would prevail if the Trinidad terms were adopted. Additional measures, therefore, are needed, such as a higher percentage of debt reduction, as advocated in the Dutch proposal; inclusion of post cutoff-date debt; equivalent measures to be taken by non-Paris Club creditors; rescheduling of multilateral debt; commercial bank debt reduction; and increased ODA flows.

Debt-relief measures must be tailored to the specific needs of debtor countries. For severely indebted, low-income countries, the writing off of two thirds of Paris Club debt, as proposed by the United Kingdom, should be treated as a benchmark. Creditor countries should consider granting additional concessional debt relief, when necessary, in order to contribute to the removal of the debt overhang.

Debt-Service Ratios of SPA Countries, Before and After Implementation of the U.K. Proposal
YearBeninCentral Africa RepublicChadGuineaGuinea-BissauMadagascar
YearMalawiMaliMauritaniaMozambiqueNigerSao Tome and Principe
Source: Author’ s calculations based on the World Bank Reporting Debtor System.
Source: Author’ s calculations based on the World Bank Reporting Debtor System.

This paper was prepared with the assistance of Anne Miroux, Cape Kasahara, and Lucette Fagedet.


The following repayment formula underlies the Trinidad terms:

Let D1 be the debt outstanding at the beginning of Period 1 (which follows the the grate period of five years); P1 the initial debt-service payment, r, the interest rate (and also the discount rate); g, the rate of growth of payments (which can be made equal to the projected rate of growth of export earnings); and N, the repayment period (20 years).

If all payments art made at the end of each period, then

From the above, the value of P1 is given by


Under the Trinidad terms, the debt-service profile varies with the growth rate of payments (g, as in footnote 1). Chart 1 shows the debt-service profiles on the basis of two hypothetical values of g: 5 percent and 8 percent. Given a constant repayment period of 20 years, the lower the value of g, the higher are the payments during the first few years. If g is equal to the export growth rate, a paradoxical situation might occur: the lower is the export growth rate—that is, the lower the debtor’ s servicing capacity, the higher are the initial payments. An alternative formula could be used whereby g is higher than the export growth rate, while the repayment period is kept unchanged. This would enable debtor countries with low growth in export earnings to make smaller payments in the earlier part of the repayment period. Another alternative to be appplied in those circumstances would be to lengthen the repayment period while g equals the export growth rate.


As of February 1991, 23 countries have received SPA support: Benin, Burundi, Chad, Central African Republic, The Gambia, Ghana, Guinea, Guinea-Bissau, Kenya, Madagascar, Mali, Malawi, Mauritania, Mozambique, Niger, Sao Tome and Principe, Senegal, Somalia, Tanzania, Togo, Uganda, Zaïre, and Zambia.


In Paris Club reschedulings, the cutoff date usually precedes the first agreement by three to six months and is kept in successive agreements.


Of the 22 SPA countries covered in Table 1, Burundi, The Gamiba, Ghana, and Kenya have not been included in the projections because they are not likely to require Paris Club reschedulings in the foreseeable future.

    Other Resources Citing This Publication