Fund Concessional Financial Support for Low-income Countries—Responding to the Pandemic

International Monetary Fund

doi:10.5089/9781513589633.007.A001

Author:

International Monetary Fund

International Monetary Fund
Search for other papers by International Monetary Fund in
Current site
Google Scholar
PubMed

Publication Date:: 22 Jul 2021

eISBN:: 9781513589633

ISBN:: 9781513589633

Language:: English

Keywords:: financial support; PRGT credit; EA criteria; lending capacity; investment strategy; COVID-19; Income

Download PDF (927.5 KB)

1. The IMF’s current framework for providing financial support to Low-Income Countries (LICs) came into effect in January 2010.1 Access to the Fund’s concessional facilities, financed via the Poverty Reduction and Growth Trust (PRGT), is available to countries that are assessed to be PRGT-eligible by the Fund’s Executive Board.2 Modifications to several aspects of the framework (such as limits on access) were made at various points between 2013 and 2020, but the basic structure of the framework has remained substantially unchanged over time.

Abstract

I. Introduction

1. The IMF’s current framework for providing financial support to Low-Income Countries (LICs) came into effect in January 2010.^¹ Access to the Fund’s concessional facilities, financed via the Poverty Reduction and Growth Trust (PRGT), is available to countries that are assessed to be PRGT-eligible by the Fund’s Executive Board.^² Modifications to several aspects of the framework (such as limits on access) were made at various points between 2013 and 2020, but the basic structure of the framework has remained substantially unchanged over time.

2. A comprehensive review of the PRGT’s lending framework and its financial underpinning was completed in May 2019. At that time, the Board approved a one-third increase in the overall limits on access (expressed as a share of quota) to PRGT resources and a similar increase in the limits on access to the Fund’s emergency financing (EF) instruments.^³ Other reforms introduced included changes to the blending rules, increases in the maximum length of ECF (Extended Credit Facility) and SCF (Standby-Credit Facility) lending arrangements, and targeted increases in access to financing under EF instruments to accommodate the special circumstances of fragile/conflict-affected states (FCS) and countries vulnerable to large natural disasters. Staff assessed that the reform package would be generally consistent with the self-sustained PRGT financing framework, with risks evenly balanced over the coming decade. Staff noted that the evolution of lending capacity would need to be monitored carefully, and policies reviewed periodically to ensure that lending capacity remained in line with the PRGT’s base envelope for annual lending of SDR 1¼ billion on average on a long-term self-sustained basis.

3. The COVID pandemic has hit LICs hard, with many experiencing significant declines in per capita income that will be not be speedily reversed. The scope for providing policy support for economic recovery was tightly constrained in most LICs by limited fiscal space and binding financing constraints, exacerbated in several cases by high pre-crisis debt levels. With very low vaccination rates, the outlook for recovery is subject to significant downside risks, with staff projections pointing to a significant divergence in economic performance between LICs and higher income countries, with the former falling behind unless there is a broad-based international effort to accelerate recovery and boost development efforts over the next 3–4 years.^⁴

4. The Fund responded to the pandemic with a series of temporary increases to access limits that facilitated an unprecedented surge in IMF emergency lending to both LICs and emerging market economies (EMs). PRGT disbursements to LICs during 2020 amounted to SDR 6.8 billion—compared with an annual average of less than SDR 1 billion during 2017–2019—while total new lending to LICs, including non-concessional loans from the Fund’s General Resources Account (GRA) as part of blended financing, reached SDR 9.3 billion.^⁵ In all, 53 of 69 PRGT-eligible countries received financial assistance from the Fund in 2020, including via debt service relief from the Catastrophe Containment and Relief Trust (CCRT). This surge in financial support helped LICs cope with the initial shock of the pandemic and ensuing slowdown in global economic activity.

5. Looking ahead, most LICs face severe economic challenges and balance of payments (BoP) needs as they seek to recover from the pandemic shock and resume progress towards their development objectives. The majority are expected to seek multiyear program support with substantial financial assistance from the Fund to help tackle their BoP difficulties. This will come on top of significant levels of pre-existing debt to the Fund, incurred either in tackling economic difficulties prior to the pandemic or as an emergency response to the pandemic.^⁶ This is a very different environment from the circumstances prevailing prior to the pandemic, calling for adjustments to the lending policy framework to allow the Fund respond to the exceptional needs of its poorest members with customized economic programs and financial support.

6. The Fund’s concessional financing through the PRGT is designed to be self-sustaining, with an endowment that generates sufficient investment returns to subsidize lending at zero or near-zero interest rates.^⁷ While the PRGT was assessed to be adequately financed to meet LICs’ future borrowing needs prior to the pandemic, the large surge in lending levels since the onset of the pandemic and the expectation of continued high lending levels in the near term means the PRGT is now significantly underfunded. The endowment will thus need a substantial injection of fresh funds if it is to sustain reasonable levels of lending to LICs over the medium-to-longer term (see Section IV for a detailed analysis).

7. This paper lays out a package of policy reforms and a funding strategy to ensure that the Fund has the capacity to respond to LICs’ exceptional needs during the pandemic and recovery. Specifically, the paper proposes: i) enhancements to the lending framework that would make it fit-for-purpose to help meet LIC financing needs through the pandemic crisis and recovery period and ii) a funding strategy that would rebuild the PRGT’s finances to ensure sustainability of the Fund’s concessional lending over the longer-term. The focus of attention in the paper is on the next 3–4 years, with the expectation being that a comprehensive review of concessional financing and policies will be undertaken in 2024–25—by which time the exceptional uncertainty surrounding projections of Fund lending and of LIC financing needs should have abated.

8. The remainder of the paper is organized as follows. Section II reviews the evolution of concessional financing levels and policies since the onset of the pandemic and looks at prospects for demand for Fund financial support from LICs through end-2021. Section III discusses potential modifications to the PRGT lending architecture to help meet LIC financing needs during this period, recognizing that there are many other official sector vehicles for providing such support, and outlines an integrated package of proposed reforms. Sections IV and V provide detailed projections of demand for Fund financing from LICs through 2024 under alternative scenarios and the associated PRGT resource needs and outline a proposed two-stage funding strategy for addressing these needs. Section VI discusses proposed interest rates for PRGT lending (which are reviewed every two years); Sections VII and VIII discuss developments in regard to the financing of debt relief provided by the Fund and enterprise risks. Finally, Section IX identifies key issues for discussion. Proposed decisions to implement staff proposals will be circulated separately in a supplement.

II. IMF Lending to Lics: Recent Experience

9. In the years immediately prior to the pandemic (2017–19), lending to LICs averaged some SDR 1.1 billion per annum, with concessional (PRGT) loans accounting for about five-sixths of the total. Access levels and annual disbursements had been gradually increasing, notably in 2019, but repayments on prior loans—consistent with the revolving nature of Fund lending—meant that the (SDR) stock of credit outstanding was increasing at a modest 3 percent per annum (1 percent in the case of PRGT credit). Ninety percent of the new financing was provided via Fund arrangements, with the remainder coming from emergency financing via the RCF (four requests, one also involving financing from the RFI) during these three years.^⁸

10. The onset of the COVID-19 pandemic saw LIC demand for Fund financial support grow dramatically, with the bulk of new financing being provided via the EF instruments (Figure 1): lending to LICs during March–December 2020 amounted to SDR 9.1 billion (SDR 6.56 billion from the PRGT), of which almost 90 percent was provided via the RCF and RFI. The surge in emergency financing was facilitated by the temporary doubling of annual limits on access to the EF facilities on April 6, 2020.^⁹ By end-year, 47 LICs had received financial support through these facilities (with 6 countries receiving two EF disbursements), with 3 more countries receiving financial support through disbursements under new or augmented UCT programs. The stock of outstanding credit to LICs at end-2020 amounted to SDR 15.7 billion (SDR 12.4 billion from the PRGT), twice the amount outstanding at end-2019.

11. The pattern of Fund lending to LICs has evolved over the course of the pandemic period, both in terms of volumes and use of facilities (Figure 2):

New lending to LICs in March-July 2020 surged to SDR 7.97 billion (PRGT plus GRA), with 90 percent of this amount delivered via 47 disbursements to 44 countries under the EF instruments; the remainder was provided through 2 new and 7 pre-existing UCT programs.^¹⁰
Following this initial unprecedented spike, the pace of lending slowed but remained elevated relative to pre-crisis levels. New lending provided during August–December amounted to SDR 1.1 billion, split almost equally between 6 disbursements under the EF instruments and 10disbursements under UCT programs (including 1 new arrangement, with Afghanistan). Lending to LICs in this 5-month period was similar in scale to average annual lending levels in 2017–19.
During the first six months of 2021, new lending amounted to SDR 2.51 billion, with SDR 2.1 billion of this coming from disbursements under UCT programs, including five new arrangements with Kenya, Madagascar, Senegal, Uganda and Sudan, and the remainder from five disbursements under the EF instruments (three involving countries that had already received EF disbursements in 2020).

12. The pick-up in lending is set to continue in the remainder of 2021, reflecting new program requests coming to the Board in the coming months (see below) and disbursements on some of the SDR 5.14 billion in undrawn commitments under the 15 arrangements currently in place. Discussions on new arrangements are at an advanced stage with 8 countries,^¹¹ with 2 of these expected to come to the Board by end-July: all cases involve arrangements of around 3 years in length, with proposed access ranging from 86 percent of quota to 280 percent of quota and an average access level of 146 percent of quota.^¹² Active discussions are also well underway in several other cases, albeit with timing for completion being less predictable given the need to reach understandings with creditors on a debt restructuring framework. Discussions on further UCT-quality programs will begin in some months’ time. Staff also anticipate additional requests for emergency financing through end-2021, including for pandemic-related financing needs.^¹³

13. The Fund has complemented scaled up emergency financing with broad-based debt flow relief to its poorest members through the CCRT. This involved a reform of the CCRT to allow 29 countries to qualify and immediately receive grants to meet debt service payments falling due to the Fund. Following modifications to the CCRT approved by the Board in late-March 2020, the 29 eligible countries have received grant support in the amount of SDR 520 million, covering all debt service payments to the IMF falling due from mid-April 2020 through mid-October 2021. A further SDR 160 million in grants, covering debt service from mid-October 2021 through mid-April 2022, is expected to be made available if there are sufficient resources in the Trust. Donors have disbursed or pledged some SDR 575 million to the CCRT since March 2020. The Fund also mobilized large internal and donor resources to support arrears clearance and debt relief in Somalia and Sudan.

14. An expected general SDR allocation of $650 billion would provide further support to LICs. The allocation, which is expected to be approved by the Fund’s Board of Governors in August 2021, would make available potential financing of $21 billion (about SDR 14.6 billion) to LICs.^¹⁴ Countries may decide to use the allocation to augment international reserves or for other purposes, including to meet budgetary financing needs or support debt management operations. As a supplement to this support, IMF members with strong economic positions are considering options for channeling a portion of their SDR holdings to finance additional low-cost loans to other members, including LICs.

III. Modifying PRGT Facilities and Lending Safeguards

15. The basic architecture of the PRGT facilities worked broadly as intended during the 2010–2019 period and has served member countries well during the pandemic. Given the exceptional conditions prevailing during much of 2020, the generalized shift from providing financial support to LICs via arrangements under the PRGT to support from the EF instruments allowed for a pragmatic and speedy response to the onset of the pandemic. With the initial shock largely passed, this shift is being reversed, with multiyear arrangements supported under the ECF set to be the main vehicle for providing financial support over the next three–four years.

16. With LICs now facing severe financing challenges, the limits on access to PRGT resources are likely to push many LICs with sound economic programs into seeking support through the Fund’s non-concessional lending facilities, financed from the IMF’s GRA. While LICs have the right to access GRA resources under the same conditions as any other IMF member, such loans come with higher floating interest rates, shorter maturities, and policy requirements not tailored to resolution of protracted balance of payments problems as is the case with loans extended under the ECF. Given that PRGT access limits are set to become binding for a significant number of LICs, reliance on additional GRA financing is likely to become more common over time.

17. Expanding lending to countries that already have significant outstanding debt to the Fund comes with some risks, although the policies in place to manage credit risk are important mitigating factors. Access to additional low-cost (zero-interest) financing from the PRGT provides clear benefits to borrowers but also adds to the already significant levels of non-restructurable debt that the country owes.^¹⁵ Should the country face a situation in which it cannot service its external debts, the space for resolving debt problems via a negotiated debt restructuring is correspondingly reduced and the risk of running arrears to the Fund (or the World Bank)—usually a very difficult situation for countries to emerge from—is heightened. Providing high levels of Fund credit to a country thus needs to be linked to careful scrutiny of public debt sustainability over the medium-term.^¹⁶

18. The analysis in this section is divided into two parts, the first focusing on proposals to adjust limits on norms and access to PRGT resources and associated safeguard measures, the second on proposals to simplify blending polices (which affect how higher-income LICs can access the PRGT). As explained in Annex III, non-blend countries can currently access PRGT resources up to the relevant limits without seeking access to the GRA; blend countries can access PRGT resources only in conjunction with access to GRA resources.^¹⁷ Countries are expected to shift from non-blend to blend status as income per capita levels rise and/or access to international financial markets increases, and from blend status to graduation from PRGT eligibility as rising income levels and/or expanded market access reach threshold levels.^¹⁸

Proposed Changes to PRGT Access Limits and Associated Safeguards

19. The limits on normal access to PRGT resources (100/300 percent of quota) were set at the conclusion of the 2018–19 Review of Facilities for LICs.^¹⁹ These limits were expected to remain in place through the next regular review of LIC facilities and financing on a five-year cycle, but have been increased on a temporary basis during the pandemic, with the current limits (245/435 percent of quota) set to expire in the coming weeks. Separately, the subset of non-blend countries that are eligible for EA may request up to 33/100 percent of quota in additional PRGT funding if the program request meets the relevant EA criteria.

20. There is a strong case for raising limits on access to PRGT resources given the large projected external financing needs of many LICs in the coming years. This would provide LICs with greater access to low-cost-financing from the Fund and would limit the number of non-blend LICs that are required to access the GRA. There are a substantial number of countries with credit outstanding to the PRGT in excess of 150 percent of quota, many of whom can be expected to request sizeable UCT-quality arrangements as they seek to navigate their way through the pandemic period and its aftermath (see Text Table 1). Absent changes to the PRGT access limits, many of these countries, among the poorest of the Fund’s LIC members, would likely be required to request financing from GRA instruments, which would be less appropriate for their needs.

^{Text Table 1.}

PRGT Credit Outstanding as Percent of Quota^* As of June 16, 2021)

^{^*}Excludes blend countries.

^{Text Table 1.}

PRGT Credit Outstanding as Percent of Quota^* As of June 16, 2021)

Country	% of Quota
Chad	295%
Madagascar	241%
Mali	223%
Burkina Faso	213%
Malawi	211%
Niger	208%
Central African Republic	192%
Mauritania	192%
Sierra Leone	180%
Grenada	177%
Rwanda	169%
Togo	169%
Guinea	160%
Mozambique	158%

^{^*}Excludes blend countries.

^{Text Table 1.}

PRGT Credit Outstanding as Percent of Quota^* As of June 16, 2021)

Country	% of Quota
Chad	295%
Madagascar	241%
Mali	223%
Burkina Faso	213%
Malawi	211%
Niger	208%
Central African Republic	192%
Mauritania	192%
Sierra Leone	180%
Grenada	177%
Rwanda	169%
Togo	169%
Guinea	160%
Mozambique	158%

^{^*}Excludes blend countries.

21. Staff propose the following set of principles to guide the setting of access limits:

Non-blend countries seeking access to PRGT resources at levels viewed as “normal access” for other member countries should be able to borrow the entire amount from the PRGT. Normal access here is access that would not trigger application of the GRA EA framework or Policy Safeguards for High Combined Credit to GRA-PRGT Resources (PS-HCC).
Poorer LICs are eligible for access to PRGT resources in excess of normal access limits: this access should no longer be subject to hard caps, but the proposed arrangements must meet the PRGT EA criteria.
LICs that meet the income criterion for blending, which is based on the International Development Association (IDA) operational cutoff, should not be eligible for access to PRGT resources above the normal access limits.

22. These principles entail some important departures from current concessional lending rules, where i) access limits are set at levels that constrain demand for PRGT resources to align with the available PRGT resource envelope and ii) there are hard caps on all countries’ access to concessional resources.^²⁰ They also involve significant continuities: only poorer LICs are eligible for EA, while access levels for all financing requests continue to be based on case-by-case assessment of program design and country circumstances, applying long-established criteria (see ¶28 below).

23. The financial cost to the PRGT of raising the access limits in this manner is expected to be modest. While higher access limits create more room for PRGT borrowing, this does not necessarily translate into higher borrowing in individual cases; access will continue to be determined on the basis of case-by-case assessment, taking account of such factors as BoP needs, the strength of the program, and capacity to repay/debt vulnerabilities. Using the Baseline projections for Fund lending through 2024 (see ⁋51 below), returning to the 2019 access limits and caps after the temporary increases now in place expire would affect only about a dozen countries and reduce PRGT lending by about SDR 1½ billion through 2024—which can be contrasted with a Baseline projection of PRGT lending of SDR 21.5 billion during 2020–24. For the countries affected in such a scenario, those with programs that meet the policy requirements of the GRA (see Annex 1) would be in a position to replace the reduced PRGT lending with GRA resources.

24. Applying these principles under current circumstances yields the proposals on access limits outlined in Box 1: these proposals take account of the temporary access limit increases currently in place. Going forward, non-transitory changes to PRGT access limits would be taken up in the context of the regular reviews of LIC facilities, given the need to align the entire package of PRGT policies with financial self-sustainability of the PRGT over time. Access limits (which are expressed as a share of quota) would also be revisited in the context of any general quota increase.

^{Box 1.}

PRGT Access Limits—Staff Proposals

Background

The GRA exceptional access (EA) framework applies to all requests for financing from the GRA that exceed threshold levels for annual and cumulative access. The threshold levels were set at 145/435 percent of quota in February 2016.^¹ The threshold level for annual access was increased on a temporary basis to 245 percent of quota in July 2020: this temporary increase is set to expire at end-2021, when the threshold would revert to 145 percent.^²

Policy safeguards for countries seeking financial support from the Fund involving high levels of combined GRA-PRGT exposure were introduced in September 2020.^³ The threshold levels for combined exposure that trigger application of these safeguards are the same as those that trigger application of the EA framework in the GRA—currently 245/435 percent of quota, with the annual threshold set to decline to 145 percent of quota at end-December 2021.

Normal access to PRGT resources

Staff propose that the limits on normal access to the PRGT be set at 145/435 percent of quota, with the limit on annual access temporarily increased to 245 percent through end-December 2021. The limits on normal access to the PRGT were temporarily increased from 100/300 to 245/435 percent of quota in March 2021, with the current limits (245/435 percent of quota) set to expire in the coming weeks. The staff proposal implies that:

the limit on normal cumulative access would (absent a new Board decision) continue at 435 percent of quota until the next full review of LIC facilities, expected to be conducted in 2024–25;
the limit on normal annual access would (absent a new Board decision) continue at 245 percent of quota until end-2021, after which it would decline to 145 percent of quota until the next full review of LIC facilities.
changes to the thresholds triggering application of the EA framework in the GRA or the Policy Safeguards for High Combined Credit (PS-HCC) before the next full review of LIC facilities would not affect PRGT access limits ahead of that review.

The proposed increases in access limits would apply to new financing requests and to existing arrangements as of the date of the effectiveness of the proposed changes, with the exception of arrangements that were grandfathered when the PS-HCC policy was adopted, which will remain subject to the PRGT EA thresholds and criteria in place at the time of the approval of these arrangements. This is to ensure adequate safeguards and even-handed treatment for exceptional/high access across all PRGT borrowers. In the event of an augmentation under such an arrangement, the grandfathering from the application of the PS-HCC policy would end and the new PRGT access rules would apply.^⁴

Exceptional access to PRGT resources

LICs that meet the relevant eligibility criteria can request EA (access above the normal limits) to PRGT resources (see Annex II). Eligibility is currently limited to countries that have not had sustained past access to international financial markets^⁵ and have income at or below the prevailing operational cutoff for assistance from IDA. Available access is currently subject to a hard cap of 33.3/100 percent of quota on top of the normal access limits. The staff proposal is that:

A country is eligible for PRGT EA only if it does not meet the proposed income threshold for blending (discussed below). It will not be disqualified from EA on the basis of market access.
Access to PRGT resources for countries eligible for EA is not subject to hard caps; a financing request may be approved in amounts exceeding the normal access limits if the PRGT EA criteria are satisfied.

^¹ The thresholds determining EA under the GRA were set in February 2016: see “Review of Access Limits and Surcharge Policies” (IMF (2016a), March 29, 2016). ^² All temporary changes to access limits introduced since the beginning of the pandemic are expected to be reviewed after the 2021 Annual Meetings. ^³ See IMF (2020e). ^⁴ Only one existing arrangement (with Ethiopia) is subject to the current PRGT EA criteria and was grandfathered from the application of the PS-HCC in September 2020. ^⁵ Countries with per capita GNI below 80 percent of the IDA operational cutoff are not precluded from EA on the basis of market access.

Norms

25. Access norms have been a feature of PRGT facilities design since the overhaul of the LIC facilities architecture in 2010. Norms have played a nuanced role in influencing access levels in PRGT arrangements: they are neither ceilings on nor floors to access and should not be viewed as an entitlement, but play a useful guiding role in setting access in cases where it is difficult to accurately determine the BoP need.^²¹ This role has been particularly useful in the context of repeated arrangements with LICs making steady progress in addressing protracted BoP problems, but less relevant in cases where countries face pressing BoP needs triggered by shocks or crises.

26. Under the current system, access norms are linked to the initial stock of credit outstanding: 120 percent of quota for a 3-year ECF when credit outstanding is below 100 percent of quota, 75 percent of quota when credit is between 100 and 200 percent of quota, and undefined if outstanding credit exceeds 200 percent of quota. Past practice in reviews has been to increase these parameters (both the access level and associated credit ranges) in line with any increase in normal PRGT access limits.

27. Staff proposes a simplification of the access norms, with a unified access norm set at 145 percent of quota for any three-year ECF arrangement.^²² The basis for differentiated norms— to tilt the use of scarce resources towards countries that have made less use of them—does not align well with current circumstances, where the pandemic has created large and diverse financing needs across countries: the current priority is to tackle the most pressing needs, rather than allocating resources on the basis of pre-existing exposure (which, in many countries, is elevated because of pandemic-related emergency financing). Staff also sees simplification as strengthening the signaling function of norms, providing a clear uniform starting point (but not end-point) for discussion of appropriate access levels in PRGT-supported programs. Setting the unified norm at 145 percent of quota is broadly in line with the proposed increase in normal limits on access to the PRGT.^²³ The role of norms in PRGT facilities will be reassessed at the time of the next comprehensive review of LIC facilities in 2024/25.

Safeguards

28. The Fund relies on a multilayered framework to mitigate and manage credit risk across all its lending operations, including program design and access policies. The approach to setting access under an arrangement (PRGT or GRA) entails a case-by-case determination, where access has to be justified on the basis of a rigorous assessment informed by the standard access policy criteria, including the size of the balance of payments need, the strength of the member’s economic program, capacity to repay the Fund, and the track record of using Fund credit in the past. This case-by-case approach will remain critical to ensuring adequate safeguards for PRGT resources.

29. As noted above, Fund programs that result in high levels of debt for member countries involve risks for both the borrowing country and the Fund, implying a need for careful assessment of the sustainability of the country’s debt position over time. In recognition of this issue, new measures were introduced in March 2021 to enhance staff’s assessment of debt sustainability in LICs.^²⁴ These included: i) disclosure requirements and deeper analysis of debt composition and dynamics in staff reports, building on the disclosure requirements in the newly-modified Debt Limits Policy (DLP), which takes effect on June 30, 2021, and including cross-country comparisons of outstanding and projected Fund credit relative to key economic metrics, and ii) a requirement, in all cases of countries at high risk of/in debt distress (regardless of proposed access levels), that program objectives include the achievement of a concrete reduction in debt vulnerabilities over the course of the program and beyond. See Annex VI for elaboration on these points.

Procedural Safeguards

30. The PRGT lending framework has high access procedural safeguards, which are applied when a financing request entails proposed access such that i) access to PRGT resources over any 36-month period would exceed 180 percent of quota (“flow trigger”), or ii) aggregate exposure to the PRGT, net of repayments, would exceed 225 percent (“stock trigger”). The objective of these procedures is to ensure enhanced Board oversight of lending proposals involving high levels of access to PRGT resources, achieved via an early informal staff consultation with the Executive Board on a country case that should occur once management agrees that a new or augmented financing request involving high access could be appropriate. The information and process requirements for these consultations, which were upgraded in May 2019, are described in Box 2.

31. Given the impact of the wave of pandemic-linked lending in 2020, these triggers were modified on March 22, 2021 as follows:^²⁵

With the majority of non-blend countries having obtained exceptional emergency financing in the first months of the pandemic, the flow trigger was set at 240 percent of quota through end-2023, by which time all the financing provided in response to the initial shock would have dropped out of the “36 month” calculation.
With many countries already having exposure to the PRGT close to (or above) 225 percent of quota, the stock trigger was increased to 300 percent of quota through June 30, 2021 to avoid triggering the procedural safeguards in cases involving modest new access.

32. It is proposed that these flow and stock triggers remain at the higher levels introduced in March 2021 until the next full review of LIC facilities.^²⁶ With average program size in the coming years projected to be substantially higher than pre-COVID levels (see below), the case for returning to the pre-COVID flow trigger after end-2023 is not compelling. Similarly, with exposure to the PRGT expected to increase for the majority of borrowers over the next few years, there is a strong case for maintaining the stock trigger at 300 percent to avoid triggering procedural safeguards in cases involving modest levels of new access.

^{Box 2.}

PRGT Facilities Framework: High Access Procedures

The PRGT lending framework includes procedural safeguards (’high access procedures’ or HAP) for new financing requests or augmentations involving access to concessional resources above specified levels. The HAP were introduced in 2009, when the PRGT Facilities Framework was established, with the aim of protecting PRGT resources via enhanced Board oversight of lending requests involving high access. Against the backdrop of rising debt vulnerabilities in many LICs, the 2018–19 Review of LIC Facilities (LIC FR) modified these procedures to enhance the focus on assessment of debt vulnerabilities and related risks to members’ capacity to repay the Fund.

All requests for new PRGT financing where the proposed access meets the HAP thresholds require early engagement with the Board through an informal Board meeting. Since May 2019, this engagement has been expected for all financing requests where (a) access under the ECF, SCF, and RCF exceeds 180 percent of quota over any 36-month period (“flow trigger”), or (b) outstanding credit under all concessional facilities is above (or projected to be above) 225 percent of quota (“stock trigger”), based on cumulative access for past and future scheduled disbursements net of repayments.^¹

In such cases, the Board would be presented with an initial assessment of the member’s BoP need, macroeconomic situation, and potential fiscal and debt vulnerabilities, as well as information on the proposed program and related impact on concessional resources. Specific informational requirements (which are the same as those required for EA under the PRGT) would include:

Factors underlying the large BoP need, after accounting for financing from donors.
A brief summary of the main policy measures and macroeconomic framework.
The expected strength of the program and an assessment of the capacity to repay the Fund, including an updated capacity-to-repay table.
An analysis of debt vulnerabilities, including the identification of potential data weaknesses and discussion of results from “realism” tools included in the LIC DSF.
A reference to the impact on the Fund’s concessional resources.
The likely timetable for discussion with authorities.
An SEI table.
DSA charts.

To ensure Directors’ views on access levels are appropriately reflected in the negotiations, the informal HA Board meetings should take place as soon as management concurs that a new request involving HA could be appropriate.

^¹ The thresholds that trigger the HAP were temporarily raised in response to the pandemic, as discussed in the main text.

Other Safeguards: Alignment of PRGT EA Criteria with Existing Policy Safeguards

33. There are currently two separate sets of standards that may apply to LICs requesting levels of access to Fund resources in excess of the proposed 145/435 percent of quota limits: the PRGT EA criteria and the Policy Safeguards for High Combined Credit (PS-HCC) relating to requests for high combined access to GRA and PRGT resources. The two are closely related but not identical; the PS-HCC conditions (introduced in September 2020) sought to adapt the standards specified in the GRA EA criteria to accommodate the distinctive features of LICs and of the LIC debt sustainability framework, while the PRGT EA criteria (introduced in 2009) focused on limiting EA to poorer LICs with comparatively strong adjustment programs, with programs with countries at high risk of (or in) debt distress expected to include a debt restructuring operation.^²⁷

34. To simplify the policy framework governing EA requests involving PRGT resources, it is proposed that the PRGT EA criteria be modified to align them with the PS-HCC requirements specified in IMF (2020e), with one significant exception. As noted above, countries would be eligible for PRGT EA only if they do not meet the income threshold for blending at the time when a new financing request is made: in all other respects, the PRGT EA criteria and the PS-HCC would be substantively similar, as illustrated in Annex II. Staff does not see this proposed change in the PRGT EA criteria as being material from a risk perspective: it eliminates the formal requirement that, for countries at high risk of/in debt distress, EA should be made available only in support of programs that include debt relief or debt restructuring operations, but the more fundamental requirement that the program achieve moderate risk of debt distress within the program period remains.

Proposed Changes to Blending Policies

35. PRGT-eligible countries are divided into two groups: i) blend countries, who can access concessional financing from the Fund only in conjunction with GRA resources and ii) non-blend countries, who can access PRGT resources up to the relevant access limits, needing to tap GRA resources only if their financing requests exceed these limits. Moving from non-blend to blend status has significant implications for a LIC, with Fund financial support now provided on less concessional terms and with programs required to meet the policy requirements of the GRA as well as the PRGT.

36. Blend status is determined by GNI per capita, access to international financial markets, and severity of debt vulnerabilities. In broad terms, LICs with GNI per capita above the IDA operational cutoff (currently $1,185) or with significant access to international financial markets and income above 80 percent of the IDA cutoff are required to blend if debt vulnerabilities are assessed to be contained (low or moderate risk of debt distress); countries assessed to be at high risk of debt distress that meet both the income and market access criteria (including market access on a forward-looking basis) are also required to blend; all other countries are not required to blend. These rules are described in detail in Annex III.

37. Staff sees a strong case for targeted reforms to the blending framework to make it both more robust and less complex:

Robustness of blending status: Countries can flip back and forth too easily between blend and non-blend status, in a manner inconsistent with the important implications of the shift to blend status for a country and that can also create operational difficulties.
- Income threshold: A low/moderate debt risk country can move to blend status following a modest increase in GNI per capita to above the IDA cutoff, and then revert back to non-blend status with a modest decline in GNI per capita (or an upward shift in the IDA cutoff, which moves modestly from year to year).^²⁸ A request for an arrangement in year one would have to be a blend; a request the following year could involve PRGT access only.
- Market access threshold: Countries at high risk of debt distress with income above the IDA cutoff and substantial past market access are expected to blend if they are assessed to have prospective market access. Assessing prospective market access requires judgment at the time of a financing request; this will move in line with significant shifts in market sentiment. Thus, Kenya was viewed as a non-blend case when it asked for EF in May 2020 but as a blend case when it requested Fund financing in April 2021.^²⁹
Complexity: The blending rules have become quite complex over time, impairing the visibility of key principles in some areas. For example, access to PRGT resources in a blended arrangement is guided by the 1:2 ratio on the PRGT-GRA mix (1:1 prior to July 2015)—but this is capped at the applicable norm for an arrangement, which can be i) 120 percent of quota, ii) 75 percent, or iii) undefined, depending on the initial level of PRGT credit outstanding (see Box 3, footnote 3). A simpler rule would allow full operation of the 1:2 principle in situations of normal access.

38. Reforms of the blending rules are proposed in three areas: a) the income criterion for blending, b) the impact of debt vulnerabilities and financial market access on blend status, and c) the mix of GRA and PRGT resources applied in blended arrangements.

39. The income criterion for blending would operate as follows:

Countries are deemed to meet the income threshold for blending when GNI per capita has exceeded the IDA operational cutoff by at least 5 percent for two consecutive years.
Having met the income threshold, the country continues to meet it provided that income per capita does not fall below 95 percent of the IDA operational cutoff: should income per capita fall below this level, the country no longer meets the income threshold.
Countries that do not meet the income threshold for blending are not required to blend, irrespective of market access.

The case for these specific parameter choices and the implications of applying these rules for the coming year (through end-June 2022) are discussed in Annex III.

40. This approach would substantially reduce the likelihood of a premature/soon-reversed shift to blend status, by raising the standard for meeting the income criterion and reducing the income level at which blend status, once achieved, would be lost. This asymmetry between the income threshold for achieving blending status and the threshold for re-entry to non-blend status also features in the income criterion for PRGT-eligibility and for similar reasons—to support robust (but not fully irreversible) graduation from PRGT eligibility. Specifying that countries who do not meet the income criterion are not required to blend simplifies the framework by eliminating cases where countries with income below the IDA cutoff must blend if they meet a market access criterion: the only country affected by this change is Tanzania, which would no longer be required to blend.^³⁰

41. Debt vulnerabilities and access to international financial markets would influence blend status as follows:

Countries that meet the income criterion for blending are required to blend unless debt vulnerabilities limit their access to international financial markets.
Countries are deemed to face such limits on their access to international financial markets if they are i) in debt distress or ii) at high risk of debt distress and a) have had limited past access to international financial markets or b) are small/micro-states.
Countries are assessed to have had limited past access to markets if they do not meet the criterion of “durable and substantial access to international financial markets as defined in the first test of market access in the PRGT eligibility decision.”^³¹

42. This set of proposals offers a simple intuitive approach that delivers a more robust determination of blend status than is currently the case. It would remove the role of prospective market access in determining blend status (currently of relevance for countries at high risk of debt distress), thereby eliminating a difficult judgment call that shifts with market sentiment and a country’s borrowing plans. It maintains the principle that countries in debt distress are not expected to blend and that better-off countries with contained debt vulnerabilities are expected to blend. It makes the reasonable assumption that countries at high risk of debt distress with little prior access to markets should not be viewed as having significant access; and it recognizes that small/micro states face particular challenges in accessing international markets given scale effects (e.g., bond issues typically have to be of substantial size to cover fixed transactions costs and ensure sufficient trading liquidity).^³² The implication of these rules for blend status is discussed in Annex III.

43. The formula for determining the mix of PRGT and GRA resources received by blend countries (PBs) would be as follows:

The funding mix made available to a blend country would follow the 1:2 ratio, with PRGT access capped at 145 percent of quota per arrangement.

A blend country requesting arrangements with combined PRGT/GRA access that would not trigger the Policy Safeguards (i.e., access not exceeding 435 percent of quota) would receive the 1:2 mix: a country seeking cumulative access of 435 percent or more would be capped at 145 percent of quota from the PRGT. Access to PRGT credit would also be subject to the limit on normal cumulative access to the PRGT of 435 percent (a limit that is very unlikely to become binding as it would require a series of successive high access arrangements).

“All-PRGT” versus “1:2” Financing Terms for Presumed Blenders: A Possible Future Reform Option

44. There are good arguments for differentiating between the terms on which concessional financing is provided to the poorest countries and the terms on which it is provided to LICs that are substantially better-off. Concessional resources are scarce: differentiated terms allow the lender to provide financing on more generous terms to the poorest countries than the terms provided to the better-off, while still providing loans to better-off countries at very attractive, albeit less generous, terms. Hardening the terms of concessional lending as countries move up the income ladder also prepares them for graduation from PRGT eligibility, after which point all borrowings from the Fund will come on regular GRA terms, linked to market rates.^³³ This logic underlies the hardening of lending terms when countries move from “IDA-only” to “IDA gap” status at the World Bank—and the hardening of terms when countries move from non-blend to blend status at the Fund.

45. Fund policies achieve a hardening of terms for “blend countries” indirectly—not by means of dual pricing within the PRGT but rather by requiring that blend countries borrow from both the PRGT and the GRA. This delivers hardened financing terms—a (weighted) average of PRGT and GRA financing terms—but has the side-effect of pushing blend countries into the GRA, where the policy requirements differ from those of the PRGT. In particular, GRA arrangements are not intended to deal with protracted BoP problems (in contrast with the ECF), but rather to address short-to-medium term BoP difficulties. Countries such as Cambodia or Comoros meet the current (and proposed) criteria for blending—but it is difficult to argue convincingly that they have broken through the development challenges that underpin protracted BoP problems and should face the same GRA policy requirements as middle-income countries responding to BoP shocks.

46. One feature of the current approach is that blend LICs are treated in a similar fashion to non-LICs, with the one exception that they obtain one-third of any Fund financing on PRGT terms. The distinction between non-blend LICs and blend LICs is much sharper than this, both in terms of cost of financing and required use of the GRA—a factor that partly motivated the attention to blending policy in this paper. Blend countries cannot, for example, rely on a protracted BoP problem standard to request multiyear financing, and would need to meet EFF/GRA qualification requirements, including regarding the pace of expected resolution of BoP needs under the program.

47. The objective of lending on harder terms to better-off LICs than to poorer LICs could be achieved more directly, and without requiring GRA program standards, through dual pricing within the PRGT, with all PRGT-eligible countries being able to meet their financing needs entirely via the PRGT.^³⁴ This can be achieved in a manner that allows for more flexible program design for countries now required to blend (e.g., by allowing all-ECF financing) while generating modest subsidy savings for the PRGT (Annex V and Box 3 in Section V).

48. However, such a reform would result in a substantial increase in PRGT credit outstanding and a sharp decline in reserve coverage in the PRGT (Section V, Box 3). Taking this issue further would thus require tackling policy and financial issues and assessing the legal implications that will take more time to resolve: work will continue on these issues, including a strategy to bolster reserve coverage of PRGT loans.

IV. Lending Scenarios and Financing Needs

49. Demand for PRGT financing is expected to remain high over the coming years as LICs gradually recover from the pandemic. Following the unprecedented scale of IMF emergency support (SDR 8.4 billion, of which SDR 6.2 billion from the PRGT) in response to the outbreak of the COVID-19 pandemic, many LICs are now moving to multiyear financial arrangements. Total new lending commitments under ECF or blended arrangements have already reached SDR 2.5 billion (of which SDR 1.2 billion from the PRGT) through May 2021, with a large pipeline of additional programs likely to materialize over coming months and years. These programs are designed to meet the considerable financing needs created by the crisis (e.g., revenue losses, health expenditures) and create additional policy space to underpin a sustainable economic recovery. Demand for PRGT financing is expected to remain elevated through the second half of the decade, given the expected significant scarring from the pandemic, with many LICs likely to seek successor arrangements.

50. Based on the policy proposals in this paper, staff projects total PRGT lending to reach around SDR 21 billion during the pandemic and its immediate aftermath (2020–24). However, this projection is subject to significant uncertainty around economic developments and demand over the coming years. Several factors are particularly difficult to predict at this juncture: (i) the number of LICs that will eventually request program support, (ii) the size of individual programs, which depends on BoP needs, policy strength, and capacity to repay, and (iii) the pace of recovery, which depends on external factors (availability of donor financing, export demand, global financial conditions) and domestic ones (health challenges, constrained policy space). To capture these uncertainties, staff has conducted a thorough country-by-country analysis to construct a benchmark scenario with accompanying sensitivity analysis (Annex VII summarizes the methodology):

The “Baseline” scenario assumes that nearly two-thirds of LICs seek program support (in line with historical peak years) during the pandemic period (2020–24). Access per arrangement is calibrated to reflect the exceptionally high financing needs, with average access scaled up to almost twice the level observed in recent years, while differentiating across countries by individual quota, GDP, debt vulnerability, and Fund credit exposure.
To capture uncertainty around the Baseline, a “Low Case” and “High Case” are calibrated by assuming a lower/higher share of LICs request programs. The High Case could be considered a tail event, constructed to “stress-test” PRGT resources in the event of an unprecedented high share of LICs seeking program support and per-country access levels at nearly three times historical levels.
Without prejudging the appropriate longer-term target for PRGT lending capacity (to be discussed in the second stage review in 2024/25), staff assumes that a base self-sustained lending envelope of at least SDR 1.65 billion per year is preserved, which would maintain access in real terms compared to pre-crisis levels. This would accommodate room for many LICs to request successor arrangements in the second half of the decade (Annex VIII describes the main building blocks of the PRGT self-sustained financing model).

51. The combination of new Fund lending and the planned general SDR allocation would cover a significant share of the BoP needs of LICs estimated recently by staff.^³⁵ Under the Baseline, lending to LICs during the pandemic years would be more than four times the historical average, with total lending commitments of about SDR 34 billion (including SDR 21 billion from the PRGT) during 2020–24. PRGT credit outstanding would peak in 2025/26 at about SDR 22 billion, more than three times the pre-pandemic level, before gradually declining in the post-crisis decade as the share of LICs with programs in place reverts to historical levels and some countries move toward blending and PRGT graduation (Figures 3A and B). This will provide significant additional support to LICs at a particularly difficult time and can help catalyze grants and highly concessional loans from other development partners during and after the pandemic.

52. The surge in lending to LICs will significantly increase country-level exposure to Fund credit, underscoring the need to carefully scrutinize capacity-to-repay in individual cases (Figures 3C, D, and E). Under the Baseline, the typical non-blend LIC would borrow about 1½ percent of GDP from the PRGT annually during the crisis period (about twice the historical level), resulting in peak PRGT credit of 5–8 percent of GDP by 2025/26 (around 11 percent of government debt), before declining thereafter. Debt service would peak at around 2–3 percent of exports annually during 2025–30, compared to about 1 percent for the PRGT historically. Under the High Case scenario, non-blend LICs could have credit peaking at 6–12 percent of GDP.

53. The unprecedented demand for Fund concessional financing is creating a large resource gap in the PRGT (Figure 3F). The financial costs of the pandemic derive from (i) the unprecedented surge in RCF support in 2020 that will amortize over 10 years; (ii) the gradual shift toward multiyear program support, which will be committed and disbursed over the coming years, and amortized beyond the middle of the next decade; and (iii) elevated demand expected for the post-crisis decade as a result of somewhat more LICs seeking successor arrangements, higher access levels compared to pre-pandemic years, and delayed transitions into graduation and presumed blending, as a result of longer-term economic scarring.

54. The reforms to access and blending policies proposed in this paper are estimated to have a fairly modest impact on PRGT finances. Even under existing policies, PRGT crisis lending would reach unprecedented high levels, with credit peaking at around three times the pre-pandemic level. For illustration, reinstating the pre-pandemic PRGT access limits and caps (after the temporarily higher access limits expire) would affect about a dozen LICs for which access limits are most binding, and reduce PRGT lending by no more than SDR 1½ billion during 2021–24—and much of this reduced PRGT lending could be offset by “top up” borrowing from the GRA by these LICs.^³⁶

V. Funding Strategy

55. The PRGT’s endowment has been built over several decades, relying on a mix of internal and donor contributions. The current framework for financing the Fund’s concessional lending relies on loan resources periodically provided by members, credit risk mitigated by the Reserve Account, and interest rate subsidization from grant resources provided by members and Fund internal resources. Several fundraising rounds to secure loan and subsidy resources have been completed since then with the generous voluntary support of member countries. To date, members have voluntarily provided about SDR 5.3 billion to the framework’s subsidy accounts and made close to SDR 55.5 billion available in loan resources. The Fund contributed about SDR 5.5 billion in internal resources, mostly derived from gold sales (see Annexes IX and X for details).

56. Staff proposes a two-stage funding strategy to cover the resource costs created by pandemic-related PRGT lending while preserving the long-term sustainability of the endowment-based financing model. In light of the significant uncertainties around potential demand for IMF concessional financing at this juncture, preliminary informal consultations suggest that most Executive Directors support mobilizing additional PRGT resources in two stages: (i) a medium-term fundraising effort to finance crisis-related lending while preserving the PRGT’s capacity to subsidize lending in the longer term, to be followed in due course by (ii) a post-pandemic review to consider a long-term solution to PRGT self-sustainability, including through possible use of internal resources.

57. “Stage one” of the strategy would aim to mobilize: (i) a further SDR 12.6 billion in PRGT loan resources, and (ii) SDR 2.8 billion in new subsidy resources. This would close the resource gaps created by the pandemic-related financial support to LICs while preserving a base self-sustained subsidization capacity for post-crisis concessional lending.

The additional loan resources (which require an increase in the cumulative PRGT borrowing limit from SDR 55.5 billion currently to SDR 68 billion) would increase the total loan mobilization round to almost SDR 30 billion, sufficient to cover the demand scenarios discussed above through 2024, including in the High Case. Staff would approach both current and potential new lenders to contribute these additional loan resources, on top of the SDR 17 billion already secured (Annex IX). Another round of PRGT loan mobilization would follow in 2024/25 to cover lending in the second half of the decade (part of “stage two” of the funding strategy). “Channeling” of SDRs by contributors, including from the proposed new allocation, could facilitate these loan mobilization efforts.
The new subsidy resources would cover all pandemic-related lending under the Baseline while leaving a residual (post-crisis) self-sustained capacity of SDR 1.65 billion per year, sufficient to preserve access levels in real terms relative to pre-pandemic levels and allow for continued significant program support in the post-pandemic decade and beyond. The actual post-crisis residual self-sustained lending capacity might be somewhat higher or lower depending on the realized lending levels, to the extent they deviate from the Baseline.

58. Staff proposes a suspension of PRGT administrative cost reimbursement through FY2026 to generate SDR 0.5 billion in subsidy resources and boost reserve coverage.^³⁷ Annual reimbursement from the PRGT Reserve Account to the GRA was made a key part of the Fund’s New Income Model in 2008. The temporary suspension of reimbursement, however, is one of the recognized contingency measures under the PRGT’s three-pillar framework when self-sustained capacity falls short of the target envelope (Annex X). The proposed suspension would retain these resources in the PRGT Reserve Account, which has the dual benefit of adding to the PRGT’s endowment for subsidization purposes and improving the reserve coverage ratio. The proposed suspension would also slow the accumulation of Precautionary Balances in the GRA by an equivalent amount, although it is not expected to significantly delay reaching the current SDR 25 billion target.

59. Bilateral subsidy contributions of SDR 2.3 billion will be sought in a broad, burden-shared fundraising campaign involving economically stronger member countries based on their respective quota shares. Solidarity with low-income members and burden sharing among donors have been crucial to the PRGT’s financing framework since its inception. Based on this approach and the proven generosity of the Fund’s membership, IMF management would request bilateral contributions from a similar group that was asked to contribute to the recent CCRT grant mobilization campaign, in particular Financial Transactions Plan (FTP) members and other advanced and G20 countries, excluding any countries requiring IMF BoP support in the last three years (Table 1). To ensure the PRGT reform and funding package is fully financed, management would request bilateral subsidy pledges in the coming months, even if these pledges are still subject to domestic procedures.

^{Table 1.}

Proposed Burden Sharing for SDR 2.3 Billion in New PRGT Subsidy Resources

(In SDR million unless otherwise noted)

^¹Staff estimates of cumulative contributions (i.e., grants, returns on members’ deposits and implicit contributions) made to the PRGT and its predecessors under all fundraising efforts since 1987, including income earned on outstanding balances of the contributions and excluding amounts transferred to the MDRI-II Trust in January 2006.

^²All contributions are voluntary. Indicative contributions are calculated based on quota shares of 61 economically stronger member countries, including those participating in the Financial Transaction Plan (FTP) and G-20 and European Union members that have not used Fund resources for BoP needs over the last 3 years.

^³Of the 190 members of the Fund, 154 provided subsidy contributions to the PRGT.

^{Table 1.}

Proposed Burden Sharing for SDR 2.3 Billion in New PRGT Subsidy Resources

(In SDR million unless otherwise noted)

		Number of countries in the group	Percent share in total member quota	Cumulative PRGT subsidy contributions as of April 30, 2021^¹	Illustrative new contributions request based on SDR 2.3 billion target and quota shares ^²
All members		190	100.00	5 ,304	*2,300*
FTP members		50	83.01	4 ,874	2,172
	G-7	7	43.47	3 ,077	1,138
	Other advanced	22	18.20	1 ,290	476
Other FTP members		21	21.34	508	558
Non-FTP members		11	4.89	125	128
Total from 61 members		61	87.90	4 ,999	*2,300*
Total from other members^³		93	12.10	305	…

^³Of the 190 members of the Fund, 154 provided subsidy contributions to the PRGT.

^{Table 1.}

Proposed Burden Sharing for SDR 2.3 Billion in New PRGT Subsidy Resources

(In SDR million unless otherwise noted)

		Number of countries in the group	Percent share in total member quota	Cumulative PRGT subsidy contributions as of April 30, 2021^¹	Illustrative new contributions request based on SDR 2.3 billion target and quota shares ^²
All members		190	100.00	5 ,304	*2,300*
FTP members		50	83.01	4 ,874	2,172
	G-7	7	43.47	3 ,077	1,138
	Other advanced	22	18.20	1 ,290	476
Other FTP members		21	21.34	508	558
Non-FTP members		11	4.89	125	128
Total from 61 members		61	87.90	4 ,999	*2,300*
Total from other members^³		93	12.10	305	…

^³Of the 190 members of the Fund, 154 provided subsidy contributions to the PRGT.

60. To provide donors with flexibility, various options for bilateral subsidy contribution schemes are available, with resources pledged upfront and disbursed over time. In the current low interest rate environment, PRGT subsidy resources will not run out any time soon. The Fund can thus accommodate options that involve a gradual accumulation of subsidy contributions so long as the agreed PRGT funding package is backed up by sufficient upfront pledges. Individual donors could choose one or a combination of methods for delivering their pledged subsidy contributions. These include (also see Annex XI):

Budgetary grants. For many donors, this will be the most straightforward approach. Individual donors could disburse pledged budgetary grants upfront or in future years, possibly in annual tranches, depending on domestic budgetary procedures. It would also be possible to provide PRGT subsidy contributions over a longer period, say 10 years, based on subsidized loan or deposit/investment agreements (see below).
Donating SDRs or interest earnings. Outright donations of SDR holdings are possible but typically constrained by domestic institutional frameworks (e.g., central banks’ legal mandates) and entail ongoing costs for donors that remain responsible for SDR charges on their SDR allocation, as well as potential permanent costs should the Fund ever decide to cancel SDRs. To the extent that donations of SDRs require budgetary appropriations, many donors may prefer providing a budgetary grant in their country’s own or other currency. Some donors may, however, be in a position to contribute part of the interest earned on their SDR holdings (or from their interest earnings on GRA lending), which can deliver subsidy resources over time (Table 2).
Providing PRGT loans at below the SDR rate. Such loans, which can be in currencies or members’ SDR holdings, provide savings to the Trust on its subsidization expenses and can be counted as a source of implicit subsidy grant contributions. One way to operationalize such loan agreements would be to introduce a cap on interest paid to lenders at a level below the projected SDR rate. While the implicit subsidy contribution would be low in the current interest rate environment, it can deliver a significant benefit over time as rates rise, with the implicit subsidy provided by the loan to the PRGT co-moving with the cost of subsidization of PRGT lending. For illustration, a SDR 1 billion loan to the PRGT provided at a fixed 5 basis points and disbursed over three years would generate an expected NPV subsidy contribution of about SDR 100 million over 13 years assuming a gradual normalization of interest rates (Table 2).
Investing resources in the Trust (in eligible investments under the PRGT’s investment strategy or BIS obligations) in the context of deposit or investment agreements where the member’s principal is invested and generates net investment returns that are used as subsidy resources. The member eventually receives back the principal at maturity or another terminating event (see Annex XI). Such investments could be done in currencies or from a member’s SDR holdings for a period of time (e.g., 10–20 years or until such time as the pledge contribution has been met) with all or part of the net earnings on the investments retained in the Trust as subsidies. Investment agreements entail a certain degree of investment risk—realized investment returns could turn negative, particularly over shorter horizons, leading to a potential loss in the principal. Incorporating some flexibility in the maturity schedule for the investment agreement could help manage investment risk by linking the maturity to the attainment of the pledged contribution. The speed at which the overall subsidy contribution accumulates depends both on realized investment returns and on the remuneration received by the contributor—for instance, a flat remuneration at 5 basis points would generate significantly higher subsidy contributions over a given investment horizon than an investment that is remunerated at the prevailing SDR rate (Table 3).

^{Table 2.}

Subsidy Value of a SDR 1 Billion 10-Year Contribution to the PRGT Under Alternative Methods^¹

(In SDR million)

^¹Unless otherwise noted, estimated for 10-year contribution period starting in 2022, end-2020 NPV discounting at the assumed SDR rate, and remuneration of non-grant contributors at the floor SDR rate of 5bp.

^²Assumes gradual normalization of interest rates to a steady state of 3 percent.

^³Based on stylized ECF commitment with disbursements starting in 2022 and repayments completed by end 2033.

^⁴A member could instruct the IMF to transfer from its SDR holdings an amount of interest earned above 5 bp.

^⁵Assuming annual return on PRGT assets at SDR rate plus 90 bp margin.

^⁶Based on SDR 1 billion of credit outstanding over ten year period and zero rate paid by PRGT borrowers.

^{Table 2.}

Subsidy Value of a SDR 1 Billion 10-Year Contribution to the PRGT Under Alternative Methods^¹

(In SDR million)

Contribution Method		Baseline Interest Assumptions^²		Interest Rates 100bp Above Baseline
		Nominal	NPV	Nominal	NPV
Grant provided in 10 annual tranches		1,000	946	1,000	889
Implicit subsidy loan to PRGT provided at 5bp rate^³		110	102	182	157
Interest on member’s SDR holdings in excess of 5bp^⁴		151	138	251	216
Earnings on investment pooled with PRGT assets^⁵
	a. Remunerated at 5bp	268	230	398	306
	b. Remunerated at SDR rate	94	80	94	72
*Memorandum item*
Subsidy cost of SDR 1 billion PRGT credit outstanding^⁶		156	143	256	221

^²Assumes gradual normalization of interest rates to a steady state of 3 percent.

^³Based on stylized ECF commitment with disbursements starting in 2022 and repayments completed by end 2033.

^⁴A member could instruct the IMF to transfer from its SDR holdings an amount of interest earned above 5 bp.

^⁵Assuming annual return on PRGT assets at SDR rate plus 90 bp margin.

^⁶Based on SDR 1 billion of credit outstanding over ten year period and zero rate paid by PRGT borrowers.

^{Table 2.}

Subsidy Value of a SDR 1 Billion 10-Year Contribution to the PRGT Under Alternative Methods^¹

(In SDR million)

Contribution Method		Baseline Interest Assumptions^²		Interest Rates 100bp Above Baseline
		Nominal	NPV	Nominal	NPV
Grant provided in 10 annual tranches		1,000	946	1,000	889
Implicit subsidy loan to PRGT provided at 5bp rate^³		110	102	182	157
Interest on member’s SDR holdings in excess of 5bp^⁴		151	138	251	216
Earnings on investment pooled with PRGT assets^⁵
	a. Remunerated at 5bp	268	230	398	306
	b. Remunerated at SDR rate	94	80	94	72
*Memorandum item*
Subsidy cost of SDR 1 billion PRGT credit outstanding^⁶		156	143	256	221

^²Assumes gradual normalization of interest rates to a steady state of 3 percent.

^³Based on stylized ECF commitment with disbursements starting in 2022 and repayments completed by end 2033.

^⁴A member could instruct the IMF to transfer from its SDR holdings an amount of interest earned above 5 bp.

^⁵Assuming annual return on PRGT assets at SDR rate plus 90 bp margin.

^⁶Based on SDR 1 billion of credit outstanding over ten year period and zero rate paid by PRGT borrowers.

61. To facilitate bilateral subsidy contributions and reinforce reserve coverage, staff proposes the creation of two new PRGT accounts that can receive member contributions—a “Subsidy Reserve Account” (SRA) and a “Deposit and Investment Account” (DIA). The SRA would have the dual purpose of holding and investing PRGT subsidy resources while also providing an additional backstop to the PRGT Reserve Account (RA) to help manage credit risk. The DIA would allow members to channel SDRs (or currencies) for generating investment returns that could be used as PRGT subsidy resources. Resources held in both accounts would be invested alongside the balances in the existing PRGT subsidy accounts and the RA.^³⁸

62. The proposed SRA would complement existing subsidy accounts and could receive bilateral grants or investment returns contributed by members to finance PRGT subsidization. Members providing subsidy resources as part of the “stage one” SDR 2.3 billion bilateral fundraising campaign could choose whether to place their contribution in the SRA or any of the four existing PRGT subsidy accounts. The SRA would invest these subsidy resources alongside the pool of PRGT assets based on a Board-approved investment strategy. The main purpose of the SRA would be to provide an additional flexible vehicle for subsidizing PRGT lending while also enhancing reserve coverage by serving as a second-line backstop for the RA in the event of arrears, providing a medium-term funding bridge toward the longer-term self-sustained PRGT endowment model. Specifically:

SRA resources would be used for subsidizing PRGT lending after resources in the existing subsidy accounts are exhausted, and before resources in the RA become the sole funding source for subsidization under the self-sustained PRGT endowment model (where the RA has become large enough to generate investment returns that cover all subsidy needs over the long run).
In addition, the SRA would be designated as a “second line” backstop for reserve coverage in the (highly unlikely) event that resources in the RA were to be depleted as a result of very large arrears. Resources contributed to the SRA would thus have the dual benefit of bolstering the PRGT’s self-sustained subsidization capacity (similar to grants provided to other subsidy accounts) while at the same time mitigating (tail) credit risks by increasing effective reserve coverage of credit outstanding.^³⁹
As with other contributor-funded accounts in the PRGT, it is proposed that any amendments to the provisions of the SRA that would adversely affect the interests of contributors would be subject to the consent of creditors, who would retain the right to receive their pro-rata contributions back in the event of changes that they did not agree with.^⁴⁰

63. The proposed DIA would become the main vehicle for borrowing SDRs or currency from members with the objective of generating investment returns for PRGT subsidization. Members wishing to contribute subsidy resources via a long-term investment in the PRGT can already enter into deposit or investment agreements. The purpose of the DIA would be to centralize any such new resources in a separate account, which could support larger-scale investments by facilitating liquidity management and an encashment regime.

Members’ investments in the DIA would be remunerated at a fixed or floating rate, between zero and the SDRi, depending on contributor preferences. The investment returns (above the agreed rate of remuneration) attributed to a contributor would be transferred as a subsidy contribution to the SRA at the final maturity of the member’s investment agreement.
Investment agreements would generally be long term (e.g., 10–20 years), with the final maturity depending on the total amount placed in the DIA, the member’s target subsidy contribution, and the interest remuneration paid to the member (lower flat rates would generate the targeted amount of subsidy resources more quickly). As discussed above, investors would bear some degree of investment risk, which can be managed by building flexibility into the maturity schedule of the investment agreement.^⁴¹
The resources could be pooled and invested alongside other PRGT assets, based on a Board-approved investment strategy (the current one aims to generate a long-term return of the SDR rate plus 90 basis points). An alternative would be to develop a separate investment strategy specifically for DIA assets, assuming that the pool of resources is sufficiently large and the risk-return profile is considered to be substantially different from the PRGT subsidy accounts and RA. Staff will develop an appropriate investment strategy following consultations with potential contributors to the DIA. This will be discussed with the Executive Board in the context of the ongoing Review of the Investment Account and Trust Asset Investment Strategy.
SDRs or currencies placed in the DIA could retain their reserve asset status if they can be encashed upon the representation that the relevant member is experiencing a BoP need,^⁴² subject to a commitment that the contributor reconstitutes the investment once the relevant member no longer has a BoP need. An encashment buffer could be created by participating contributors depositing an additional amount of SDRs equivalent to 20–30 percent of their invested resources into the DIA (remunerated at the SDRi), which would be kept available to allow any participating contributor to encash its DIA assets quickly if needed. An alternative would be to invest DIA resources based on an investment strategy that allocates a certain share of the portfolio in sufficiently liquid assets that can support encashment, which would however result in lower expected investment returns and hence a reduced subsidy contribution for a given principal amount and investment period. DIA investments would be best placed by members with strong BoP and reserve positions and low risk of encashment needs.
If donors choose a combination of methods for delivering their pledged subsidy contributions, the DIA could facilitate the “stage one” subsidy fundraising campaign by providing contributing members the option of contributing up to half of their prospective subsidy contribution through a long-term investment agreement. If all contributors pursued this option, it would result in around SDR 10 billion in DIA investments.
To the extent that the DIA can facilitate channeling of SDRs, which will be in ample supply if the proposed SDR allocation is approved, it may also be a useful vehicle for generating additional subsidy resources as part of stage two of the funding strategy.

64. The first stage of the funding strategy would fully finance crisis-related resource needs and help preserve adequate reserve coverage under the Baseline. The suspension of PRGT reimbursement to the GRA through FY26 would retain additional resources in the RA, which would increase gradually over time based on investment returns. Reserve coverage of PRGT credit outstanding would decline from currently 32 percent to an average of 22–34 percent in 2025–29 (depending on demand for PRGT loans and the level of subsidy contributions channeled to the proposed SRA), before gradually increasing thereafter. For illustration, and assuming that about half of PRGT subsidy contributions are channeled to the SRA, the reserve coverage ratio would remain above 20 percent throughout even if PRGT lending is 20 percent above the Baseline (Figure 4). If a significantly higher lending trajectory were to materialize, or the reform option of providing “all-PRGT” financing to presumed blenders is implemented (see Box 3), an interim review of concessional financing and policies would be called for, including to assess options for bolstering credit protections (see below).

65. The adequacy of PRGT resources would need to be carefully monitored throughout the first phase of the funding strategy. The two-stage funding strategy is designed to be robust enough to allow for some variation around Baseline lending and fundraising assumptions, as well as deviations from assumed interest rates and investment returns. For instance, under the Baseline scenario, a shortfall of one-third in bilateral contributions would reduce the self-sustained capacity in the post-crisis decade by more than 10 percent, to below SDR 1.5 billion per year. Such a shortfall would increase the amount of needed contributions to be raised in the second stage of the proposed funding strategy by about SDR 0.7 billion—which is the minimum amount that would need to be raised to achieve a self-sustained lending envelope of at least SDR 1.65 billion a year. However, under the Low Case, this illustrative shortfall would not undermine the PRGT’s self-sustainability, as it would be roughly offset by the lower subsidy needs estimated for this scenario (see last block of Table 3).

66. Annual reviews by the Executive Board could trigger contingency measures when the evolution of lending and/or fundraising points to substantial risks to PRGT resources that would warrant urgent remedial actions prior to the second stage review in 2024/25. Annual reviews will include: (i) an update of lending developments and demand projections, (ii) an update on loan and subsidy resource mobilization, as well as investment and interest developments, (iii) an assessment of the PRGT’s lending capacity and resource outlook, (iv) an assessment of credit risks, Fund exposure across LICs, and reserve coverage outlook, and (v) possible options for adapting the funding strategy if warranted. In the event of a significant fundraising shortfall, an exceptionally high lending trajectory, and/or a significant deterioration in multiple borrowers’ capacity to repay the Fund, it may be necessary to take remedial measures to ensure adequate subsidy and loan resources, and/or credit protections. IMF management may then propose possible contingency measures ahead of the second stage review. These could include:

Additional bilateral fundraising efforts, led by IMF management and supported by the Executive Board; the focus could be on any countries that have not yet pledged/disbursed a quota-based share for the stage one fundraising target;
A decision to extend the suspension of reimbursements to the GRA for PRGT administrative costs for a number of years beyond FY2026; while it would take some time to have a material impact on reserve coverage, it would be a lasting one, and provide additional subsidization capacity;
Seeking member support for a “gold pledge” that would provide a backstop for possible future credit losses and possibly also a restoration of subsidization capacity under certain conditions; while a gold pledge could be potentially implemented more quickly than outright gold sales, it would require the same broad support, including 85 percent of the total voting power for the Executive Board decision and parliamentary procedures where applicable;
A coordinated effort to secure government guarantees from a group of advanced countries as an ultimate backstop against possible credit losses, in the event that reserve coverage is deemed insufficient;
A Board decision to consider a distribution of GRA reserves to facilitate contributions to the PRGT, contingent on a minimum threshold of pledges being reached, once precautionary balances have reached their medium-term target;
Recalibration of access limits and norms; any reductions in room for PRGT access would need to consider possible spillovers of LICs’ demand into top-up financing through the GRA;
A review of the PRGT interest framework that could lead to higher, though still concessional lending rates.

67. A decision on the appropriate longer-term PRGT envelope would be taken up at the second stage of the funding strategy, as part of the next full Review of Concessional Financing and Policies scheduled for 2024/25. As is customary, this would include a review of policies and possible reforms, the financial situation of the endowment under different policy and demand scenarios, and a broad range of possible funding options. A central question for the review will be the appropriate longer-term self-sustained lending capacity of the PRGT. The additional funding needs at stage two will depend on the actual crisis-period lending levels, the outlook for demand for concessional resources, which is highly uncertain, and any further policy reforms.

A self-sustained lending envelope of SDR 1.65 billion annually would essentially maintain access per country at the pre-pandemic level in real terms, allowing for future periodic upward revisions to access limits and norms to avoid eroding access relative to nominal GDP, consistent with the approach taken during the 2018/19 LIC facilities review.^⁴³ This envelope would accommodate supporting LICs with longer-term scarring from the pandemic, several of which would be expected to seek successor arrangements in the second half of the decade. However, if realized lending during the pandemic turns out to be at high levels, such an envelope could constrain the scope for a smooth transition of LICs as they reduce their exposure to Fund credit in the post-pandemic period.
The second stage review will consider the merits of a larger self-sustained PRGT lending envelope. For instance, a lending capacity of SDR 2.4–3.0 billion a year would make room for per-country PRGT access levels broadly in line with GRA arrangements for EM countries, and limit the risk of LICs resorting to “top up” GRA borrowing. Such an increased steady-state lending capacity could help smooth the transition of LICs reducing their exposure to Fund credit in the post-pandemic period. For instance, if lending evolves as assumed under the Baseline, mobilizing additional subsidy resources of SDR 3.4 billion in stage two (on top of the SDR 2.8 billion mobilized in stage one) would increase the long-term self-sustained lending capacity from SDR 1.65 billion to SDR 2.4 billion a year.
The more Fund lending that materializes during the crisis years, the larger will be the need to supplement PRGT finances at the second stage. For instance, in the extreme case that near-term demand evolves along the High Case trajectory, the residual self-sustained annual lending capacity would decline to around SDR 1.3 billion by end-2024 (assuming the fundraising target of the stage one funding campaign has been met). Such reduced lending capacity would severely constrain the Fund’s ability to support LICs in the longer term and would justify mobilizing additional subsidy resources, potentially in the range of around SDR 5–7½ billion to increase the self-sustained lending capacity to SDR 2.4–3.0 billion a year.

68. Voluntary “channeling” of SDRs would facilitate the mobilization of additional PRGT loan resources, which could range from SDR 24–34 billion for the remainder of this decade (2021–29), and significantly more if blenders were to receive “all PRGT” financing.^⁴⁴ These ranges assume a significant spike in crisis lending through 2024, and another round of loan mobilization in 2024/25 to cover concessional lending over the second half of the decade. The higher end of the range assumes a scaling up of the PRGT’s longer-term lending capacity. If the Board were to endorse a move toward the reform option of “all PRGT” financing for presumed blenders, loan resource needs could rise to around SDR 42–59 billion.

Total SDRs channeled to the PRGT over the coming years are projected in the range of SDR 20– 35 billion, of which SDR 12–20 billion in the near term. These estimates include possible channeling of SDRs as investment resources for the DIA.
The overall range for the decade could increase to around SDR 33–54 billion if the additional blending reform is implemented, which would require large additional PRGT loan resources.

69. Use of IMF internal resources should be carefully considered during the “stage two” review, especially if the Board were to pursue a significantly larger PRGT lending envelope and associated endowment. This could include a limited sale of IMF gold, which could be used to boost the RA and also generate investment returns for subsidization, or alternatively a distribution of IMF reserves conditional on securing a critical threshold of commitments from members to contribute equivalent amounts for PRGT subsidies. Both options would need to be carefully assessed against their impact on the Fund’s balance sheet. Broad support across the membership, and sufficient time for implementation, would be required for the use of either of these options.

^{Box 3.}

Financing Implications of Alternative Loan Terms for Presumed Blenders

The reform option discussed in Section III would provide presumed blenders with “all PRGT” lending at a differential (but still concessional) interest rate from non-blenders, rather than the current 2:1 mix of GRA and PRGT. This would have important implications for subsidy and loan resources, as well as reserve coverage. To illustrate:

- Under Baseline demand projections, PRGT credit outstanding would peak at SDR 27.5 billion, rather than SDR 22.5 billion, reflecting the shift from blended to all-PRGT financing for blenders.
- The reform would increase PRGT loan resource needs by SDR 26 billion for the remainder of the decade, which could be facilitated by SDR “channeling.”
- Under Baseline demand assumptions, the reform would generate net subsidy savings to the PRGT of around SDR 40 million during 2023–25, and SDR 500 million through 2034.
- Reserve coverage would reach a trough of 17 percent under the reformed Baseline. To keep reserve coverage above 20 percent, the Reserve Account would need to be augmented by more than SDR 1 billion.

Providing “all-PRGT” loans to presumed blenders would have a number of other financial implications. It would introduce a floating rate for concessional lending, which could have potential implications for financial reporting under IFRS-9 and donors’ accounting of the amount of ODA provided to the PRGT. System changes would be required to track different loan terms under the same facility depending on the borrower’s blend status at the time of financing approval. Under the proposed interest formula (two-thirds of the GRA rate of charge), the net interest paid by presumed blenders to the PRGT when global rates are low would have to be tracked and attributed in financial reporting.

VI. PRGT Interest Rate Review

70. In line with the approved PRGT interest rate mechanism, staff proposes keeping zero interest rates for all PRGT credit outstanding under the ECF, SCF, and RCF. The PRGT interest rate mechanism, adopted in 2009 and most recently modified in 2019, established the interest rate across PRGT facilities and links them to the global interest rate, which have been historically low. The current interest rate mechanism has served borrowers well, allowing for zero rates on all facilities. Applying the PRGT’s existing interest rate setting mechanism and thresholds and based on the average SDR interest rate over the most recently observed 12-month period (0.08 percent^⁴⁵), staff proposes a continuation of a zero interest rate for the ECF, the SCF as well as for the RCF until July 2023. Staff also proposes that the next review of PRGT interest rates will be held according to the usual schedule in two years by July 2023. This could include a discussion of interest rates for presumed blenders if the Executive Board decides to consider the reform option discussed above. See Annex XII for background on the current interest mechanism.

VII. Debt Relief Financing

Catastrophe Containment and Relief Trust (CCRT)

71. Additional resources are needed to fund the mandate of the CCRT. The CCRT was created in 2015 to provide grants for debt relief to eligible low-income members hit by catastrophic natural disasters or fast-spreading public health disasters. In March 2020, the Executive Board adopted a set of reforms to the CCRT to provide immediate debt service relief to the poorest and most vulnerable member countries affected by the COVID-19 pandemic. In April 2020, the Fund approved debt service relief for all 29 eligible countries of up to two years until April 2022, to be disbursed in tranches and subject to resource availability. So far, the Fund has disbursed SDR 520 million in debt service relief in three tranches through mid-October 2021, helping free up scarce financial resources for vital emergency health, social, and economic support to mitigate the impact of the COVID-19 pandemic.

72. Additional grants are urgently needed to deliver the fourth and final tranche of COVID- related CCRT relief. In April 2020, the IMF launched an urgent fundraising effort of SDR 1 billion (US$1.4 billion) to provide the needed resources to cover two years of relief (SDR 679 million) and address the initial CCRT underfunding (SDR 200–275 million). To date, grant pledges of about half of the target (SDR 575 million) have been secured from 18 contributors (Annex XIII). Substantial additional grant resources are needed to unlock the fourth and final tranche of debt service relief (estimated at SDR 160 million), ending in April 2022.

Heavily Indebted Poor Countries (HIPC) Initiative

73. The HIPC Initiative is nearly completed. The Fund has provided SDR 2.6 billion in debt relief to 38 of the 39 eligible countries.^⁴⁶ The protracted arrears cases (Liberia, Somalia, Sudan) were not included in the original costing of the HIPC Initiative to avoid undermining the Fund’s financial capacity as a result of debt forgiveness.^⁴⁷ Consequently, the Fund needed to mobilize the necessary resources to cover the Fund’s share of debt relief for these countries.

74. In March 2020 and June 2021, the IMF and the World Bank jointly committed to provide HIPC and “beyond-HIPC” debt relief to Somalia and Sudan, respectively. A large share of the membership provided generous and timely contributions to both Somalia’s and Sudan’s financing packages, which utilized distributions from internal resources and new cash grants.

For Somalia, 121 countries pledged an equivalent of SDR 280.1 million to finance the IMF’s share of debt relief, estimated at SDR 252.9 million. This included cash grants from 13 member countries in addition to their contributions from internal Fund resources as part of Somalia’s financing package. The European Commission also provided a grant of EUR 9 million, bringing total grants for Somalia to SDR 114.2 million. Of the total pledged amount, SDR 237.3 million have materialized to date.
For Sudan, 120 countries pledged SDR 1,059 million to finance the IMF debt relief, which is estimated at SDR 992 million. This included cash grants from 8 member countries in addition to their contributions from internal Fund resources as part of Sudan’s financing package. The European Commission also provided a grant of EUR 12 million, bringing total grants for Sudan to SDR 136.5 million. As of end-April 2021, the balance in the PRG-HIPC Trust stood at SDR 258 million.^⁴⁸

VIII. Enterprise Risks

75. The proposals in this paper seek to mitigate multiple enterprise risks, with risks created by the proposals consistent with current Board-approved risk acceptance levels under baseline scenarios. The proposals to increase the PRGT’s access limits could mitigate strategic risks and credit/repayment risks arising from emergency lending to date by providing more headroom under existing or follow-up UCT programs on PRGT terms, reducing the need for LICs to access GRA resources. As noted in paragraph 6 above, the proposed higher access limits themselves have a relatively modest impact on PRGT resource needs in stage one of the fundraising strategy, with the primary driver of resource needs being the unprecedented surge in lending levels over the course of 2020–24. PRGT credit/repayment risks may rise by increasing the threshold for EA, although the application of staff’s proposed safeguards/associated higher scrutiny should serve to mitigate this risk. Under the Baseline demand scenario and implementation of “stage one” of the funding strategy (suspension of reimbursement to the GRA and new bilateral subsidy contributions, including to the proposed SRA), the PRGT’s reserve coverage ratio would remain well above 20 percent during the anticipated peak in PRGT credit outstanding, and is expected to gradually increase to above 40 percent over the longer term. The impact of suspending reimbursement to the GRA for PRGT administrative expenses through FY2026 will slow the accumulation of the Fund’s precautionary balances, although there would not be a significant delay in terms of reaching the Board-approved target of SDR 25 billion.

76. Residual liquidity and credit risks remain, so the adequacy of PRGT resources would need to be carefully monitored throughout the first phase of the funding strategy. The two- stage funding strategy is designed to be robust enough to allow for some variation around baseline lending and fundraising assumptions, as well as deviations from assumed interest rates and investment returns. In addition, the proposed annual reviews of PRGT resources could trigger contingency measures when the evolution of lending and/or fundraising point to substantial risks to PRGT resources that would warrant urgent remedial actions prior to the second stage review in 2024/25. In the event of a significant fundraising shortfall, an exceptionally high lending trajectory, and/or a significant deterioration in multiple borrowers’ capacity to repay the Fund, remedial measures will likely be needed to ensure adequate subsidy and loan resources, and/or credit protections as described in the previous section.

IX. Issues for Discussion

77. Do Directors support the proposed reforms to enhance the PRGT lending framework, including:

a. increases in the normal access limits for the PRGT?
b. elimination of hard caps on PRGT exceptional access?
c. safeguards on debt sustainability/capacity to repay as endorsed by the Board in March 2021 and further elaborated in Annex VI?
d. thresholds for triggering the High Access Procedures?
e. alignment of the PRGT EA criteria with the criteria under the policy on Policy Safeguards on High Combined Credit?
f. simplification of the specification of norms?

78. Do Directors support the proposals on modifying the blending rules, including:

a. adjusting the income threshold used in determining blend status?
b. simplifying the role of market access and debt vulnerabilities in determining blend status?
c. simplifying the rules for determining the mix of PRGT and GRA resources in arrangements for presumed blenders?

79. Do Directors see merit in continued work to explore reforms that would introduce a dual interest rate mechanism in the PRGT while allowing PRGT-eligible countries to meet all their financing needs from the PRGT?

80. Do Directors support the two-stage funding strategy, with (i) a medium-term fundraising effort to finance crisis-related lending while preserving the PRGT’s capacity to subsidize lending in the longer-term, to be followed in due course by (ii) a long-term solution to PRGT self-sustainability?

81. Do Directors support a “stage one” medium-term fundraising effort to mobilize (i) SDR 12.6 billion in additional loan resources, which will require an increase in the PRGT cumulative borrowing limit from SDR 55.5 billion currently to SDR 68 billion and (ii) SDR 2.8 billion in new subsidy resources, including a suspension of reimbursement to the GRA through FY26 and SDR 2.3 billion via a burden-shared bilateral grant fundraising campaign, providing donors flexibility in terms of timing and method of contributions?

82. Do Directors agree with the proposal for the Executive Board to review annually concessional resources and progress with stage one fundraising?

83. Do Directors support the creation of two new PRGT accounts that can receive member contributions—a “Subsidy Reserve Account” (SRA) and a “Deposit and Investment Account” (DIA)?

84. Do Directors agree that the PRGT interest mechanism remains appropriate, and that PRGT interest rates on all facilities will be set at zero through end-July 2023?

Annex I. PRGT Facilities-Selected Features

1. PRGT-eligible countries, currently 69 in number, can be divided into subgroups based on the conditions under which they can access PRGT resources (see Annex III for a full listing):

Blend countries (shorthand for “presumed blenders”) can access PRGT resources only in conjunction with GRA resources, with the financing combined in a mix of 1:2 (PRGT: GRA) subject to a cap on the level of access to PRGT resources. Blend countries hence must meet the policy requirements of the GRA as well as the PRGT. A blend country is thus similarly positioned to a GRA-only country, except that it can access to up to one-third of its Fund financing on (more favorable) PRGT terms.
Non-blend countries are required to access the GRA only if they are seeking Fund resources above the relevant limits on PRGT access: an access request in excess of these limits moves them to a position where their economic program must meet the policy requirements of the GRA (and PRGT). This group can be further divided into i) countries eligible to seek EA to PRGT resources and ii) countries not eligible to seek EA.

2. PRGT lending facilities (available only to LICs) differ from the regular GRA facilities (available to all members) in some important dimensions:

Lending terms for the PRGT’s concessional lending facilities are more generous than those provided under the corresponding GRA facility.
The PRGT’s main lending vehicle, the ECF, has distinct policy requirements from its GRA counterpart.
- Programs supported under the GRA are designed to resolve the member’s BoP problems during the program period. Specifically, policy actions needed to resolve a member’s BoP problem should be undertaken during the program period; implementation of these policies should lead to a strengthening of the member’s BoP before repurchases begin such that repurchases from the Fund can occur without strain.
- The ECF is available to LICs that face protracted BoP problems (where underlying macroeconomic imbalances are expected to be resolved over an extended period): the purpose of programs supported under the ECF is to enable PRGT-eligible members with protracted BoP problems to make significant progress toward a stable and sustainable macroeconomic position.
- All PRGT financing instruments are expected to support policies that lead toward a sustainable macroeconomic position that is consistent with strong and durable poverty reduction and growth.^¹
Given that the ECF arrangement is designed to address a member’s “protracted BoP problem,” a member may require a series of ECF-supported programs to resolve BoP difficulties, rather than in a single arrangement (as is normally expected to be the case for GRA-supported programs).
Access to PRGT resources is subject to hard limits. All 69 countries are eligible for access up to the normal limits, set at 100/300 percent of quota in May 2019.^² A subgroup of 28 countries, all with incomes below the IDA operational threshold (currently $1,185), are eligible for EA, subject to hard limits of 133/400 percent of quota, provided that the programs for which they are seeking support meet the PRGT EA criteria.

Annex II. The PRGT Exceptional Access Criteria and Policy Safeguards for High Combined Credit Exposures

The PRGT EA Criteria specify safeguards that must be satisfied for use of PRGT resources above the normal access limits. PS-HCC exposure, which were approved in September 2020, specify safeguards for a member to receive combined PRGT and GRA access above the GRA exceptional access limits. The PS-HCC build on the respective PRGT and GRA EA frameworks and are broadly aligned with the GRA EA criteria.

1. The PRGT EA criteria are outlined in column 1 of the table below. Requests for PRGT financing that exceed 100/300 percent of quota (temporarily increased to 245/435 through end-June 2021) can be accommodated only if these criteria are met. The language on “reduce the risk of debt distress to a moderate level or low level” in criterion 2 has been interpreted, outside the HIPC process, to mean achieving a projected assessment of moderate/low risk of debt distress within three years of program approval.

2. The PS-HCC are outlined in column 2 of the table below. Requests for Fund financing that would entail combined access to GRA and PRGT resources in excess of 145/435 percent of quota (temporarily increased to 245/435 through end-December 2021) can be accommodated only if these criteria are met.

3. Differences between the PS-HCC and the PRGT EA criteria include:

PRGT EA 1 requires that the country experience an exceptionally large BoP need that cannot be resolved within the normal limits. PS-HCC 1 instead requires that the country is experiencing or alternatively has the potential to experience exceptional BoP pressures (on the current or capital account) that cannot be met within the normal limits.
PRGT EA 2 requires that the member have a comparatively strong adjustment program and ability to repay the Fund. PS-HCC 4 requires that the member’s program have a reasonably strong prospect for success, taking account of both program policies and the institutional and political capacity to deliver them. The operational distinction between these conditions is not substantial: under both formulations, a program involving exceptional levels of access would be expected to be strong and have reasonably strong prospects for success.
For countries at high risk of/in debt distress, PRGT EA 2 requires that the program being supported i) include the provision of debt relief or restructuring by creditors and ii) is projected to reduce the risk of debt distress to a moderate/low level, with the interpretation to date being that (outside the HIPC process) this must be achieved within a three-year time-frame. By contrast, PS-HCC 2 does not require a debt-restructuring as part of the program, while the time-frame for achieving moderate/low risk of debt distress (for a new program) is “within 36 months from Board approval of the financing request or within the period of a newly approved arrangement (whichever is longer).” The removal of the debt restructuring requirement contained in PRGT EA 2 from PS-HCC 2 allowed greater flexibility as to how the improved debt position would be achieved; adapting the three-year time frame to accommodate cases where programs would exceed three years was seen as an operationally pragmatic adjustment.

The proposed new specification of the PRGT EA criteria is contained in column 3.

^{Annex II Table 1.}

Current and Proposed PRGT Exceptional Access Criteria

^{Annex II Table 1.}

Current and Proposed PRGT Exceptional Access Criteria

	Current PRGT EA Criteria	HCCE (High Combined Credit Exposure) Safeguards	Proposed PRGT EA Criteia
Criterion 1	Countries that experience an exceptionally large balance of payments need that cannot be met within the normal limits.	The member is experiencing or has the potential to experience exceptional balance of payments pressures on the current account or capital account, resulting in a need for Fund financing that cannot be met without giving rise to a combined access to PRGT and GRA resources in amounts exceeding the thresholds that apply as limits in the GRA.	The member is experiencing or has the potential to experience exceptional balance of payments pressures on the current account or capital account, resulting in a need for Fund financing that cannot be met within the normal limits.
Criterion 2	The member has a comparatively strong adjustment program and ability to repay the Fund. This criterion would generally not be met for countries with a high risk of debt distress or those that are in debt distress as defined under the joint Bank-Fund DSA, unless expected debt relief or restructuring is projected to reduce the risk of debt distress to a moderate level or low level (IMF Policy Paper “A New Architecture of Facilities for Low Income Countries” June 26, 2009, Footnote 62).	Risks to the sustainability of public debt are adequately contained. This is evidenced by For members for whom use of the Bank-Fund Debt Sustainability Framework for Low Income Countries (the “LIC-DSF”) is warranted: A rigorous and systematic analysis indicates that there is a high probability that the member’s public debt is sustainable in the medium term. This is generally considered to be met for countries that are assessed under the LIC-DSF as having low or moderate overall risk of public debt distress; or Where the member’s public debt is not assessed to be sustainable with high probability, combined access above the proposed thresholds will only be made available if the combination of the member’s policies and financing from sources other than the Fund, which may include debt restructuring, restores public debt sustainability with high probability (i.e., to a point where application of the LIC-DSF would yield a rating of low or moderate overall risk of public debt distress) (i) within 36 months from Board approval of the financing request or within the period of a newly approved arrangement (whichever is longer) or (ii) within the remaining period of an arrangement, in cases where the Board approves an augmentation or rephasing request. For members for whom use of the MAC DSA is warranted: the debt sustainability requirements for providing exceptional access to GRA resources are met.	Risks to the sustainability of public debt are adequately contained. This is evidenced by A rigorous and systematic analysis indicates that there is a high probability that the member’s public debt is sustainable in the medium term. This is generally considered to be met for countries that are assessed under the LIC-DSF as having low or moderate overall risk of public debt distress; or Where the member’s public debt is not assessed to be sustainable with high probability, access above the proposed thresholds will only be made available if the combination of the member’s policies and financing from sources other than the Fund, which may include debt restructuring, restores public debt sustainability with high probability (i.e., to a point where application of the LIC-DSF would yield a rating of low or moderate overall risk of public debt distress) (i) within 36 months from Board approval of the financing request or within the period of a newly approved arrangement (whichever is longer) or (ii) within the remaining period of an arrangement, in cases where the Board approves an augmentation or rephasing request.
Criterion 3	Countries that have GNI per capita at or below the prevailing operational cutoff for assistance from IDA and have not had sustained past access to international financial markets (if GNI per capita is below 80 percent of the IDA cutoff, market access does not preclude EA).		Countries that do not meet the income criterion for presumed blending at the time when a new financing request (including augmentation/rephasing) is made.
Criterion 4		The policy program of the member provides a reasonably strong prospect of success, including not only the member’s adjustment plans but also its institutional and political capacity to deliver that adjustment.	The policy program of the member provides a reasonably strong prospect of success, including not only the member’s adjustment plans but also its institutional and political capacity to deliver that adjustment.

^{Annex II Table 1.}

Current and Proposed PRGT Exceptional Access Criteria

	Current PRGT EA Criteria	HCCE (High Combined Credit Exposure) Safeguards	Proposed PRGT EA Criteia
Criterion 1	Countries that experience an exceptionally large balance of payments need that cannot be met within the normal limits.	The member is experiencing or has the potential to experience exceptional balance of payments pressures on the current account or capital account, resulting in a need for Fund financing that cannot be met without giving rise to a combined access to PRGT and GRA resources in amounts exceeding the thresholds that apply as limits in the GRA.	The member is experiencing or has the potential to experience exceptional balance of payments pressures on the current account or capital account, resulting in a need for Fund financing that cannot be met within the normal limits.
Criterion 2	The member has a comparatively strong adjustment program and ability to repay the Fund. This criterion would generally not be met for countries with a high risk of debt distress or those that are in debt distress as defined under the joint Bank-Fund DSA, unless expected debt relief or restructuring is projected to reduce the risk of debt distress to a moderate level or low level (IMF Policy Paper “A New Architecture of Facilities for Low Income Countries” June 26, 2009, Footnote 62).	Risks to the sustainability of public debt are adequately contained. This is evidenced by For members for whom use of the Bank-Fund Debt Sustainability Framework for Low Income Countries (the “LIC-DSF”) is warranted: A rigorous and systematic analysis indicates that there is a high probability that the member’s public debt is sustainable in the medium term. This is generally considered to be met for countries that are assessed under the LIC-DSF as having low or moderate overall risk of public debt distress; or Where the member’s public debt is not assessed to be sustainable with high probability, combined access above the proposed thresholds will only be made available if the combination of the member’s policies and financing from sources other than the Fund, which may include debt restructuring, restores public debt sustainability with high probability (i.e., to a point where application of the LIC-DSF would yield a rating of low or moderate overall risk of public debt distress) (i) within 36 months from Board approval of the financing request or within the period of a newly approved arrangement (whichever is longer) or (ii) within the remaining period of an arrangement, in cases where the Board approves an augmentation or rephasing request. For members for whom use of the MAC DSA is warranted: the debt sustainability requirements for providing exceptional access to GRA resources are met.	Risks to the sustainability of public debt are adequately contained. This is evidenced by A rigorous and systematic analysis indicates that there is a high probability that the member’s public debt is sustainable in the medium term. This is generally considered to be met for countries that are assessed under the LIC-DSF as having low or moderate overall risk of public debt distress; or Where the member’s public debt is not assessed to be sustainable with high probability, access above the proposed thresholds will only be made available if the combination of the member’s policies and financing from sources other than the Fund, which may include debt restructuring, restores public debt sustainability with high probability (i.e., to a point where application of the LIC-DSF would yield a rating of low or moderate overall risk of public debt distress) (i) within 36 months from Board approval of the financing request or within the period of a newly approved arrangement (whichever is longer) or (ii) within the remaining period of an arrangement, in cases where the Board approves an augmentation or rephasing request.
Criterion 3	Countries that have GNI per capita at or below the prevailing operational cutoff for assistance from IDA and have not had sustained past access to international financial markets (if GNI per capita is below 80 percent of the IDA cutoff, market access does not preclude EA).		Countries that do not meet the income criterion for presumed blending at the time when a new financing request (including augmentation/rephasing) is made.
Criterion 4		The policy program of the member provides a reasonably strong prospect of success, including not only the member’s adjustment plans but also its institutional and political capacity to deliver that adjustment.	The policy program of the member provides a reasonably strong prospect of success, including not only the member’s adjustment plans but also its institutional and political capacity to deliver that adjustment.

Annex III. Blending Policies and Eligibility for Exceptional Access: Current and Proposed

1. PRGT-eligible countries are divided into two groups: i) blend countries, who can access concessional financing from the Fund only in conjunction with GRA resources and ii) non-blend countries, who can access PRGT resources up to the relevant access limits, needing to tap GRA resources only if their financing requests exceed these limits. The current rules determining the set of blend countries and the applicable mix of PRGT and GRA resources for individual financing requests from these countries are outlined in Annex III Box 1.

2. The general principles guiding the blending framework would remain substantially unchanged under the reforms proposed in this paper. These can be summarized as: i) countries are required to blend only if they meet a per capita income threshold; ii) countries that meet the income threshold are required to blend unless debt vulnerabilities impair their access to international financial markets; iii) countries required to blend that request Fund financing receive PRGT and GRA resources in a 1:2 mix.

3. Proposed technical changes to the manner in which these principles are applied are intended to make the framework both more robust and less complex.

The Income Threshold

4. The income threshold for blending has been based on whether GNI per capita lies above or below the IDA cutoff level. The data (expressed in US dollars) are produced by the World Bank, with updates released each year at the beginning of the Bank’s fiscal year (July 1).

5. Data on GNI per capita is volatile, with annual fluctuations in real GDP being augmented by the volatility of exchange rates, notwithstanding the use of smoothing mechanisms.^¹ Deciding that a country meets the income threshold for blending on the basis of data for a single year thus creates significant risks that countries with income per capita levels close to the IDA cutoff level (which itself moves each year) can shift back and forth between blend and non-blend status. This is undesirable given the important operational implications of shifting between non-blend and blend status.

6. The paper proposes to modify the specification of the income threshold to limit the risk of a premature/soon-reversed shift to blend status, raising the income requirements for meeting the threshold while tightening the conditions under which the threshold would no longer be met.^² Specifically, it is proposed that:

A country is deemed to meet the income threshold for blending when GNI per capita has exceeded the IDA operational cutoff by at least 5 percent for two consecutive years.^³
Having met the income threshold, the country continues to meet it provided that income per capita does not fall below 95 percent of the IDA operational cutoff. Should income per capita fall below this level, the country no longer meets the income threshold.

7. In deciding on the proposed parameters, staff has analyzed the evolution of GNI per capita across LICs from 2000 through 2019 and can identify only four cases where a country, having recorded GNI per capita at least 5 percent above the IDA cutoff level for two successive years, later fell below the IDA cutoff level. These include: i) Solomon Islands (meeting the 5 percent margins in 2000–01, falling below the IDA cutoff from 2002–11); ii) Sudan (meeting the margins in 2011–17, falling below the IDA cutoff in 2018–19); iii) Tajikistan (meeting the margins in 2013–15, falling below the cutoff level in 2016–19); and iv) Yemen (meeting the margins in 2012–14, falling below the cutoff level in 2015–19). Three of the four cases reflect large declines in income levels linked to serious internal conflict—a situation involving fundamental changes in economic circumstances that clearly warrant re-classification to lower-income non-blend status. Increasing the margins used to set the income threshold to 7½ percent or 10 percent (or to 10 percent for the most recent year) would not significantly change the picture: for example, the margins over the cutoff for Tajikistan were 9 percent and 11 percent respectively in 2013–14. Since the 2000–2019 period covered periods of substantial volatility across LICs, we conclude that the proposed rules are quite robust.

8. The size of the shock to income levels in many (but not all) LICs in 2020 has been exceptionally large, albeit with partial rebounds likely for many in 2021. Applying the proposed rules, the number of countries that do not meet the income threshold for blending would increase by five—Kyrgyz, Lesotho, Myanmar, Zambia, and Zimbabwe. Under the current rules, all except Myanmar (whose 2020 income level still exceeds the IDA cutoff but by less than 5 percent) would also have been classified as not meeting the income threshold. Separately, an upgrade of the national income accounts in Haiti has yielded a large increase in the measured level of national income: Haiti would meet the income threshold under current rules (with 2020 GNI per capita above the threshold) but not under the new rules (2020 GNI per capita is not 5 percent above the IDA cutoff).

9. The list of countries that would meet/not meet the proposed income threshold for blending for the period through end-June 2022 is provided in Annex III Table 1, which also shows how the outcome differs from the results if one applied the current rules.^⁴

^{Annex III Table 1.}

Which Countries Meet the Income Threshold for Blending?

Notes: 1. 29 countries that exceeded the IDA cutoff by at least 20 percent in 2019 and 2020 are not shown here. 2. 23 countries that fell below the IDA cutoff by at least 10 percent in 2019 and 2020 are not shown here. 3. 6 countries for which data is incomplete/not available are not shown here: Eritrea, Somalia, South Sudan, Tonga, Micronesia, and Marshall Islands. For the latter three, 2019 GNI per capita was >120 percent of the IDA cutoff. 4. Assessment for July 2020 is based on GNI per capita data for 2019, issued by the World Bank on July 1, 2020. Assessment for July 2021 is based on GNI per capita data for 2021, issued by the World Bank on July 1, 2021.

^{Annex III Table 1.}

Which Countries Meet the Income Threshold for Blending?

	GNI per capita/IDA cutoff		Current Rules		Proposed Rules
	2019	2020	July 2020	as of July 2021	as of July 2021
Zambia	122.4	98.8	Yes	No	No
Senegal	122.3	118.7	Yes	Yes	Yes
Comoros	119.8	120.3	Yes	Yes	Yes
Myanmar	117.3	104.6	Yes	Yes	No
Zimbabwe	117.3	90.5	Yes	No	No
Lesotho	114.8	91.3	Yes	No	No
Benin	105.5	106.2	Yes	Yes	Yes
Kyrgyz Republic	104.6	96.3	No	No	No
Nepal	92.0	98.8	No	No	No
Tanzania	91.1	89.6	Yes	Yes	No
Haiti	66.7	103.7	No	Yes	No

^{Annex III Table 1.}

Which Countries Meet the Income Threshold for Blending?

	GNI per capita/IDA cutoff		Current Rules		Proposed Rules
	2019	2020	July 2020	as of July 2021	as of July 2021
Zambia	122.4	98.8	Yes	No	No
Senegal	122.3	118.7	Yes	Yes	Yes
Comoros	119.8	120.3	Yes	Yes	Yes
Myanmar	117.3	104.6	Yes	Yes	No
Zimbabwe	117.3	90.5	Yes	No	No
Lesotho	114.8	91.3	Yes	No	No
Benin	105.5	106.2	Yes	Yes	Yes
Kyrgyz Republic	104.6	96.3	No	No	No
Nepal	92.0	98.8	No	No	No
Tanzania	91.1	89.6	Yes	Yes	No
Haiti	66.7	103.7	No	Yes	No

The Market Access Threshold

10. Countries that meet the income criterion for blending are currently required to blend unless debt vulnerabilities limit their access to international financial markets; this general approach would be maintained with some technical simplifications. Specifically, debt vulnerabilities are deemed to limit a country’s access to international financial markets when:

the country is assessed to be in debt distress or
the country is assessed to be at high risk of debt distress and a) has had limited past access to external financial markets or b) is a small/micro-state.

Countries are assessed to have had limited past access to markets if they do not meet the established criterion of “durable and substantial access to international financial markets.”^⁵

11. The list of countries that could be required to blend under existing rules and/or under the proposed rules is outlined in Annex III Table 2.^⁶ Three countries previously required to blend drop off the list because of sharp declines in national income in 2020 (Lesotho, Kyrgyz Republic, Myanmar). Of the eight countries previously assessed to be potential blend countries, three (Cameroon, Ghana, Kenya) are now required to blend, given the level of confirmed past market access; three do not have to blend, given their small island status (Cabo Verde, Dominica, and Maldives); and the blend status of two countries (Lao PDR, Papua New Guinea) would depend on validation of the data on past external borrowings with country authorities.

^{Annex III Table 2.}

Blend Status of Potential Presumed Blenders^1/ ^2/

^{^1/}Excludes all countries that fail to meet the income threshold in 2021

^{^2/}Countries required to blend under the current rules and under the proposed rules are not shown in the table.

^{^3/}Blend status dependent on staff assessment of prospective market access

^{^4/}Blend status dependent on validation of scale of debt issuance in previous 5 years.

^{Annex III Table 2.}

Blend Status of Potential Presumed Blenders^1/ ^2/

	Current Rules	Proposed Rules	Comment
Cabo Verde	^/3	No	Small State
Cameroon	^/3	Yes
Dominica	^/3	No	Small State
Ghana	^/3	Yes
Kenya	^/3	Yes
Lao P.D.R	^/3	^/3
Maldives	^/3	No	Small State
Papua New Guinea	^/3	^/4

^{^1/}Excludes all countries that fail to meet the income threshold in 2021

^{^2/}Countries required to blend under the current rules and under the proposed rules are not shown in the table.

^{^3/}Blend status dependent on staff assessment of prospective market access

^{^4/}Blend status dependent on validation of scale of debt issuance in previous 5 years.

^{Annex III Table 2.}

Blend Status of Potential Presumed Blenders^1/ ^2/

	Current Rules	Proposed Rules	Comment
Cabo Verde	^/3	No	Small State
Cameroon	^/3	Yes
Dominica	^/3	No	Small State
Ghana	^/3	Yes
Kenya	^/3	Yes
Lao P.D.R	^/3	^/3
Maldives	^/3	No	Small State
Papua New Guinea	^/3	^/4

^{^1/}Excludes all countries that fail to meet the income threshold in 2021

^{^2/}Countries required to blend under the current rules and under the proposed rules are not shown in the table.

^{^3/}Blend status dependent on staff assessment of prospective market access

^{^4/}Blend status dependent on validation of scale of debt issuance in previous 5 years.

Eligibility for Exceptional Access to PRGT Resources

12. As of end May 2020, there were 28 countries eligible for EA to PRGT resources, based on GNI per capita levels in 2019 that lay below the IDA cutoff level. Absent any changes to this specification of eligibility, this number would increase to 32 with the release of GNI per capita data for 2020 on July 1, 2021, with the addition of Lesotho, Kyrgyz Republic, Zambia, and Zimbabwe.

13. With the proposal that eligibility for EA to PRGT resources be based on meeting the income threshold for blending, the number of countries eligible for EA would increase to 34, with the inclusion of Myanmar and Tanzania.^⁷

^{Annex III Box 1.}

Current Blending Policy

A. Blending. PRGT-eligible countries are expected to receive a blend of PRGT and GRA resources when they meet the following criteria:

For countries at low or moderate risk of debt distress (as assessed by the most recent joint Bank-Fund LIC Debt Sustainability Analysis (DSA)), blending is presumed if either (i) per capita income is above 100 percent of the International Development Association (IDA) operational cutoff; or (ii) the country has sustained past and prospective access to international financial markets and a per capita income that exceeds 80 percent of the IDA operational cutoff. The criterion for sustained past market access in such cases would be met if the country has issued or guaranteed eligible external debt during at least two of the past five years in a cumulative amount equivalent to at least 25 percent of the member’s quota.^¹ The criterion for prospective market access in such cases is assumed to be met based on established past market access and limited debt vulnerabilities as evidenced by low or moderate risk of debt distress.
For countries at high risk of debt distress, but not in debt distress, blending is presumed where (i) per capita income is above 100 percent of the IDA operational cutoff; (ii) the country has issued or guaranteed eligible external debt in at least three out of the past five years in a cumulative amount equivalent to at least 50 percent of its quota; and (iii) the country has prospective market access. The assessment of prospective market access in such cases would require judgment based on such factors as the evolution of debt vulnerabilities in the context of the program DSA, the evolution of sovereign spreads and credit ratings over time, program assumptions on commercial financing, and the scale and evolution of nonresident holdings of domestic-currency debt. The quality of public debt data—including coverage of public sector entities outside the central government and of publicly guaranteed debt, and transparency on terms and conditions—would also be an important factor in the assessment, given the threat to prospective market access from any significant debt surprises.^²

B. Access limits. When financing is blended under a PRGT arrangement and an arrangement under the GRA, total access is determined based on the standard criteria, implying that total access should be comparable across country cases with similar balance of payments needs, program strength, and outstanding Fund credit, irrespective of whether the Fund’s financial assistance comes in the form of blended or PRGT-only resources. The blending rules stipulate a 1:2 mix of PRGT and GRA resources, with access to concessional resources capped at the norm^³ (or equivalent) applicable to unblended arrangements. All access above the norm needs to be met from the GRA.

^¹ Accessing international financial markets refers to the issuance or guarantee by a public debtor of external bonds in international markets or disbursements under external commercial loans contracted or guaranteed by a public debtor in such markets: see “PRGT eligibility paper 2020.” ^² The potential for countries at high risk of debt distress to be classified as PBs was introduced in 2019 in the context of the 2018–19 Review of Facilities for Low-Income Countries: the reform was motivated by the observance of large-scale issuance of debt on international financial markets by countries at high risk of debt distress. ^³ High access norms (120 percent of quota for a 3-year ECF) apply if PRGT credit outstanding is less than 100 percent of quota. Low access norms (75 percent of quota) apply if PRGT credit outstanding is between 100 and 200 percent of quota. Norms are not applicable if PRGT credit outstanding >200 percent of quota. In such cases, access is guided by consideration of the cumulative access limit of 300 percent of quota under PRGT facilities (400 percent of quota in exceptional access cases), the expectation of future need for Fund support, and the repayment schedule.

^{Annex III Table 3.}

Blend Status of PRGT-Eligible Countries Under Proposed Reform

Source: World Bank, World Development Indicators.

^{^1/}Blend status dependent on validation of scale of debt issuance in previous five years.

^{Annex III Table 3.}

Blend Status of PRGT-Eligible Countries Under Proposed Reform

Countries not required to blend			Presumed Blenders
Country	2020 GNI per capita (US $)	Risk of Debt Distress (end-May 2021)	Country	2020 GNI per capita (US $)	Risk of Debt Distress (end-May 2021)
St. Lucia	8,790	High	Moldova	4,570	Low
Grenada	8,740	In debt distress	Bhutan	2,860	Moderate
St.Vincent and the Grenadines	7,340	High	Vanuatu	2,780	Moderate
Dominica	6,870	High	Lao PDR ^1/	2,480	High
Maldives	6,830	High	Solomon Islands	2,300	Moderate
Tuvalu	5,820	High	Côte d’Ivoire	2,280	Moderate
Samoa	4,070	High	Papua New Guinea ^1/	2,260	High
Djibouti	3,320	High	Ghana	2,230	High
Cabo Verde	3,060	High	Honduras	2,200	Low
Kiribati	3,010	High	Bangladesh	2,010	Low
São Tomé and Principe	2,070	In debt distress	Nicaragua	1,850	Moderate
Congo, Rep.	1,830	In debt distress	Timor-Leste	1,830	Low
Mauritania	1,640	High	Kenya	1,760	High
Myanmar	1,260	Low	Uzbekistan	1,670	Low
Haiti	1,250	High	Cameroon	1,500	High
Nepal	1,190	Low	Cambodia	1,490	Low
Zambia	1,190	In debt distress	Comoros	1,450	Moderate
Kyrgyz Republic	1,160	Moderate	Senegal	1,430	Moderate
Lesotho	1,100	Moderate	Benin	1,280	Moderate
Zimbabwe	1,090	In debt distress
Tanzania	1,080	Low
Tajikistan	1,060	High
Guinea	1,020	Moderate
Togo	920	Moderate
Ethiopia	890	High
Mali	830	Moderate
Uganda	800	Low
Burkina Faso	790	Moderate
Rwanda	780	Moderate
Guinea-Bissau	760	High
Gambia, The	750	High
Chad	660	High
Sudan	650	In debt distress
Malawi	580	Moderate
Congo, Dem. Rep.	550	Moderate
Niger	530	Moderate
Liberia	530	Moderate
Central African Republic	510	High
Afghanistan	500	High
Sierra Leone	490	High
Madagascar	480	Moderate
Mozambique	460	In debt distress
Somalia	310	In debt distress
Burundi	270	High
Eritrea	-	In debt distress
Marshall Islands	-	High
Micronesia	-	High
South Sudan	-	High
Tonga	-	High
Yemen	-	In debt distress

Source: World Bank, World Development Indicators.

^{^1/}Blend status dependent on validation of scale of debt issuance in previous five years.

^{Annex III Table 3.}

Blend Status of PRGT-Eligible Countries Under Proposed Reform

Countries not required to blend			Presumed Blenders
Country	2020 GNI per capita (US $)	Risk of Debt Distress (end-May 2021)	Country	2020 GNI per capita (US $)	Risk of Debt Distress (end-May 2021)
St. Lucia	8,790	High	Moldova	4,570	Low
Grenada	8,740	In debt distress	Bhutan	2,860	Moderate
St.Vincent and the Grenadines	7,340	High	Vanuatu	2,780	Moderate
Dominica	6,870	High	Lao PDR ^1/	2,480	High
Maldives	6,830	High	Solomon Islands	2,300	Moderate
Tuvalu	5,820	High	Côte d’Ivoire	2,280	Moderate
Samoa	4,070	High	Papua New Guinea ^1/	2,260	High
Djibouti	3,320	High	Ghana	2,230	High
Cabo Verde	3,060	High	Honduras	2,200	Low
Kiribati	3,010	High	Bangladesh	2,010	Low
São Tomé and Principe	2,070	In debt distress	Nicaragua	1,850	Moderate
Congo, Rep.	1,830	In debt distress	Timor-Leste	1,830	Low
Mauritania	1,640	High	Kenya	1,760	High
Myanmar	1,260	Low	Uzbekistan	1,670	Low
Haiti	1,250	High	Cameroon	1,500	High
Nepal	1,190	Low	Cambodia	1,490	Low
Zambia	1,190	In debt distress	Comoros	1,450	Moderate
Kyrgyz Republic	1,160	Moderate	Senegal	1,430	Moderate
Lesotho	1,100	Moderate	Benin	1,280	Moderate
Zimbabwe	1,090	In debt distress
Tanzania	1,080	Low
Tajikistan	1,060	High
Guinea	1,020	Moderate
Togo	920	Moderate
Ethiopia	890	High
Mali	830	Moderate
Uganda	800	Low
Burkina Faso	790	Moderate
Rwanda	780	Moderate
Guinea-Bissau	760	High
Gambia, The	750	High
Chad	660	High
Sudan	650	In debt distress
Malawi	580	Moderate
Congo, Dem. Rep.	550	Moderate
Niger	530	Moderate
Liberia	530	Moderate
Central African Republic	510	High
Afghanistan	500	High
Sierra Leone	490	High
Madagascar	480	Moderate
Mozambique	460	In debt distress
Somalia	310	In debt distress
Burundi	270	High
Eritrea	-	In debt distress
Marshall Islands	-	High
Micronesia	-	High
South Sudan	-	High
Tonga	-	High
Yemen	-	In debt distress

Source: World Bank, World Development Indicators.

^{^1/}Blend status dependent on validation of scale of debt issuance in previous five years.

Annex IV. Evolution of PRGT Access Limits

This annex discusses the evolution of PRGT access limits since 2015, including the changes implemented in the context of the pandemic.

1. Limits on access to PRGT resources have served in the main to ration access to scarce concessional resources: they also have helped to mitigate credit risk in cases where LICs eligible for EA request support in excess of normal access limits, given the stronger policy requirements of the EA criteria. These limits have been raised periodically to address erosion relative to established metrics of demand—including GDP, trade, and external financing needs—and to respond to higher financing needs arising from global economic developments.

A member’s total access under all concessional facilities is subject to “global” annual and cumulative limits. This includes the ECF, SCF, and RCF. The annual access limit refers to disbursements in any 12-month period, on a rolling basis. The cumulative access limit refers to total outstanding Fund concessional credit at any point in time, after accounting for projected disbursements and repayments. There are normal limits on access and exceptional limits (hard caps) on access, with EA available only to the poorest LICs.
In addition to global limits on access under all PRGT facilities, access to the RCF is subject to annual and cumulative sub-ceilings. These sub-ceilings are differentiated across “the regular,” “exogenous shocks,” and “large natural disasters” windows of the RCF. Purchases under the RFI count towards the applicable RCF annual and cumulative sub-ceilings.

2. There have been three significant changes to PRGT access limits since establishment of the PRGT facilities framework in 2009:

2015–17:^¹ Annual and cumulative access limits were raised by 50 percent across all facilities (including the RCF) on July 1, 2015 to address the erosion of access levels relative to trade, capital flows, and GDP since 2009–10 and to make available additional support to the poorest LICs in the context of supporting efforts to achieve the Sustainable Development Goals (SDGs). The increases in access limits, expressed as a share of quota, were reduced by one-half in January 2016 as the 14th General Review of Quotas became effective, leaving access in SDR terms for most countries unaffected by the quota increase. Additionally, in May 2017, a large natural disaster window, allowing higher annual access levels than other windows, was created under the RCF and RFI.
2018–19 Review of Facilities for LICs:^² Annual and cumulative access limits and norms were raised by one-third across all lending instruments in May 2019. The increase was intended to offset access erosion and restore access limits in relation to GDP and trade to the levels achieved when generalized access increases had occurred in 2009 and 2015, and preserve the potential financing contribution of Fund program engagement in LICs. The cumulative RCF/RFI access limits for disbursements associated with large natural disasters were raised by an additional one-third to provide room to support members hit by a large natural disaster that already had significant outstanding RCF/RFI exposure.
2020–21 Temporary modifications to access limits in response to COVID-19:^³
- i. In April 2020, with the onset of the pandemic, limits on annual and cumulative access under the RCF exogenous shocks window were increased from 50 percent of quota and 100 percent of quota to 100 percent and 150 percent, respectively, with similar increases introduced for the RFI. The new limits applied initially for a six-month period and were later extended through end-2021.
- ii. In July 2020, the normal annual access limit (NAAL)^⁴ on use of PRGT resources was raised from 100 to 150 percent of quota through April 6, 2021. The increase was intended to make room for higher access – without triggering the application of the EA framework – for countries that had used up much of the annual borrowing space under the NAALs due to COVID-19 related EF.
- iii. In March 2021, PRGT global annual and cumulative access limits were temporarily raised through end-June 2021: the NAAL from 150 to 245 percent of quota, the normal cumulative access limit (NCAL) from 300 to 435 percent of quota.^⁵ These increases were intended to create more room to provide concessional financing to LICs—and to avoid requiring LICs with high outstanding exposure to seek Fund support through the GRA—in the uncertain environment created by the pandemic, pending a wider discussion of the Fund’s concessional finances and policies. On June 25, 2021, staff proposed a temporary extension of the increased limits to July 31, 2021.
- iv. In June 2021, annual and cumulative access limits under the Large Natural Disaster (LND) window of the RCF and RFI were increased by 50 percent of quota, from 80 percent of quota and 133.33 percent of quota to 130 percent and 183.33 percent, respectively, until end-2021. The increase was in line with the April 2020 increase of normal and cumulative access limits under the RFI and RCF exogenous shocks window, and was intended to allow for augmented access by countries vulnerable to LNDs as was the case in the pre-pandemic period.

^{Annex IV Table 1.}

PRGT Global Access Limits (in percent of quota)

^{Annex IV Table 1.}

PRGT Global Access Limits (in percent of quota)

	Completion of 14th quota review (January 2016)	Review of Facilities for LICs (May 2019)	Temporary modification to annual access limits until end-April 2021 (July 2020)	Temporary modification to access limits until end-June 2021 (March 2021)	Proposed access limits
*Cumulative access limits*
All PRGT facilities-normal	225	300	300	435	435
All PRGT facilities-exceptional	300	400	400	535	No hard cap
*Annual access limits*
All PRGT facilities-normal	75	100	150	245	Annual limit is at 245 percent of quota until end-2021, after which it would decline to 145 percent of quota.
All PRGT facilities-exceptional	100	133.33	183.33	278.33	No hard cap
*Norms*
3-year ECF – High access	90	120	120	120	Norm is set at 145,
- Low access	56.25	75	75	75	independent of the stock of credit outstanding.

^{Annex IV Table 1.}

PRGT Global Access Limits (in percent of quota)

	Completion of 14th quota review (January 2016)	Review of Facilities for LICs (May 2019)	Temporary modification to annual access limits until end-April 2021 (July 2020)	Temporary modification to access limits until end-June 2021 (March 2021)	Proposed access limits
*Cumulative access limits*
All PRGT facilities-normal	225	300	300	435	435
All PRGT facilities-exceptional	300	400	400	535	No hard cap
*Annual access limits*
All PRGT facilities-normal	75	100	150	245	Annual limit is at 245 percent of quota until end-2021, after which it would decline to 145 percent of quota.
All PRGT facilities-exceptional	100	133.33	183.33	278.33	No hard cap
*Norms*
3-year ECF – High access	90	120	120	120	Norm is set at 145,
- Low access	56.25	75	75	75	independent of the stock of credit outstanding.

^{Annex IV Table 2.}

Access Limits to Emergency Financing Instruments (in percent of quota)

^{Annex IV Table 2.}

Access Limits to Emergency Financing Instruments (in percent of quota)

	Completion of 14th quota review (January 2016)	Large Natural Disasters-Enhancing the Financial Safety Net (May 2017)	Review of Facilities for LICs (May 2019)	Temporary modifications to access limit until end-2021 (April 2020, October 2020, March 2021, June 2021)
*Cumulative access limits*
RCF (regular window)	75	75	100	100
RCF/RFI (large natural disasters window)	N.A.	75	133.33	183.33
RCF (exogenous shocks window)/ RFI (regular window)	75	75	100	150
*Annual access limits*
RCF (regular window)	18.75	18.75	50	50
RCF/RFI (large natural disasters window)	N.A.	60	80	130
RCF (exogenous shocks window)/ RFI (regular window)	37.5	37.5	50	100

^{Annex IV Table 2.}

Access Limits to Emergency Financing Instruments (in percent of quota)

	Completion of 14th quota review (January 2016)	Large Natural Disasters-Enhancing the Financial Safety Net (May 2017)	Review of Facilities for LICs (May 2019)	Temporary modifications to access limit until end-2021 (April 2020, October 2020, March 2021, June 2021)
*Cumulative access limits*
RCF (regular window)	75	75	100	100
RCF/RFI (large natural disasters window)	N.A.	75	133.33	183.33
RCF (exogenous shocks window)/ RFI (regular window)	75	75	100	150
*Annual access limits*
RCF (regular window)	18.75	18.75	50	50
RCF/RFI (large natural disasters window)	N.A.	60	80	130
RCF (exogenous shocks window)/ RFI (regular window)	37.5	37.5	50	100

Annex V. A Dual Interest Rate Mechanism in the PRGT

Under current policies, all loans extended from the PRGT carry the same interest rate, based on an interest rate mechanism that yields zero or near-zero rates, depending on prevailing SDR rates. This Annex illustrates how a potential dual interest rate structure could be introduced in the PRGT to allow all LICs to meet their entire financing needs from within the PRGT. The proposal here would provide benefits to countries that would otherwise need to blend PRGT and GRA resources, while providing a modest reduction in subsidy outlays by the PRGT. Such a reform would lead to a substantial increase over time in PRGT credit outstanding, PRGT loan resource needs, and a significant reduction in the reserve coverage ratio unless new resources to boost reserve coverage are made available.

1. The PRGT would have two interest rates: one (RA), set in accordance with the interest rate mechanism, which currently implies a zero rate for all PRGT facilities; and a second (higher) interest rate (RB) linked to, but less than, the GRA rate of charge (the SDR interest rate (SDRi) + 100 basis points).

2. The blending criteria discussed in the main text would now serve as the criteria for determining which of the two interest rates a country would incur when borrowing from the PRGT. Countries currently required to blend PRGT and GRA resources (henceforth “intermediate interest rate countries” (IIRs)) would now be eligible to meet their entire financing needs from the PRGT, with RB as the interest rate. LICs that are not IIRs would face unchanged borrowing conditions.

3. To see how this would affect IIRs and PRGT finances, let RB be set at two-thirds of the current rate of charge (the average interest rate paid on a 1:2 PRGT-GRA blending mix). Under this approach:

IIRs would benefit in that i) they would no longer be required to meet the policy requirements of the GRA, which are less suitable for the needs of LICs than the conditions for borrowing under an ECF;^¹ ii) the repayment periods would be somewhat more generous than with current blended arrangements; and iii) they would not face GRA interest rate surcharges.^² Each of these features fits better the needs of LICs, which typically face protracted BoP problems, than does the current approach of mixing GRA and PRGT funding sources.
Scarce PRGT subsidy resources would be conserved, with IIR country borrowers generating income for the PRGT endowment when SDRi < 2.0 percent and requiring fewer subsidy resources than the 1:2 PRGT-GRA funding at any level of SDRi.^³ The improvement in the income position of the PRGT endowment would come at the cost of reduced income for the GRA.

4. The need for PRGT loan resources would increase significantly, since all IIRs’ financing needs would now be met from the PRGT (rather than one-third with blending). Higher lending volumes from the PRGT would imply a substantial increase in credit outstanding and a decline in the reserve coverage ratio (see Box 3 in Section V).

5. Adopting this alternative approach to the current blending framework would require several changes to PRGT rules and design, including:

the introduction of two sets of interest rates within the PRGT, with the group of countries not eligible for the lower interest rate being determined by what are now the blending rules;
reforms to allow the transfer of interest income (net of the cost of payments to the loan provider) to either the subsidy or reserve account of the PRGT;
transitional arrangements to exempt IIR countries with outstanding PRGT credit from higher interest rates on outstanding loans; and
a mechanism to bolster the reserve coverage ratio, discussed in Section VI of this paper.

6. Preliminary analysis of the legal changes needed to address the first three of these issues suggests that the changes could be approved by the Executive Board with a majority of the votes cast.

7. Separate from these legal changes, lending at the IIR would need to be consistent with the general purpose of the PRGT to provide loans on concessional terms.^⁴ While concessionality is not defined in the PRGT Instrument, the following arguments could be offered:

Grant Element Approach:

Under the debt limits policy, the concessionality of a loan is measured by the associated grant element (GE), calculated using a discount rate determined during the periodic reviews of the Bank-Fund LIC-DSF: this rate has been 5 percent since October 2013.
A zero-interest rate loan under the RCF/ECF has a GE of some 32 percent: a zero-interest rate on SCF terms has a GE of some 26 percent. By contrast, a loan on RCF/ECF maturities with an interest rate equal to 2/3*(SDRi +1.0) has a GE of about 28 percent at today’s (very low) SDRi rate, declining to 20 percent at SDRi= 2 percent and falling further as SDRi rises.
Lending on concessional terms for purposes of the PRGT instrument could be defined as loans with a GE that exceeds some minimum threshold level: given ECF/RCF maturities, this would determine a maximum level of the interest rate paid by IIRs (Rcon) that would meet the concessionality requirement
The IIR would be set in the context of the biennial review of interest rates—for example, as the lesser of [%*(SDRi +1.0)] and Rcon. Rising levels of SDRi would eventually shift the interest rate to Rcon, at which point the financial benefits to the PRGT cited above would decline/disappear.

Benefit to the Borrower:

Lending from the PRGT at an intermediate-level interest rate can be viewed as concessional on the basis that it provides the borrower with funding on more attractive terms than the GRA and on more attractive terms than the current 1:2 blended arrangements.

Annex VI. Analysis of Debt Sustainability and Capacity to Repay

1. Debt vulnerabilities have been increasing in LICs for several years, with the onset of the pandemic—and the associated weakening of tax bases and export receipts—adding new pressures.^¹ As of end-May 2021, some 42 percent of LICs were assessed to be at high risk of experiencing debt distress with a further 14 percent in debt distress: the comparable numbers for end-2016 were 26 percent and 7 percent, respectively.

2. Staff analytical and policy tools have been modified in response to these developments. An upgrade of the framework for analyzing debt sustainability in LICs (the LIC-DSF) was introduced in mid-2018, with a modified statistical methodology to improve prediction accuracy, new tools for assessing the realism of underlying macroeconomic projections, and greater scope to use customized scenarios adapted to the specific country context.^² A joint IMF-WB multipronged approach to addressing emerging debt vulnerabilities has been under implementation stage since late-2018.^³ A review of the Fund’s Debt Limits Policy, which guides how debt conditionality is deployed in Fund programs, was completed in October 2020, with reforms introduced to improve debt disclosure and allow better adjustment of debt conditionality to country circumstances: the new policy takes effect on June 30, 2021.^⁴

3. The proposed increase in normal access limits in the PRGT would affect the approach taken to debt sustainability in certain circumstances—specifically, for programs with access requests in excess of the current normal access limits of 100/300 percent of quota but below the new limits of 145/435.^⁵ For the 28 countries currently eligible for PRGT EA, a request for access to PRGT resources within this range would no longer require meeting the PRGT EA criteria (with its requirement to reduce debt vulnerabilities to low/moderate risk): for the 20-plus non-blend countries that are not currently eligible, access in this range could now be met entirely from the PRGT (and hence without meeting the policy requirements of the GRA).

4. The March 2021 staff paper on temporary increases in access levels specified new requirements intended to bolster scrutiny of debt sustainability and capacity to repay the Fund, applying to requests for arrangements with access above the current normal access limits (100/300) and to all requests for arrangements from countries at high risk of, or in, debt distress.^⁶

5. In all such cases, program documents are expected to include discussion of:^⁷

the structure of public external debt and its projected evolution over time, focusing on the amount and shares of debt owed to the Fund and other senior creditors, informed by tables showing two distinct breakdowns of public external debt: i) de facto senior debt (debt to the IMF; debt to the World Bank and other international financial institutions; known collateralized debt) and other debt and ii) multilateral versus official bilateral versus private debt.^⁸
the evolution of projected Fund debt and debt service relative to key economic metrics over the course of the repayment period as compared with other PRGT programs, supported by a set of standardized charts provided by the Finance Department (see below). Where financing requests would result in comparatively elevated levels of key capacity to repay indicators, the staff report would examine the severity of the implied risks and explain how program design—including access, phasing, and conditionality—seeks to mitigate these risks.

6. For countries at high risk of debt distress or in debt distress, the core program objectives should include the achievement of a concrete reduction in debt vulnerabilities over the course of the program and beyond.^⁹ Reducing debt vulnerabilities would typically involve reducing breaches of thresholds for the four key indicators in the LIC-DSF over the program period under the baseline scenario:^¹⁰ staff do not propose a mechanical approach to assessing the projected improvement in debt vulnerabilities, favoring instead an overall assessment of the strength of the program and any assurances from creditors on new concessional financing or restructuring of existing claims.

7. To facilitate a comparative assessment of projected levels of debt and debt service to the Fund in a proposed program, staff in the Finance Department have developed a methodology that allows graphical comparison of the evolution of key debt metrics under the program with a control group of PRGT arrangements.^¹¹ The relevant metrics would include the projected stock of Fund credit outstanding relative to i) quota, ii) GDP, and iii) the aggregate level of public and publicly-guaranteed (PPG) external debt; and projected annual debt service to the Fund relative to i) fiscal revenues (excluding grants), ii) exports of goods and services, iii) all debt service on PPG debt, and iv) the level of gross international reserves.^¹² For an illustration focused on four key metrics, see Annex VI Figure 2. The comparative assessment will be based on the baseline scenario underpinning the proposed program but could also include information on downside scenarios included in program documents, and realism checks whenever warranted. Guidance will be provided to staff on how to reflect information on Fund debt metrics in capacity to repay assessments.

Annex VII. Methodology for Estimating PRGT Resource Needs

1. The lending scenarios presented in this paper draw on staff’s analysis of crisis-related demand for PRGT resources and the outlook for the post-pandemic decade. Staff’s scenarios comprise (i) an in-depth country-by-country analysis of the potential demand ranges for the crisis period (2020–24) and (ii) illustrative lending envelopes for the post-pandemic decade (2025–34), together with an analysis of the PRGT’s self-sustainability. Key features are summarized below.

2. All lending scenarios build on the access and blending policy changes discussed in this paper, including an across-the-board 45 percent increase in normal annual and cumulative access limits to 145 and 435 percent of quota, respectively, a unified access norm of 145 percent of quota per 3-year arrangement, and the removal of a hard cap on PRGT access for countries that do not meet the proposed income threshold for blending on the IDA cutoff. These access policies are assumed to remain unchanged until the next LIC Facilities and Financing Review in 2024/25. The temporary access increases for the RCF and normal annual PRGT access are assumed to expire at end-2021.

3. All lending scenarios apply these access rules to demand projections by country, based on existing credit exposures and previous disbursements for each country. Subject to applicable caps, blenders are assumed to be financed by PRGT/GRA at a 1:2 ratio.

4. The scale of financing under multiyear financing arrangements during the pandemic period (through 2024) is calibrated to reflect the exceptionally high financing needs, with average access scaled up to almost twice the level observed in recent years (and at three times in the High Case). Access per arrangement is differentiated by the country’s degree of debt vulnerabilities, blend status, and pre-existing credit exposure to the Fund, and subject to the access limits proposed in the paper. Given the wide range of quota/GDP ratios among LICs and to capture potential BoP needs, access is a function of both quota and GDP, subject to lower and upper limits.

5. The near-term projections make an allowance for new EF and augmentations of pre-existing arrangements subject to applicable sub-limits on access. All scenarios also accommodate some degree of Fund financing for health-related and vaccine needs.

6. Total PRGT lending is ultimately dependent on demand, i.e., on the share of LICs seeking Fund financial support. The Baseline through 2024 assumes that almost two-thirds of eligible LICs (on a quota-weighted basis) request program support, which is in line with peak demand years during the global financial crisis and the percentage of LICs that requested financial support in 2020. The Low Case assumes 40 percent (the longer-term average) and the High Case assumes about 80 percent, which would be unprecedented. The range between the Low and High Cases reflect the large uncertainty around economic developments and demand for Fund financing over the coming years.

7. The longer-term demand projection allows for periodic increases in nominal access levels per country in line with GDP growth, partly offset by savings from transitions from non-blending to blending, and from blending to graduation. The Baseline and Low Case projection for the post-crisis decade (2025–34) projects average annual demand of SDR 1.65 billion based on an assumption that future access increases are calibrated to preserve access in real terms relative to pre-pandemic levels. The longer-term envelope for the High Case is calibrated to per-country access in line with GRA arrangements for emerging market countries, implying SDR 3 billion of average annual lending in the post-crisis decade.

8. The demand projections are then fed into a capacity/supply model, which translates projected commitments into disbursements and credit outstanding, calculates subsidization costs and the evolution of investment returns on PRGT assets, and produces an estimate of the PRGT’s self-sustained lending capacity at different points in time based on existing subsidy and reserve account resources. The combination of demand scenarios and supply analysis then provides a complete picture of the PRGT’s lending capacity and the resource implications discussed in this note for each scenario.

Annex VIII. The PRGT Financing Model

1. PRGT lending is supported by an endowment-based financing model that relies on loan and subsidy resources. Loan resources are provided through bilateral agreements with members and on-lent by the PRGT on a passthrough basis to LICs. Loan resources are typically remunerated by the PRGT at the SDR interest rate and secured through the PRGT’s RA and are mobilized periodically depending on expected resource needs. Balances in the PRGT’s subsidy accounts and the RA (SDR 3.9 billion and SDR 4.1 billion respectively as of end-April 2021) as well as any investment earnings on these account—are used to cover the interest cost of PRGT loans, so that borrowers can benefit from concessional (currently zero) interest rates. Under the self-sustained model, the resources in PRGT subsidy accounts would be gradually drawn down to zero, while balances in the RA would grow over time by the amount of investment returns on the RA balance, until returns on its assets would subsidize PRGT lending in perpetuity.

2. The framework for the PRGT provides for the annual reimbursement of the GRA for the expenses of conducting the business of the PRGT. This reimbursement is an integral part of the Fund’s new income model approved in 2008 based on the principle that the GRA should not cross-subsidize the PRGT’s activities. However, it was explicitly recognized that reimbursement can be temporarily suspended when the resources in the PRGT are insufficient to meet expected demand.

The reimbursement of the PRGT can be waived notwithstanding that the PRGT contains Special Disbursement Account (SDA) resources derived from gold sales profits. According to Article V, Section 12(i), the SDA needs to reimburse the GRA for expenses in administering resources of the Special Disbursement Account (SDA). The PRGT has SDA resources in the Reserve Account and the General Subsidy Account. Reimbursement has been waived for FY1998–FY2004 (to fund contributions for HIPC Initiative) and FY2005–FY2012 to provide contributions for PRGT. The suspension was taken on the understanding that reimbursable expenses would only arise where PRGT principal loan disbursements would be funded with SDA resources, and that absent such use, the Fund may decide that the GRA bear the cost of the PRGT. As indicated in the FY2020 and FY2021 income papers, staff has started a review of the reimbursement practices under various trusts funded with SDA resources, the completion of which has been delayed due to the many urgent Covid-19 response related priorities. While staff had aimed at completing the review as part of this review of concessional financing, this was not possible and staff will present it at the next possible opportunity.

3. The three-pillar strategy was adopted in 2012 to make the PRGT’s lending self-sustaining without the need for periodic subsidy mobilization. Following the transfer of windfall profits from gold sales and additional bilateral grants, the PRGT’s subsidy resources were considered adequate for a sustained level of lending in perpetuity without the need for regular subsidy contributions from the Fund’s membership. The three-pillar strategy consists of (i) a base envelope of SDR 1¼ billion in annual PRGT lending capacity, which is expected to cover concessional lending needs during normal periods; (ii) contingent measures that can be invoked when average financing needs exceed the base envelope by a substantial margin for an extended period, including additional bilateral fund-raising, suspending reimbursement of the GRA for PRGT administrative expenses for a limited period, and modifying access, blending, interest rate, and eligibility policies to reduce the need for subsidy resources; and (iii) a principle of self-sustainability under which future modifications to LIC facilities would be expected to ensure that demand for IMF concessional lending can be reasonably met with the available resources.

4. The adequacy of resources under the self-sustained PRGT is assessed annually. The analysis includes several elements: (i) short-term demand projections and sensitivity analyses derived from country desk surveys; (ii) a demand model that is used to project medium- to longer-term concessional lending, based on specific policy assumptions (e.g., access, blending, graduation) and plausible demand scenarios derived from historical patterns; (iii) an assessment of available PRGT loan resources under different near- to medium-term demand scenarios; and (iv) a capacity (“supply”) model that calculates the PRGT’s self-sustained lending capacity based on available subsidy resources and projected demand. This analysis informs staff’s assessment of the adequacy of the overall framework, the affordability of any policy refinements, and the potential need for corrective contingency measures, including possible loan mobilization or subsidy fundraising.

5. In the May 2019 Reviews of LIC Facilities and Concessional Financing, the Fund adopted several reforms that were calibrated to be consistent with the self-sustainability of the PRGT. The reform package included a generalized one-third increase in access limits and norms. Together with other policy changes, this was projected to result in average annual demand of SDR 1.0–1.7 billion over the next decade. Based on this demand range, the PRGT’s self-sustained annual lending capacity would reach a range of SDR 1.1–1.4 billion by 2028, symmetric around the target of SDR 1¼ billion. Loan resources were deemed sufficient to cover the PRGT’s lending operations well into the next decade. The review noted that the evolution of the PRGT’s self-sustained capacity would require careful monitoring given downside risks.

6. The COVID-19 pandemic is a major negative shock that hit hard all LICs and led to unprecedented demand for concessional financing. Even under current policies, the crisis has already eroded the self-sustained annual lending capacity to the lower end of the above range. Without fresh subsidy resources, the lending capacity is projected to fall well below the range under any plausible scenarios featuring larger PRGT lending over the medium- to longer-term to better meet LICs’ evolving financing needs. Bringing the lending capacity up to the “new normal” while preserving the self-sustained endowment model would require significant injections of new subsidy resources as discussed above.

Annex IX. PRGT Loan Resource Mobilization

1. The fast-track PRGT loan mobilization round launched in April 2020 has secured about SDR 17 billion from 16 PRGT lenders. Members responded quickly to the call for urgently needed new loan resources. The resources provided so far are expected to cover loan needs under current policies. However, as mentioned in the main text, additional loan resources (almost SDR 13 billion) would be needed on top of the resources secured so far to cover crisis-related demand under all scenarios, including the High Case.

2. The new loan agreements include several improved features, including expanded use of SDRs; a broadened lender base compared to the previous round (e.g., Australia, Germany); de-earmarking to allow use for all PRGT facilities; unification of lenders’ interest rate at SDRi; easing of drawing limits; and extended drawdown period (2024–29).

^{Annex IX Table 1.}

New PRGT Loan Resources Effected under the 2020 Round1

(As of June 23, 2021)

Source: Finance Department. 1 All agreements are for the benefit of the General Loan Account, remunerated at the SDR interest rate (with the exception of the UK loan capped at 0.05%) and expire at end-2029. With the exception of Germany, all loans are denominated in SDRs. 2 To be available in two equal tranches. 3 SDR equivalent of EUR 3 billion at the exchange rate of January 11, 2021 when the agreement became effective. The actual value of the loan will be calculated at the exchange rate at the time of drawings and net of operational expenses incurred by Germany.

^{Annex IX Table 1.}

New PRGT Loan Resources Effected under the 2020 Round1

(As of June 23, 2021)

Country	SDR Million	USD Million	Modality	Media	Type of Agreement	Encashment
Japan 2	3,600	5,143	Augmentation	SDR	NPA	Yes
Germany 3	2,534	3,619	New agreement	EUR	Loan Agreement	No
France	2,000	2,857	New agreement	SDR	Loan Agreement	Yes
UK	2,000	2,857	Augmentation	SDR	NPA	Yes
China	1,000	1,429	New agreement	SDR	NPA	Yes
Italy	1,000	1,429	New agreement	SDR	Loan Agreement	Yes
Spain	750	1,071	Augmentation	EUR	Loan Agreement	Yes
Australia	500	714	New agreement	SDR	Loan Agreement	Yes
Brazil	500	714	Augmentation	USD	NPA	Yes
Canada	500	714	Augmentation	USD	Loan Agreement	No
Netherlands	500	714	New agreement	SDR	Loan Agreement	No
Sweden	500	714	New agreement	USD	Loan Agreement	Yes
Switzerland	500	714	New agreement	EUR	Loan Agreement	No
Norway	400	571	New agreement	USD	Loan Agreement	Yes
Belgium	350	500	New agreement	EUR	Loan Agreement	No
Denmark	300	429	New agreement	EUR	Loan Agreement	No
Total	16,934	24,191

^{Annex IX Table 1.}

New PRGT Loan Resources Effected under the 2020 Round1

(As of June 23, 2021)

Country	SDR Million	USD Million	Modality	Media	Type of Agreement	Encashment
Japan 2	3,600	5,143	Augmentation	SDR	NPA	Yes
Germany 3	2,534	3,619	New agreement	EUR	Loan Agreement	No
France	2,000	2,857	New agreement	SDR	Loan Agreement	Yes
UK	2,000	2,857	Augmentation	SDR	NPA	Yes
China	1,000	1,429	New agreement	SDR	NPA	Yes
Italy	1,000	1,429	New agreement	SDR	Loan Agreement	Yes
Spain	750	1,071	Augmentation	EUR	Loan Agreement	Yes
Australia	500	714	New agreement	SDR	Loan Agreement	Yes
Brazil	500	714	Augmentation	USD	NPA	Yes
Canada	500	714	Augmentation	USD	Loan Agreement	No
Netherlands	500	714	New agreement	SDR	Loan Agreement	No
Sweden	500	714	New agreement	USD	Loan Agreement	Yes
Switzerland	500	714	New agreement	EUR	Loan Agreement	No
Norway	400	571	New agreement	USD	Loan Agreement	Yes
Belgium	350	500	New agreement	EUR	Loan Agreement	No
Denmark	300	429	New agreement	EUR	Loan Agreement	No
Total	16,934	24,191

Annex X. PRGT Subsidy Resources and Reserve Account

1. The PRGT’s endowment has been built over several decades, relying on a mix of internal resources and donor contributions. The current framework for financing the Fund’s concessional lending, with loans provided by members on market terms, credit risk mitigated by the Reserve Account, and interest rate subsidization from grant resources, was first established in 1987^¹ when the ESAF Trust, the predecessor of the current PRG Trust, replaced the early Trust Fund.^² Several fundraising rounds to secure loan and subsidy resources relying on members’ generosity have been completed since then. To date, members have voluntarily provided about SDR 5.3 billion to the framework’s subsidy accounts and made close to SDR 55.5 billion available in loan resources. The Fund contributed about SDR 5.5 billion in internal resources, mostly through the recycling of resources originating from the 1976–80 gold sales^³ and non-reimbursement of the GRA.^⁴ The funding sources varied:

The PRGT’s subsidy accounts have been mostly funded by bilateral contributions from economically stronger members, predominantly in the form of grants. Several members provided significant contributions through concessional loans remunerated at below market rate (allowing the Trust to save about SDR 0.3 billion in subsidy resources). The 2012–13 distributions of windfall gold sales profits facilitated bilateral contributions of about SDR 2.2 billion from a wider base of 152 members. The Fund also contributed some of the Trust Fund reflows to the PRGT’s subsidy accounts, including SDR 148 million transferred from the Reserve Account in lieu of non-reimbursement of the GRA during FY2010–12.
The PRGT’s Reserve Account has been fully funded from resources originating from the 1976–80 gold sales. The Account’s current balance of SDR 4.1 billion includes income earned over time on its balances and about SDR 324 million retained from non-reimbursement of the GRA during FY2005–09 and FY2021. Table 1 provides a summary of contributions provided to the PRGT in the past.

^{Annex X Table 1.}

Historical Contributions to the PRGT

(As of end-April 2021)

^¹Staff estimates of cumulative contributions (i.e., grants, returns on members’ deposits and implicit contributions) made to the PRGT and its predecessors under all fundraising efforts since 1987, including income earned on outstanding balances of the Trust and excluding contributions transferred to the MDRI-II Trust in January 2006.

^²Cash basis.

^³Includes transfers from SDA and income earned on balances.

^⁴Amount transferred from members’ contributions to the ESF-PRGF Trust’s subsidy account in January 2006.

^{Annex X Table 1.}

Historical Contributions to the PRGT

(As of end-April 2021)

			In SDR million
Total contributions to PRGT subsidy accounts			6 ,715
	Bilateral contributions from members^¹		5,304
	of which:
		from 2012–13 windfall gold sales profits distribution^²	2,188
		implicit subsidies^²	326
	Contributions from the Fund (SDA)		1,411
		of which GRA non-reimbursement2	148
Reserve Account (RA) balance^³			4 ,115
	of which GRA non-reimbursement^²		324
*Memorandum items*^²
SDA Contributions to the PRG-HIPC Trust			1,167
	of which GRA non-reimbursement		366
SDA Contributions to the MDRI-I Trust			1,500
Bilateral contributions to the MDRI-II Trust^⁴			1,120

^¹Staff estimates of cumulative contributions (i.e., grants, returns on members’ deposits and implicit contributions) made to the PRGT and its predecessors under all fundraising efforts since 1987, including income earned on outstanding balances of the Trust and excluding contributions transferred to the MDRI-II Trust in January 2006.

^²Cash basis.

^³Includes transfers from SDA and income earned on balances.

^⁴Amount transferred from members’ contributions to the ESF-PRGF Trust’s subsidy account in January 2006.

^{Annex X Table 1.}

Historical Contributions to the PRGT

(As of end-April 2021)

			In SDR million
Total contributions to PRGT subsidy accounts			6 ,715
	Bilateral contributions from members^¹		5,304
	of which:
		from 2012–13 windfall gold sales profits distribution^²	2,188
		implicit subsidies^²	326
	Contributions from the Fund (SDA)		1,411
		of which GRA non-reimbursement2	148
Reserve Account (RA) balance^³			4 ,115
	of which GRA non-reimbursement^²		324
*Memorandum items*^²
SDA Contributions to the PRG-HIPC Trust			1,167
	of which GRA non-reimbursement		366
SDA Contributions to the MDRI-I Trust			1,500
Bilateral contributions to the MDRI-II Trust^⁴			1,120

^¹Staff estimates of cumulative contributions (i.e., grants, returns on members’ deposits and implicit contributions) made to the PRGT and its predecessors under all fundraising efforts since 1987, including income earned on outstanding balances of the Trust and excluding contributions transferred to the MDRI-II Trust in January 2006.

^²Cash basis.

^³Includes transfers from SDA and income earned on balances.

^⁴Amount transferred from members’ contributions to the ESF-PRGF Trust’s subsidy account in January 2006.

2. Solidarity with low-income members and burden sharing among donors have been crucial to the PRGT’s framework since its inception. Under past fundraising rounds, bilateral contributions were provided mostly by economically stronger members and typically in proportion to their quota shares. Based on this approach and the proven generosity of the Fund’s membership, the proposed concessional financing package relies on a mix of internal and donor resources.

In the first stage of the two-stage funding strategy, staff now propose to suspend reimbursement to the GRA for PRGT administrative expenses through FY2026, which would retain an additional SDR 0.5 billion in the PRGT Reserve Account and support the PRGT’s self-sustained concessional lending capacity.
In addition, to fully cover the subsidy costs created by COVID-related PRGT lending to LICs, voluntary bilateral subsidy contributions totaling SDR 2.3 billion will be requested from 61 members considered to be in a comparatively stronger economic position, specifically those that currently participate in the IMF’s Financial Transaction Plan (FTP), plus non-FTP members that belong to the G20 or EU (except those that have used Fund resources for BoP needs over the past three years). This group, which is similar to the one recently approached under the CCRT fundraising campaign, accounts for about 88 percent of current IMF quotas and in the past have supplied about 94 percent of total bilateral contributions to the PRGT’s subsidy accounts.
Table 2 provides a breakdown by country of cumulative PRGT subsidy contributions, including through implicit subsidization and investment returns. It also includes, for illustration, an indicative breakdown of how the SDR 2.3 billion subsidy gap could be closed through voluntary bilateral contributions based on the quota shares of these member countries.
To provide donors with flexibility, various options for bilateral subsidy contribution schemes are available as discussed in the main text, with resources pledged upfront and disbursed over time.
A decision on the appropriate longer-term PRGT envelope would be taken up at the second stage of the funding strategy. Possible additional use of IMF internal resources will be considered during the “stage two” review in 2024/25.

^{Annex X Table 2.}

Bilateral Contributions to the PRGT

(In SDR million unless otherwise noted)

^¹Staff estimates of cumulative contributions (i.e. grants, returns on members’ deposits and implicit contributions) made to the PRGT and its predecessors under all fundraising efforts since 1987, including income earned on outstanding balances of the contributions and excluding amounts transferred to the MDRI-II Trust in January 2006.

^{Annex X Table 2.}

Bilateral Contributions to the PRGT

(In SDR million unless otherwise noted)

Country			Percent share in total member quota	Cumulative PRGT subsidy contributions as of April 30, 2021^¹	Illustrative new contributions request based on SDR 2.3 billion target and quota shares ^²
All members			100.00	5 ,304	*2,300*
FTP members			83.01	4 ,874	*2,172*
	G-7		43.47	3,077	*1,138*
		Canada	2.32	287	61
		France	4.24	390	111
		Germany	5.60	313	146
		Italy	3.17	257	83
		Japan	6.48	695	169
		United Kingdom	4.24	539	111
		United States	17.44	596	456
	Other advanced		18.20	1,290	*476*
		Australia	1.38	72	36
		Austria	0.83	93	22
		Belgium	1.35	107	35
		Czech Republic	0.46	24	12
		Denmark	0.72	67	19
		Estonia, Republic of	0.05	1	1
		Finland	0.51	41	13
		Israel	0.40	-	11
		Korea	1.80	90	47
		Lithuania, Republic of	0.09	2	2
		Luxembourg	0.28	18	7
		Malta	0.04	2	1
		Netherlands	1.84	210	48
		New Zealand	0.26	11	7
		Norway	0.79	74	21
		Saudi Arabia	2.10	100	55
		Singapore	0.82	27	21
		Slovak Republic	0.21	5	6
		Slovenia, Republic of	0.12	2	3
		Spain	2.00	78	52
		Sweden	0.93	146	24
		Switzerland	1.21	121	32
Other FTP members			21.34	508	*558*
		Algeria	0.41	17	11
		Botswana	0.04	2	1
		Brazil	2.32	-	61
		Brunei Darussalam	0.06	3	2
		Chile	0.37	2	10
		China	6.41	135	168
		India	2.76	80	72
		Kuwait	0.41	19	11
		Malaysia	0.76	40	20
		Mauritius	0.03	1	1
		Mexico	1.87	42	49
		Oman	0.11	6	3
		Peru	0.28	0	7
		Philippines	0.43	6	11
		Poland, Republic of	0.86	-	23
		Qatar	0.15	2	4
		Russian Federation	2.71	113	71
		Thailand	0.68	24	18
		Trinidad and Tobago	0.10	1	3
		United Arab Emirates	0.49	9	13
		Uruguay	0.09	5	2
Non-FTP members			4.89	125	*128*
	Advanced economies		1.80	74	47
		Cyprus	0.06	2	2
		Greece	0.51	36	13
		Ireland	0.73	20	19
		Latvia, Republic of	0.07	2	2
		Portugal	0.43	15	11
Other non-FTP members			3.08	51	81
		Bulgaria	0.19	7	5
		Croatia, Republic of	0.15	2	4
		Hungary	0.41	-	11
		Indonesia	0.98	5	26
		Romania	0.38	9	10
		Turkey	0.98	29	26
Total from 61 members			87.90	4 ,999	*2,300*
Total from other members			12.10	305	…

^{Annex X Table 2.}

Bilateral Contributions to the PRGT

(In SDR million unless otherwise noted)

Country			Percent share in total member quota	Cumulative PRGT subsidy contributions as of April 30, 2021^¹	Illustrative new contributions request based on SDR 2.3 billion target and quota shares ^²
All members			100.00	5 ,304	*2,300*
FTP members			83.01	4 ,874	*2,172*
	G-7		43.47	3,077	*1,138*
		Canada	2.32	287	61
		France	4.24	390	111
		Germany	5.60	313	146
		Italy	3.17	257	83
		Japan	6.48	695	169
		United Kingdom	4.24	539	111
		United States	17.44	596	456
	Other advanced		18.20	1,290	*476*
		Australia	1.38	72	36
		Austria	0.83	93	22
		Belgium	1.35	107	35
		Czech Republic	0.46	24	12
		Denmark	0.72	67	19
		Estonia, Republic of	0.05	1	1
		Finland	0.51	41	13
		Israel	0.40	-	11
		Korea	1.80	90	47
		Lithuania, Republic of	0.09	2	2
		Luxembourg	0.28	18	7
		Malta	0.04	2	1
		Netherlands	1.84	210	48
		New Zealand	0.26	11	7
		Norway	0.79	74	21
		Saudi Arabia	2.10	100	55
		Singapore	0.82	27	21
		Slovak Republic	0.21	5	6
		Slovenia, Republic of	0.12	2	3
		Spain	2.00	78	52
		Sweden	0.93	146	24
		Switzerland	1.21	121	32
Other FTP members			21.34	508	*558*
		Algeria	0.41	17	11
		Botswana	0.04	2	1
		Brazil	2.32	-	61
		Brunei Darussalam	0.06	3	2
		Chile	0.37	2	10
		China	6.41	135	168
		India	2.76	80	72
		Kuwait	0.41	19	11
		Malaysia	0.76	40	20
		Mauritius	0.03	1	1
		Mexico	1.87	42	49
		Oman	0.11	6	3
		Peru	0.28	0	7
		Philippines	0.43	6	11
		Poland, Republic of	0.86	-	23
		Qatar	0.15	2	4
		Russian Federation	2.71	113	71
		Thailand	0.68	24	18
		Trinidad and Tobago	0.10	1	3
		United Arab Emirates	0.49	9	13
		Uruguay	0.09	5	2
Non-FTP members			4.89	125	*128*
	Advanced economies		1.80	74	47
		Cyprus	0.06	2	2
		Greece	0.51	36	13
		Ireland	0.73	20	19
		Latvia, Republic of	0.07	2	2
		Portugal	0.43	15	11
Other non-FTP members			3.08	51	81
		Bulgaria	0.19	7	5
		Croatia, Republic of	0.15	2	4
		Hungary	0.41	-	11
		Indonesia	0.98	5	26
		Romania	0.38	9	10
		Turkey	0.98	29	26
Total from 61 members			87.90	4 ,999	*2,300*
Total from other members			12.10	305	…

Annex XI. PRGT Investment Strategy and Options for Contributors

1. The PRG Trust Instrument allows for borrowing from official lenders (i) for the purpose of on-lending to eligible PRGT borrowers and (ii) in order for subsidy accounts to benefit from net investment earnings on the proceeds of loans extended at a concessional interest rate. Such borrowings are based on bilateral agreements with lenders and can be done in both SDRs and currencies. The investment risk in case of borrowing for investment is carried by lenders.^¹

2. Members have used this vehicle for providing subsidy resources since 1988 with the principal of investments varying from SDR 1.5 million to SDR 135 million. In the context of reviewing the investment strategy for Trust assets in July 2017, the Executive Board approved the following two options for investments by members who wish to lend to the PRGT for the purpose of contributing income earned on investments:^²

(i) Their investments may be pooled with the PRG Trust assets and share the same risk and return profile.^³ The current long-term investment return target for PRGT assets is to achieve 90 bps over the SDR rate, however, the realized investment returns are subject to high uncertainty, including that returns may turn negative, particularly over shorter horizons, leading to a potential loss in the principal of investments. The risk of losses at a 10-year maturity is currently estimated at 11 percent. To avoid credit risk to lenders, the maturity date of investments could be linked to reaching the amount of the pledged contribution in NPV terms. Such investments would continue to be liquid with the possibility of encashment/early repayment from the proceeds of the liquidation of the investment itself, if needed.

(ii) In case the risk related to investing in PRGT assets is not acceptable, contributors may elect to invest their resources in BIS obligations, managed separately from the PRGT assets. These obligations consist primarily of deposits with a maximum maturity of 12 months. The lower risk associated with BIS investments comes at the cost of lower returns, which are unlikely to significantly exceed the SDR rate. As currently observed, returns on BIS obligations are below the SDR rate.

3. Currently there are eight active agreements for the benefit of the PRGT, three of which for investment in the Trust’s assets and five in BIS obligations. Members also contribute through deposit and investment agreements for the benefit of the PRG-HIPC and the CCR Trusts.

4. Due to the low interest rate environment in recent years, subsidization through income earned on investments has been slow in providing pledged subsidy resources and required extension of agreements with the contributing members. Annual rates and returns on investments from 2010-present are provided below.

^{Annex XI Table 1.}

SDR Rate and Returns on PRGT Assets

(In percent)

^¹Reserve Account assets until end-2017 and pooled investments since then.

^{Annex XI Table 1.}

SDR Rate and Returns on PRGT Assets

(In percent)

	2011	2012	2013	2014	2015	2016	2017	2018	2019	2020	Jan-May 2021	Average (2002–20)
SDRi Rate	0.40	0.11	0.08	0.08	0.05	0.08	0.53	0.93	1.00	0.20	0.06	1.20
Return on PRGT Assets^¹	1.64	1.03	0.23	0.55	0.51	0.64	0.38	0.66	4.54	4.63	1.80	2.28
Return on BIS Deposits								0.84	1.06	0.47	0.03

^¹Reserve Account assets until end-2017 and pooled investments since then.

^{Annex XI Table 1.}

SDR Rate and Returns on PRGT Assets

(In percent)

	2011	2012	2013	2014	2015	2016	2017	2018	2019	2020	Jan-May 2021	Average (2002–20)
SDRi Rate	0.40	0.11	0.08	0.08	0.05	0.08	0.53	0.93	1.00	0.20	0.06	1.20
Return on PRGT Assets^¹	1.64	1.03	0.23	0.55	0.51	0.64	0.38	0.66	4.54	4.63	1.80	2.28
Return on BIS Deposits								0.84	1.06	0.47	0.03

^¹Reserve Account assets until end-2017 and pooled investments since then.

5. Large scale lending in SDRs for investment purposes would require operational arrangements for converting SDRs into currencies through VTAs and managing risks related to exchange rate fluctuations and currency conversion costs.

Annex XII. PRGT—Review of Interest Rate Structure

1. This Annex reviews in greater detail the interest rate charged on borrowing from the PRGT. The background section describes the PRGT interest rate mechanism and its application to date. The following section discusses the implications of the existing interest rate mechanism in the current economic environment. The final section sets out the proposed interest rate on PRGT lending based on the application of the PRGT’s existing interest rate setting mechanism, as modified in 2019, of a zero rate for the ECF, the SCF, and the RCF for the period August 2021–July 2023.

Background

2. The PRGT interest rate mechanism was adopted in 2009 as part of a comprehensive reform of the IMF’s concessional facilities. Prior to the reforms, the Fund’s concessional lending was traditionally extended at a uniform rate of 0.5 percent. The objective of the 2009 reforms was to increase the flexibility of IMF support to LICs and better tailor assistance to countries’ diverse needs given their heightened exposure to global volatility.^¹ The interest rate structure and adjustment mechanism aimed to balance the following objectives: (i) increase concessionality of PRGT financing, especially in the context of low global interest rates at the time; (ii) preserve the Trust’s scarce resources; (iii) avoid permanently zero interest rates; (iv) tailor financial terms to LICs’ needs and capacity; and (v) limit fluctuations in concessionality of PRGT instruments and subsidy costs.

3. The mechanism links interest rates on PRGT lending to global interest rates (Table 1). As conceived in 2009, interest rates on PRGT credit provided under different facilities are set for the upcoming two years in the context of biennial reviews; their level is linked to the average SDR interest rate over the most recently observed 12-month period; and the rate charged on SCF loans was initially set at 25 basis points above that for the ECF, as SCF users were expected to have somewhat higher capacity to service debt than ECF users, and reflecting also differences in the duration of their BoP financing and adjustment needs. The reform package in the parallel LIC Facilities Review in 2019 included a modification of the interest rate mechanism to align the SCF rate structure with that of the ECF (Table 1). The change made the SCF more concessional by (i) setting the SCF rate at zero when the SDR rate is below 2 percent and (ii) reducing the SCF rate by 0.25 percent when it is above 2 percent.

^{Annex XII Table 1.}

Evolution of Interest Rate Mechanism for the Fund’s Concessional Facilities¹

(In percent)

^¹This is based on the average SDR rate over the most recently observed 12-month period.

^²An Interest Rate Mechanism for the Fund’s Concessional Facilities was approved by the Executive Board in July 2009; it was in effect for January 7, 2010-July 2015.

^³A temporary waiver of interest payment for PRGT-eligible members on all outstanding concessional loans was approved by the Executive Board in July 2009, became effective on January 7, 2010 and further extended in December 2011, December 2012, and December 2014 until end-December 2016, after which the mechanism would apply.

^⁴The interest rate charge on RCF lending was set permanently to zero as of July 2015.

^⁵In December 2016, IMF Executive Board approved a new Interest Rate Mechanism and set zero rates on all low-income country lending facilities through end-2018, which was subsequently extended to end-June 2019.

^⁶In June 2019, IMF Executive Board approved a modification of the Interest Rate Mechanism and set zero rates on all low-income country lending facilities through end-June 2021, which was subsequently extended to end-July 2021.

4. Since the interest rate mechanism was first established in 2009, no interest has been charged on PRGT credit. In 2009 as part of a comprehensive reform of the IMF’s concessional facilities, the Executive Board granted interest waivers on all outstanding concessional credit during 2010–16 when LICs faced considerable headwinds from the global economic environment The interest rate mechanism has been in operation since then, resulting in zero rates on credit under all three PRGT facilities on the basis of the prevailing low global interest rates.

5. In July 2015, the PRGT interest rate mechanism was modified to enhance support for PRGT-eligible countries in fragile situations or hit by natural disasters. As part of the Fund’s response to the UN-sponsored dialogue on policies to promote financing of the 2030 Development Agenda, the Board approved an increase in concessionality of fast-disbursing support under the RCF by setting the interest rate levied on RCF financing permanently at zero, while preserving the PRGT interest rate mechanism for the SCF and ECF.

6. In October 2016, the PRGT interest rate mechanism was amended to accommodate periods of very low interest rates worldwide. A new threshold was proposed whereby both the ECF and the SCF rate would be set at zero when the average SDR rate over the most recently observed 12-month period was less than or equal to 0.75 percent (Table 1). This proposal in effect kept all PRGT interest rates under the mechanism at zero percent through December 2018, while incurring only modest subsidy costs for the PRGT. In addition, the interest rate charges on outstanding legacy balances under the ESF—which are not included under the PRGT interest rate mechanism—were waived until December 2018. Most Directors expressed the view that the merits and implications of unifying the interest rate structure for the ECF and SCF should be examined as one element of the forthcoming review of the LIC Facilities.^² In December 2018, the Executive Board postponed the deadline for the next interest rate review to no later than June 30, 2019 so that the PRGT interest rate mechanism could be assessed in light of the findings of the parallel LIC Facilities Review.

7. In May 2019, the Board approved an amendment of the interest rate mechanism. Specifically, the Board approved reforms to align the interest rates on SCF loans with those on ECF loans to modestly increase the degree of concessionality of PRGT financing under the SCF, with moderate subsidy costs financed within the PRGT’s self-sustaining financing envelope.^³ Interest on RCF credit would remain permanently at zero, as decided in 2015. The Board also decided that the existing zero percent interest rates under the ECF and SCF continue to be applied to outstanding balances of PRGT loans through June 30, 2021 and the waiver of interest rate charges on outstanding legacy balances under the ESF was further extended until their full repayment in October 2020.^⁴

PRGT Interest Rates in the Current Economic Context

8. Reflecting recent trends in global interest rates, the SDR interest rate has decreased to average 0.08 percent over the last 12 months, which remains low by historical standards (Figure 1). The SDR rate declined from 0.23 percent at the time of the first interest rate waiver in January 2010 to 0.05 percent in September 2014 and remained at this level until September 2016. Since then, the SDR rate steadily increased to reach 1.14 percent by end-March 2019, then declined gradually until the onset of the pandemic, after which it fell sharply. It is currently 0.05 percent.

9. Given low global interest rates, the degree of concessionality of PRGT loans remains below the traditional benchmark of 35 percent—as it has been since the introduction of the current facilities architecture in 2010. The average grant element in PRGT loans is currently estimated to be 32 percent for ECF and RCF and 26 percent for SCF.^⁵

10. The global economic outlook for LICs has substantively worsened since the previous review and since the onset of the pandemic, with significant downside risks. Global growth slowed to -3.3 percent in 2020, and LICs GDP growth is projected to be 4.3 percent in 2021. LICs have been hit harder and are expected to suffer more significant medium-term losses and to face substantial risks, including from limited access to vaccines, limited fiscal space to mount major health care policy responses or support livelihoods, high and rising debt levels, climate change and potent natural disasters.^⁶ The bulk of the PRGT credit outstanding is under the RCF, following the increase in emergency financing in response to the COVID-19 pandemic, as well as credit outstanding under the ECF (Figure 2).

PRGT Interest Rates for August 2021–July 2023

11. With the application of the interest rate mechanism as modified in 2019, no interest would be charged on PRGT credit until July 2023. With the 12-month average SDR rate currently well below the 2 percent threshold, the interest rate on SCF and ECF credit would be zero for the period August 2021–June 2023. As decided by the Executive Board in 2015, the RCF interest rate would remain permanently at zero. Consistent with Section II, Paragraph 4(b) of the PRGT Instrument, the next review of the PRGT interest rate structure would be completed by July 2023.

Annex XIII. CCRT Grant Mobilization

^{Annex XIII Table 1.}

New Contributions to the CCRT

(As of June 23, 2021)

^¹On April 5, 2021, the EU disbursed SDR 141 million to the CCRT as the first installment of its pledge.

^²Disbursed as part of the 2015–17 fundraising campaign.

^{Annex XIII Table 1.}

New Contributions to the CCRT

(As of June 23, 2021)

		CCRT grants
Country		In millions of SDRs	In millions of original currency (if appl.)	In millions of US$	Current status of contribution
	European Union^¹	152.5	€ 183	217.8	Partially disbursed
	UK	135.8	£150	194.0	Disbursed
	Japan	73.4	$100	104.8	Disbursed
	Germany	66.2	€ 80	94.6	Disbursed
	France	33.7	€ 40	48.1	Pledge
	Spain	20.9	€ 25	29.9	Pledge
	Netherlands	20.8	€ 25	29.7	Disbursed
	Switzerland	19.5	CHF 25	27.8	Disbursed
	Norway	14.5	NOK 180	20.7	Disbursed
	Singapore	12.2	$17.6	17.5	Pledge
	Greece	7.6	$11.0	10.9	Pledge
	China	5.6		8.0	Disbursed
	Mexico^²	2.9	$4	4.2	Disbursed
	Philippines	2.8	$4	4.0	Pledge
	Sweden	2.4	SEK 30	3.5	Disbursed
	Bulgaria	1.9		2.7	Disbursed
	Luxembourg	1.7	€ 2	2.4	Disbursed
	Malta	0.6	$0.8	0.8	Disbursed
Total		574.9		821.3
Target		SDR 1 billion		US$1.4 billion

^¹On April 5, 2021, the EU disbursed SDR 141 million to the CCRT as the first installment of its pledge.

^²Disbursed as part of the 2015–17 fundraising campaign.

^{Annex XIII Table 1.}

New Contributions to the CCRT

(As of June 23, 2021)

		CCRT grants
Country		In millions of SDRs	In millions of original currency (if appl.)	In millions of US$	Current status of contribution
	European Union^¹	152.5	€ 183	217.8	Partially disbursed
	UK	135.8	£150	194.0	Disbursed
	Japan	73.4	$100	104.8	Disbursed
	Germany	66.2	€ 80	94.6	Disbursed
	France	33.7	€ 40	48.1	Pledge
	Spain	20.9	€ 25	29.9	Pledge
	Netherlands	20.8	€ 25	29.7	Disbursed
	Switzerland	19.5	CHF 25	27.8	Disbursed
	Norway	14.5	NOK 180	20.7	Disbursed
	Singapore	12.2	$17.6	17.5	Pledge
	Greece	7.6	$11.0	10.9	Pledge
	China	5.6		8.0	Disbursed
	Mexico^²	2.9	$4	4.2	Disbursed
	Philippines	2.8	$4	4.0	Pledge
	Sweden	2.4	SEK 30	3.5	Disbursed
	Bulgaria	1.9		2.7	Disbursed
	Luxembourg	1.7	€ 2	2.4	Disbursed
	Malta	0.6	$0.8	0.8	Disbursed
Total		574.9		821.3
Target		SDR 1 billion		US$1.4 billion

^¹On April 5, 2021, the EU disbursed SDR 141 million to the CCRT as the first installment of its pledge.

^²Disbursed as part of the 2015–17 fundraising campaign.

References

International Monetary Fund, 2009, “A New Architecture of Facilities for Low-Income Countries and Reform of the Fund’s Concessional Financing Framework—Supplementary Information” (Washington).
- Search Google Scholar
- Export Citation
International Monetary Fund, 2015, “Financing for Development: Enhancing the Financial Safety Net for Developing Countries,” IMF Policy Paper 2020/044 (Washington).
- Search Google Scholar
- Export Citation
International Monetary Fund, 2016a, “Review of Access Limits and Surcharge Policies” (Washington).
- Search Google Scholar
- Export Citation
International Monetary Fund, 2016b, “IMF Executive Board Modifies PRGT Interest Rate Mechanism and Approves Zero Rates on All Low-Income Country Lending Facilities Through end-2018,” Press Release No. 2016/448 (Washington).
- Search Google Scholar
- Export Citation
International Monetary Fund, 2016c, “Financing for Development: Enhancing the Financial Safety Net for Developing Countries—Further Considerations” (Washington).
- Search Google Scholar
- Export Citation
International Monetary Fund, 2017a, “Large Natural Disasters—Enhancing the Financial Safety Net for Developing Countries” (Washinaton).
- Search Google Scholar
- Export Citation
International Monetary Fund, 2017b, “Investment of Temporary Resources to Generate Income to Contribute to PRG, PRG-HIPC and CCR Trust” (Washinaton).
- Search Google Scholar
- Export Citation
International Monetary Fund, 2018a, “Guidance Note on the Bank-Fund Debt Sustainability for Framework for Low Income Countries “ (Washinaton).
- Search Google Scholar
- Export Citation
International Monetary Fund, 2018b, “Macroeconomic Developments and Prospects in Low-Income Developing Countries: 2018” (Washinaton).
- Search Google Scholar
- Export Citation
International Monetary Fund, 2018c, “2018 Review of Facilities for Low-Income Countries” (Washinaton).
- Search Google Scholar
- Export Citation
International Monetary Fund, 2019a, “2018–19 Review of Facilities for Low-Income Countries—Reform Proposals,” IMF Policy Paper No. 2019/014 (Washington).
- Search Google Scholar
- Export Citation
International Monetary Fund, 2019b, “Poverty Reduction and Growth Trust—Review of Interest Rate Structure” (Washington).
- Search Google Scholar
- Export Citation
International Monetary Fund, 2020a, “The Evolution of Public Debt Vulnerabilities in Lower Income Economies,“ IMF Policy Paper No. 20/003 (Washington).
- Search Google Scholar
- Export Citation
International Monetary Fund, 2020b, “Eligibility to Use the Fund’s Facilities for Concessional Financing, 2020,“ IMF Policy Paper No. 2020/016 (Washington)
- Search Google Scholar
- Export Citation
International Monetary Fund, 2020c, “Enhancing the Emergency Financing Toolkit-Responding to the COVID-19 Pandemic” (Washington).
- Search Google Scholar
- Export Citation
International Monetary Fund, 2020d, “Temporary Modification to the Fund’s Annual Access Limits,” IMF Policy Paper 2020/036 (Washington).
- Search Google Scholar
- Export Citation
International Monetary Fund, 2020e, “Policy Safeguards for Countries Seeking Access to Fund Financial Support that would Lead to High Levels of Combined GRA-PRGT Exposure,” IMF Poiicy Paper No. 2020/039 (Washington).
- Search Google Scholar
- Export Citation
International Monetary Fund, 2020f, “Review of Enhanced Access Limits under the Rapid Credit Facility and Rapid Financing Instrument,” IMF Policy Paper 2020/044 (Washington).
- Search Google Scholar
- Export Citation
International Monetary Fund, 2020g, “Update on the Joint IMF-WB Multipronged Approach to Address Debt Vulnerabilities,“ IMF Policy Paper No. 2020/066 (Washington).
- Search Google Scholar
- Export Citation
International Monetary Fund, 2020h, “Reform of the Policy on Public Debt Limits in IMF-Supported Programs,” IMF Policy Paper 2020/061 (Washington).
- Search Google Scholar
- Export Citation
International Monetary Fund, 2021a, “Temporary Extensions and Modifications of Access Limits in the Fund’s Lending Facilities” (Washington).
- Search Google Scholar
- Export Citation
International Monetary Fund, 2021b, “The Acting Chair’s Summing Up – Temporary Extensions and Modifications of Access Limits in the Fund’s Lending Facilities, Press Release (Executive Board Assessment)” (Washington).
- Search Google Scholar
- Export Citation
International Monetary Fund, 2021c, “Macroeconomic Developments and Prospects in Low-Income Countries, 2021,” IMF Policy Paper No. 2021/020 (Washington).
- Search Google Scholar
- Export Citation
World Economic Outlook (WEO), 2021, WEO 2021, IMF, April 2021.
- Search Google Scholar
- Export Citation

The current framework comprises three concessional lending facilities under the PRGT—the Extended Credit Facility (ECF), which provides medium-term support to LICs with protracted balance of payments problems; the Standby Credit Facility (SCF) to help members deal with short-term balance of payment needs; and the Rapid Credit Facility (RCF) to provide rapid financing with limited conditionality to help members deal with urgent balance of payment needs—and one non-financial instrument, the Policy Support Instrument (PSI). In addition, PRGT-eligible members have access to the General Resources Account, as well as to the non-financial Policy Coordination Instrument (PCI).

The terms “low income countries” and “PRGT-eligible countries” are used interchangeably throughout this paper. For a full discussion of the determinants of PRGT eligibility, see “Review of Eligibility to Use the Fund’s Facilities for Concessional Financing, 2020,” IMF Policy Paper 20/016 (IMF, 2020b).

The Fund’s emergency financing instruments include the RCF, available only to LICs, and the Rapid Financing Instrument (RFI), available to all IMF member countries.

See “Macroeconomic Developments and Prospects in Low-Income Countries, 2021,” IMF Policy Paper 21/020 (IMF, 2021c) for discussion of the economic impact of the pandemic and projections of LICs’ likely financing needs for 2021–25.

LICs at relatively high levels of income per capita are required to blend concessional loans from the PRGT with non- concessional loans from the GRA facilities. The GRA facilities can be used by all Fund members.

More than half the debt owed to the PRGT at end-2020 was borrowed during the preceding 12 months, with repayments falling due from late-2025 through 2030.

To provide concessional financing, the PRGT borrows from member countries (currently 18) with which it has loan agreements, paying the SDR interest rate on these loans; it lends to LICs at a lower interest rate, creating a subsidy cost. PRGT lending has carried a zero-interest rate since the new framework was introduced in 2010.

Financial support from the Fund is typically provided via multiyear arrangements (or programs) in which the quality of economic policies being supported under the arrangement must meet the standards of upper credit tranche (UCT) conditionality. Financial support provided through the EF facilities takes the form of single disbursements; there is no ex post conditionality and the policy framework being supported is not required to meet UCT standards. We use the terms “arrangements” and “UCT-quality programs” interchangeably in this paper.

See Annex IV for a discussion of the various temporary changes made to access limits in both the GRA and the PRGT since March 2020.

The two new UCT arrangements (with The Gambia and Somalia) were approved in March, with negotiations completed before the scale of the pandemic had become apparent.

In these cases, Management has approved a staff policy position and proposed access levels and staff do not see any significant obstacles to completing program discussions and bringing the program to the Board speedily.

Access levels under the six arrangements already approved in the past nine months ranged from 80 percent of quota to 305 percent of quota, underscoring the diversity of countries’ financing needs.

Staff expect to see one request for EF reaching the Board before end-July—a request from St. Vincent and the Grenadines for assistance in responding to a major (volcanic eruption). More requests could occur where urgent health-related spending, creating BoP needs, is required before a program can be put in place.

An SDR provides the right to request, and be provided with, the equivalent amount in strong currencies from the Fund. Countries pay the SDR interest rate (currently at its minimum level of 0.05 percent per annum) on the amount of their SDR allocation that has been used (i.e., on the difference between their allocation of SDRs and their current holdings of SDRs). There is no set period for countries to rebuild their holdings of SDRs: many LICs have made active use of almost all of their existing allocation of SDRs.

Loans extended by the World Bank and some other multilateral creditors are also afforded preferred creditor status and hence are effectively non-restructurable.

The linkage between the provision of Fund financial support and sustainability of a country’s debt position applies to all potential borrowers, not just LICs. The Fund may provide support to countries with very high levels of debt, as long as such debt is sustainable on a forward-looking basis, which may require an agreement with creditors on a debt restructuring sufficient to render debt sustainable over time. Efforts to delay restructuring where debt is clearly unsustainable are almost always harmful to the borrowing country.

The terms “blend” and “non-blend” countries is used throughout this paper as a shorthand for the technical terms “presumed blender” and “non-presumed blender” (see Annex III).

Graduation from PRGT eligibility is not automatic, but also requires a judgment that the country does not face serious short-term vulnerabilities.

The notation “A/B percent of quota” refers to annual and cumulative access limits, respectively. To simplify the exposition, the discussion focuses in the main on the “permanent” (non-transitory) access limits, which are reviewed at regular intervals (now a five-year cycle): the temporary changes made to access limits are described in Annex IV.

How these principles can be reconciled with PRGT financial self-sustainability over the medium-to-longer term will need to be examined in the context of the post-crisis LIC facilities review and the analysis of PRGT financing needs.

See “Financing for Development: Enhancing the Financial Safety Net for Developing Countries—Further Considerations” (IMF (2016c), October 24, 2016).

The access norm for ECF arrangements longer than 3 years would be based on the length of the arrangement and the annual access norm under the three-year ECF arrangement. Consistent with earlier practice, the norm for access under a 18-month SCF would be set equal to that of the 3-year ECF arrangement, again varying proportionately with the length of the SCF arrangement, up to the amount allowable under a 2-year SCF arrangement (193.33 percent of quota).

Increases of 45 percent to the access norm for low credit outstanding (120 percent of quota) and higher credit outstanding (75 percent) would yield an unweighted average of 141.4 percent, rounded to 145 percent of quota.

See IMF (2021a) and IMF (2021b); the new requirements are permanent in nature.

Ibid.

The specification of the flow and stock triggers could be revisited in the period between now and the next such review if Directors, in reviewing the development of PRGT finances, called for such a move.

The case for aligning the PS-HCC conditions to the standards underpinning the GRA EA criteria is discussed in IMF PP 2020/039 (IMF, 2020e).

For example, the Kyrgyz Republic moved to blend status in July 2019, based on 2018 GNI per capita; GNI per capita had also exceeded the IDA operational cutoff in 2014 but fell below it in 2016–17.

Similarly, Ghana was viewed as a non-blend case when it requested EF in April 2020 but would now likely be treated as a PB given its large dollar sovereign bond issue in late-March 2021.

This simplification would also improve consistency across instruments for supporting LICs: Tanzania is currently required to blend and is not eligible for PRGT EA, but is eligible for assistance from the CCRT (for which some 20 non- blend countries are not eligible). Tanzania’s 2019 income per capita was 91 percent of the IDA cutoff, falling to 89.6 of the IDA cutoff in 2020.

See Decision No. 14521-(10/3), as amended, and Eligibility to Use the Fund’s Facilities for Concessional Financing, 2020 (IMF, 2020b). This criterion is met if the country has issued or guaranteed eligible external debt in at least three of the past five years in a cumulative amount equivalent to at least 50 percent of its quota. The second test (if there were convincing evidence that the sovereign could have tapped international financial markets on a durable and substantial basis) does not apply. As discussed in the 2020 Review of PRGT Eligibility (IMF Policy Paper 20/016), staff assessment as to whether this criterion is met requires validation of the debt data (taken from the World Bank’s International Debt Statistics) with country authorities. The 2020 eligibility review introduced refinements to the methodology for assessing past market access (the first test noted above), which apply also to assessment of market access under PRGT blending and EA policies.

The exception for small/micro-states affects Cabo Verde, Dominica, and Maldives, all of which are at high risk of debt distress but meet the criterion of substantial past market access. The specification of market access, which also plays an important role in driving PRGT graduation decisions, will likely be subject to further analysis during the upcoming review of PRGT eligibility, due at the Board in Q1 2022.

Borrowing from the GRA will still likely be on more generous terms than those faced by the majority of PRGT or IDA graduates in the bond markets.

All PRGT-eligible countries would still have full rights to access the GRA, but none would be required to do so.

The recent paper “Macroeconomic and Developments Prospects in Low-Income Countries—2021” (IMF Policy Paper No. 2021/020 (IMF, 2021c)) estimated LICs’ total external financing needs for 2021–25 at US$450 billion, of which some US$150–200 billion could be met via new borrowing. Fund lending in the baseline/high case scenarios would cover somewhere between one-quarter and one-half of the available borrowing space.

Under the High Case, the impact would be more pronounced, with about half of eligible LICs affected, reducing PRGT lending by up to SDR 4¼ billion, and lowering the subsidy resource gap by up to SDR 0.3 billion.

The reimbursement of the PRGT can be waived notwithstanding the fact that the PRGT contains Special Disbursement Account (SDA) resources derived from gold sales profits. For details, see Annex VIII.

The necessary amendments to the PRGT to introduce the accounts could be adopted by a majority of the votes cast and would not require the consent of current contributors to the PRGT.

The SRA proposal has several advantages over the (alternative) option of contributors directly placing grants in the RA: (i) resources derived from SDA resources are not co-mingled with contributor resources; (ii) grant contributions would more quickly be used for their primary purpose of subsidization (before the RA); and (iii) grants would be used only for coverage of arrears after the resources in the RA are exhausted.

Existing protected provisions are listed under Section IX of the PRGT Instrument.

Borrowed resources would fall under the existing investment authority to invest temporary resources from members in which the contributing member bears the investment risk. See Annex XI and “Investment of Temporary Resources to Generate Income to Contribute to PRG, PRG-HIPC and CCR Trusts.“ IMF (2017b)

Where a contributor is not a member but rather an institution of a member, the representation of BOP need would be made by the contributor based on the BOP situation of the relevant member.

See “2018–19 Review of Facilities for Low-Income Countries—Reform Proposals” (IMF Policy Paper No. 19/014, IMF (2019a)).

This would be on top of the SDR 17 billion already mobilized so far. The estimates take into account the assumed encashment buffer of 20 percent.

Covering the period from June 2020 to May 2021.

Eritrea has yet to start the HIPC qualification process.

See The G-8 Debt Cancellation Proposal and Its Implications for the Fund.

This amount includes income earned on members’ deposits agreed during 1997–2001, 14 of which have been recently extended beyond the original maturity date and continue to benefit the PRG-HIPC Trust. Five of these deposits were repurposed for the benefit of the CCRT and three for the benefit of the PRGT.

PRGT-supported programs with an initial duration of more than two years are expected to be informed by an explicit Poverty Reduction and Growth Strategy (PRGS).

The notation “A/B percent of quota” refers to annual and cumulative access limits, respectively. Unless otherwise stated, the access limits cited are the access limits set during the 2018–19 Review of LIC Facilities: several limits have been increased on a temporary basis since March 2020.

See https://datahelpdesk.worldbank.org/knowledgebase/topics/19373-data-compilation-methodology.

A similar approach is taken in regard to the specification of the income criterion for PRGT eligibility.

The starting point for applying this test would be the years 2019–2020.

Estimates for income levels for 2019 are from the data release by the World Bank on July 1, 2020, with estimates for 2020 taken from the data to be released on July 1, 2021.

This criterion is met if the country has issued or guaranteed eligible external debt in at least three out of the past five years in a cumulative amount equivalent to at least 50 percent of its quota; a staff assessment on whether this is met requires validation of the debt data (from the World Bank’s International Debt Statistics) with country authorities.

Estimates for income levels for 2019 are from the data release by the World Bank on July 1, 2020); estimates for 2020 are based on the data to be released by the Bank on July 1, 2021.

See paragraph 35 of the main text.

See “Financing for Development: Enhancing the Financial Safety Net for Developing Countries,” (IMF (2015), July 2015) and “Large Natural Disasters—Enhancing the Financial Safety Net for Developing Countries,” (IMF (2017a), May 2017).

See “2018–19 Review of Facilities for Low-Income Countries—Reform Proposals,” (IMF (2019a), June 2019).

See “Enhancing the Emergency Financing Toolkit-Responding to the COVID-19 Pandemic,” (IMF, (2020c), April 2020), “Temporary Modification to the Fund’s Annual Access Limits,” (IMF (2020d), July 2020), “Review of Enhanced Access Limits under the Rapid Credit Facility and Rapid Financing Instrument,” (IMF (2020f), October 2020), and “Temporary Extensions and Modifications of Access limits in the Fund’s Lending Facilities,” (IMF (2021a), March 2021).

The increase in the NAAL was accompanied by an increase in the Exceptional Annual Access Limit (EAAL) by 50 percent of quota, to 183.33 percent, for the same period.

The EAAL and the exceptional cumulative access limit (ECAL) were increased by similar absolute amounts through June 30, 2021. Access norms in the PRGT have remained unchanged since May 2019.

An arrangement supported under the GRA is expected to ensure that BoP difficulties are resolved before repayments to the Fund begin: an arrangement supported under the ECF is expected to help countries with protracted BoP problems to make significant progress towards a stable and sustainable macroeconomic position. The GRA thus implicitly requires a faster pace of adjustment than would be expected under the ECF.

IIR countries would also avoid the 0.5 percent one-off charge on each drawing of GRA resources.

A PRGT loan of 1 SDR to an IIR country would generate annual net revenues equal to the interest earned [%*(SDRi+ 1.0)] less the funding cost [SDRi], or (0.67-⅓*SDRi), which > 0 if SDRi < 2.0. The net income to the PRGT of a blended loan of 1 SDR is -⅓*SDRi (the funding cost of the PRGT share).

The PRGT Instrument specifies that the Trust shall assist in fulfilling the purposes of the Fund by providing loans on concessional terms to low-income developing members that qualify for assistance under the Instrument.

For analysis of pre-pandemic trends, see “Macroeconomic Developments and Prospects in Low-Income Developing Countries: 2018,” IMF (2018b), March 2018, and “The Evolution of Public Debt Vulnerabilities in Lower Income Economies,” IMF (2020a), February 2020.

See “Guidance Note on the Bank-Fund Debt Sustainability for Framework for Low Income Countries,” IMF (2018a), February 2018.

See “Update on the Joint IMF-WB Multipronged Approach to Address Debt Vulnerabilities,” IMF (2020g), December 2020.

See “Reform of the Policy on Public Debt Limits in IMF-Supported Programs,” IMF Policy Paper 2020/61, IMF (2020h).

This abstracts from the temporary increases in access levels that are set to expire shortly.

These requirements were discussed in “Temporary Extension and Modifications of Access Limits in the Fund’s Lending Facilities,” IMF (2021a), March 12, 2021.

Guidance and templates to implement these requirements are under preparation.

The new Debt Limits Policy requires reporting of these details, but not an explicit discussion in program documents.

While staff teams will usually make a reduction in debt vulnerabilities an important component of program design in cases where countries at high risk of/in debt distress, this has been a requirement only where PRGT EA or high combined GRA-PRGT access is involved

This need not necessarily involve steady reductions in all indicators for which thresholds are breached: for example, debt service ratios could still spike in individual years, reflecting a bunching of debt repayments (say, a Eurobond issue with a single bullet repayment).

The control group could be as broad as all PRGT arrangements during 2010–2020, but could also be customized to focus on a specific set of comparator cases, including to better reflect country-specific characteristics or the type of arrangement (e.g., fragile states, emergency financing, UCT-quality programs), while ensuring uniformity of treatment across programs. The underlying data on control groups will be updated regularly (e.g., once a year).

The methodology also allows for comparison of peak levels of debt service indicators with the peak levels of these indicators in a subset of cases in the control group (e.g., the top quartile of observations for the indicator), facilitating the identification of key stress points.

The so called ESAF Trust established in December 1987 was allowed to borrow from donor countries to on-lend to eligible members on concessional terms. It was supported by newly created reserve and subsidy accounts financed from repayments of Trust Fund’s loans, and further supplemented by bilateral contributions from members.

The Trust Fund established in 1976 was providing concessional lending to eligible members on revolving basis.

During 1976–80 the Fund sold 25 million ounces of gold it acquired in the conduct of its operations. These sales generated profits of US$4.6 billion, of which US$1.3 billion was distributed directly to 104 developing country members. The remainder of the profits, together with interest income and other transfers to the Trust (about SDR 3 billion in total) constituted the resources of the Trust Fund. For description of 1976–1980 gold sales and funding of the Trust see the IMF’s Annual Report 1980, pp. 85–89.

Please see further details in Annex VIII.

In line with Section IV, paragraph 3(b) of the PRGT Instrument, the repayment of principal and any payment of interest on borrowings for investment shall be made exclusively from the proceeds of liquidation of the investment and the earnings thereon.

Decision No. 16253-(17/70), adopted July 28, 2017. Under the PRGT instrument, Section VII, paragraph 3(a), the Executive Board would need to approve any new investment strategy for investing the PRGT subsidy accounts’ borrowed resources beyond the two investment options adopted under Decision No. 16253.

Under the current investment strategy, the PRGT portfolio has been structured to maintain its reserve-like properties, provide security to the PRGT loan providers, and ensure liquidity in the event of unexpected needs through its large allocation to short-term fixed-income instruments. The eligible asset classes reflect a moderately diversified portfolio. The target asset allocation assumes 45 percent in liquid and short-term fixed-income instruments, 30 percent in corporate bonds, 5 percent in emerging market government bonds, and 20 percent in publicly listed equities.

See A New Architecture of Facilities for Low-Income Countries and Reform of the Fund’s Concessional Financing Framework—Supplementary Information, IMF (2009), July 20, 2009.

See “IMF Executive Board Modifies PRGT Interest Rate Mechanism and Approves Zero Rates on All Low-Income Country Lending Facilities through end-2018,” Press Release No. 16/448, IMF (2016b), October 6, 2016.

For SCF arrangements treated as precautionary, no interest is charged. An availability fee of 0.15 percent applies at the end of each six-month period on available but undrawn credit.

See Decision No. 16521-(19/42), adopted May 24, 2019, and Poverty Reduction and Growth Trust—Review of Interest Rate Structure (IMF, 2019b). Staff has recently proposed for the Executive Board’s approval on lapse of time basis: (i) to postpone the review of the PRGT interest rate mechanism scheduled for June 2021 to July 2021, (ii) to continue having the interest rates of zero percent per annum applicable to the outstanding balances of PRGT loans under the ECF and SCF through July 31, 2021; and (iii) to extend the waiver of the interest rate charge on the outstanding legacy balances under the ESF through July 31, 2021.

See 2018 Review of Facilities for Low-Income Countries, IMF (2018c), August 1, 2018. Since October 11, 2013, a unified discount rate of 5 percent is used to calculate the grant element of individual loans.

See World Economic Outlook (IMF, April 2021).

Save
Cite
Email this content

Share Link

Copy this link, or click below to email it to a friend
Email this content
or copy the link directly:

https://www.elibrary.imf.org/view/journals/007/2021/053/article-A001-en.xml
The link was not copied. Your current browser may not support copying via this button.

Link copied successfully

Collapse
Expand

Fund Concessional Financial Support For Low-Income Countries—Responding To The Pandemic

Author:

International Monetary Fund

Volume/Issue:: Volume 2021: Issue 053

Publisher:: International Monetary Fund

ISBN:: 9781513589633

ISSN:: 2663-3493

Pages:: 155

DOI:: https://doi.org/10.5089/9781513589633.007

Keywords
I. Introduction
II. IMF Lending to Lics: Recent Experience
III. Modifying PRGT Facilities and Lending Safeguards
- Proposed Changes to PRGT Access Limits and Associated Safeguards
- Proposed Changes to Blending Policies
- “All-PRGT” versus “1:2” Financing Terms for Presumed Blenders: A Possible Future Reform Option
IV. Lending Scenarios and Financing Needs
V. Funding Strategy
VI. PRGT Interest Rate Review
VII. Debt Relief Financing
- Catastrophe Containment and Relief Trust (CCRT)
- Heavily Indebted Poor Countries (HIPC) Initiative
VIII. Enterprise Risks
IX. Issues for Discussion
Annex I. PRGT Facilities-Selected Features
Annex II. The PRGT Exceptional Access Criteria and Policy Safeguards for High Combined Credit Exposures
Annex III. Blending Policies and Eligibility for Exceptional Access: Current and Proposed
- The Income Threshold
- The Market Access Threshold
  - Eligibility for Exceptional Access to PRGT Resources
Annex IV. Evolution of PRGT Access Limits
Annex V. A Dual Interest Rate Mechanism in the PRGT
Annex VI. Analysis of Debt Sustainability and Capacity to Repay
Annex VII. Methodology for Estimating PRGT Resource Needs
Annex VIII. The PRGT Financing Model
Annex IX. PRGT Loan Resource Mobilization
Annex X. PRGT Subsidy Resources and Reserve Account
Annex XI. PRGT Investment Strategy and Options for Contributors
Annex XII. PRGT—Review of Interest Rate Structure
- Background
  - PRGT Interest Rates in the Current Economic Context
  - PRGT Interest Rates for August 2021–July 2023
Annex XIII. CCRT Grant Mobilization
References

View raw image

^{Figure 1.}

Fund Financing to LICs, 2013–2020

View raw image

^{Figure 2.}

Level of Financing Disbursed: January 2020–May 2021

(In billions of SDR)

View raw image

^{Figure 3.}

Access Level, and Implications for PRGT Resources and Fund Credit Exposure

(In percent of the indicated variable)

View raw image

^{Figure 4.}

PRGT Reserve Coverage Ratio Under Illustrative Scenarios, 2017–34^¹

(In percent)

View raw image

Projected PRGT Credit Outstanding, 2020–34^¹

(In SDR million)

View raw image

^{Annex VI Figure 1.}

PRGT-Eligible Countries: Risk of Debt Distress ^1/

(Percent)

View raw image

^{Annex VI Figure 2.}

Country X: Fund Credit Indicators Compared to All PRGT Arrangements

(In percent of the indicated variable)

View raw image

^{Annex XII Figure 1.}

SDR Interest Rates, January 2008–May 2021

(In percent)

View raw image

^{Annex XII Figure 2.}

PRGT Credit Outstanding, as of end-May 2021

(In millions of SDRs)

Foresee

^{Annex XIII Table 1.}

New Contributions to the CCRT

(As of June 23, 2021)

		CCRT grants
Country		In millions of SDRs	In millions of original currency (if appl.)	In millions of US$	Current status of contribution
	European Union^¹	152.5	€ 183	217.8	Partially disbursed
	UK	135.8	£150	194.0	Disbursed
	Japan	73.4	$100	104.8	Disbursed
	Germany	66.2	€ 80	94.6	Disbursed
	France	33.7	€ 40	48.1	Pledge
	Spain	20.9	€ 25	29.9	Pledge
	Netherlands	20.8	€ 25	29.7	Disbursed
	Switzerland	19.5	CHF 25	27.8	Disbursed
	Norway	14.5	NOK 180	20.7	Disbursed
	Singapore	12.2	$17.6	17.5	Pledge
	Greece	7.6	$11.0	10.9	Pledge
	China	5.6		8.0	Disbursed
	Mexico^²	2.9	$4	4.2	Disbursed
	Philippines	2.8	$4	4.0	Pledge
	Sweden	2.4	SEK 30	3.5	Disbursed
	Bulgaria	1.9		2.7	Disbursed
	Luxembourg	1.7	€ 2	2.4	Disbursed
	Malta	0.6	$0.8	0.8	Disbursed
Total		574.9		821.3
Target		SDR 1 billion		US$1.4 billion

^¹On April 5, 2021, the EU disbursed SDR 141 million to the CCRT as the first installment of its pledge.

^²Disbursed as part of the 2015–17 fundraising campaign.

Topics

Countries

Series

About

Resources

Abstract

I. Introduction

II. IMF Lending to Lics: Recent Experience

III. Modifying PRGT Facilities and Lending Safeguards

Proposed Changes to PRGT Access Limits and Associated Safeguards

Norms

Safeguards

Procedural Safeguards

Other Safeguards: Alignment of PRGT EA Criteria with Existing Policy Safeguards

Proposed Changes to Blending Policies

“All-PRGT” versus “1:2” Financing Terms for Presumed Blenders: A Possible Future Reform Option

IV. Lending Scenarios and Financing Needs

V. Funding Strategy

VI. PRGT Interest Rate Review

VII. Debt Relief Financing

Catastrophe Containment and Relief Trust (CCRT)

Heavily Indebted Poor Countries (HIPC) Initiative

VIII. Enterprise Risks

IX. Issues for Discussion

Annex I. PRGT Facilities-Selected Features

Annex II. The PRGT Exceptional Access Criteria and Policy Safeguards for High Combined Credit Exposures

Annex III. Blending Policies and Eligibility for Exceptional Access: Current and Proposed

The Income Threshold

The Market Access Threshold

Eligibility for Exceptional Access to PRGT Resources

Annex IV. Evolution of PRGT Access Limits

Annex V. A Dual Interest Rate Mechanism in the PRGT

Annex VI. Analysis of Debt Sustainability and Capacity to Repay

Annex VII. Methodology for Estimating PRGT Resource Needs

Annex VIII. The PRGT Financing Model

Annex IX. PRGT Loan Resource Mobilization

Annex X. PRGT Subsidy Resources and Reserve Account

Annex XI. PRGT Investment Strategy and Options for Contributors

Annex XII. PRGT—Review of Interest Rate Structure

Background

PRGT Interest Rates in the Current Economic Context

PRGT Interest Rates for August 2021–July 2023

Annex XIII. CCRT Grant Mobilization

References

Share Link

Table of Contents

Headings

Figures

Metrics

Foresee