Debt Relief Home

Documents about the HIPC Debt Initiative

Country Documents

Publications on Debt Relief

World Bank HIPC Information

Use the free Adobe Acrobat Reader to view pdf files.

Heavily Indebted Poor Countries (HIPC) Initiative Perspectives on the Current Framework and Options for Change - Supplement on Costing
Prepared by the Staffs of the World Bank
and International Monetary Fund

April 13, 1999


1. This supplement presents updated estimates of the costs of the HIPC Initiative under the current framework, and in a range of possible alternative frameworks.1 It updates the estimates provided in the note "HIPC Initiative--Tentative Costing of Illustrative Scenarios" but broadens that analysis by providing creditor-specific information and additional estimates, based on the latest country information available to staff (Box 1).2

2. As in earlier costing exercises, a number of important caveats apply. The cost estimates rely on key assumptions, and on macroeconomic projections and debt numbers which have mostly not been fully reconciled between creditors and debtor governments (Box 2). Therefore, the estimates need to be interpreted with caution and should be seen as subject to a substantial margin of uncertainty. In making estimates for the current framework, staff has aimed to provide realistic but conservative estimates of costs; thus, in cases where a choice of targets or timing was required, the option implying a higher cost was used. The costing exercise is not intended to prejudge the results of the country-specific tripartite debt sustainability analyses, or the Boards' decisions regarding the eligibility of individual countries to qualify for assistance under the HIPC Initiative, the NPV of debt-to-export targets, the decision points, or completion points to be set for those countries. Finally, the inclusion of alternatives to the current HIPC Initiative in the present note does not imply any endorsement by staff or management, nor does the exclusion of other proposals imply a rejection.

Box 1. The Database

  • The costing analysis is based on the most recent country-specific debt sustainability analyses (DSAs) presented to the Boards, supplemented in some cases by more recent information prepared by Bank and Fund staff.

  • Country coverage for these costing estimates is based on the original list of 41 HIPCs, excluding Nigeria which is not eligible to borrow on IDA-only terms from the World Bank, and adding Malawi on the basis of its recent DSA. The countries are: Angola, Benin, Bolivia, Burkina Faso, Burundi, Cameroon, Central African Republic, Chad, Dem. Rep. of Congo, Rep. of Congo, Côte d'Ivoire, Equatorial Guinea, Ethiopia, Ghana, Guinea, Guinea-Bissau, Guyana, Honduras, Kenya, Laos, Liberia, Madagascar, Malawi, Mali, Mauritania, Mozambique, Myanmar, Nicaragua, Niger, Rwanda, São Tomé and Príncipe, Senegal, Sierra Leone, Somalia, Sudan, Tanzania, Togo, Uganda, Vietnam, Yemen, Zambia.

  • Ghana, Kenya, Laos, Liberia, Malawi, Somalia, and Sudan have never received a concessional rescheduling from the Paris Club. Countries must make full use of traditional debt-relief mechanisms to be eligible for debt relief under the HIPC Initiative.

  • Liberia, Somalia, and Sudan are included in some aggregate cost estimates, but not in creditor-specific costing breakdowns because of the relatively poor database and uncertainty regarding the treatment of their large arrears.

  • Nine countries have not yet met the entry requirement to begin qualifying for the HIPC Initiative by having in place Fund- and Bank-supported adjustment and reform programs in the period since September 1995: Angola, Burundi, Democratic Republic of Congo, Equatorial Guinea, Liberia, Myanmar, São Tomé and Príncipe, Somalia, and Sudan. These countries would need to begin such programs before end-2000 to qualify for the Initiative (see the Review and Outlook paper--footnote 2).

  • Formal DSAs are unavailable or out-of-date for: Burundi, Democratic Republic of Congo, Liberia, Myanmar, and Somalia. For these countries information is especially unreliable, and cost estimates involve considerable uncertainty.

  • It is likely that some proposals for change of the HIPC Initiative could extend eligibility to countries outside of this group of 41 countries. Staff has not, however, included other countries for this analysis, both because the debt data are for the most part poor, and because most other countries have not received concessional reschedulings from Paris Club creditors and in many cases the full use of traditional debt relief mechanisms would be sufficient for debt sustainability.

Timing of decision and completion points

3. The baseline for timing of decision points which has been assumed for the purposes of these costing estimates is the earliest that might be proposed under the current framework of the Initiative (Table 1). Based on actual performance under reform programs, some decision points may be reached later. All countries found eligible for HIPC assistance are assumed to have completion points three years following the decision point under the baseline scenario. In alternative scenarios, the performance period is shortened according to the various proposals made.

Box 2. Assumptions used for Cost Estimates

  • Macroeconomic framework: The estimates are based on the medium-term macro-economic frameworks developed by staff country teams, including especially the export and debt data and projections. These data are subject to wide margins of uncertainty.

  • Timing of assistance: It is assumed that all countries which are potentially eligible for HIPC Initiative assistance actually request such assistance, and that assistance is delivered without delays (meaning no slippages in implementing economic and social reform programs).

  • Debt sustainability targets: In proposals (including the current framework) in which a range of NPV of debt-to-export targets or NPV of debt-to-revenue targets is suggested, the costing estimates assume that all countries receive the lowest available target except countries which (i) have already reached the decision point, (ii) have had preliminary HIPC discussions at the Boards. The latter assumptions are relaxed in scenarios which cost retroactive application of new criteria. All countries are assumed to receive assistance based on whichever window provides the largest amount of relief.

  • Burden-sharing: For those countries which have already received commitments of assistance, the division of costs by creditor group reflects the amounts of assistance shown in each decision point document (unless retroactivity is applied). For prospective cases, burden sharing is assumed to be fully proportional after full application of traditional debt relief mechanisms (67 percent NPV reduction on eligible debt under Naples terms). For simplicity, proportional burden sharing is also assumed in scenarios with retroactivity, thus excluding the additional, exceptional costs undertaken by the Bank and Fund for Mozambique.


4. Total costs for the HIPC Initiative, as currently structured, are now estimated at US$12½ billion in 1998 NPV terms, up by 29 percent from the estimates made in August 1998 (Table 2).3 4 As presaged in the Tentative Costing note, lower commodity prices have affected the export base for many potentially eligible countries. Lower projected exports (which may also include some export volume changes) contribute roughly two-thirds of this increase. Further, market interest rates used to discount debt service streams have generally declined, contributing to a higher estimate of the present value of debt.5 Finally, in some countries, revised estimates of the outstanding debt have affected assistance.

5. It is now estimated that 29 countries could potentially be eligible for debt relief under the current framework of the HIPC Initiative (Table 3). In comparison with the 23 countries expected in last year's costing estimates, Cameroon, Chad, Republic of Congo, Guinea, Malawi, and Sierra Leone are added. Of the 29, three are estimated to receive HIPC assistance under the fiscal criteria: Côte d'Ivoire and Guyana, which have already reached their decision points, and Republic of Congo.6

6. Baseline costs for selected country groups indicate that all of the increase in costs is attributable to the group of countries which have met the entry requirement, but have not yet reached the preliminary document stage (Table 2). Large increases in costs were estimated for Republic of Congo, Nicaragua, and Zambia, mostly on the basis of revisions to the outstanding debt stocks, while new export projections have also contributed. The wide variation in the estimates of debt stocks reflects again the poor quality of the data. Cost estimates generally improve with economic and institutional conditions, particularly as the country approaches the decision point, when a reconciled, loan-by-loan DSA is prepared.

7. Following the practice of earlier analyses, country-specific costings are not presented here, as they might create misleading expectations when in many cases the basis for these individual estimates is not firm and could change substantially. Tentative costing estimates will be provided in individual DSAs as they are prepared. Country-specific indicators including the debt burden of each country with respect to the export and fiscal base at the completion point are shown in Table 4. This table also presents the status of each country with respect to the fiscal thresholds at the decision point.7

Costs by creditor group

8. Under the current framework, the shares of HIPC Initiative costs for bilateral and multilateral creditors are estimated to remain roughly equal. The World Bank's costs would be US$2.4 billion, up from US$2.0 billion in the 1998 exercise. The IMF's costs have increased from US$0.8 billion to US$1.2 billion in the current estimates (excluding Liberia, Somalia, and Sudan).8

9. In some cases, the assistance provided by Paris Club creditors through Lyon terms (up to 80 percent NPV reduction on eligible debt) and the assumption of comparable treatment from other bilateral and commercial creditors may not provide sufficient assistance to reach the bilateral creditors' required share of assistance.9 This has already occurred in the case of Mozambique. Based on current cost estimates, a similar situation will arise in varying degrees for Burundi, Guinea-Bissau, Madagascar, Malawi, Nicaragua, Madagascar, Mauritania, Myanmar, Rwanda, São Tomé and Príncipe, Sierra Leone, and Zambia.10 But it should be noted that the breakdown into pre- and post-cutoff date debt is often uncertain until a loan-by-loan reconciliation is done, and this information is thus subject to considerable uncertainty.


10. Tables 5-7 present costings of a number of possible alternative frameworks for the HIPC Initiative, and Table 8 presents country eligibility under each framework. As in the Tentative Costing note, the costings are presented excluding retroactive treatment of countries which have already reached their decision point, with an indication of the component due to retroactivity shown separately.11

11. Staff has attempted to encompass a number of the proposals described in the main paper within these estimates, although there are elements of each proposal which are not readily costed--such as the political or governance tests some would apply for HIPC Initiative assistance. Relief on ODA claims would provide further debt relief under many of these proposals; nominal costings of proposals made by major creditors with respect to ODA are discussed below.

Breakdown of costs by creditor

12. Cost sharing among creditors for a group of alternative proposals are shown in Tables 9 and 10, excluding and including retroactive costs for countries which have already reached the decision point, respectively. As in the current framework, multilateral and bilateral creditors share costs approximately evenly in proposals using the baseline timing assumptions of the current framework. Bilateral creditors take on a slightly larger share of assistance if delivery is brought forward by eliminating the second stage. This occurs because new multilateral lending is projected to be greater than new bilateral lending during the second stage, and thus multilaterals would hold a larger share of claims at a later completion point. The country-specific distribution of assistance is also likely to have an effect on this result. For reference, the total stock of claims in nominal terms of the IMF and World Bank on the 41 countries considered here was US$9.4 billion and US$39.2 billion, respectively, as at end-February 1999 (Tables 11 and 12).

Debt relief on ODA claims

13. A number of proposals from creditor countries (and others) have suggested additional action on ODA claims. While such claims have been forgiven in the past by several creditors, either completely or partially, it is estimated that approximately US$24 billion in ODA claims on the 41 HIPCs remains outstanding to DAC members, and roughly the same amount to non-DAC bilateral creditors, both stated in face-value terms. The nominal value of full cancellation of ODA claims by all creditors would potentially be on the order of US$40-50 billion.12 Under traditional debt relief mechanisms (Naples terms) as well as Lyon terms, ODA claims are rescheduled over 40 years with 16 years' grace period at no higher than the original concessional interest rates. Such treatment, on an illustrative basis, is estimated to lead to a profile of debt service having a grant element of about 25 percent (and an unchanged nominal value).13 Cancellation of ODA debt service for a generation (30 years), as suggested by France, could--again on an illustrative basis--reduce the present value of ODA by two-thirds.

14. The impact of ODA debt relief on the cost of assistance under the HIPC Initiative depends on the timing of such action. First, if it is delivered prior to the completion point, it reduces the overall cost of the HIPC Initiative--as previous ODA cancellation has done. Second, if it is delivered as part of HIPC assistance, it reduces the need for bilateral creditors to provide debt relief on other types of claims. Finally, if it is delivered after the completion point as suggested by some creditors, it would be additional to HIPC Initiative assistance and lead to lower debt outcomes below HIPC target levels.14 15

Box 3. The Nominal Amount of HIPC Initiative Debt Relief

For countries which have reached the decision point, staff has provided a rough translation from HIPC Initiative relief--expressed in NPV terms--into nominal debt service relief, i.e., the undiscounted sum of the future debt service that would not have to be paid as a result of HIPC Initiative assistance. The ratio of nominal debt service relief to NPV relief measured at the completion point has typically been about 2:1. Total debt relief in 1998 NPV terms would be subject to a slightly larger ratio, since, for most countries, the NPV reduction delivered at the completion point is discounted back to 1998 terms. However, this ratio depends on the profile of delivery of assistance, which differs for each multilateral creditor and is often not known with precision before the completion point. For instance, the ratio was lower than 2:1 for Bolivia, since its debt relief was frontloaded to help smooth an early hump in debt servicing.


15. The cash flow of assistance under the HIPC Initiative has attracted a good deal of interest, as noted in the Options paper before the Boards. It is particularly hard to measure cash flow, however, since different creditors provide their assistance over different time profiles. The cash flow effects of HIPC Initiative assistance using the set of countries which have reached the decision point is shown in Table 13. On average, assistance under the HIPC Initiative is estimated to reduce debt service due in the period following the completion point by 18 percent (within a wide range). In comparison to the debt service paid prior to HIPC debt relief, the reduction is about 2 percent on average, within a wide range, and some countries are expected to experience an increase in debt service due even after HIPC assistance. In part this reflects the fact that debt service due to Paris Club creditors rises with a stock-of-debt operation compared with a flow rescheduling under which interest falling due is also rescheduled. Furthermore, the real burden of future debt service is expected to decline as it would represent a shrinking proportion of the export or fiscal base as HIPCs continue to grow after the completion point.16

16. An approximate range of the early profile of nominal debt service relief for a given NPV reduction is given in the final column of Table 13. On average, HIPCs which have reached the decision point are expected to receive annual nominal debt service relief representing 9 percent of the NPV value of HIPC assistance at the completion point in the early years. Higher frontloading (13 percent per year) was provided for Bolivia in light of the high debt service burden in the years just following the completion point. Less frontloading was provided to Mozambique (5 percent per year), with a relatively large share of debt owed to bilateral creditors. (The debt service due to bilateral creditors under Naples terms is already very low in the early years, so the additional relief provided by bilateral assistance under the HIPC Initiative is small.)

17. The cash flow impact of proposals to change the HIPC Initiative will be developed when the options have been narrowed further. The profile of delivery of multilateral assistance, including the possibility of advancing some assistance into the second stage, will need to reflect the availability of multilateral institutions' financing.


18. The time profile of commitments of HIPC assistance under the current framework is shown in Table 14, based on the assumed earliest decision points presented in Table 1. On this basis, 24 percent of total commitments (in 1998 dollar terms) were made by end-1998. A large number of countries are costed on the basis of decision points assumed to be in 1999, and on that basis nearly 50 percent of total commitments would be made during this year. Past experience suggests that at least some countries would be expected to be delayed from this earliest possible timing. The time profile of commitments of assistance under the four alternative scenarios from Table 9, while not shown, reflect approximately the same annual shares as shown in Table 14.


19. It is beyond the scope of this paper to consider the financing of various alternative proposals to change the HIPC Initiative framework. However, even the current framework is now estimated to be some US$2.8 billion more costly than last year's assistance, and this alone will require identification of new resources for multilateral creditors.17 Any discussion of changing the Initiative needs to be informed by a careful consideration of the source of additional financing. Based on the guidance on directions for change provided by the Boards of the Bank and Fund, and by the Development and Interim Committees, staff of the two institutions will provide further information on the financial constraints and possible sources of additional support.

1 Descriptions of the proposals made by various commentators and member countries of the IMF and World Bank are provided in the main paper (EBS/99/52, April 2, 1999, and IDA/SecM99-155), and a supplement providing the specific proposals received in the consultation process has been forwarded to the Boards.
2 EBD/99/32 and IDA/R99-19 (February 16, 1999). The tentative costing paper used the database and analytical framework developed for "The Initiative for Heavily Indebted Poor Countries--Review and Outlook," (EBS/98/152 and IDA/SecM98-480, August 25, 1998).
3 This estimate uses a rate of 6 percent to discount country-specific costs at the respective completion points to 1998 NPV terms. The NPV of debt at the completion point, and thus assistance at the completion point, are calculated using currency-specific discount rates (Commercial Interest Reference Rates published by the OECD).
4 Excluding Liberia, Somalia, and Sudan. Total costs including rough estimates for these three countries are US$19 billion, in comparison with the US$16 billion estimated in the Tentative Costing note using the 1998 database. Potential costs for Liberia, Somalia, and Sudan remain highly uncertain, and thus are shown only in Table 5. However, firm estimates of these costs are not provided here because the data and underlying assumptions remain highly uncertain. As in last year's estimates, HIPC Initiative costs for these three countries are calculated assuming that arrears to multilateral creditors have been cleared with a concessional refinancing at the decision point. Such an approach has not been agreed, and financing would need to be identified. In this scenario, debt at the completion point, and HIPC Initiative costs, would be lower than in an alternative scenario in which arrears were cleared nonconcessionally.
5 Lower interest rates reduce the effective concessionality of fixed-rate debt, thus increasing its present value.
6 Honduras is not eligible under the current framework because it does not meet the 20 percent revenue-to-GDP threshold in the fiscal criteria (see Table 4). Honduras' debt situation will be presented to the Boards shortly.
7 Similar information was provided in "Cap Paper for the Preliminary HIPC Initiative Documents for Bolivia, Burkina Faso, Cote d'Ivoire, and Uganda," (EBS/97/59 and IDA/SecM97-104), April 1, 1997, and "HIPC Initiative-Guidelines for Implementation," (EBS/97/75 and IDA/R97-35), April 21, 1997.
8 This is equivalent to SDR 1.4 billion on the "as-needed basis" used in financing analysis for the ESAF-HIPC Trust.
9 As noted in the main paper, several creditors have suggested that this constraint should be relaxed.
10 The cost estimates include assistance above 80 percent NPV reduction where this is necessary for proportional burden sharing.
11 Retroactivity is defined such that countries which would have been eligible for earlier timing or lower targets under new rules, would have their debt situation reevaluated on the basis of the latest actual data in 1999 (as of end-1998). Thus, for instance, Uganda's debt situation is reevaluated as of 1998 data, rather than based on the 1997 data used is at its completion point.
12 The data sources do not allow the calculation of the NPV of these claims. These estimates of face-value amounts are derived from some information provided directly by creditors, supplemented by further information based on the World Bank Debtor Reporting System data, and are not reconciled with the other data presented in this paper. Information on ODA claims remains limited; some amounts included may represent rescheduled claims originally extended on commercial terms.
13 This estimate reflects a simulation based on outstanding claims of major Paris Club creditors and market-based interest rates. The DAC uses a fixed 10 percent discount rate to determine the minimum 25 percent grant element of ODA loans. In contrast, the grant element under the HIPC Initiative is lower because it is based on market-related discount rates, and also varies considerably across currencies.
14 Japan has indicated it will provide additional relief on ODA claims after the completion point in exceptional cases. Canada and Norway have indicated they are prepared to do so on a case-by-case basis for commercial claims.
15 From a broader perspective, overall additionality could be affected by possibly related decisions on aid flows.
16 The country-specific information underlying these results is presented in Annex 1, Chart 1, page 46 of the main options paper.
17 For example, as a result of the higher estimated baseline cost of the HIPC Initiative, the total financing requirement for the IMF’s ESAF-HIPC Trust for its contribution under the HIPC Initiative and for interim ESAF subsidies has increased from US$2.1 billion to US$2.5 billion in NPV terms at end-1998. Thus far the Trust has received bilateral contributions from nine countries of approximately US$75 million; there are positive indications of bilateral contributions for another US$1.1 billion. Taking into account direct contributions from the IMF of US$0.1 billion, the residual financing requirement for the ESAF-HIPC Trust would be US$1.2 billion. Under the alternative frameworks shown on Table 9, the residual financing requirement would be between US$1.4 billion and US$1.9 billion.With regard to the current ESAF, subsidy resources are sufficient, and the Fund has received pledges for loan resources from Canada, France, and Germany of around US$1.3 billion--more than half-way towards the target of an additional US$2.1 billion in loan resources.