2018 Review of Facilities for Low-Income Countries

International Monetary Fund. Strategy, Policy, &amp; Review Department; International Monetary Fund. Finance Dept.

doi:10.5089/9781498310871.007.A001

Author: International Monetary Fund. Strategy, Policy, &, Review Department, and International Monetary Fund. Finance Dept.

Publication Date:: 01 Aug 2018

eISBN:: 9781498310871

Language:: English

Keywords:: PP; financing; Fund; lending; LIC facilities; financing needs; PRGT resource; PRGT credit; financial support; Fund financing; debt service; Natural disasters; Balance of payments need; Debt sustainability; Global

Download PDF (2.2 MB)

The Fund is facing strong demand for financing from low-income countries (LICs).

Abstract

The Fund is facing strong demand for financing from low-income countries (LICs).

Introduction

1. The set of Fund facilities for low-income countries (LICs) took its present form following a comprehensive overhaul of the previous toolkit in 2009 (IMF, 2009a and 2009c). While all Fund members can access facilities financed through the General Resources Account (GRA), only the subset of the membership—mostly LICs—deemed eligible to access the resources of the Poverty Reduction and Growth Trust (PRGT) can make use of the concessional facilities.¹ The concessional toolkit has been designed to reflect the specific development challenges and weak institutional capacity of many LICs, while including instruments adapted to the needs of countries on a path towards graduation (see Box 1).

2. The Fund has periodically increased access levels under its concessional facilities to meet LICs’ evolving financing needs, subject to the availability of resources. Access limits were doubled in 2009 as part of the initial overhaul, and a further 50 percent across-the-board increase was instituted in 2015.² In 2012, a three-pillar strategy was adopted to make the PRGT’s concessional lending capacity self-sustaining over the long term (see Supplement);³ a key principle is that any proposals to modify LIC facilities need to be consistent with the PRGT’s self-sustainability.

3. The economic challenges facing LICs have evolved since 2010, with increasing diversity in economic conditions across countries and rising debt vulnerabilities in many cases (see Box 2).⁴ Commodity price shocks, policy slippages, and a relative decline in official aid flows have contributed to rising public and external debt levels as well as financing needs in many countries: close to half of all LICs are now assessed to be at high risk of experiencing debt distress or already in distress.⁵ Many small developing states are suffering from natural disasters with increasing frequency and intensity, against a backdrop of sluggish growth, while several countries have experienced internal conflict or heightened social tensions, disrupting economic activity. At the same time, many LICs—particularly those with diversified export structures—have enjoyed robust growth in recent years with contained vulnerabilities, and have had only episodic need for Fund financing. Around a third of LICs have accessed international bond markets over the past decade.⁶

4. The core question for the review is whether the toolkit remains well aligned with the evolving needs of the membership over the medium term, taking account of the need to maintain the PRGT’s self-sustaining lending capacity. Demand for Fund support is likely to remain strong in the coming years, whether it be from LICs with Balance of Payments (BoP) needs linked to financing policies to meet the Sustainable Development Goals (SDGs), from those undertaking adjustment programs to restore sustainable fiscal and external positions, or from those seeking to recover from natural disasters or escape from economic and social fragility. A range of policies exist to safeguard the Fund’s resources while responding to members’ needs; key policies relating to conditionality and program design are being assessed in parallel in the ongoing Review of Conditionality (IMF, 2018d).⁷ A new debt sustainability framework for LICs (IMF, 2018a) will also come into effect from July 2018, and help strengthen the analysis of debt vulnerabilities in designing new programs. Access policies (including phasing, limits, and norms) play an additional important safeguards role, by limiting the amount of concessional resources made available to any single member.⁸ Balancing members’ needs with safeguard considerations inevitably involves trade-offs.⁹

5. The Review is taking a two-step approach. This first paper takes stock of the experience with the LIC facilities and identifies areas where there may be scope for innovations and refinements, with the main options summarized in a final section. Based on feedback from Executive Directors, staff will prepare a second paper with specific proposals.

6. The remainder of this paper is organized as follows. The next section analyzes the experience with the LIC facilities and seeks to identify areas where design features could be improved. The ensuing section discusses the main areas where reforms warrant consideration, with a focus on access limits and norms, the degree of concessionality of Fund support, increasing flexibility of the Standby Credit Facility (SCF) in particular for precautionary support, the future of the Policy Support Instrument (PSI), and the provision of support to countries vulnerable to large natural disasters as well as those in fragile situations.

Box 1.

Key Features of the LIC Toolkit

The Fund’s concessional facilities took their current general form in 2009 (see supplement). The Extended Credit Facility (ECF) provides financial assistance to countries with protracted BoP problems, in support of economic programs that enable them to make significant progress toward a stable and sustainable macroeconomic position. The SCF provides financing to LICs that no longer face protracted BoP problems, supporting economic programs to tackle (actual or potential) short-term BoP needs; like the Stand-By Arrangement (SBA), it can be used on a precautionary basis. The Rapid Credit Facility (RCF) provides rapid low-access financing via a single drawing to countries where there is an urgent BoP need but a UCT-quality program is not feasible (or not necessary); beneficiaries have included countries hit by natural disasters and those in fragile/post-conflict situations. The PSI provides non-financial policy support to countries with a broadly stable macroeconomic position.

The facilities broadly mirror those in the GRA, though with some differences. Lending to LICs is at lower (currently zero) interest rates and longer maturities than in the GRA. The adjustment requirement for an ECF-supported arrangement is that “significant progress” be made towards a sustainable position, rather than attainment of such a position. The SCF has a maximum term of two years, compared with three years under the SBA. There are no incentives for early repayment in the concessional facilities, but rather hard access limits to contain demand.

The toolkit for LICs has continued to evolve. One new feature, added in 2010, was the Post Catastrophe Debt Relief (PCDR) trust, which allowed the Fund to join debt relief efforts when countries are hit by catastrophic natural disasters, and which was later repurposed as the Catastrophe Containment and Relief Trust (CCRT) to also cover countries facing pandemics. An important milestone was the creation in 2012 of a self-sustaining PRGT financing framework.

Box 2.

Rising Debt Vulnerabilities and Increasing Financial Risks to the PRGT

Debt vulnerabilities have risen in many LICs in recent years (IMF, 2018b). Some 44 percent of LICs faced significant debt-related challenges at end-2017, up from 25 percent in 2013. As discussed in IMF (2018b), there are downside risks attached to this assessment of debt vulnerabilities: fiscal adjustment and growth projections may fail to materialize, boosting debt burdens, while significant gaps in public debt data in many countries may be concealing accumulation of liabilities outside the current debt measurement framework. Liabilities to the Fund remain contained in most cases (below 4 percent of GDP), except in a handful of countries with large Fund quotas relative to GDP.

Rising debt vulnerabilities entail heightened financial risks to the PRGT. ^1/ Demand for PRGT resources has been elevated in recent years, with new commitments averaging SDR 1½ billion during 2015–17. The share of PRGT credit outstanding to countries at high risk of, or in, debt distress has risen sharply since 2013, reaching 30 percent at end-2017: PRGT credit to countries with high debt vulnerabilities amounted to 52 percent of the PRGT reserve account balance, up 10 percentage points from a 2009 peak. ^2/ As discussed in the recent update on IMF concessional financing, default on obligations to the PRGT by a single medium-to-large PRGT borrower at high risk of/in debt distress would reduce the PRGT’s self-sustained lending capacity by some SDR 15––68 million (IMF, 2018c). The concentration of PRGT credit across all borrowers has remained broadly unchanged in recent years.

Though demand for concessional resources is expected to moderate somewhat in 2018, financial risks to the PRGT may well increase further—even without a change in access limits policy—with potentially large program requests in the pipeline and fewer PRGT-eligible members presumed to use blended PRGT/GRA financing due to heightened debt vulnerabilities.

^1/ Similarly, lending to non-PRGT-eligible countries with high debt burdens creates credit risk for the GRA.^2/ The reserve account provides the ultimate backstop for the PRGT in the event of arrears or liquidity strains.

Stocktaking of Experience

7. This section examines the experience with usage of LIC facilities to help identify pressure points and potential gaps in the lending framework. It builds on the assessment conducted at the time of the last review, which generally found the toolkit to be working well, and looks at a range of issues raised by external and internal stakeholders.

A. Patterns of Use of the Facilities¹⁰

The four main instruments have been used flexibly to provide support to some three-quarters of PRGT-eligible countries (see Figure 1), under a very diverse set of circumstances. The ECF has been used extensively; the SCF has been used less frequently, typically on a precautionary basis; the RCF has provided emergency financing to a mix of countries hit by natural disasters or in fragile/post-conflict situations; and the PSI has been used repeatedly by a handful of African economies.

8. The ECF remains the workhorse facility, providing medium-term support to a wide range of LICs in varying situations. Since 2010, half of all PRGT-eligible countries have used the ECF at least once.¹¹ The ECF has been used to meet the BoP needs of LICs with large and protracted development needs that are scaling-up infrastructure and improving social safety nets. It has also provided support for countries recovering from natural and health disasters (see Box 3), for commodity exporters coping with price and output volatility, and for countries adjusting to political shocks and facing increased security spending.¹² ECF-supported programs have also been an important element of Fund engagement with fragile states.

9. Usage of the SCF has been more limited, mainly to provide financial support on a precautionary basis or, in some cases, to address short-term financing needs of PSI users. Initial concerns that the SCF might be overused, with members deterred from using the ECF have not materialized (IMF, 2009b). Since 2010, seven countries have requested support under the SCF,¹³ five of them on a precautionary basis, although no SCF remained in place as of end-May 2018. The SCF has also been used to provide short-term financial support to PSI users, via both disbursing and precautionary arrangements.¹⁴ Only one standalone disbursing SCF has been approved since 2010. For all SCFs, the sustainability requirements are more stringent than for an ECF,¹⁵ while the maturity structure of lending is shorter. Under normal global interest rates, the SCF would carry a higher interest rate than the ECF; in the event, interest rates on all Fund concessional lending have been set at zero during 2010–18.

10. Precautionary support has accounted for one-seventh of financing arrangements with LICs since 2010. The more limited use of precautionary support under the PRGT compared with the GRA partly reflects the fact that many LICs still face protracted BoP needs, for which the SCF is not designed.¹⁶ But some features of the SCF may also be limiting demand among those LICs that would qualify, such as the maximum two-year duration for SCFs and the more restrictive sub-limits on access.

11. The RCF has provided rapid low-access financing in a wide range of emergency situations, including for fragile states and after natural disasters. Some 17 countries have requested support under the RCF since 2010, including 7 repeat users. Financing has been provided to countries hit by natural disasters (including health epidemics)—13 requests by 10 countries—as well as to fragile states that were not yet able to implement UCT-quality arrangements (10 requests by 6 countries). The RCF has also been used twice to assist countries hit by other shocks (specifically, the global financial crisis and a terms-of-trade shock). RCF requests peaked in 2015, including two requests in response to the Ebola epidemic (see Box 3).¹⁷ Since 2010, six LICs have used the RCF only (and no other Fund facility) and all but one of their requests were in response to natural disasters. At end-2017, outstanding RCF credit (including its predecessor, the Exogenous Shocks Facility-Rapid Access Component, ESF-RAC) was about 9 percent of total PRGT credit outstanding.

12. The PSI has continued to provide policy support to several countries, while facilitating access to short-term financing. Since 2010, six countries—all from sub-Saharan Africa—that had achieved broadly stable and sustainable macroeconomic positions requested support under the PSI, and all but one have used the instrument repeatedly. The PSI has supported countries that are expanding development spending, without requiring Fund financing, while facilitating access to SCF financing when users were hit by shocks. Some LICs that previously used the PSI have expressed interest in support under the Policy Coordination Instrument (PCI).¹⁸

13. While posing operational challenges, the fixed review cycle of the PSI does not seem to have created undue pressure to complete reviews prematurely. The PSI limits the amount of time that a review can be delayed and allows for non-completion of review. This exerts pressure to complete reviews to avoid sending a negative “off-signal,” and the near-universal completion of PSI reviews has occasionally led to concerns about bias towards recommending completion. To assess such concerns, staff has compared program performance under PSIs with ECFs and SCFs with countries with broadly similar characteristics to those seeking PSIs.¹⁹ Reviews have on average been completed with around 90 percent of the quantitative targets met, with no significant difference between PSIs and ECFs/SCFs, although PSI reviews have been completed with a somewhat lower share of structural benchmarks met than ECF/SCF reviews.²⁰

Box 3.

Addressing BoP Needs Linked to the Ebola Outbreak

In 2014–16 several West African countries experienced the largest Ebola outbreak in history. The first outbreak was reported in Guinea in March 2014, which spread to other countries. Guinea, Liberia, and Sierra Leone were the most severely affected.

The Fund provided financial support to these three countries in several waves. The initial support was released shortly after the outbreak and augmented as the number of victims increased. Total support was about US$380 million, of which US$280 million came through existing lending facilities and the remainder as debt relief from the CCRT.

First round: In September 2014, a total of US$130 million of emergency financial assistance was provided to Guinea (RCF, 25 percent of quota), Liberia (ECF augmentation, 25 percent of quota), and Sierra Leone (ECF augmentation, 25 percent of quota). ^1/
Second round: In February 2015, Guinea (ECF augmentation, 42.1 percent of quota) and Liberia (RCF, 25 percent of quota) received financial support, followed by Sierra Leone (ECF augmentation, 50 percent of quota) in March 2015.
Debt relief: The CCRT was created in February 2015 and provided debt flow relief to the countries most affected by Ebola. All three countries received debt flow relief of 20 percent of quota.

In November 2015, Sierra Leone received a third augmentation of its ECF arrangement (45 percent of quota) to address the impact of Ebola and a sharp drop in commodity prices.

The Fund also supported other countries indirectly hit by the Ebola outbreak. In December 2014, Côte d’Ivoire received an ECF augmentation (40 percent of quota) to address additional BoP needs generated by its Ebola prevention plan, and in April 2015, The Gambia received an RCF disbursement (25 percent of quota) to address BoP needs following a decline in tourism due to the outbreak as well as the commodity price shock.

^1/ All numbers are in percent of quota at the time of program augmentations and RCF requests.

14. The duration and type of Fund-supported arrangements have varied markedly across countries.²¹ Forty-nine countries (some two-thirds of the total) had used at least one UCT-quality program during 2010–17, of which one-half had program engagement during at least six of the last eight years—accounted for by five countries making repeated use of the PSI, with most of the remaining countries being fragile states. Since 2010, nine countries have received blended PRGT-GRA support, whereas no PRGT-eligible countries have received standalone GRA support, consistent with the aim of the 2009 reform to design a toolkit that would adequately address the financing needs of PRGT-eligible countries.

15. Use of Fund arrangements has varied significantly across regions and country groups:²²

Regional pattern. LICs in sub-Saharan Africa (AFR), accounting for half of all PRGT-eligible countries, have been the heaviest users. At the other end of the spectrum, LICs in the Asia-Pacific region (APD), accounting for about a quarter of PRGT-eligible countries, have been the least frequent users of Fund-supported arrangements. To date, AFR countries have been the only users of the PSI.
Country type. Fragile states and countries heavily dependent on commodity exports have generally used Fund-supported arrangements more frequently than other countries, mainly tapping the ECF. Countries with a more diverse export structure have also been frequent users, across the spectrum of arrangements. Small states have been the least frequent users of Fund arrangements.²³
Non-use. Since 2010, about a quarter of LICs have not used the LIC facilities at all. The main reasons for non-use appear to have been the adequacy of domestic policy buffers or availability of alternative external financing sources.

16. LICs are much more likely to have repeated program engagement with the Fund than GRA-only members.²⁴ As of end-2017, 14 LICs had had at least 7 years of non-precautionary Fund arrangements in the preceding 10 years, compared with only 3 GRA-only members. The greater prevalence of longer-term engagement is consistent with the design of the LIC facilities, which explicitly allows for circumstances where countries with protracted BoP needs will require more than one Fund arrangement to fully address these difficulties. By contrast, GRA arrangements are expected to fully address BoP problems during the program period; making repeated use of arrangements is typically considered a sign of sluggish progress in resolving economic problems. The average duration of GRA programs has converged over time towards that of PRGT programs, reflecting the substantial adjustments needed in many GRA-only countries to address imbalances after the global financial crisis.

17. Despite growing economic challenges—and the scope under the current toolkit for longer ECF programs—the average length of the ECF has shown little change over time. Only one country (Madagascar, 2016) has requested an ECF with an initial duration exceeding the minimum three-year duration (for 40 months). While program extensions are common, only two countries (Malawi and Liberia), also long-term program users, have extended their ECF-supported programs to the maximum duration of five years. Given the frequency of longer-term engagement by many LICs, the prevalence of consecutive three-year programs is noteworthy—reflecting provisions in the PRGT that ECFs should normally have a three-year duration, as well as factors potentially limiting demand for longer-term arrangements including government time horizons, constrained planning capacity, and uncertainties concerning the longer-term outlook.

B. Access and Demand

The 2015 increase in access norms and limits provided additional room for the Fund to respond to requests for high access programs in recent years. Use of blending has increased, but is likely to become less frequent under the current blending policy as rising debt vulnerabilities reduce the number of countries presumed to blend. The Fund continues to have a significant catalytic effect on average, although the impact varies significantly with country circumstances.

18. Overall demand for Fund support has been strong and broad-based in recent years. During 2010–17, overall lending commitments to PRGT-eligible countries averaged SDR 1.4 billion per year, ranging between a peak of SDR 2.4 billion in 2016 and a trough of SDR 0.1 billion in 2013. Annual commitments from the PRGT alone averaged SDR 1.1 billion over the period, with the remainder coming from the GRA through blended arrangements. Disbursements to LICs have been lower and more stable than commitments, averaging SDR 0.9 billion per annum during 2010–17, of which SDR 0.8 billion came from the PRGT. More than half of the GRA resources made available to LICs were for precautionary support.

19. The average level of access to PRGT resources has risen since the 2015 increase in access limits.²⁵ The 2015 across-the board increase of access limits and norms created space for the Fund to respond to LICs’ evolving needs with concessional financing during the past three years. Compared with the counterfactual (see chart), the increased PRGT cumulative access limits allowed the Fund to provide higher concessional support and insurance to countries in diverse circumstances, including those hit by the 2014 commodity price shock and those facing increased security threats, limiting the need to compress other spending. Average annual access in ECFs and SCFs has increased since mid-2015 (see Figure 2); front-loaded access has occurred in only a few cases.

20. Access has been above the norm in several recent ECF arrangements, and has in a few cases come close to applicable access limits. Several LICs have recently requested high-access concessional financing. While most countries are still well below overall cumulative PRGT access limits, a handful of small island economies have approached the cumulative access limit under the RCF (but not to concessional financing provided under other facilities) following natural disasters. There has also been a shift towards greater variance around the norm, partly reflecting greater clarity on the role of norms, which are intended to provide indicative guidance on access but are neither ceilings nor floors.²⁶ Between 2010 and early 2013, half of all ECF arrangements received access at the norm (see Figure 3). From 2013 until November 2016, access under the ECF displayed greater variation around the norm, with about one-third of ECF arrangements at the norm. Since November 2016, access under most new ECF arrangements has been above the norm.

Use of Blending

21. The Fund’s blending policy targets scarce concessional resources to the poorest and most vulnerable LICs. There is a presumption that better-off PRGT-eligible countries receive blended PRGT and GRA resources, rather than relying fully on concessional Fund financing. In 2015, the funding mix in blended programs was adjusted—from a 1:1 ratio to 1:2—to offset in large part the impact on demand for resources of the generalized access increase (IMF, 2015c), while concessional access for presumed blenders remained capped at the applicable norm. Blending is considered a first step towards graduation from PRGT-eligibility, and criteria for presumed blenders are set accordingly—specifically, countries are presumed to blend if their per capita income exceeds the prevailing International Development Association (IDA) operational cutoff, or if per capita income exceeds 80 percent of this threshold and the country has sustained past and prospective market access. However, countries meeting the income or market access thresholds that are also at high risk of debt distress or in debt distress are not expected to blend, providing them with cheaper funding to reduce their debt service burdens.

22. The use of blending increased through 2016. Some 12 blended arrangements or outright disbursements have been approved since 2010, roughly equally split between the ECF, SCF and RCF—involving a geographically diverse range of countries and both precautionary and drawing arrangements. Their share in the number of overall arrangements (or outright disbursements) doubled between 2010–13 and 2014–17 to nearly one-fifth. Since end-2015, three blended arrangements have been approved, accounting for over 40 percent of all Fund resources provided to LICs (PRGT and GRA); some two-thirds of the blended funding came from the GRA, illustrating how blending has helped conserve scarce PRGT resources for other users.

23. However, the increasing prevalence of debt vulnerabilities suggests that, given current policies, fewer countries will be presumed to blend in the period ahead. While 33 members meet either the income or market access criteria for blending, the number of presumed blenders has fallen from 25 in 2015 to 15 in 2017 (returning to 2010 levels). In terms of quota, presumed blenders now account for one-third of PRGT-eligible countries, down from nearly half in 2015. This could have a noticeable impact on PRGT resources in case of new financing requests. Due to increased debt vulnerabilities, two large LICs have, in recent years, received purely concessional financing, rather than a blend of concessional and GRA resources.

Catalytic Effect of Fund Support

24. The combination of elevated financing needs and the scarcity of PRGT resources reinforces the need for Fund financing to catalyze additional donor/multilateral development bank (MDB) support for LIC borrowers. As past work has shown, Fund programs addressing policy or exogenous shocks have a significant catalytic impact on both the size and modalities of Official Development Assistance (ODA), with the greatest impact on LICs experiencing substantial macroeconomic imbalances or large shocks.²⁷ Staff analysis indicates that the Fund continues to play an important catalytic role in LICs, though its magnitude varies markedly depending on country circumstances (see Box 4).

Box 4.

Catalytic Role of the Fund’s Concessional Lending

During the past decade, ODA flows to LICs have declined relative to LICs’ GDP. Over this period, countries receiving concessional Fund financing attracted on average larger inflows than others, though with significant variation across countries.

The Fund’s concessional lending is expected to catalyze donor support. To support LICs in achieving their development goals, MDBs and bilateral development partners usually provide highly concessional long-term financial assistance, including grants; many LICs also tap resources on market or near-market terms, including through sovereign bond issues. The Fund’s financial support, in contrast, provides short-to-medium term financial support to help smooth external and fiscal adjustment over time (IMF, 2009a). Fund-supported programs are expected to play a catalytic role, leveraging the Fund’s scarce subsidy resources.

The literature suggests that Fund financing can have a significant impact on the size of ODA. Bal Gündüz and Crystallin (2014), using a propensity score matching approach, find that Fund programs addressing policy or exogenous shocks have a significant catalytic impact on the size and modalities of ODA. Results in Bird and Rowlands (2007) suggested a strong positive association between Fund-supported programs and donor assistance to LICs, while the literature overview of Steinwald and Stone (2008) suggested that the catalytic impact is not uniform across countries. Similarly, Kuruc (2018) finds a strong catalytic effect of Fund programs for fragile states, but not for other LICs.

During the past decade, ODA flows to LICs have declined in real terms, though not for all countries. ^1/ While ODA flows rose in nominal terms from an average of US$ 41 billion in 2000–09 to US$ 61 billion in 2010–16, they declined significantly in GDP terms for the median LIC during the same period, from 7.2 to 5.5 percent of GDP. However, the decline in ODA flows was not universal over this period, as flows to fragile countries and small states increased.

ODA flows were larger for countries receiving concessional Fund financing, suggesting a catalytic impact. The median LIC with a Fund facility received about 3 percent of GDP more in non-Fund concessional financing than the median LIC without a facility. At the same time, the size of ODA assistance during the time of Fund financing differed across country groups. An event study, similar to Kuruc (2018), suggests that ODA flows picked-up in the year of the first Fund disbursement and further increased in the following year. However, the impact varies widely, with a very large confidence interval. The impact is larger for fragile states and commodity exporters, and smaller for small states (though only a few received Fund financial support during this period).

The size of Fund financing relative to other donor assistance varies significantly with countries’ needs and circumstances. Donor conferences are an important vehicle to coordinate donor pledges with Fund support, including for countries hit by disasters, emerging from conflict, and following significant political changes. For example, following the 2015 earthquake in Nepal, an RCF disbursement addressed an urgent BoP need and while Fund financing was very small compared with the overall aid pledged in a donor conference (less than 2 percent), Fund support was expected to help firm up those pledges.

The Fund’s catalytic role also changes over time. An example is Sierra Leone, which received exceptionally high donor support during the 2014–15 Ebola crisis in the context of Fund support (see Box 3). Donor financing has subsided since then, and the Fund’s share of overall financing was higher in the new 2017 ECF than in previous years. In the context of Malawi’s 2012 ECF request, donors signaled their intention to restart and/or increase their support; in the event, donors suspended budget support in the wake of the “cashgate” corruption scandal in late-2013, resuming support with a lag after the Fund program was brought back on track. When Malawi was hit by weather-related shocks in 2015–16, the ECF augmentation again catalyzed significant donor support. Fund program engagement has also helped to catalyze donor support and facilitate market access for countries seeking support only on a precautionary basis (e.g., Honduras and Kenya).

^1/ ODA flows are defined as the sum of grants and loans.

C. Poverty Reduction Strategies

The provision of financial support through the Fund’s concessional facilities has long been linked to the pursuit of poverty reduction in LICs, as articulated in a poverty reduction strategy. The features required of such a strategy were substantially simplified in 2015: this has eased the documentary burdens on borrower countries while maintaining the assessed quality of such strategies in most cases.

25. Poverty reduction strategies are an important element of PRGT-supported programs of a medium-term duration. Programs supported by the PRGT aim to assist LICs in achieving a stable and sustainable macroeconomic position consistent with strong and durable poverty reduction and growth.²⁸ Following the World Bank’s decision to drop the operational role of Poverty Reduction Strategy Papers (PRSP), the Fund simplified the process for documenting the link between a country’s poverty reduction strategy and a PRGT program in 2015 (IMF, 2015b). While maintaining a clear link and preserving national ownership of the process, the new Economic Development Document (EDD)—which is required for ECF and PSI support—provides flexibility in the scope and coverage of poverty reduction strategies to reflect different country circumstances. It can involve an existing national development plan or strategy document, or a newly prepared document on the country’s poverty reduction strategy (PRS). “Good practice guidelines” are available, for example, to follow a participatory process in developing EDDs, but are not mandatory. Staff views on the PRS are included in program documents, as is a World Bank assessment of the EDD.

26. The new approach has significantly reduced the burden on country authorities, while the quality of poverty reduction strategies has largely been maintained. Since the adoption of the new policy, 17 countries have produced EDDs, mostly linked to ECF-supported programs. Two-thirds of these EDDs were existing national poverty reduction strategies and one-third were newly prepared documents, illustrating that the 2015 reforms have led to a material reduction in burdens on country authorities. Staff analysis of the EDDs and accompanying World Bank assessments found that most documents covered the core standards required for EDDs, although some lacked detail on policy advice or appeared to have unduly ambitious targets that did not provide useful guidance to policy-makers. Country authorities and mission chiefs viewed the PRS/EDD requirements as broadly appropriate in most cases. A few mission chiefs expressed concern that the requirements remain too onerous for small or fragile states and felt that the burden on these countries should be lowered.

D. Takeaways

27. The Fund’s concessional facilities have been extensively used by LICs and have provided flexibility in supporting countries under very diverse circumstances. Key lessons as to how the facilities have functioned in practice include:

Three-quarters of LICs have received support via at least one of the facilities during 2010–17, underscoring the strong demand from LICs for Fund policy and financial support.
One-quarter of LICs have had financial arrangements (SCF/ECF) during at least six of the past eight years, underscoring the central role of continued Fund support in these cases, both for policy and financial support, as well as to catalyze donor support.
Countries seeking financial support from the Fund viewed the ECF as the favored instrument, with one half of LICs having used the facility at least once.
The SCF has been used predominantly as a vehicle for obtaining precautionary support from the Fund, either directly or indirectly (as the financing “arm” of the PSI should adverse circumstances materialize). Initial concerns that the SCF might prove more attractive than the ECF for accessing financing did not materialize; in the event, the SCF has been used less than originally envisaged.
The RCF has been used to meet urgent financing needs in a range of circumstances, including in countries hit by natural disasters or those affected by fragility or emerging from conflict—in the former, typically as a once-off form of support, in the latter, typically as a first step to obtaining sustained Fund financial support via the ECF. The RCF also provided flexibility to deliver urgent financial support when disbursements under an existing arrangement were held up.
Access levels under financial arrangements have been rising in recent years, both in SDR terms and relative to norms, with high access programs linked to sharp commodity price declines (as in Chad) or to financing challenges in frontier markets (as in Ghana).
Blended arrangements have played a significant role in meeting financing needs while containing demand for concessional resources, but they are likely to become less common under existing rules, as rising debt vulnerabilities reduce the number of countries presumed to blend.

Options for Refinement

28. The review of experience suggests that the structure of the concessional facilities remains broadly appropriate—an assessment echoed in the consultations with stakeholders. The ECF has been the main vehicle for providing sustained financial and policy support to LICs with protracted BoP problems; the SCF has provided precautionary support; the RCF has been used to provide emergency low-access financing in a range of circumstances; and the PSI has provided effective policy support and signaling instrument for a small group of countries without current financing needs. There is no obvious need for a new instrument, although the viability of the PSI over time will be tested given the introduction of the PCI.

29. There is, however, scope for updating access policies and refining aspects of the various instruments to better respond to LICs evolving needs while maintaining adequate safeguards for Fund resources. The consultations with stakeholders (see Supplement) pointed to concerns with i) the adequacy of access to concessional resources; ii) the limits on provision of precautionary financial support; iii) the extent to which the toolkit met the specific needs of fragile states and of countries vulnerable to natural disasters; iv) the interest rate structure across the concessional facilities; and v) the implementation of the EDD policy. The scope for refinements in these areas is explored below.

A. Access Policies

Access policies must strike a balance between meeting the growing financing needs of members and maintaining the PRGT’s self-sustaining nature, including by containing credit risks.

Access Norms and Limits

30. Access policies provide a framework for balancing members’ financing needs against available resources and appropriate safeguards. Access to Fund resources is determined on a case-by-case basis using the standard criteria, including BoP need, program strength, capacity to repay the Fund, informed by debt sustainability analysis, Fund credit outstanding and record of past use (IMF, 2016e). The extent to which these resources can be made available from the PRGT, rather than the GRA, depends on PRGT access policies, including annual and cumulative access limits, access norms, and exceptional access rules, as well as blending policies. Given the limited availability of concessional resources, PRGT access policies help in allocating these resources between users of the PRGT financing facilities, while blending policies aim at targeting scarce subsidy resources to the poorest and most vulnerable members (see Section B).²⁹ Access policies also play an important safeguard function for the Fund and contributors to the PRGT, with cumulative access limits constraining overall exposure of the PRGT to a single borrower and annual access limits constraining the extent of front-loading of programs, limiting exposure risk by more closely linking disbursements to sustained program implementation.³⁰

31. LICs have the right to access GRA resources on the same conditions as any other Fund member (IMF, 2016d). Given the financial benefits from borrowing on concessional terms, staff advises PRGT-eligible members considering Fund financial support to borrow from the PRGT up to the applicable limits before seeking additional financing through GRA resources. Generally, all PRGT-eligible countries can access concessional financing up to the normal cumulative limit under the PRGT (225 percent of quota); for countries qualifying for exceptional access, the limit is 300 percent of quota. Should additional financing be warranted, this could be filled through GRA financing.

32. PRGT access norms and limits are projected to erode significantly relative to rising potential BoP needs. Over the next five years, GDP and trade volumes for LICs are projected to grow on average by around 8–9 percent annually in nominal US$ terms, while gross financing needs³¹ are expected to increase by 12 percent annually as countries borrow to support economic expansion and some finance a scaling-up in investment. As a corollary, effective access limits and norms will continue to erode over the period to the next facilities review, set for 2023. The Fund has ensured that, over time, access to PRGT (and GRA) resources has increased in line with members’ economic growth and financing needs, through changes in quotas or in access policies. However, any quota increase that may be agreed as part of the 15^th quota review is unlikely to become effective for some time.

33. While credit outstanding is currently well below the cumulative PRGT access limit for most LICs, the existing limits could become more constraining for a number of members in the coming years. By end-2020, under arrangements approved through end-2017, PRGT credit outstanding is projected to exceed 112.5 percent of quota (half of the cumulative limit) for eight countries, exceed 150 percent of quota (two-third of the cumulative limit) for three countries, while one will approach the normal limit of 225 percent of quota. For these members, access to concessional financing could be constrained in the event of successor arrangements or shocks leading to augmentation requests. Absent an increase in the normal cumulative access limit, it will become more likely over time that some members will request exceptional access under the PRGT (which is capped at 300 percent of quota) or to blend in GRA resources.

34. Changes in access policies need to be consistent with the resource envelope available under the three-pillar strategy for the self-sustained PRGT. The PRGT’s sustainable average annual lending capacity is currently estimated at SDR 1.31 billion, close to the target capacity of SDR 1¼ billion (IMF, 2018c). The assessment of the PRGT’s financial sustainability assumes periodic upward adjustments in access limits and norms (in SDR terms), broadly offset by eventual graduation of higher-income countries from PRGT eligibility.³² This estimate of annual lending capacity is robust to a range of scenarios and while capacity estimates are sensitive to a variety of factors affecting the PRGT endowment, the most recent assessment implies that a moderate generalized increase in access levels (e.g., by 25–35 percent) can be accommodated within the three-pillar self-sustaining PRGT framework, with the exact amounts dependent on further analysis, including of the quantitative impact of any reforms adopted. The follow-up Board paper, which will include a concrete proposed package of reforms, will quantify the resource impact of an increase in access norms and limits, as well as other changes to the facilities, under various scenarios.

35. Against this backdrop, there is a case for considering a generalized increase in access levels, consistent with the self-sustained PRGT, to accommodate LICs’ growing financing needs. An across-the-board increase, as done in July 2015, would help to partly offset the gradual erosion of access limits and norms, relative to conventional metrics (gross financing needs, value of external trade, and dollar GDP) since then and projected during the next five years. An across-the-board approach would maintain proportionality between PRGT access norms, annual and cumulative limits on total PRGT access, normal and exceptional limits, and annual and cumulative sub-limits on financing under the RCF. It would also reduce the risk that LICs would resort to supplemental GRA financing. The size of the increase that can be financed while ensuring the self-sustaining nature of PRGT financing would depend in part on the financial impact of other reforms introduced as part of this review.

36. There is also a case for revisiting the criteria for obtaining exceptional access to PRGT resources to ensure that countries with strong policies and very high financing needs can be adequately supported. The exceptional access policy allows for access levels up to one-third higher than the regular access limits, in cases where the country has a comparatively strong program and ability to repay the Fund; it is available only to poorer LICs, with the proposed program being subject to a higher level of scrutiny and specific procedural requirements to provide added safeguards.³³ Whereas exceptional access under the GRA is not uncommon—about a quarter of programs approved since 2010—exceptional access under the PRGT has been untapped since the 2009 reforms and is expected only in “exceptional circumstances.”³⁴

37. A number of reforms could be considered to facilitate use of exceptional access under appropriate circumstances. For one, exceptional access could be made available for all comparatively strong programs that are deemed to restore debt sustainability, with or without a debt restructuring operation.³⁵ In addition, the group of countries eligible could be expanded beyond poorer LICs without sustained past and prospective access to capital markets to include all countries that do not yet meet the market access or income criteria for presumed blending,³⁶ or more widely to include all countries that are not presumed to blend. Presumed blenders would continue to be expected to obtain access through the GRA to supplement PRGT financing. The foregoing proposals would require the amendment of the exceptional access criteria set out in the PRGT Instrument. Exceptional PRGT access is already subject to enhanced procedural scrutiny, including an early informal Board meeting prior to formal consideration;³⁷ further safeguards could be introduced, such as a requirement that individual program requests be accompanied by an assessment of the risk the program poses to the finances of the PRGT.³⁸

38. Alongside proposals to increase access levels, it would be important to ensure strong safeguards against credit risk and concentration, especially in light of rising debt vulnerabilities in many LICs. The primary safeguards against credit risk are robust program design, careful phasing of disbursements, and rigorous assessment of capacity to repay the Fund. The ongoing 2018 Review of Conditionality and Design of Fund-Supported Programs is examining the adequacy of program design and how it can be strengthened, while implementation of the new Debt Sustainability Framework for LICs (IMF, 2018a) will sharpen the analysis of debt vulnerabilities that should inform both program design and assessment of capacity to repay. In addition, staff will look at options to enhance procedural safeguards in cases of high debt vulnerabilities and potentially elevated credit risk in the follow-up Board paper on LIC facilities.

39. One option to further strengthen the safeguarding of PRGT resources would be to keep the cumulative normal access limit unchanged, while raising all other norms and limits. If combined with the full set of reforms to the exceptional access policy discussed above, this approach would ensure that countries that are heavy borrowers from the PRGT would benefit from the increase in all other norms and limits, but would need to meet the policy requirements (and higher scrutiny) of the exceptional access policy, including regarding the need for addressing debt vulnerabilities, when seeking access in excess of the current normal cumulative access limit of 225 percent of quota.

B. Financing Terms and Blending Policies

Financing terms and blending policies help target scarce subsidy resources toward poorer and more vulnerable LICs.

Financing Terms

40. In light of low global interest rates, the degree of concessionality of PRGT loans has been lower than envisaged when the interest rate mechanism was introduced in 2009. The average grant element in PRGT loans is currently estimated to be below 30 percent, compared with a traditional benchmark of 35 percent.

41. As global interest rates normalize, the SCF rate, and eventually the ECF rate, will rise (modestly) above zero. Under the PRGT interest rate mechanism,³⁹ interest rates for the ECF and SCF are expected remain at zero through 2018, while the RCF interest rate is set permanently at zero. As global interest rates rise, the interest rate mechanism will increase the SCF rate by 25 basis points compared to the ECF rate (starting when the SDR rate exceeds 0.75 percent), and both rates would increase by another quarter point once the SDR rate exceeds 2.0 percent. The mechanism’s design reflects several objectives, including (i) preserving the Fund’s scarce concessional resources; (ii) tailoring financing terms to the needs and capacity of LICs; and (iii) limiting fluctuations in concessionality and subsidy costs (IMF, 2009c). The SCF was tailored to meet short-term adjustment and financing needs—with sustainability achieved over the life of the program—as compared with more protracted needs of countries using the ECF. Furthermore, countries using the SCF generally have stronger macroeconomic positions than ECF users and somewhat higher capacity to service debt. Reflecting these differences and limited subsidy resources, the SCF interest rate was set somewhat higher than for the ECF. Shorter maturity and grace periods also reduce the SCF’s concessionality compared to the ECF.

42. There is scope for revisiting the parameters of the PRGT interest rate mechanism to ensure an adequate degree of concessionality of PRGT loans. In particular, the concessionality of the SCF remains lower than previously envisaged given the low interest environment. Given this, and the fact that earlier concerns about possible overuse of the SCF have not materialized, consideration could be given to aligning the SCF interest rate with that of the ECF.⁴⁰ This would imply that the first adjustment in the interest rate on both SCF and the ECF would not occur until the SDR rate exceeds 2 percent. From a resource perspective, such a change is likely to create only modest additional demands on subsidy resources – although this would need to be assessed in detail in conjunction with other reform proposals, including possible changes to the SCF and to blending policy (see paragraphs 44 and 47).

Blending Policies

43. Blending policy is the main tool for targeting scarce concessional resources to the poorest and most vulnerable countries, while continuing to provide more advanced LICs with lower financing costs than they will face on graduation from PRGT eligibility. Existing blending policy targets PRGT resources to (i) the poorest LICs (below the IDA operational cutoff for per capita income) and (ii) LICs at high risk of debt distress (to provide the least costly source of Fund financing when their debt service burdens are considerable). In the context of growing debt vulnerabilities in recent years, including among countries with income above the IDA cutoff, the second criterion has led to a decline in the number of PRGT-eligible countries that are presumed blenders; the group of countries at high risk of debt distress that would otherwise be presumed to blend represents about one-sixth of the aggregate LIC quota base. These countries obtain full access to PRGT resources (up to the various limits), instead of the 1:2 PRGT-GRA funding mix expected of presumed blenders.

44. Given the scarcity of PRGT subsidy resources, there may be a case for reconsidering the rules for presumed blending. One option would be to remove the exemption from the blending presumption for relatively higher-income LICs that are at high risk of debt distress (often because of their own weak macroeconomic policies); this would help further target scarce concessional resources to the poorest LICs. Alternatively, as blending is linked to moving towards graduation from PRGT-eligibility,⁴¹ one could introduce additional income thresholds (above the IDA operational cut-off level) above which debt vulnerabilities would not provide an exemption from blending, although staff analysis indicates that this would not necessarily generate significant savings.⁴² While such changes would enhance the longer-term resource envelope for poorer LICs, it would increase (via blending) the average cost of Fund resources for countries with high debt vulnerabilities; it could also potentially call for faster adjustment in these countries given the tighter requirements of GRA policies.⁴³

C. Precautionary and Policy Support

Given persistently high global risks and volatility, improving the Fund’s ability to meet LICs’ demand for contingent financing and policy support would help limit the need for costly self-insurance.

Precautionary Support

45. As many LICs’ linkages to international markets continue to deepen, their macroeconomic management challenges will progressively resemble those of emerging market economies. Increasing integration into international capital markets and trade flows has raised some LICs’ exposure to global risks and volatility (see Box 5), while the deepening of domestic financial markets, even if a welcome development, has increased the impact of financial shocks. The need to mitigate these risks is likely to increase demand for state-contingent Fund financing. Ensuring adequate precautionary support for LICs will limit the need for costly self-insurance and free up resources for, say, investment and social spending. Climate change and more frequent large natural disasters could also shift demand from emergency financing to ex ante insurance, including because of the scope it provides to request immediate disbursements in the event of a shock.

46. Several concessional options are available for LICs with potential BoP needs, but no significant actual needs. The appropriateness of these options depends inter alia on the underlying adjustment and financing needs, particularly the time needed to resolve macroeconomic imbalances. The SCF can be a genuine precautionary instrument for countries facing potential short-term BoP needs and which can resolve potential macroeconomic imbalances during the arrangement. A low-access ECF can be considered for a country with a protracted BoP problem but minimal present or prospective needs, and if needs increase, the arrangement could, in principle, be augmented. A country receiving support under the PSI can also rapidly access support under the RCF or SCF when needed. Under a precautionary SCF, drawing rights gradually accumulate and are available upon request, while augmenting a low-access ECF or requesting financing under an arrangement in the context of a PSI will likely take longer. Therefore, a country exposed to shocks likely to result in urgent financing needs, such as volatile portfolio flows, is normally served best by a precautionary SCF.⁴⁴

47. As countries become integrated and exposed to global markets, the existing design of the SCF may need to be made more flexible to meet countries’ evolving needs. As discussed above, usage of the SCF has been more limited than envisaged, partly reflecting design constraints introduced to avoid excessive recourse to the SCF relative to the ECF. Options for providing enhanced access to precautionary financing for LICs include:

Increasing the maximum duration of the SCF to three years.⁴⁵ Currently, precautionary SCF arrangements cannot be longer than two years, although there are no limits on successive arrangements provided they remain precautionary; under the GRA, three-year precautionary SBAs are available. Aligning the duration of the SCF with the SBA would increase flexibility for potential users to phase adjustment and reform programs;⁴⁶ it would also facilitate blending, as phasing of access would not need to be engineered to soften the abrupt decline in accumulated access rights when an SCF expires earlier than the concurrent SBA.⁴⁷, ⁴⁸ This proposal to increase the maximum duration of the SCF would require modifications to the PRGT Instrument; staff sees this change as being consistent with the purpose of the SCF.
Abolishing access sub-limits for precautionary SCFs. Currently, access under a precautionary SCF is restricted relative to a disbursing SCF arrangement.⁴⁹ These additional sub-limits were introduced to discourage tying up PRGT loan resources for contingent financing, while potential BoP needs were expected to be small for LICs. With increased integration into global capital markets, some LICs may potentially face larger BoP needs (e.g., in case of capital outflows). Abolishing the access sub-limits for SCFs that are precautionary at approval, thereby aligning access limits with drawing SCFs, would allow for higher and more front-loaded access and enhance the SCF’s insurance function.

Box 5.

Precautionary Support for LICs

The original arguments used when the toolkit was created to motivate the importance of precautionary support for LICs have strengthened over time. ^1/ While FDI flows were broadly stable during the last decade, other capital flows have been more volatile. Remittances remain high, but have declined in recent years. The exposure to global risks and volatility has increased as more LICs—around a third of the total as of end-2017—have tapped international capital markets and on a more significant scale. And shocks, as illustrated by the recent commodity price shock, can lead to a sharp decline in overall capital flows.

^1/ See IMF (2012), Box 1.

48. Increasing the flexibility of the SCF could increase PRGT demand. To the extent that the proposed changes lead to additional demand for precautionary Fund support, beyond a shift from disbursing arrangements, this would increase the demand for PRGT loan resources,⁵⁰ but without necessarily absorbing more subsidy resources over the longer term. The net impact on demand for subsidy resources is difficult to predict, even qualitatively. Having a precautionary arrangement in place increases the likelihood of actual disbursements if a country is hit by shocks (where the alternative is a country that does not have a precautionary arrangement in place when hit by such shocks). At the same time, the catalytic effects of such a precautionary arrangement on donor support, coupled with likely enhanced policies under an arrangement, could reduce the need for disbursements in the future.

49. The options highlighted above, if adopted, would not lead to a significant change in usage of the ECF. Most LICs continue to face protracted BoP problems and would thus not qualify for SCF support. And, as noted earlier, all but one of the SCFs approved to date were precautionary or concurrent to PSIs, meeting needs for which an ECF cannot be used. The changes proposed would make the SCF a more attractive instrument—but for countries seeking ongoing financing, the longer maturities on ECF financing and potentially longer program periods should still tilt the balance towards the ECF.

Policy Support

50. The PSI has proven to be a popular instrument with several African LICs, helping to underpin strong policies. It has enabled LICs with strong institutions that have achieved a broadly stable and sustainable macroeconomic position to signal their continued commitment to strong policies and their “graduation” from Fund financial support, helping them in securing continued donor support. At the same time, these countries have occasionally used the option provided by adhering to Upper-Credit Tranche (UCT) policy standards to access Fund financial support on a temporary basis from the SCF, concurrently to the PSI.

51. The advent of the PCI, a non-financial instrument available to all Fund members, calls into question whether the PSI is still needed. As with the PSI, the PCI can be used to signal a country’s commitment to UCT policies without a financing arrangement. In contrast to the PSI, eligibility for the PCI is not restricted to pre-qualifying PRGT-eligible countries. The PCI also has greater flexibility on the timing of reviews and does not require formal waivers when quantitative targets are missed.

52. Given feedback from country authorities, staff considers that the PSI should be maintained for the time being, potentially with some modalities streamlined in line with the PCI. The PSI has served several countries well and two PSIs remain in place. Introducing the more flexible review modalities and review-based conditionality to the PSI could enhance its attractiveness

53. Consideration could therefore be given to streamlining the following modalities in line with the PCI (IMF, 2016c and 2017b):

A more flexible review schedule. While concerns about a bias towards completing reviews under the PSI seem overstated, a more flexible review schedule could help overcome some operational challenges. Creating a limited time buffer around reviews can provide extra time if there are delays in program implementation, while maintaining regular updates on program performance.
Review-based conditionality. While preserving the UCT standard and the Executive Board’s judgment on whether to complete a review, a review-based approach to monitoring program conditionality could alleviate stigma and streamline the review process. As under the PCI, monitoring would focus on assessing whether policies are on track to meet the specified program objectives, rather than emphasizing small deviations from specific targets. Staff would continue to assess whether deviations from program targets do not undermine the achievement of program objectives—namely whether any deviation from a quantitative target is minor and temporary, or if sufficient corrective actions have been taken.⁵¹

D. Support for LICs Vulnerable to Natural Disasters

The Fund can provide emergency financial assistance to countries hit by natural disasters, via the RCF. In exceptional circumstances, it can use a trust fund to provide relief on debt payments falling due to the Fund. Vulnerable countries can also seek program/precautionary assistance from the Fund, but the small states most at risk of experiencing large disasters have not usually done so. Steps to promote ex ante resilience building in vulnerable countries are being covered in a separate work-stream.

54. Natural disasters represent a significant, and likely growing, tail risk for many LICs. A record number of natural disasters occurred in the early 2000s—with LICs particularly exposed to floods, epidemics, storms and droughts—and the frequency and intensity of weather-related disasters is expected to continue rising given climate change. LICs are also vulnerable to large disasters (those causing damage exceeding 20 percent of GDP) with 6 percent of disasters in LICs and 25 percent of those in small developing states breaching this threshold, compared with just 0.4 percent of disasters in more developed economies— although truly catastrophic events remain rare.⁵² The impact of large natural disasters on small states can be magnified by weak institutions and capacity, as well as more constrained access to financing.⁵³

55. The Fund has a number of tools for providing assistance to countries hit by natural disasters. Since 2010, there have been some 26 disaster-related lending operations with LICs— roughly equally split between RCFs and program augmentations—just over a third of which were accounted for by small states. The tools include:

Emergency assistance. The RCF provides speedy but limited financial assistance without requiring a UCT program. Under the exogenous shocks window, the annual access limit at 37.5 percent of quota is somewhat higher than average annual access under an ECF at the high-access norm (30 percent of quota) and significantly higher than the low-access norm (18.75 percent of quota). Under the large natural disaster window, annual access can be up to 60 percent of quota if the disaster damage exceeds 20 percent of GDP. As part of RCF requests, staff prepares an updated post-disaster macroeconomic framework to help catalyze donor support and, in most cases, RCF are approved within four months of a natural disaster. Small states have accounted for around 60 percent of emergency assistance requests since 2010.
Program support. Countries hit by natural disasters can request a medium-term arrangement, such as an ECF, to support adjustment and reconstruction, while an existing arrangement can be augmented if a natural disaster exacerbates BoP needs. As an ex ante measure, a country can also request precautionary support under an SCF and gradually accumulate access rights that can be exercised if a disaster strikes and a BoP need arises.⁵⁴ Since 2010, about half of the Fund’s financial support to LICs in response to natural disasters has been provided via augmentations of existing arrangements—although small states have relied almost exclusively on the RCF in such circumstances.

56. In exceptional circumstances, the Fund can also join in debt relief efforts after major disasters. If countries are hit by catastrophic disasters, the Fund can provide debt flow and stock relief through the Post-Catastrophe Relief (PCR) window of the Castastrophe Containment and Relief Trust (CCRT).⁵⁵, ⁵⁶ Eligibility is limited to the poorest and most vulnerable LICs.⁵⁷ Catastrophic disasters are defined as those that (i) directly affect at least a third of the population; and (ii) destroy more than a quarter of the country’s productive capacity or cause damage exceeding 100 percent of GDP. The CCRT can provide debt flow relief on debt service falling due to the Fund in the two years following the disaster; it can provide debt stock relief—including full repayment or clearance of eligible debt owed to the Fund—only as part of a concerted international debt relief effort.

57. Responding to concerns raised by stakeholders, staff considered options to better align the Fund’s toolkit with the potentially large and diverse needs of countries coping with natural disasters. The cumulative RCF access limit caps the availability of emergency concessional financing; countries prone to natural disasters can quickly reach this limit, either following a single large disaster or repeated disasters, and could benefit from an increase in this limit. The massive hurricane damage suffered by Dominica in September 2017 raised questions about the appropriateness of the eligibility criteria for debt relief from the CCRT and the relevant disaster thresholds.⁵⁸ Therefore, staff considered whether access to coordinated debt relief under the CCRT for countries hit by catastrophic disasters could be broadened. Finally, given that the immediate aftermath of a natural disaster is challenging for LICs, especially small states, ways to bridge the time between a disaster and the provision of emergency financing are explored.

Raising RCF Access Limits

58. The cumulative RCF limit provides an upper bound on the availability of emergency concessional financing. Countries prone to natural disasters can exhaust the space available under the RCF, either following a single large disaster (e.g., Vanuatu) or repeated disasters.⁵⁹ Dominica is an example where repeated tapping of emergency assistance has brought RCF credit outstanding close to its limit, leaving little room to provide emergency financing when a large hurricane struck the island in late 2017. An increase in the cumulative RCF access limit would especially benefit countries close to the limit, prone to recurrent disasters, and facing exogenous shocks associated with climate change. The staff access proposal envisages that the cumulative RCF access limit, as well as the annual sub-limits under the exogenous shocks and large natural disaster windows, would increase by the same proportion as the annual access limits/norms for the ECF (see paragraph 35).

59. Any modification to RCF limits needs to strike a balance between user needs and safeguard considerations. The absence of ex post conditionality has raised moral hazard concerns, i.e., that countries favor accessing Fund resources through single drawings rather than Fund-supported arrangements with UCT-quality policies.⁶⁰ For example, Dominica can tap additional Fund resources through an ECF given its post-disaster protracted BoP needs; an augmentation under an ECF could be considered if struck by another natural disaster.⁶¹ Fund policy has been to ensure that access under the RCF is significantly less than under a UCT facility (e.g., the cumulative sub-limit on RCF support is currently one-third of the normal cumulative access limit to PRGT resources).

Widening Access to the CCRT

60. A widening of access to the PCDR window of the CCRT, currently limited to the poorest and most vulnerable LICs, could be considered if funding needs were adequately addressed. Dominica, hit by a massive hurricane in late 2017, was not eligible for debt relief from the CCRT, as its income per capita exceeded applicable income thresholds. In keeping with the spirit of providing debt relief only for catastrophic disasters, staff has explored a few options for widening access to debt relief under the PCR window that emerged in consultations with stakeholders: (i) introducing an additional income threshold for microstates, (ii) allowing access to all PRGT-eligible countries, and (iii) lowering the disaster threshold. The CCRT currently has funds of only about SDR 145 million, and is some SDR 200 million below the required target under current policies, because of both the initial funding shortfall and because limits on debt flow relief were not halved relative to quota following the 14^th Quota Review. Adoption of changes would require additional fundraising to support what is already an under-funded instrument.

Introducing an additional, higher income threshold for microstates. The income criteria for CCRT eligibility are in principle aligned with the entry criteria for PRGT-eligibility, except for microstates which become PRGT-eligible if their per capita income is less than five times the IDA cutoff but CCRT-eligible only if per capita income is less than twice the IDA cutoff. Based on natural disasters during the last two decades, no additional disaster would have qualified for CCRT funding if the income criterion for micro-state eligibility been aligned with the current PRGT income criterion.
Widening access to all PRGT-eligible countries. Eligibility for the CCRT is currently limited to the poorest LICs: those with per capita income less than the IDA cutoff (twice the cut-off in the case of small states). Very large disasters occur rarely, although their frequency is expected to increase with climate change. During 2000–2017, only three disasters caused damage exceeding 100 percent of GDP (Grenada in 2004, Haiti in 2010, and Dominica in 2017); making debt relief under the CCRT available to all PRGT-eligible countries (at current disaster cost thresholds) over this period would have added only two disasters in microstates to the qualifying disasters (since Haiti was provided with debt relief via the precursor of the CCRT—the PCDR).⁶²
Lowering the disaster threshold. Based on current CCRT eligibility, the disaster threshold would need to be lowered to 30 percent of GDP for any disaster during the last two decades, beyond Haiti in 2010, to qualify: it would cover Nepal’s earthquake in 2015.⁶³ If, at the same time, an additional income threshold was introduced for microstates, Samoa’s disaster in 2012 would also have qualified and Tonga’s in 2001 would be very close (with 29 percent of GDP damage).⁶⁴

61. Combining substantive changes in eligibility with reductions in the disaster threshold would likely result in markedly higher demands on the already underfunded CCRT. Absent an elastic supply of donor support over time, an expansion of eligibility would eventually squeeze the support made available to current CCRT-eligible countries, including under the pandemics window. To address this concern, consideration could be given to making the effectiveness of any such changes conditional upon the achievement of a new fundraising target. The modifications to eligibility discussed above that affect the purpose of the CCRT would require the consent of contributors.⁶⁵

Short-Term Relief on Debt Service

62. In responding to members’ needs in the challenging aftermath of a natural disaster, emergency Fund financing can be provided very quickly, though not always immediately. While RCF financial support can be and has been provided very rapidly, in some cases it may be preferable to first work with the authorities to put in place a post-disaster macroeconomic policy framework (e.g., to help coordinate donor support), which could take several months in the absence of a pre-existing program. In this interim period, a cash-constrained country could incur technical arrears to its creditors, and the Fund could face reputational costs if the country needs to make a large payment to the Fund. Some stakeholders thus favor options to bridge the time between a disaster and an RCF disbursement, and delay debt service payments for disaster-struck countries.

63. Introducing explicit state-contingent features into PRGT lending, such as hurricane clauses, would pose serious challenges for the Fund. Some stakeholders have suggested allowing for an automatic rescheduling of debt service payments to the Fund if members experience natural disasters beyond specific pre-determined triggers. However, the concessional financing framework under the PRGT instrument prohibits the rescheduling of debt service, and significant reforms would be needed to change this.⁶⁶ Given the pass-through of PRGT loan resources to eligible borrowers, a change of the instrument would have to be financed either by amending all PRGT loan agreements with donors or through interim payments from the PRGT reserve account. Either option could adversely affect future PRGT loan mobilization efforts. Introducing state-contingent features could also weaken the Fund’s preferred creditor status and trigger calls for postponing debt service in other situations (including the GRA)—with wider implications for the Fund’s financing model and the reserve asset character of members’ claims on the Fund.

64. There may be alternative ways to help members cover debt service falling due in the immediate aftermath of a large natural disaster. A donor-funded administered account could be established to cover the debt service to the Fund in the immediate aftermath of a prespecified disaster event. Alternatively, the CCRT could be expanded to include a window to provide such financing, with use to be authorized by a Board decision (rather than specification of ex ante triggers);⁶⁷ this would require a re-purposing of the Trust, including consent of donors. Beyond providing the member with some immediate relief and fiscal space in the aftermath of a disaster, such a provision could in principle have a catalytic effect by encouraging other creditors to consider standstills.⁶⁸ The legal aspects and fundraising needs of these options will require further assessment when the options are fleshed out.

Ex ante Financial Assistance

65. In addition to possible reforms to the Fund’s ex-post financial assistance, staff is also considering how best to help members build resilience ex ante. In countries that experience frequent natural disasters, policies to prepare in advance can reduce the human and economic costs if disasters occur (IMF, 2016f). This can involve strengthening planning capacity, calibrating financing strategies (including insurance), and developing targeted infrastructure investment. Such measures require expertise on policies and institutional frameworks outside the Fund’s competence, and close collaboration with other institutions will be needed. Work is therefore underway, in collaboration with the World Bank, to explore how best to support resilience building of this nature in the context of Fund engagement with countries vulnerable to disasters.

E. Supporting LICs in Fragile Situations

66. LICs in fragile situations are frequent users of the Fund’s financing facilities. Over half of all PRGT-eligible countries are assessed to be fragile, and in any given year nearly half of them have had some form of Fund financial support in place. The circumstances and needs of fragile countries vary markedly: those emerging from conflict have different financing and policy support needs than, say, small states with weak institutional capacity or countries emerging from international isolation.⁶⁹ Reflecting the challenges of emerging from fragility, many fragile countries have had prolonged engagement with the Fund, often with interruptions as programs go off track.⁷⁰

67. Fund financial support for fragile states is generally modest compared with overall financing needs, but Fund engagement plays an important policy support and catalytic role. The IEO evaluation on fragile states (IEO 2018) found that the presence of an IMF arrangement helped catalyze substantial additional support from multilateral institutions and bilateral donors—a welcome feature, given that fragile states usually have limited space for accumulating debt (even on RCF terms) and instead need grants and/or highly concessional loans (such as IDA terms).

68. Fund financing for fragile states is typically provided through either the RCF or an ECF-supported arrangement.⁷¹ The first instrument is used when countries are not able to implement an economic program with policies that meet UCT standards; the limit on annual access via the regular window of the RCF (18¾ percent of quota) is modest, reflecting the absence of (i) ex post conditionality and (ii) upper credit tranche quality policies. The Fund’s strategy for engagement with fragile states (IMF, 2011) envisages repeated use of the RCF (ideally coupled with a Staff-Monitored Program (SMP) to build a track-record) if substantial time is needed to build the capacity and political support to move to an ECF. Four fragile LICs have received consecutive disbursements under the RCF and several fragile states have made a transition from the RCF to ECF-supported programs, albeit with mixed results in some cases.

69. Stakeholders consulted for this review generally welcomed the Fund’s extensive engagement with fragile states, but many flagged the scale of Fund financing and the length of arrangements as issues for reconsideration. IEO (2018) echoes these views, arguing that the greatest reform need is to reduce the gap between the limited access/non-UCT policies/once-off nature of the RCF and the higher access/UCT policies/three-year framework of the ECF. They suggest, as options, i) raising the annual access limit under the RCF (perhaps divided into two disbursements a year); and ii) introducing a short UCT arrangement as a bridge to an eventual ECF arrangement (IEO (2018), page 53).⁷²

70. Staff sees a need for a reaffirmation of the Fund’s strategy of using repeated RCFs in conjunction with SMPs for countries that need more time to build capacity to implement UCT-policies, accompanied by enhanced guidance on implementation.⁷³ As to the appropriate limits on access to RCF drawings, an increase in annual and cumulative access limits under the RCF is justified, in line with the general expansion of access limits proposed above.⁷⁴

As noted in the preceding section, staff sees merit in increasing the cumulative limit on RCF drawings by the same proportion as the increase in annual access limits/norms for the ECF.
The annual access limit under the “regular window” of the RCF could be increased by the same proportion as the cumulative limit on RCF drawings.
Following the IEO, the argument can be made that the limit on annual access is modest (at half the level provided under the exogenous shocks window of the RCF), leaving the Fund with very little flexibility to calibrate support to specific country circumstances—particularly post-conflict situations.
Financial support for countries in fragile situations that do not have the capacity to implement coherent economic programs should ideally be provided in the form of outright grants (rather than burdening an already-stressed country with debt)—implying that Fund lending should not be a lead source of financing in such cases. But there may be circumstances where significant Fund disbursements are warranted to help tackle emergency situations.
To provide greater flexibility, the annual access limit under the regular window of the RCF could be set equal to the limit for the exogenous shocks window. To safeguard resources, this increase could be coupled with the requirements that i) an individual disbursement should not exceed half the (new) annual access limit; and ii) annual access should not exceed one-half of the annual access limit on more than one occasion.⁷⁵

71. In fragile states emerging from conflict and/or facing large uncertainties, staff sees merit in making full use of the flexibility inherent in the ECF to focus attention on immediate objectives—essential stabilization and capacity-building measures, “quick wins” to build public support, priority actions to maintain social stability. In such a context, policies for the first year of the program would be spelt out in detail, while medium-term program objectives would be specified in broad terms, with the understanding that they would be adapted flexibly to evolving circumstances. As the program proceeds and uncertainty is reduced, a more concrete medium-term reform agenda could be fleshed out, drawing on lessons learned from experience. The need for an explicit poverty reduction strategy could also be deferred until there is sufficient state capacity to flesh out a country-owned strategy (see paragraphs 76–77).

72. The argument can be made that, in situations of very high uncertainty and sociopolitical instability, fitting a short term-focused program into a three-year ECF arrangement dilutes focus and is not aligned with policy-makers’ time horizons. As implied by the IEO, a short-term arrangement may be better suited to deliver on immediate priorities and ensure country ownership. Such a short-term arrangement (e.g., covering 12–18 months) could act as a bridge towards a regular ECF, which would be deployed as uncertainty regarding the medium-term outlook declined and it became feasible to have a meaningful discussion with country authorities on a three-year program. There are circumstances where the highest priorities are short term in nature and policy discussions on medium-term objectives and policies may be a distraction: examples include (i) post-conflict or near-conflict situations where an appropriately designed short-term arrangement can deliver very large payoffs (relative to the costs of falling into conflict); and (ii) situations of uncontrolled inflation, where a tightly focused stabilization program that succeeds would represent an essential step forward towards a sustainable position. In such circumstances, countries with adequate capacity and political will may find that a short-term UCT quality arrangement is better suited to their needs than a longer-term program.⁷⁶

73. Modifying the allowable length of the ECF to include arrangements of 12–18 months in length would represent a significant structural change in the architecture of LIC facilities. The ECF and its predecessors, which has long been the primary vehicle for Fund financial support to LICs, provides medium-term support to countries with protracted BoP problems.⁷⁷ The 2009 architecture reform ensured that the two UCT-quality facilities (the SCF and ECF) did not overlap— they are delineated by the length of the BoP need, which is judged to be either “short term” or “protracted.” The reform also preserved the longstanding presumption that protracted BoP problems are best addressed via a three-year or longer program, in part because it grounds policy advice within the framework of a medium-term policy framework and PRS. A short-term ECF (e.g., 12–18 months) could create the risk of potential overlap between the SCF and the ECF in circumstances where the delineation between short-term and protracted balance of payments needs is not clear-cut. A more important concern is that it could encourage countries with protracted needs to seek shorter-term financing in “bad times,” eschewing a medium-term program in the cyclical upswing (“good times”) when policy buffers need to be re-built and deeper reforms should be undertaken. Any short-term ECF would therefore need to be clearly ring-fenced to ensure that it was usable only in the special circumstances cited above, and not as a vehicle to be tapped by all countries with protracted BoP needs.

74. Given that the introduction of a shorter-term ECF represents a significant change in the architecture of LIC facilities, it would very likely require the consent of contributors to the PRGT. It would also require consideration of the impact of such a reform for potential future fundraising efforts. Whether such a change could garner the necessary broad support remains to be determined.

75. Some stakeholders have argued that, in fragile states where institutional capacity is very weak, achieving a sustained improvement in macroeconomic conditions requires more time than the three-year period normally covered by an ECF. An example cited is the challenge of building a national tax system (policies and administration) that will produce a transformative increase in domestic revenue mobilization—a process that typically involves an extended multi-year reform program. In such situations, there would be a case for a longer-term program (say, five to six years), in which a central focus of the program would be on taking the multiple steps needed to achieve a fundamental up-grading of the tax system. A longer-term program of this nature would likely make sense in situations where the challenge is overcoming consistently sluggish growth and weak institutions, rather than coping with political and social instability; with significant instability, use of repeated three-year ECFs, wrapped within a longer-term strategy (such as a five-year plan), would likely be the more sensible option.

F. Links from PRGT Support to Poverty Reduction

76. Routes to strengthening the links between PRGT-supported programs and poverty reduction can be explored.

Strengthen the EDD process, providing more support to countries with weak capacity. As gaps in EDDs usually reflect insufficiently detailed policy advice or unrealistic targets, the Fund would need to work closely with the World Bank and other donors to identify options, including capacity building and technical assistance, to strengthen the strategy documents. Such external assistance would help reduce the burden on countries with weak capacity (notably, small or fragile states), while taking cognizance of the need to avoid undercutting national ownership.
Allow greater flexibility on the timing of EDDs for countries with weak capacity, backing this move with targeted technical assistance.
The links between poverty reduction objectives and PRGT-supported programs by drawing more on the work of the World Bank. For example, the Fund could draw more closely on the World Bank’s Systematic Country Diagnostics (SCDs) and Country Partnership Frameworks (CPFs) to inform the design of PRGT programs.

77. There is also a case for standardizing the use of the EDD across PRGT arrangements (the SCF, ECF, and PSI). A simple unified approach would be to specify that: “An EDD is expected to be produced in all cases where the Fund-supported program has a term of at least two years.” An EDD should be in place early on, normally by the time of the 1^st review—but exceptions to this rule could be made in circumstances where capacity to produce an EDD is scarce and better deployed to other ends—e.g., as in the case of a country emerging from conflict, where state capacity is at a premium.

Issues for Discussion

Lessons from Experience

Do Directors agree with the main conclusions of the stock-taking exercise?
that the basic architecture of LIC facilities remains broadly appropriate, and has generally been effective in tailoring Fund support to the diverse needs of LICs;
that the ECF remains the cornerstone of the LIC facilities architecture, and has generally been effective in delivering support in a wide range of situations to countries facing protracted balance of payments problems;
that the SCF is a valuable component of the LIC facilities toolkit, albeit used primarily as a vehicle for obtaining precautionary support rather than as a disbursing arrangement;
that the RCF has effectively met emergency financing needs without undermining demand for Fund arrangements or creating excessive financial risks to the PRGT.

Access Policies

Staff sees scope for increasing access limits and norms to meet growing financing needs of LICs, with the feasible scale of the increase to be determined by the objective of preserving the self-sustaining nature of the PRGT, taking account of the likely impact of other proposed reforms to the facilities. Staff also sees merit in revisiting the exceptional access policy to ensure that countries with comparatively strong programs and very high financing needs can receive the necessary financial support, subject to appropriate safeguards (particularly in cases where debt levels are elevated).

Do Directors agree with the proposal of a generalized increase in access limits and norms in a manner that is compatible with the self-sustaining financing of the PRGT?
Do Directors see merit in revisiting the exceptional access policy to ensure that countries with strong programs have access above the normal access limit, subject to heightened scrutiny of program quality and risks to the PRGT?
Do Directors support keeping unchanged the normal cumulative limit on access to PRGT resources as a tool to contain credit risk to the PRGT?

Blending Policies and Financing Terms

Staff sees merit in revisting the PRGT interest rate policy to align the SCF interest rate to the ECF interest schedule, which would enhance the degree of concessionality. Given the scarcity of PRGT resources and increasing debt levels among comparatively higher-income LICs, there is a case for reconsidering aspects of the blending policy, with a view to better targeting subsidy resources to the poorest countries.

Do Directors support aligning the SCF interest rate to the ECF interest rate schedule?
Do Directors favor reconsidering the current exemption from blending requirements for countries that meet the income and/or market access criteria for blending when their risk of debt distress is (re-)classified as “high”?

Adjustments to the Facilities—the SCF

With the more advanced LICs increasingly integrated into global markets, and thus exposed to potentially larger shocks, staff sees merit in loosening the constraints on providing precautionary and short-term financial support under the SCF.

Do Directors agree that the constraints on use of precautionary/short-term support should be eased by:
- aligning the maximum length of the SCF with that of the SBA (three years);
- eliminating the sub-limits on access for precautionary arrangements;
- aligning EDD requirements for arrangements under the SCF, ECF, PSI in excess of two years.

Adjustments to the Facilities—the PSI

Given its track record, staff considers that the PSI has been a useful tool for LICs that have achieved broadly stable and sustainable external positions:

Do Directors see merit in retaining the PSI, separately from the newly established PCI, as a policy support instrument specific to LICs?
Do Directors see merit in modifying the features of the PSI by introducing a more flexible review schedule and review-based conditionality, as in the PCI?

Support for Countries hit by Natural Disasters

Given the likely increase in the impact/frequency of natural disasters, staff sees merit in raising RCF access limits in line with other access limits and norms, and in intensifying efforts to help vulnerable countries build ex ante resilience. Staff notes that the CCRT is not adequately funded and that expanding the scope of support provided would require an additional injection of donor funds. Staff points to options for establishing a mechanism to provide short-term financing to cover debt service falling due after major disasters, bridging the time needed to disburse emergency financing, including under the RCF.

Do Directors agree that RCF access limits should be raised in line with other access limits to help address urgent financing needs after natural disasters?
Do Directors favor any particular adjustments to the current eligibility and access criteria for the CCRT? Do they see merit in expanding eligibility to cover all PRGT-eligible small and micro-states? Do they agree that the effectiveness of any such adjustments should be conditioned on additional fundraising to ensure that the CCRT is adequately funded?
Do Directors see merit in exploring a discretionary or automatic mechanism to provide countries hit by large natural disasters with financing from a trust fund or administered account to meet debt service falling due to the Fund in the wake of the shock?

Support for Countries in Fragile Situations

Staff sees merit in reaffirming the role that repeated use of the RCF can play in providing policy support to countries that need more time to build capacity for a UCT-quality arrangement while raising the regular annual RCF limit in line with other access limits; there is also a case for enhancing the flexibility of Fund financial support by a proportionally larger increase in the annual limit, with additional safeguards on use. Staff also sees merit in making full use of the flexibility of the ECF to focus on short-term priorities in situations of heightened uncertainty and socio-political fragility. The IEO proposal to introduce a shorter-term arrangement within the ECF could have benefits in certain situations, especially post conflict, but also drawbacks; it would require substantial changes in the design of the LIC facilities architecture, very likely requiring support from PRGT contributors.

Do Directors support the Fund’s current approach of relying on repeated RCFs, together with SMPs, to assist countries not yet able to implement UCT-policies?
Do Directors see merit in increasing the annual access limit to the regular window of the RCF by proportionally more than the increase in other access limits?
Do Directors see merit in modifying the current policy on length of ECF arrangements to:
Allow for a short-term ECF arrangement, ring-fenced for countries with protracted BoP problems where uncertainties are elevated, such as post-conflict situations.
Allow the term of a regular ECF to be flexible between three and five years, depending on country circumstances, such as reform priorities.

References

Bal Günduz, Yasemin, and Masyita Crystallin, 2014, “Do IMF-Supported Programs Catalyze Donor Assistance to Low-Income Countries?” IMF Working Paper No. 14/202 (Washington: International Monetary Fund).
- Search Google Scholar
- Export Citation
Bird, Graham and Dane Rowlands, 2007, “The IMF and the Mobilisation of Foreign Aid,“ The Journal of Development Studies, Vol. 43 (5).
- Search Google Scholar
- Export Citation
Independent Evaluation Office, 2018, “The IMF and Fragile States—An Evaluation by the Independent Evaluation Office,” Independent Evaluation Office (Washington).
- Search Google Scholar
- Export Citation
International Growth Centre (IGC), 2018, “Escaping the Fragility Trap,” Commission on State Fragility, Growth, and Development (London: London School of Economics and Political Science (LSE) and University of Oxford’s Blavatnik School of Government).
- Search Google Scholar
- Export Citation
International Monetary Fund, 2009a, “A New Architecture of Facilities for Low-Income Countries” (Washington).
- Search Google Scholar
- Export Citation
International Monetary Fund, 2009b, “The Chairman’s Summing Up: The Fund’s Facilities and Financing Framework for Low-Income Countries” (Washington).
- Search Google Scholar
- Export Citation
International Monetary Fund, 2009c, “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, 2011, “Macroeconomic and Operational Challenges in Countries in Fragile Situations” (Washington).
- Search Google Scholar
- Export Citation
International Monetary Fund, 2012, “Review of Facilities for Low-Income Countries” (Washington).
- Search Google Scholar
- Export Citation
International Monetary Fund, 2014, “Honduras—Request for a Stand-by Arrangement and an Arrangement under the Standby Credit Facility,“ IMF Staff Report (Washington).
- Search Google Scholar
- Export Citation
International Monetary Fund, 2015a, “Proposal to Enhance Fund Support for Low-Income Countries Hit by Public Health Disasters” (Washington).
- Search Google Scholar
- Export Citation
International Monetary Fund, 2015b, “Reform of the Fund’s Policy on Poverty Reduction Strategies in Fund Engagement with Low-Income Countries—Proposals” (Washington).
- Search Google Scholar
- Export Citation
International Monetary Fund, 2015c, “Financing for Development—Enhancing the Financial Safety Net for Developing Countries” (Washington).
- Search Google Scholar
- Export Citation
International Monetary Fund, 2016a, “2016 Poverty Reduction and Growth Trust—Review of the Interest Rate Structure” (Washington).
- Search Google Scholar
- Export Citation
International Monetary Fund, 2016b, “The Acting Chair’s Summing Up Poverty Reduction and Growth Trust—Review of Interest Rate Structure” (Washington).
- Search Google Scholar
- Export Citation
International Monetary Fund, 2016c, “Adequacy of the Global Financial Safety Net—Considerations for Fund Toolkit Reform” (Washington).
- Search Google Scholar
- Export Citation
International Monetary Fund, 2016d, “Financing for Development: Enhancing the Financial Safety Net for Developing Countries: —Further Considerations” (Washington).
- Search Google Scholar
- Export Citation
International Monetary Fund, 2016e, “The Chairman’s Summing Up Financing for Development: Enhancing the Financial Safety Net for Developing Countries—Further Considerations” (Washington).
- Search Google Scholar
- Export Citation
International Monetary Fund, 2016f, “Small States’ Resilience to Natural Disasters and Climate Change—Role for the IMF” (Washington).
- Search Google Scholar
- Export Citation
International Monetary Fund, 2017a, “Large Natural Disasters—Enhancing the Financial Safety Net for Developing Countries” (Washington).
- Search Google Scholar
- Export Citation
International Monetary Fund, 2017b, “Adequacy of the Global Financial Safety Net—Proposal for a New Policy Coordination Instrument” (Washington).
- Search Google Scholar
- Export Citation
International Monetary Fund, 2017c, “The Acting Chair’s Summing Up Adequacy of the Global Financial Safety Net— Proposal for a New Policy Coordination Instrument” (Washington).
- Search Google Scholar
- Export Citation
International Monetary Fund, 2018a, “Guidance Note on the Bank-Fund Debt Sustainability Framework for Low-Income Countries” (Washington).
- Search Google Scholar
- Export Citation
International Monetary Fund, 2018b, “Macroeconomic Developments and Prospects in Low-Income Developing Countries—2018” (Washington).
- Search Google Scholar
- Export Citation
International Monetary Fund, 2018c, “Update on the Financing of the Fund’s Concessional Assistance and Debt Relief to Low-Income Countries” (Washington).
- Search Google Scholar
- Export Citation
International Monetary Fund, 2018d, “2018 Review of Conditionality and Design of Fund-Supported Programs” (forthcoming) Washington.
- Search Google Scholar
- Export Citation
Kuruc, Kevin, 2018, “The IMF and Fragile States: Assessing Macroeconomic Outcomes,” Independent Evaluation Office of the International Monetary Fund, IEO Background Paper No. 18–01/05.
- Search Google Scholar
- Export Citation
Steinwand, Martin and Randall W. Stone, 2008, “The International Monetary Fund: A Review of the Recent Evidence,” Review of International Organizations, Vol. 3 (2), pp. 123–49.
- Search Google Scholar
- Export Citation

PRGT eligibility is assessed every two years. The terms LICs and PRGT-eligible countries are used interchangeably in this paper, although PRGT eligibility is not determined by income levels alone.

Access levels in SDR terms for some better-off LICs (whose quota share declined) later declined somewhat when the 14 th quota review became effective in early 2016 (as access limits in quota terms were then cut by 50 percent).

The PRGT’s self-sustained lending framework operates using an endowment model based on loan and subsidy resources derived from bilateral contributions from members and proceeds from past gold sales. Annual investment income earned from the reserve account balance is expected to be sufficient over time to cover the subsidy needs for annual PRGT lending and projected administrative costs.

Recent developments in LICs are discussed in IMF (2018b). Appendix IV of the paper covers public debt developments in 13 middle income small states that are PRGT-eligible but not classified as Low-Income Developing Countries (LIDCs).

This is a marked turnaround from 2009 when most LICs were on course for manageable debt profiles, helped by the debt relief that close to one-half of the countries received under the Heavily Indebted Poor Countries (HIPC) initiative.

The pattern of Fund engagement tends to differ between PRGT-eligible and other member countries. Non-PRGT-eligible countries make more episodic use of Fund resources and are more prone to capital flow “sudden stops,” generally involving higher and more frontloaded access. While this is less common for LICs, the number of so-called frontier markets is increasing. See Box 2 in IMF (2016d).

Article V, Section 3(a) requires the Fund to establish adequate safeguards for the temporary use of its resources. Conditionality refers to the various policies that require, as a condition for the use of the Fund’s resources, a member to implement measures that enable it to resolve its balance of payments difficulties and repay the Fund.

The PRGT reserve account, which stood at the equivalent of 60 percent of outstanding credit as of end-2017, provides the ultimate backstop for the PRGT in the event of arrears or liquidity strains. The reserve account is also central to the sustainability of the PRGT’s operations and lending capacity, as it will replace the subsidy account once its resources have been exhausted.

This paper covers a number of issues arising in the Fund’s risk management framework including strategic and reputational risks as well as the adequacy of Fund resources.

This section covers the 76 countries that were PRGT-eligible for some period after April 2010; there are now 70 PRGT-eligible countries.

Nearly half of the 49 countries that had at least one UCT-quality program have relied exclusively on the ECF— some 22 countries, with an average 4¼ years of program engagement.

Consistent with the Fund’s policies on military spending, staff has monitored developments in security outlays and worked with authorities to adjust macroeconomic frameworks as needed.

In aggregate, 11 countries have used the SCF or its precursor, the Exogenous Shocks Facility-High Access Component (ESF-HAC), since 2010, including 4 countries which already had an ESF-HAC in place as of January 2010 and a further 7 countries that have requested support under the SCF since January 2010.

Members can request support under a PSI only if they have no need for Fund financing. If, during the program, a short-term financing need arises, they can request financial support under an SCF or RCF. In periods of increased uncertainty or risk, PSI users can request concurrent precautionary support under the SCF. To date, PSI users have solely relied on the SCF, including on a precautionary basis.

Under an SCF arrangement, a member is expected to aim at achieving, maintaining, or restoring a stable and sustainable macroeconomic position consistent with strong and durable poverty reduction and growth; under an ECF arrangement, the expected standard is to make significant progress towards it.

The Fund has typically support countries with protracted BoP problems by disbursing ECF arrangements. These can be augmented in the event of shocks, thus providing an element of insurance alongside regular BoP support.

The RCF requests by Ebola-hit countries arose from technical considerations. All three Ebola-hit countries had disbursing ECF arrangements in place when the epidemic hit; RCF requests were used in two cases to ensure speedy disbursement of funding while providing adequate time to complete ECF program reviews.

The PCI is a non-financial instrument available to all Fund members, see IMF (2017b).

Potential qualification for the PSI is used to identify comparator countries that have comparably strong starting positions for the programs being compared. The main indicators used include: strong growth in real per capita GDP; low underlying inflation; adequate international reserves; external and domestic debt sustainability; and strong institutional capacity. All but 2 of the 11 PSIs approved since 2010 would have met the criteria and 13 comparator ECF/SCF arrangements were selected on this basis.

Reviews under the PSI included on average a slightly higher number of structural benchmarks than in comparator country ECF/SCF reviews, which might influence completion rates.

Fund-supported arrangements cover the ECF, SCF, and PSI, while excluding drawings under the RCF.

See footnote 21.

7 of the 20 PRGT-eligible small states have entered into Fund arrangements, with a further 5 making drawings at least once under the RCF.

This paper uses this term GRA-only to refer to members that are not PRGT-eligible.

By contrast, under the Ghana ECF-supported program approved in April 2015—prior to the increase in access limits—access was capped at 180 percent of quota, approaching the prevailing cumulative limits, and the authorities covered much of their financing needs by raising non-concessional funds on the international bond market.

Access norms are neither a ceiling nor a floor on the level of access provided; they should help to inform the assessment of access levels but should not be misconstrued as access limits or entitlements (see supplement for an elaboration on the role of access norms).

See, for example, Bal Gündüz and Crystallin (2014).

The ongoing 2018 Review of Conditionality and Design of Fund-Supported Programs will analyze the impact of program engagement on macroeconomic performance and social spending (IMF, 2018d).

The blending policy, for example, places tight limits on access to concessional resources for LICs that are relatively well off and deemed to be less in need of these scarce resources than poorer countries. Access norms are a reference point for distribution of access across current users and provide an indicative level of financing when BoP needs are subject to significant uncertainties. The self-sustained lending envelope and the three-pillar strategy aim at preserving resources for future users of concessional resources.

The relationship between the annual and cumulative access limits (1:3) is such that a three-year ECF arrangement at the annual access limit for each year also hits the cumulative access limit for a country with no PRGT credit outstanding.

Defined as the sum of the current account balance without grants, amortization, arrears clearance and gross reserve accumulation.

Access limits and norms are assumed to increase with nominal GDP in LICs, currently 24 percent every three years, thereby keeping access levels over time broadly stable relative to LIC GDP, though access as share of gross financing needs might decline.

Exceptional access is available to countries with: i) an exceptionally large BoP need; ii) a comparatively strong adjustment program and ability to repay the Fund; iii) no sustained past and prospective market access, and iv) income at or below the IDA operational cutoff. The first two criteria correspond to similar criteria for exceptional access under the GRA, while the third and fourth criteria seek to target PRGT resources to the poorest members. To meet the second criterion, a country’s program and ability to repay the Fund are currently expected to be stronger than for a large majority of LICs and is generally considered not met for countries at high risk of debt distress or in debt distress unless expected debt relief or restructuring is projected to reduce the risk rating to moderate or low.

Prior to 2009, exceptional access was provided only in the context of arrears clearance (to Liberia and Zambia).

Meeting the standard of a “comparatively strong” program is more demanding than the prevailing standard that programs are “sufficiently strong;” the exact specification of the requirement would be fleshed out in the successor paper.

This would be in line with the original formulation (IMF, 2009a). Currently, countries with market access are precluded from exceptional access even if per capita income is less than 80 percent of the IDA operational threshold and they are, therefore, not presumed to blend.

Early informal Board meetings are also required for PRGT access requests exceeding 135 percent of quota in any 36-month period.

Such a requirement already exists in the case of requests for exceptional access under the GRA.

See IMF (2016a); the next review of the interest rate mechanism is expected to take place in the second half of 2018.

Most Executive Directors expressed the view that the merits and implications of unifying the interest rate structure for the SCF and ECF should be examined as part of this review of the LIC facilities, see IMF (2016b).

Countries with sufficiently high income per capita—at least twice the IDA operational cutoff (higher thresholds for small countries and microstates)—are candidates for graduation from PRGT eligibility (market access and near-term vulnerability conditions also apply).

Establishing a threshold at 150 percent of the IDA cutoff would affect some 14 countries, predominantly small states. Lowering the threshold to 125 percent would only add one large country.

Use of blending in such cases could, in principle, involve imposing a higher policy standard, as GRA programs seek to resolve the country’s BoP difficulties during the program period, while ECF programs for countries with protracted BoP problems are required to support “significant progress towards a stable and sustainable macroeconomic position.” How important this would be in practice is less clear-cut.

LICs with potential BoP needs can also request a precautionary SBA. A precautionary SBA is more flexible than any of the concessional instruments (see paragraph 46), but carries less attractive financing terms.

When the SCF was created, its use on a precautionary basis was subject to specific access limits in order to discourage tying up PRGT loan resources (on a commitment basis) for contingent support, as opposed to using these funds to meet actual financing needs in the poorest countries. The relatively infrequent use of the SCF to date suggests that these limits may have turned out to be overly restrictive.

The key distinguishing feature between a three-year SCF and a three-year ECF would be the nature of the underlying BoP problem, which under an SCF cannot be protracted and should be short-term and resolved during the arrangement.

In the case of Honduras, a three-year program, supported by a precautionary SCF-SBA arrangement, was designed, given sizable fiscal adjustment needs and a broad structural reform agenda. Given the two-year maximum duration of the SCF, the third year was exclusively covered by the GRA. See IMF (2014).

If this change is adopted, the relationship between access norms for the ECF and SCF would be re-assessed, taking account of the differing underlying needs.

For precautionary SCF arrangements, a sub-ceiling on access applies: at approval, annual access is limited to 56.25 percent of quota, while average annual access (total access divided by the length of the arrangement in years) cannot exceed 37.5 percent of quota.

While loan resources committed under a precautionary arrangement become in principle available again after expiration of the arrangement, the resources cannot always be used again as PRGT loan resources are available only for limited time and need to be renegotiated with lenders, to users, without eroding the integrity of the instrument. The pre-qualification criteria, including a proven track record of policy implementation, would remain a key distinguishing feature of the PSI. In practice, however, the absence of a need for Fund financing might prove to be a more important distinguishing feature for PSI users as the signal from pre-qualification criteria. If that is the case, streamlining the set of non-financial instruments around the PCI, which is available to all members, could be re-considered once there is more experience with the coexistence of the two instruments.

While Executive Directors supported the proposed review-based conditionality for the PCI, they stressed that this would not have implications for Fund financial arrangements, see IMF (2017c).

These percentages refer to natural disasters during 1950–2014.

Vulnerabilities of small developing states to natural disasters are analyzed in detail in IMF (2016f).

If a member has an on track precautionary arrangement and requests a drawing on the basis of a BOP need, it would normally receive funds in a matter of days. By contrast, RCF requests generally take three–six months to disburse (see text chart).

The CCRT has two windows. Under the PCR window assistance is provided following catastrophic disasters and is the focus of this section. The Catastrophe Containment (CC) window addresses needs arising from public health disasters, and is not covered here.

Since the CCR Trust includes Fund resources from the Special Disbursement Account (SDA), any use of these funds needs to comply with the principle of uniformity of treatment of all eligible low-income members applicable to these resources. Under this principle, Fund resources cannot be allocated on a country-specific basis but must be made uniformly available to all low-income members in a similar situation. The principle further requires that the Trust financed by Fund resources has sufficient aggregate financing to support an assessment that it can credibly provide financing beyond already identified potential recipients, see IMF (2015a).

Access to the CCRT is available only to PRGT-eligible countries with annual per capita income below the prevailing IDA operational cut-off (currently US$1,165) or twice the cut-off for small states (defined as having a population below 1.5 million). In contrast to the entry criteria for PRGT-eligibility, there is no additional income cut-off for microstates, defined as having a population below 200,000.

Dominica is a middle-income country and remains PRGT-eligible, as it does not yet meet the criteria for graduation. However, not all PRGT-eligible countries can access the CCRT, which is targeted at the poorest and most vulnerable LICs.

In 2015, Vanuatu suffered cyclone damage estimated at around 60 percent of GDP and requested Fund support under an RCF/RFI blend amounting to 100 percent of quota (50:50 PRGT-GRA mix), equal to the annual access limit under each instrument. As RFI access now counts towards the RCF access limit and with disbursements under the RCF/RFI at 72 percent of current quota, Vanuatu has almost reached the cumulative access limit of RCF credit.

This moral hazard, sometimes labeled “facilities shopping,” refers to repeated use of the RCF in lieu of adopting a conventional multiple-disbursement Fund arrangement involving UCT-quality policies with later disbursements subject to reviews and ex post conditionality. See, for example, IMF (2016d).

As of end-April 2018, Dominica’s cumulative access from the PRGT stood at 76 percent of quota, still well below the cumulative limit of 225 percent of quota.

Taking current Fund credit outstanding as a plausible estimate for the likely stock of debt at the time of a disaster, possible debt relief would at most have amounted to SDR 37 million for Grenada and Dominica.

Based on Nepal’s current Fund credit outstanding, the cost of potential debt relief could be up to SDR 61 million.

Again, drawing current Fund credit outstanding, debt relief in Samoa’s case could be up to SDR 12 million, while Tonga has no credit outstanding.

The CCRT was funded by repurposing the PCDR, surplus funds of the Fund’s own resources in the Multilateral Debt Relief Initiative (MDRI)-I Trust, and surplus funds from 36 bilateral contributors to the MDRI-II Trust, as well as additional bilateral contributions from 5 members.

Many of these issues were discussed in the previous review of LIC facilities. See IMF (2012).

See IMF (2017a) for a discussion of challenges associated with parametric definitions of disaster thresholds. The time needed for different types of disaster assessments is also crucial for ensuring a speedy response.

This approach benefits countries only to the extent that they have debt service falling due to the Fund (as with the CCRT); wider BoP needs would be addressed in the context of a request for an RCF or a Fund arrangement.

For a recent discussion of fragility issues, see IGC (2018).

The IEO, in its recent evaluation of IMF engagement with fragile states, notes that completion rates for programs with fragile states are much lower than for non-fragile states—which is not surprising, given the particular challenges faced by fragile states (see IEO, 2018).

The length of time needed to emerge from fragility and achieve a fully sustainable balance of payments position normally precludes use of the SCF.

The IEO also suggested extending the repayment periods under the concessional facilities: staff sees the current maximum maturity of Fund instruments (10 years in the RCF, ECF, and EFF) as appropriate, and that extending maturities beyond 10 years would fundamentally alter the revolving nature of Fund resources, moving into the terrain the multilateral development banks.

SMPs are normally expected to last for at least 6 months and no more than 18 months, while longer durations are not precluded. In practice, the majority of SMPs since 2010 were for 12 months and the longest for 15 months.

A higher annual limit would expand available financing for LICs with large and urgent BoP needs that are unable to implement a UCT-quality program, while a higher cumulative access limit would facilitate repeated use of this engagement model.

The second condition is needed to ensure that the RCF can be used for at least three years, given the cumulative limit on access to the RCF.

In cases where a country does not have the capacity to implement a three-year UCT-quality program, it can use repeated RCF support as a bridge. A short-term ECF would in these circumstances provide higher levels of Fund financing and a stronger signal to donors, with protections through ex-post conditionality. That said, the presumption would remain that the bulk of external official financing in fragile states should take the form of grants and highly concessional loans, rather than Fund financing.

See for example, IMF (2009a, Executive Summary) “The proposed ECF provides medium-term concessional financing to LICs with protracted balance of payments problems.”

Other Publishers

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/007/2018/039/article-A001-en.xml
The link was not copied. Your current browser may not support copying via this button.

Link copied successfully

2018 Review of Facilities for Low-Income Countries

Author: International Monetary Fund. Strategy, Policy, &, Review Department, and International Monetary Fund. Finance Dept.

Volume/Issue:: Volume 2018: Issue 039

Publisher:: International Monetary Fund

ISBN:: 9781498310871

ISSN:: 2663-3493

Pages:: 71

DOI:: https://doi.org/10.5089/9781498310871.007

Keywords
Introduction
Stocktaking of Experience
Options for Refinement
Issues for Discussion
References

View raw image

Architecture of Lending Facilities for LICs

View raw image

View raw image

Evolution of Risk of Debt Distress

(In percent of 2017 PRGT-eligible LICs with DSAs)

View raw image

PRGT Credit Outstanding by Risk of Debt Distress

(In percent)

View raw image

PRGT Credit Concentration

(In percent of total credit outstanding at year end)

View raw image

Use of Fund Facilities by PRGT-Eligible Countries

Use of LIC Facilities and Staff-Monitored Programs (SMP)

(Number of countries with facility or SMP in place, 2010–17)

View raw image

Share of Precautionary Arrangements

(In percent of total financial arrangements)

View raw image

Reasons for RCF Requests, 2000–2017

View raw image

Timing of Fund Support Following the Ebola Outbreak

(Number of cases)

View raw image

Duration of Program Engagement (ECF, SCF, and PSI)

(2010–2017)

View raw image

Intensity of Use of LIC Facilities by Region, 2010–17

(In percent of months with program in place) ^1/

View raw image

Intensity of Use of LIC Facilities by Country Group, 2010–17

(In percent of months with program in place) ^1/

View raw image

Average Length of Fund Arrangements, 2002–2017

(In years, as per program approved)

View raw image

New Lending Commitments to PRGT-Eligible Countries

(In billion SDR)

View raw image

Counterfactual: Previous PRGT Access Limits and New Arrangements

(In percent of current quota)

View raw image

PRGT Arrangements: Average Annual Access

(ECF and SCF arrangements since 2010, in percent of quota*)

View raw image

RCF Credit Outstanding

(In percent of current quota, as of December 2017)

View raw image

ECF Arrangements: Access Relative to Applicable Norms

(In percent of current quota, 2010–17)

View raw image

View raw image

ODA Flows During Periods of Fund Financing (ECF, SCF, RCF)

(Percent of GDP, median)

View raw image

ODA Flows Surrounding Fund Assistance (ECF, SCF, RCF)

(Average index=100 at first disbursement)

View raw image

PRGT Access Limits

(Relative to key economic indicators, index 2009 = 100)

View raw image

ECF Access Norm

(Relative to key economic indicators, index 2009 = 100)

View raw image

Projected PRGT Credit Outstanding at end-2020 and Income

View raw image

Private Capital Flows to LICs

(In percent of GDP)

View raw image

LICs’ Exposure to International Bonds Markets

View raw image

Disaster Distribution Across Different Country Groups

View raw image

Natural Disasters: Timeliness of RCFs

Topics

Countries

Series

About

Resources

Abstract

Introduction

Stocktaking of Experience

A. Patterns of Use of the Facilities10

B. Access and Demand

Use of Blending

Catalytic Effect of Fund Support

C. Poverty Reduction Strategies

D. Takeaways

Options for Refinement

A. Access Policies

Access Norms and Limits

B. Financing Terms and Blending Policies

Financing Terms

Blending Policies

C. Precautionary and Policy Support

Precautionary Support

Policy Support

D. Support for LICs Vulnerable to Natural Disasters

Raising RCF Access Limits

Widening Access to the CCRT

Short-Term Relief on Debt Service

Ex ante Financial Assistance

E. Supporting LICs in Fragile Situations

F. Links from PRGT Support to Poverty Reduction

Issues for Discussion

Lessons from Experience

Access Policies

Blending Policies and Financing Terms

Adjustments to the Facilities—the SCF

Adjustments to the Facilities—the PSI

Support for Countries hit by Natural Disasters

Support for Countries in Fragile Situations

References

Same Series

Other IMF Content

Other Publishers

Asian Development Bank

Food and Agriculture Organization

Inter-American Development Bank

International Labour Organization

The World Bank

Share Link

Table of Contents

Headings

Figures

Metrics

A. Patterns of Use of the Facilities¹⁰