IDA13 New Options for IDA Lending Terms International Development Association September 2001 NEW OPTIONS FOR IDA LENDING TERMS I. Introduction ..............................................................................................................................1 II. Blend Countries and Their Lending Programs.........................................................................2 Blend status............................................................................................................................2 Principles guiding lending to blend countries........................................................................3 Summary ­ case for differentiating terms of IDA lending to notional blends with per capita income below the operational cutoff .................................................................4 III. Exceptional Eligibility of Countries with per Capita Income Above the Operational Cutoff..5 Small island countries............................................................................................................5 Other countries.......................................................................................................................7 Summary ­ case for differentiating terms of IDA lending to countries with exceptional eligibility...................................................................................................8 IV. Options for Differentiation of IDA Lending Terms .................................................................9 Notional blends with per capita income below the operational cutoff ..................................9 Countries with exceptional eligibility..................................................................................10 Summary of proposed modifications of IDA terms.............................................................10 V. Implications of Hardened Lending Terms for IDA Allocations............................................13 VI. Effect of Hardened Lending Terms on IDA Finances...........................................................14 VII. Conclusions and Issues for Discussion .................................................................................15 Tables: Table 1: Notional blends in IDA13................................................................................................ 5 Table 2: Projected exceptional eligibility ­ small island countries ............................................... 6 Table 3: Projected exceptional eligibility ­ other countries .......................................................... 8 Table 4: New options for IDA lending terms .............................................................................. 11 Table 5: Indicators guiding classification of borrowers and proposed lending terms................. 12 Table 6. Terms factor for allocations of IDA on notional blend terms........................................ 13 Annex ............................................................................................................................................17 NEW OPTIONS FOR IDA LENDING TERMS I. Introduction 1. At their first IDA13 meeting on February 28-March 1, 2001, Deputies considered the discussion paper IDA Eligibility, Terms and Graduation Policies (January 2001). While current eligibility and graduation policies were found to be broadly adequate, Deputies asked for further work in two areas: (i) on IDA grants; and (ii) on hardening the terms of IDA lending to countries that are among the more creditworthy IDA borrowers but lack adequate access to IBRD lending. This paper responds to the second request. It sets out options for introducing new IDA lending terms in accordance with IDA's eligibility and graduation policies. 2. Current IDA lending terms are highly concessional and nearly the same for all borrowers. All IDA credits have a 10-year grace period, a zero interest rate, a 0.75 percent service charge on outstanding balances, and extended maturities ­ 35 years for blend borrowers and 40 years for IDA-only borrowers. This small difference in maturities results in almost identical degrees of concessionality: a grant element of 61 percent for blend borrowers and 64 percent for IDA-only borrowers.1 For the most part, IDA's current, highly concessional, lending terms continue to be appropriate. This is especially true of the IDA-only countries, which are very poor and have limited natural resources and difficult development prospects. Indeed, as is argued in a companion paper, for certain activities it may be useful to deploy IDA grants in such countries.2 On the other hand, uniformity of treatment stands in contrast to a widened range of country circumstances, including in the extent of effective access to other sources of development financing (notably from IBRD) and in capacity to service debt on harder terms. 3. This paper presents two groups of countries where new IDA lending terms could be introduced, subject to Deputies' guidance: (i) blend borrowers with per capita income below the eligibility threshold that are not able to borrow from IBRD but are able to take on debt on hardened IDA terms; and (ii) borrowers that are IDA-eligible on an exceptional basis despite their per capita income being above the eligibility threshold (normally in the context of a graduation scenario). Subject to Deputies' guidance, some of the countries that have retained access to IDA under the small island economy exception would be included in this second group. If all of the options presented in this paper were implemented, the majority of IDA clients, including most of the small island countries, would continue, however, to receive IDA credits on standard terms. 1The grant element is defined as the difference between the discounted present value of disbursements and the discounted present value of service payments, expressed as a percentage of the face value of the loan or credit. Grant element calculations in this paper are based on a discount rate of 6.3 percent and the actual average disbursement profile for IDA credits. The discount rate chosen is the value of the commercial interest reference rate (CIRR) for SDR-denominated loans in 2000; the CIRR is the minimum interest rate applicable to official financing support of export credits, in accordance with the Arrangement on Guidelines for Officially Supported Export Credits of the OECD. For more information, see "Finance and Investment" on the OECD web page http://www.oecd.org. 2Grants in IDA13, September 2001, available on the IDA13 web page http://www.worldbank.org/ida/ida13docs.html. - 2 - 4. Introducing less concessional lending terms for countries that are able to service such debt would strengthen IDA's role as a transitional instrument because it would facilitate a phasing-out of the subsidy embodied in IDA lending. IDA already pursues this important objective through a provision for a possible acceleration of repayments to IDA by countries where per capita incomes have surpassed the eligibility threshold.3 In addition, IDA's graduation policies provide that IDA lending be phased out in countries with per capita income above the so-called operational cutoff,4 or when a country is creditworthy for adequate amounts of IBRD lending. Credits on hardened terms would be an additional instrument that would complement these mechanisms in the pursuit of the same objective: to help augment the transfer of IDA resources to the poorest countries, especially the IDA-only countries at the lower end of the range of per capita incomes. Moreover, introducing hardened terms for IDA assistance to countries with exceptional eligibility would mean that setting the terms of exceptional IDA access on a case-by- case basis would no longer be necessary. 5. This paper is organized as follows. It first describes the principles guiding lending to the two groups of countries for which less concessional IDA lending terms could be considered (Sections II and III). Section IV presents options for introducing new terms of IDA assistance during IDA13 to these two groups of countries. Section V looks at the possible implications for the level of IDA allocations. Section VI shows the impact on IDA's finances. Section VII summarizes the proposed options and sets out some of the issues on which Deputies' guidance is needed before new IDA lending terms could be introduced into the IDA13 policy framework. II. Blend Countries and Their Lending Programs Blend status 6. A country is classified by World Bank management as having blend status when it is eligible to borrow both from IDA and from IBRD. In particular, such a country needs to satisfy IDA's eligibility criteria with regard to policy performance, poverty, and limited creditworthiness. At the same time, the country has to be creditworthy for a limited amount of IBRD borrowing or have the potential to become so (for example, with sustained strong performance). Blend status can be seen as a stage in the development process in which a country undergoes a transformation from IDA-only, to blend, to IBRD-only, and eventually to graduation 3An acceleration clause has been included in IDA credits since 1987. It was modified in 1996, providing for possible doubling of repayments to IDA from borrowers where per capita income has exceeded the operational cutoff for three consecutive years (see Section IV below). 4The `operational cutoff' (currently $885) guides determination of IDA eligibility and differs from the `historical ceiling' (currently $1,445). The operational cutoff was introduced in IDA8 to enable IDA to concentrate its scarce resources on the poorest countries. In IDA9, the operational cutoff was formally established as the relevant eligibility guidepost and the former eligibility threshold was named the `historical ceiling.' Both thresholds are updated annually; please refer to Estimating per Capita Income for Operational Purposes (SecM2000-625), October 30, 2000, or Estimation of Internationally Comparable per Capita Income Numbers for Operational Purposes (2001) at http://www.worldbank.org/data/databytopic/pci.doc. - 3 - from IBRD. Among the 79 countries eligible for IDA, those with blend status5 tend to be better off ­ with higher per capita incomes and more diversified economies ­ and have stronger development prospects and repayment capacity than most IDA-only countries. Principles guiding lending to blend countries 7. As with all IDA borrowers, the lending amount to all blend countries is individually determined for a three-year rolling period. The total IBRD/IDA lending volume is set in relation to the country's need for development finance, its creditworthiness, and its policy and portfolio performance. The level of IDA lending is determined through the performance-based allocation (PBA) system and is set out in the country assistance strategy (CAS), taking into account other factors such as access to alternative sources of concessional financing, and country conditions. For a number of blend countries ­ India, Indonesia, Nigeria, and Pakistan ­ the IDA allocation is effectively "capped" because of the countries' large size and to reflect IDA donors' guidance regarding their IDA allocations. 8. Assessments of country performance and creditworthiness are key determinants of World Bank (IBRD and IDA) lending to blend countries. Since good policy performance is a necessary but not a sufficient condition for creditworthiness, asymmetric changes in IBRD and IDA lending can be expected. For example, if policy performance and/or economic conditions deteriorate, both IDA and IBRD lending may be reduced, but not proportionately. If the change in the country's circumstances is sufficiently severe to mean that it is no longer creditworthy, IBRD lending would be suspended, while IDA lending may continue, making the country a de facto IDA-only country for a time. Conversely, when a country's policy performance begins to improve from a low level, IDA lending would increase in line with the country's performance- based allocation while IBRD lending may not be possible until the country has established a performance track record and any other factors affecting the country's creditworthiness have been resolved. These two scenarios help explain the lack of IBRD lending to some blend countries and illustrate that, in general, IDA lending to such countries cannot be replaced by increased lending from IBRD, because of creditworthiness constraints. Thus, a number of countries have become so-called `notional blends' since they are classified as having blend status but are de facto IDA-only borrowers with no access to IBRD borrowing due to lack of creditworthiness. 9. When a blend country has not borrowed from IBRD for an extended period, World Bank management may decide, after review, to keep its status as blend or to change it to IDA-only. For example, in FY00, a reevaluation of the economic and balance of payments prospects of Armenia, Georgia, Kyrgyz Republic and Moldova ­ which have deteriorated in the late 1990s ­ resulted in a change in their status from blend to IDA-only.6 Similarly, Cameroon, Republic of Congo, and Côte d'Ivoire were re-classified as IDA-only in the early 1990s. 5There are fourteen countries currently classified as blend borrowers: Azerbaijan, Bolivia, Bosnia, Dominica, Grenada, India, Indonesia, Nigeria, Pakistan, St. Lucia, St. Vincent and the Grenadines, Uzbekistan, Yugoslavia, and Zimbabwe. 6Two of these countries ­ Georgia and Kyrgyz Republic ­ never borrowed from IBRD, despite their blend status. - 4 - 10. IDA and IBRD have different legal and financial structures, with independent decision- making by the two institutions. In particular, since IBRD is a market-funded institution, its financial policies do not accommodate lending to countries that do not meet minimum requirements.7 There is thus very little or no scope for modulating the ratio between IDA and IBRD lending volumes to a blend borrower to achieve "blended" (harder) terms of assistance from the World Bank if the country's creditworthiness for IBRD loans is marginal or absent. 11. The policy framework for recent replenishments has provided that IDA lending to blend borrowers be concentrated mainly in certain types of operations (notably social sector investments and interventions aimed directly at poverty reduction), as these countries were expected to borrow from IBRD to meet financing needs in other areas. However, this has not proved to be a feasible option for the notional blends because they have obtained only limited, if any, lending from IBRD. The composition of IDA lending to the notional blend countries has therefore resembled the composition of IDA's assistance to IDA-only borrowers ­ for example, in Azerbaijan, Bosnia, FYR Macedonia, and Pakistan IDA-supported operations have included investments in transport and other infrastructure as well as macroeconomic and sectoral adjustment programs. Summary ­ case for differentiating terms of IDA lending to notional blends with per capita income below the operational cutoff 12. In recent years, only a few blend borrowers with per capita income below the operational cutoff8 received World Bank assistance from both IBRD and IDA ­ including, notably, India and Indonesia. A number of other countries have become notional blends ­ they are classified as having blend status but have received only very small or no lending from IBRD because their access to IBRD has been constrained by lack of creditworthiness. 13. Table 1 shows the countries that are expected to be notional blends with per capita income below the operational cutoff during IDA13. Though all of these countries satisfy IDA's eligibility criterion relating to poverty (since their per capita incomes are expected to remain below the operational cutoff), per capita income is above $450 in four of these five notional blend countries. Thus, as a group, these five notional blend countries are better off than the majority of IDA's borrowers ­ the IDA-only countries, among which two-thirds (43 countries out of 65) have per capita incomes below $450 and experience deeper and more pervasive poverty.9 7Doing otherwise could weaken the World Bank Group, including IBRD's capacity to make transfers to IDA. 8Countries with income above the operational cutoff are eligible for borrowing from IDA on an exceptional basis; during IDA13 the terms of lending to such countries would be determined by their exceptional eligibility, as set out in Section III below. 9One half of IDA-only countries actually have per capita incomes below $350 (less than $1/day). - 5 - Table 1: Notional blends in IDA13 Basic Indicators IBRD-to-IDA IDA13 Lending Per capita blend ratio Projection income Population FY99-01 projected FY03-05, SDR m (US$, 2000) (m, 2000) (CAS period) a/ Blend countries with income below operational cutoff and no or uncertain IBRD access Azerbaijan 610 8.0 0 : 100 0 : 100 80 Nigeria 260 126.9 0 : 100 0 : 100 750 Pakistan 470 138.1 43 : 57 0 : 100 700 Uzbekistan 630 24.7 n/a 50 : 50 75 Zimbabwe 480 12.1 0 : 100 0 : 100 52 Total 1, 657 Memo: Blend countries with income below operational cutoff and effective access to IBRD India 460 1,015.9 62 : 38 72 : 28 2,012 Indonesia 570 210.4 86 : 14 65: 35 315 a/ Based on projections of creditworthiness and policy performance; both are subject to uncertainty. 14. The continuing blend status of the notional blend countries reflects their stronger economic position and development prospects, compared to the majority of IDA-only countries. Yet because of their lack of access to IBRD, which may last for an extended period, these notional blend countries would continue to receive World Bank assistance embodying the very high concessionality of IDA's standard terms, unless the terms on which they receive IDA lending are hardened. Hardening of IDA terms would not only increase the equity of IDA resource allocations but also better correspond to the full range of sectors and operations that IDA is called upon to support in these countries. III. Exceptional Eligibility of Countries with per Capita Income Above the Operational Cutoff 15. A number of countries are eligible to borrow from IDA although their per capita incomes are above the operational cutoff. Among these, the small island economies are a distinct group. The other countries with exceptional eligibility are normally in the process of graduation from IDA or are eligible to borrow from IDA on a limited and exceptional basis while they lack access to IBRD lending. The rationales for possible hardening of IDA terms for lending to these two sets of countries are presented below. Small island countries 16. In 1985, six small island countries (Dominica, Grenada, St. Kitts and Nevis, St. Lucia, St. Vincent and the Grenadines, and Tonga) obtained access to IDA resources even though their per capita incomes were above the IDA-eligibility criterion. Since that time, the number of countries eligible for IDA under this "small island exception" grew to ten as one of those originally eligible graduated from IDA (St. Kitts and Nevis) and five other small island countries saw their - 6 - per capita income increase above the operational cutoff (Cape Verde, Kiribati, Maldives, Samoa, and Vanuatu). 17. The decision to introduce the small island exception was based on the following considerations. First, the six small island economies shared many of the problems of low- income developing countries including export concentration, small domestic markets, high cost of infrastructure, limited skill base, and weak institutions. In addition, it was determined that these economies suffered from certain circumstances particular to their size and geography such as vulnerability to natural disasters, isolation, lack of natural resources, and unavailability of commercial credit. Second, it was stated as a principle that a member country of the World Bank Group should not be left without access to either IBRD or IDA provided its performance is adequate. While six of the countries that are now eligible for IDA under the small island exception are IDA-only, lacking creditworthiness for IBRD, four have blend status and do borrow limited amounts from IBRD (Table 2). Table 2: Projected exceptional eligibility ­ small island countries Basic Indicators IBRD-to-IDA IDA13 Per capita blend ratio Lending income Population FY99-01 projected a/ Projection (US$, 2000) (m, 2000) SDR m IDA-only countries: Cape Verde, 950-1,660 1.7 n/a n/a 65 Kiribati, Maldives, Samoa, Tonga, Vanuatu Blend countries: Dominica, 2,690-4,070 0.4 35 : 65 64 : 36 30 Grenada, St. Lucia, St. Vincent Total 95 a/ Based on projections of creditworthiness and policy performance; both are subject to uncertainty. 18. The rationale for differential treatment of the small island countries remains valid, and was recently confirmed by the Small States Task Force, which concluded that small size does constitute a handicap and should be taken into account by aid agencies in setting policies.10 Since these countries account for only a very small share of IDA lending, the impact on overall availability of IDA resources to the poorest countries is not a relevant consideration. Rather, the issue is one of equitable treatment across IDA-eligible countries: the per capita incomes of the four blend countries in the Caribbean now exceed three times the operational cutoff. For the six IDA-only countries, the degree to which per capita income exceeds the operational cutoff is modest and is in line with the deviation at the time the small island exception was introduced. There may, therefore, be a case for differentiated treatment of the relatively better off among the small island countries. One option would be to hasten their graduation from IDA eligibility ­ but this may be difficult to accomplish in the near term, as these countries' access to alternate sources of financing will suffer from the expected decline in air travel and tourism following the 10Small States: Meeting Challenges in the Global Economy, Report of the Commonwealth Secretariat/World Bank Joint Task Force on Small States (April 2000). The report and comments on its conclusions made at the April 2000 meeting of the Development Committee are available at http://www.worldbank.org/smallstates. - 7 - current global slowdown, and security concerns following the September 2001 terrorist attack on the U.S. A second option would be to provide instead for continued access to IDA, but on hardened terms. This would mean introducing a per capita income threshold for small island countries that is higher than the operational cutoff (which guides eligibility for all other countries). Other countries 19. IDA's graduation policies envisage that once a country's per capita income surpasses the operational cutoff ($885), the transition to IBRD-only status would be associated with a progressive hardening of terms of World Bank assistance (IDA plus IBRD) and phasing-out of IDA lending. For reasons described above, IBRD may not be able to increase lending to countries that do not become creditworthy for IBRD at the same time as they begin to graduate from IDA eligibility. As a result, it may be difficult to implement a gradual decline in IDA allocations ­ and IDA assistance continues to be provided on highly concessional terms to countries that are better off than most IDA countries. Since the introduction of the lower operational cutoff in IDA9, IDA's policy framework has provided that "on a limited and exceptional basis, IDA eligibility be permitted for countries with a per capita income between the operational and historical cutoffs that are not creditworthy for IBRD borrowing."11 In line with this provision, FYR Macedonia, for example, received high IDA lending throughout the 1990s, because increased IBRD lending has been constrained by creditworthiness concerns arising in part from political risks in the subregion. Another graduating country, Bolivia, has had per capita income above the operational cutoff since 1997, and though it has acquired blend status as of July 1, 2001, its ability to take on increasing amounts of IBRD debt is constrained while its needs for development financing continue to be substantial. Table 3 shows the countries (other than the small island economies) with exceptional access to IDA during the IDA12 period and those expected to have such access during IDA13. 11Paragraph 41 of Additions to IDA Resources: Twelfth Replenishment, Report of the IDA Deputies, December 23, 1999, available at http://www.worldbank.org/ida/ida12/ida12report.pdf. - 8 - Table 3: Projected exceptional eligibility ­ other countries Basic Indicators IBRD-to-IDA IDA13 Per capita blend ratio Lending income Population FY99-01 projected f/ Projection (US$, 2000) (m, 2000) SDR m Albania a/ 1,100 3.4 n/a n/a 120 Bolivia b/ 1,000 8.3 n/a 26 : 74 91 Bosnia c/ 1,310 3.9-4.3 0 : 100 0 : 100 135 Macedonia, FYR d/ 1,710 2.0 36 : 64 100 : 0 -- Yugoslavia c/ e/ 840-990 10.6 n/a 0 : 100 268 Total 614 a/ IDA-only status. b/ Graduated from IDA-only to blend status as of July 1, 2001. c/ Blend status. d/ Graduated from blend to IBRD-only status as of July 1, 2001. Exceptional IDA access, on hardened terms (20-year maturity), in FY02. e/ Exceptional IDA access, on hardened terms (20-year maturity), till end-IDA13. Population includes Kosovo. Blend ratio does not reflect planned IBRD consolidation loans (est. to amount to $1.7 billion). f/ Based on projections of creditworthiness and policy performance; both are subject to uncertainty. Summary ­ case for differentiating terms of IDA lending to countries with exceptional eligibility 20. Introduction of hardened terms for countries with exceptional access to IDA would improve equity in the distribution of IDA assistance as it would reduce the subsidy accorded through IDA's concessionality to relatively well off countries that do not satisfy IDA's per capita income criterion. Since these are countries which donors wish to support through IDA ­ because these countries have no or limited access to IBRD ­ it would be appropriate to transparently align exceptional access with hardened terms. Consistent with IDA's eligibility and graduation policies, hardened terms could therefore apply to IDA13 lending to: · small island countries with per capita income above a sufficiently high threshold, which could be set at three times the operational cutoff for IDA eligibility ($2,655); and/or · countries (other than the small island economies) with per capita income between the operational cutoff (currently $885) and the historical ceiling for IDA eligibility (currently $1,445).12 12FYR Macedonia and Yugoslavia already have IDA access on an exceptional basis and on hardened IDA terms. For Yugoslavia, exceptional eligibility until the end of IDA13 and modified terms were proposed in Bank Membership and Proposed Financial Assistance Package (R2001-0048), approved by the Board on May 8, 2001. For FYR Macedonia, exceptional access to IDA in FY02 and modified terms were proposed in Transitional Support Strategy (IDA/R2001-0148), and were approved by the Board on September 13, 2001. - 9 - IV. Options for Differentiation of IDA Lending Terms Notional blends with per capita income below the operational cutoff 21. IDA could offer notional blends with per capita income below the operational cutoff hardened, though still highly concessional, terms. Since most of these countries are experiencing medium-term balance of payments difficulties, including liquidity problems, and since it is desirable to front-load IDA's concessionality (because the creditworthiness of these countries is expected to improve over time), maintaining an extended grace period is desirable. It is therefore proposed that hardening would be achieved by shortening the maturity of IDA credits to these countries.13 Hardened IDA terms could comprise an unchanged grace period of ten years, zero interest rate, a 25-year maturity, a standard service charge, and no acceleration clause. These notional blend terms would yield a grant element of 50 percent at the 6.3 percent discount rate and correspond to lending at an IBRD-to-IDA blend ratio of 28:72 (in this case IDA credits on standard blend terms with 35-year maturity). 22. The notional blend terms would still be highly concessional compared to the terms obtained by the blend countries with per capita incomes below the operational cutoff that in fact do borrow from IBRD. For example, in the last three years, the grant element of combined World Bank (IBRD plus IDA) assistance to India was 32 percent and to Indonesia, 21 percent. The proposed notional blend terms would not apply to active blend countries ­ such as India and Indonesia ­ where IBRD and IDA14 can provide assistance on "blended" terms. In particular, it is proposed that notional blend terms apply only to countries with an IBRD-to-IDA blend ratio of less than 30:70, based on planned lending in the current year and actual lending in the preceding two years.15 23. While these hardened terms are likely to result in debt service schedules that are suitable to most of the affected countries' circumstances and overall debt-service capabilities, IDA could, in addition, offer other options, provided that they yield an equivalent grant element (of 50 percent at a discount rate of 6.3 percent). These other options would use combinations of (i) shorter grace period and longer maturity, and (ii) longer maturity and a non-zero interest rate (in addition to the standard service charge). The repayment profiles for credits on notional blend terms are presented in Charts 1 and 2 in the Annex. The case for departing from the standard notional blend terms (10-year grace and 25-year maturity) would be made on the basis of results of creditworthiness analysis. 13In discussing options for hardening IDA terms at their meeting on February 28-March 1, 2001, Deputies did not favor (but also did not reject) introduction of a low interest rate. It is thus proposed to adjust maturity but leave open other options, should this be warranted by specific country circumstances. 14IDA lending on standard blend terms with 35-year maturity. 15Specifics of implementation arrangements have not yet been defined; for example, they could comprise a provision in new credit agreements for deceleration of repayments, should planned IBRD lending to the notional blend country in the current year exceed that needed to obtain an IBRD share of 30 percent of total lending in the three-year period. - 10 - Countries with exceptional eligibility 24. Countries eligible for IDA on an exceptional basis ­ because their per capita income is above the operational cutoff ­ would obtain access to IDA resources on the following terms: an unchanged grace period of ten years, zero interest rate, a 20-year maturity, a standard service charge, and no acceleration clause. These exceptional access terms yield a grant element of 44 percent and correspond to the terms implicit in lending at an IBRD-to-IDA blend ratio of 36:64. Included in the group of countries that would obtain IDA on these terms are countries that are IDA-only (Albania), notional blends (Bosnia, Yugoslavia) as well as active blends (Dominica, Grenada, St. Lucia, St. Vincent and the Grenadines) and potentially active blends (Bolivia); all countries with exceptional eligibility would be subject to the same exceptional access terms, regardless of their status. In the countries that do not yet have creditworthiness for IBRD lending, access to IDA resources, on hardened terms, would help to partially offset lack of access to IBRD lending. 25. For most of the countries in this group, the terms proposed (10-year grace period and 20- year maturity) would not necessarily imply a significant financial impact compared to the standard terms. Rather, it would mean a shift from hardening of IDA terms ex-post (which is already embodied in the acceleration clause in credits to all borrowers) to hardening of IDA terms ex-ante.16 For countries where creditworthiness for IBRD is effectively restored in the short to medium term and that are therefore likely to have the acceleration clause activated soon after approval of their credits, the standard blend terms (with the acceleration clause) would be in fact less concessional ex-post than the proposed exceptional access terms. Introduction of hardening of terms ex-ante would have an important signaling effect regarding the exceptional basis for IDA access for relatively well-off countries. Summary of proposed modifications of IDA terms 26. The proposed hardened IDA terms are summarized in Table 4, along with standard terms to IDA-only and blend borrowers. The last two columns show that there would be no change in payments during the 10-year grace period compared to the standard IDA terms. However, for the countries eligible on an exceptional basis, annual payments in years 11-20 would more than double because of the reduced maturity, to about the same level as payments on an IBRD loan.17 Notional blends would experience an increase in payments in years 11-20 to about halfway between the payments due on an IDA credit on present terms and those on an IBRD loan. The 16When the acceleration clause is activated, repayments double, leading to a shortening of effective maturity as the credit is repaid sooner. From 1987 to August 1996, the acceleration clause provided that the country should have per capita income above the historical ceiling for five consecutive years, and that the minimum grace period be 10 years. For later credits, the trigger is per capita income above the operational cutoff for three consecutive years, and the minimum grace period is five years; please refer to Modification to Accelerated Repayment Terms of IDA Credits (IDA/R96-145), July 23, 1996. The proposed hardening ex ante could result in the exceptional access terms being more concessional ex post, since activation of the acceleration clause at the beginning of the grace period of a credit on standard blend terms (10-year grace and 35-year maturity) would actually yield terms including 5-year grace and 20-year maturity. 17Payments on IBRD loans would be made over 12-15 years, compared to over 10 years for IDA credits, because of the shorter grace period on repayments of IBRD loans. The payment schedule for an IBRD loan is shown in Chart 2 in the Annex. - 11 - payment schedules of credits on the proposed hardened IDA terms are shown in Charts 1-3 in the Annex. Table 4: New options for IDA lending terms Payments on SDR 10 million credit, annual average Interest Service Grace Grant (million SDR) Charge Charge Maturity period element During grace After grace (%) (%) (years) (years) (%) a/ period period Exceptional Access Terms 0 0.75 20 10 44 0.04 1.04 Notional Blend Terms b/ 0 0.75 25 10 50 0.04 0.71 Memo items: Standard Blend Terms 0 0.75 35 10 61 0.04 0.44 IDA-only Terms 0 0.75 40 10 64 0.04 0.38 a/ Grant element is calculated using a discount rate of 6.3 percent. If a discount rate of 10 percent is used instead (this is the method followed by the Development Assistance Committee (DAC) of the OECD), the grant element is 61 percent credits on exceptional access terms and 66 percent for credits on notional blend terms. The grant element can be interpreted as a measure of the financial benefit received by IDA borrowers. b/ Eligible countries could also receive hardened terms using a shorter grace period, longer maturity, and/or small interest charge, as justified by results of creditworthiness analysis (normally presented in a CAS). 27. The difference between the proposed two kinds of hardened terms ­ the notional blend terms and exceptional access terms ­ was determined to take into account equity concerns and the affected countries' expected ability to take on such debt. The exceptional access terms (20- year maturity) are the same as those already approved for FYR Macedonia and Yugoslavia, and they are also nearly equivalent to standard terms with an acceleration clause activated about a decade after credit approval. Since the affected countries already have per capita incomes above the operational cutoff, with continuing growth over a decade and adequate policies, the proposed exceptional access terms seem appropriate for current and possible future cases of this kind. The notional blend terms (25-year maturity) are more concessional than the exceptional access terms, to reflect these borrowers' deeper poverty ­ their per capita incomes are expected to remain below the operational cutoff for some time and hence their time horizon for graduation from IDA eligibility is probably more distant; these terms would require more modest annual payments after the grace period. 28. If the proposals in this paper were accepted, there would be four IDA lending products available to borrowers during IDA13. These are credits on: (i) IDA-only terms; (ii) standard blend terms; (iii) notional blend terms; and (iv) exceptional access terms. At present, there are two standard lending products available (IDA-only and blend terms), along with two country - 12 - examples of hardened exceptional access terms. The introduction of the two new products ­ the notional blend terms and the exceptional access terms ­ would fill two important needs while not unduly complicating the menu of lending products offered by IDA. As is the case now, all IDA credits would remain interest-free and their concessionality would be front-loaded, thanks to a long grace period. 29. Classification of borrowers as having "IDA-only" or "blend" status would continue unchanged. The terms of lending to a specific country at a given time, however, would depend not only on this status. Additional factors determining the terms of credits would be: (i) whether a blend country with per capita income below the operational cutoff has access to adequate levels of IBRD lending (if not, it would be a notional blend); and (ii) whether a country's access to IDA is exceptional because its per capita income exceeds established eligibility thresholds. Table 5 below shows how the four IDA lending products would be offered to different countries, according to these indicators and status. There would be no change in how terms are set for IDA lending to most IDA borrowers ­ the 64 IDA-only countries where per capita incomes are expected to be below established thresholds.18 Table 5: Indicators guiding classification of borrowers and proposed lending terms Per capita income Creditworthiness Below $885 Between $885 and $1,445 (operational cutoff) (historical ceiling) Not creditworthy for IBRD · IDA-only countries: IDA-eligibility on exceptional & standard IDA-only terms limited basis a/ (40-year maturity) · Exceptional access terms · Notional blend countries: (20-year maturity) notional blend terms (25-year maturity) Partially creditworthy for · Blend countries: IDA-eligibility on exceptional & IBRD standard blend terms limited basis a/ (35-year maturity) · Exceptional access terms (20-year maturity) Creditworthy for IBRD · IBRD-only · IBRD-only a/ In addition, 10 small island states would continue to be eligible for IDA (although their per capita incomes are above the operational cutoff); five of these countries now have per capita incomes above the historical ceiling. Exceptional access terms would apply to IDA13 lending to some of these countries ­ only those with per capita incomes substantially above the eligibility threshold, while small island states with lower per capita incomes would continue to receive IDA lending on standard terms. 18During IDA13, there are projected to be 58 IDA-only countries with per capita incomes below the operational cutoff and six small island countries with per capita incomes between the operational cutoff and the proposed higher threshold applicable to small island countries. - 13 - V. Implications of Hardened Lending Terms for IDA Allocations 30. Notional blend countries with per capita incomes below the operational cutoff could note that hardening the terms of IDA lending ­ on the grounds that they can potentially regain access to IBRD ­ would be unfair. It would be equitable to combine the introduction of the hardened, notional blend terms with increased flexibility in IDA allocations. In particular, IDA allocations could be increased by a multiplier (or terms factor) so that the country could obtain a higher level of IDA commitments (on hardened IDA terms), at least partially offsetting its lack of access to IBRD lending. 31. The terms factor, used to augment a country's IDA allocation, could be set in view of the following considerations. First, to ensure that over time IDA resources are conserved for the poorest IDA countries, the factor should not yield a higher grant equivalent in IDA lending to the notional blends than to IDA-only countries with the same level of performance. Thus the factor can be no higher than 1.3 (=64/50).19 Second, the proposed flexibility in allocations should reinforce IDA's strong link between performance and lending. The terms factor should therefore serve to increase allocations to countries with strong performance, as in Table 6. For example, if the three-year allocation on standard blend terms (35-year maturity) for a notional blend country in the second highest performance quintile were $150 million, the allocation on notional blend terms (25-year maturity) would be $173 million (=150 * 1.15). Table 6: Terms factor for allocations of IDA on notional blend terms Performance quintile Lowest Low Median High Highest Terms factor 1.0 1.0 1.0 1.15 1.3 32. IDA allocations to countries with per capita income above the operational cutoff (and with either blend or IDA-only status) would not be more flexible with the introduction of exceptional access terms (10-year grace, 20-year maturity). In many of these countries, however, hardened terms would be important to help justify the length of the period of exceptional access to IDA, for example in the context of the gradual graduation scenarios envisaged for Bolivia and Bosnia- Herzegovina. 19The grant element in standard IDA-only terms (10-year grace and 40-year maturity) is 64 percent. The grant element in the proposed hardened terms (10-year grace and 25 year maturity) is 50 percent. Since for most notional blends allocations are `capped', application of the terms factor in the event of strong policy performance would still result in per capita allocations well below those of IDA-only countries with the same performance rating. The only anticipated IDA13 notional blend country where allocations are not capped is Azerbaijan. - 14 - VI. Effect of Hardened Lending Terms on IDA Finances 33. The introduction of hardened terms through reduction of maturity would have some ramifications for IDA's finances. While the nominal value of the total expected reflows would remain unchanged, the additional present value would increase IDA's overall future lending capacity. For some countries, the introduction of exceptional access terms (10-year grace, 20- year maturity) would have a virtually identical impact as the application of the acceleration clause, which provides that repayments of principal should double once a country's per capita income is above the operational cutoff for three consecutive years and the country is creditworthy for IBRD lending. The advantage for IDA finances in hardening terms ex-ante is to make this impact more certain and not dependent on full creditworthiness for IBRD which arguably sets a higher standard than that necessary to service debt on the proposed exceptional access terms. 34. Current projections of lending on notional blend terms (10-year grace and 25-year maturity) to the countries expected to be eligible amounts to SDR 1.66 billion during IDA13 (see Table 1 above). Subject to Deputies' guidance on the introduction of a terms factor, some of these countries could benefit from augmentation of their IDA allocations, if they achieve sufficiently high country performance ratings, though the policy performance of the countries in this group has been relatively poor to date. (If use of the terms factor to increase allocations is accepted, and the performance of the countries were to improve sufficiently, IDA13 lending on notional blend terms could reach about SDR 2.0 billion.) 35. Projected lending on exceptional access terms (10-year grace and 20-year maturity) is SDR 644 million during IDA13. Among the countries expected to be in this group, Yugoslavia already has access to IDA on hardened terms. Projected IDA13 lending to the other countries (Albania, Bolivia, Bosnia, Dominica, Grenada, St. Lucia, St. Vincent and the Grenadines) is SDR 376 million. Activation of the acceleration clause is most relevant to this group of borrowers but is contingent on the countries being creditworthy for IBRD lending ­ an uncertain prospect for several of these countries. Given this uncertainty, the impact of the acceleration clause has not yet been included in IDA financial projections. 36. Should Deputies agree with the proposals set out in this paper, some SDR 2.03 billion of lending in IDA13 would shift from standard terms to hardened terms: SDR 1.66 billion on notional blend terms and SDR 376 million on exceptional access terms. This level of additional lending on hardened terms, as proposed, would result in an increase in reflows in years 11-20 of about SDR 180 million per annum, which represents an approximately 5 percent increase in reflows. Introduction of hardened terms would thus appreciably help increase the future availability of IDA resources for lending to the poorest countries. - 15 - VII. Conclusions and Issues for Discussion 37. For most borrowing countries, IDA's current, highly concessional, lending terms continue to be appropriate. The range of lending products offered by IDA has not, however, kept pace with a widening range of country circumstances. There are, in particular, two groups of countries for which differentiation of terms may merit consideration during IDA13: blend borrowers that are creditworthy for hardened IDA terms but lack creditworthiness for IBRD for the time being, and countries with exceptional eligibility for IDA. 38. Countries in both of these groups tend to be considerably better off than the vast majority of IDA's borrowers ­ and they also have relatively strong development prospects and substantial credit repayment capacity. Hardening of terms of lending to such better-off countries would allow faster repayment to IDA and thus, over time, increase equity by augmenting the future availability of IDA's resources for the poorest countries. For the notional blend countries with per capita incomes below the operational cutoff, hardened terms could be combined with increased IDA allocations whenever policy performance warrants it. Thus, these countries would be helped by obtaining increased access to financing ­ partly offsetting the temporary lack of access to lending from IBRD (which sets high creditworthiness standards). For countries where per capita incomes exceed established thresholds, hardening of terms would send an appropriate signal about the exceptional basis of such countries' access to IDA. 39. The two new lending products ­ credits on notional blend terms and credits on exceptional access terms ­ would be fully consistent with IDA's graduation policies which embody the principle that IDA lending should be phased out as countries become more creditworthy (and hence can borrow more adequate amounts on non-concessional terms, including from IBRD) or as their per capita incomes rise above established thresholds. In particular: · Credits on notional blend terms constitute a lending product that would suit the needs and reflect the debt-service capacities of countries that are creditworthy for hardened IDA terms but are not, for the time being, creditworthy for IBRD. · The exceptional access terms could be a useful instruments for countries that are graduating from IDA as their per capita income rises above the eligibility threshold, but risk graduating prematurely, before establishing creditworthiness for adequate amounts of financing from other sources, including IBRD. Hardened terms to such countries would help IDA fill a potential void, and do so in a manner that is equitable vis-à-vis the majority of IDA's borrowers. - 16 - 40. Deputies' guidance is sought on introducing hardened IDA lending terms into the IDA13 policy framework, along the following lines: Notional blend countries (per capita income below the operational cutoff) · Should notional blend terms comprising zero interest rate, a service charge of 0.75 percent, 10-year grace period, 25-year maturity, and no acceleration clause be introduced for notional blends? During IDA13 this group of countries is expected to include Azerbaijan, Nigeria, Pakistan, Uzbekistan, and Zimbabwe. · Should the notional blend terms be combined with increased flexibility in IDA allocations for countries with strong policy performance (in the upper two performance quintiles)? Should this flexibility allow for up to 30 percent higher allocations (a terms factor of up to 1.3)? · Should notional blend countries be offered alternate hardened IDA terms, provided that the grant element remains 50 percent? This would allow IDA to respond to countries with atypical debt service profiles and hence enhance these countries' ability to service debt over time. Countries eligible for IDA access on exceptional basis (per capita income above the operational cutoff) · Should countries with exceptional access to IDA (because their income exceeds the poverty threshold) receive hardened terms comprising zero interest rate, a service charge of 0.75 percent, 10-year grace period, 20-year maturity, and no acceleration clause? In particular, should exceptional access terms apply to IDA lending to either or both of the following groups: (a) countries with per capita incomes between the operational cutoff and the historical ceiling for IDA eligibility (other than small islands)? During IDA13 this group of countries is expected to include Albania, Bolivia, and Bosnia (as well as Yugoslavia, which is already subject to hardened terms). Do Deputies agree that flexibility through the use of a terms factor is not appropriate for use in determining IDA allocations to these countries? (b) countries that have remained eligible for IDA under the small island country exception and where per capita incomes are significantly above the operational cutoff? If so, would a threshold of three-times the operational cutoff (currently $2,655) be appropriate? During IDA13 this group of countries is expected to include Dominica, Grenada, St. Lucia, St. Vincent and the Grenadines.20 20The other small island countries (with per capita incomes between the operational cutoff and the threshold) would continue to benefit from access to IDA on standard IDA terms. During IDA13, this group of countries is expected to include Cape Verde, Kiribati, Maldives, Samoa, Tonga, and Vanuatu. - 17 - Annex Page 1 of 2 Chart 1. Notional blend terms $ m 8 on 7 Proposed notional blend terms: 6 charges) 10-yr grace, 25-yr maturity loan/credit 5 interest, 4 equivalent 3 Combined program: 70% standard IDA blend terms (principal, million 2 and 30% IBRD $100 1 Payment 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 year Chart 2. Notional blend terms - alternatives with identical 50% grant element $ m 8 10-yr grace, 25-yr maturity on 7 6 charges) credit 5 Shorter grace,longer maturity: interest, equivalent 4 8-yr grace, 28-yr maturity Small interest rate: 3 1% p.a. (10-yr grace, million (principal, 2 35 -yr maturity) $100 1 Payment 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 year - 18 - Annex Page 2 of 2 Chart 3. - Exceptional access terms $ m Exceptional access terms: 12 10-yr grace, 20-yr maturity IBRD terms for on low-income countries 10 charges) loan/credit 8 interest, 6 equivalent Accelerated standard blend terms: 4 5-yr grace, 20-yr maturity (principal, million 2 $100 Payment 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 year