Proposal To Establish A Resilience and Sustainability Trust

1. Even as the fight against the COVID-19 pandemic and latest geopolitical developments continue to disrupt global economic activity, it is important that the international community does not lose sight of the longer-term structural challenges facing the global economy. The world is experiencing a changing climate, increasing inequality, changing demographics, and a breakneck pace of digitalization, to name just a few. The pandemic has eroded policy space in low-and vulnerable middle-income countries—with many now also facing shocks to energy and food prices—just when they need greater access to affordable, longer-term financing to address these enduring challenges.1

Abstract

1. Even as the fight against the COVID-19 pandemic and latest geopolitical developments continue to disrupt global economic activity, it is important that the international community does not lose sight of the longer-term structural challenges facing the global economy. The world is experiencing a changing climate, increasing inequality, changing demographics, and a breakneck pace of digitalization, to name just a few. The pandemic has eroded policy space in low-and vulnerable middle-income countries—with many now also facing shocks to energy and food prices—just when they need greater access to affordable, longer-term financing to address these enduring challenges.1

Section I. Background

1. Even as the fight against the COVID-19 pandemic and latest geopolitical developments continue to disrupt global economic activity, it is important that the international community does not lose sight of the longer-term structural challenges facing the global economy. The world is experiencing a changing climate, increasing inequality, changing demographics, and a breakneck pace of digitalization, to name just a few. The pandemic has eroded policy space in low-and vulnerable middle-income countries—with many now also facing shocks to energy and food prices—just when they need greater access to affordable, longer-term financing to address these enduring challenges.1

2. The August 2021 SDR allocation of $650 billion has helped address the long-term global need to supplement reserves, while also fostering economic stability. About $275 billion of the allocation went to emerging markets and developing economies (EMDEs), including about $20 billion to low-income countries (LICs). These resources are helping member countries raise reserve buffers and relax financing constraints at a time of high need.

3. To amplify the effect of the SDR allocation, many members with strong external positions have expressed interest in voluntarily channeling part of their SDR allocation to support recoveries in poor and vulnerable middle-income countries (see the G7 and the G20 communiques). On July 23, 2021, the Executive Board discussed three options to channel SDRs: (1) scaling up the Poverty Reduction and Growth Trust (PRGT); (2) establishing a new IMF-administered Resilience and Sustainability Trust (RST); and (3) channeling SDRs to prescribed holders, including the World Bank. Based on initial Board guidance and feedback, staff have focused on the first two options and continue to support interested MDBs in exploring the third.

4. Following a series of discussions on the design of a new trust with IMF members and other stakeholders, there is broad support for establishing a Resilience and Sustainability Trust. In the October 2021 Communique, the IMFC noted “a Resilience and Sustainability Trust (RST) at the IMF, to provide affordable long-term financing to support countries undertaking macro-critical reforms to reduce risks to prospective balance of payment stability, including those related to climate change and pandemics. The RST should preserve the reserve asset characteristics of the SDRs. We call upon the IMF to develop and implement the RST and collaborate closely with the World Bank in this process, and to provide technical support in exploring viable options for channeling SDRs through multilateral development banks.”

5. Against that background, this paper details a proposal for establishing the RST and puts forward the decisions to establish the RST, including the RST instrument. The proposal is in line with previous informal discussions with Executive Directors and other stakeholders including the World Bank.

Section II. The Resilience and Sustainability Trust—Guiding Principles

6. The proposed RST would complement the IMF’s existing lending toolkit by focusing on longer-term structural challenges. The RST’s goal is to enhance economic resilience and sustainability thereby contributing to prospective balance of payments stability. This will be achieved by providing eligible members affordable, longer-maturity financing to (i) support reforms (including by covering BoP costs associated with them) that reduce macro-critical risks associated with select longer-term structural challenges and (ii) augment policy space and financial buffers to mitigate the risks arising from such longer-term structural challenges. This financing would complement traditional IMF support that focuses on resolving more near-term balance of payments difficulties.

A. Legal Basis

7. An overarching role of the Fund is to help members address balance of payments issues through policy support and the provision of financial resources. Under the Articles, there are three types of financial resources the Fund can draw upon to assist members, each of which has its own legal constraints. The first is GRA resources, which constitute the bulk of these resources. The Articles contain specific provisions mandating that adequate safeguards be established for the temporary use of GRA resources (Article V, Section 3), and require that a member must be experiencing an actual balance of payments need at the time it purchases these resources. The second type of financial resources are resources held in the Special Disbursement Account (SDA). These are resources which can be used to provide balance of payments assistance on special terms to developing countries (Article V, Section 12(f)(ii)).2 SDA resources have been used (along with donor contributions) in the PRGT, CCRT and PRG-HIPC Trust, including to address “protracted” balance of payments problems under the PRGT even where the member does not have an actual balance of payments need. Third, Article V, Section 2(b) provides for the Fund’s administration of resources contributed by others, through which financing can also be provided to Fund members, subject to specific conditions.

8. Article V, Section 2(b) authorizes the Fund if requested, to provide financial and technical services to its members. This includes the establishment of administered accounts or trusts.3 Financial services under Article V, Section 2(b) (i) must be consistent with the purposes of the Fund; (ii) must not be “on the account of the Fund”, meaning the GRA cannot bear any risk of loss in respect of the provision of these services; and (iii) cannot impose obligations on members without their consent.4 Beyond these limitations, the Fund has broad flexibility in providing financing under Article V, Section 2(b), including as regards the purposes of the financing (which can be broader than the purposes of GRA financing specified in Article 1(v), for example), and its terms (e.g., flexibility for conditions and maturities that are different than those applicable to GRA financing, which has included grant financing and a 20-year repayment option under the PRG-HIPC trust).5

9. The proposed RST meets the conditions applicable under Article V, Section 2(b):

  • The RST would be consistent with the Fund’s balance of payments financing focus but also the specific purpose under Article l(iii) to promote exchange stability. The latter has informed, in the surveillance context, the concept of “balance of payments stability”, which is defined as a member’s balance of payments position that does not, and is not likely to, give rise to disruptive exchange rate movements.6 In this context, under the Integrated Surveillance Decision, the Fund, in carrying out its surveillance mandate, should focus on those policies of members that can significantly influence present or prospective balance of payments and domestic stability, recognizing that systemic stability (a stable system of exchange rates) is best served by each member promoting its own balance of payments and domestic stability.7 The RST would complement the existing policy advice provided by the Fund under Article IV on members’ policies that influence their balance of payments stability by providing financial assistance to eligible members to help them implement policies designed to strengthen prospective balance of payments stability. Such financial assistance would also assist eligible members in addressing the balance of payments impact of such reforms as they are implemented.

  • The financial risk from operations conducted through the RST—as envisaged in its proposed financial architecture—would be borne entirely by the contributors to the Trust and would not expose the GRA to any risk of loss.8

  • Finally, financial contributions to the RST, and requests by eligible members for financing under the RST, would be completely voluntary. Financing would be provided following such a request and Board approval of the request.

B. Conceptual and Lending Toolkit Considerations

Economic Considerations

10. Certain longer-term structural challenges can cause large, negative macroeconomic and financial outcomes if not addressed. While larger and more affluent economies generally have the tools to address and withstand such challenges, other countries often lack the necessary policy space and may thus suffer severe economic setbacks with long-term implications. The importance of proactively addressing longer-term challenges and of securing ample policy space was acknowledged in the 2021 Comprehensive Surveillance Review and most recently evidenced in the COVID-19 pandemic when advanced economies were able to provide stimulus in multiples of what low and middle-income countries were able to afford. Similarly, on climate change—as noted in the 2021 IMF Strategy to Help Members Address Climate Change Related Policy Challenges and the World Bank’s 2019 Action Plan on Climate Change Adaptation and Resilience—there are significant economic and social costs as the frequency and magnitude of disasters rise, particularly affecting lower-income and smaller countries. Global warming is also driving other trends that are less visible and disruptive in the short term but carry large long-term economic risks.

11. Shortfalls in addressing these challenges can have negative implications for future fiscal, monetary, financial, and external policies and stability. For example, realized and future economic losses from climate change could translate into revenue losses, spending pressures and higher borrowing costs, further constraining fiscal space and exacerbating debt sustainability concerns in vulnerable countries, and imposing fiscal constraints exactly when greater redistribution may be needed. Additional trade, monetary and financial implications may also arise if the long -term shock materializes as businesses are disrupted, losses mount, and pressures rise on balance sheets of households, corporates and financial institutions; the COVID-19 pandemic provides a sobering example of these processes at the global scale. This could result in exchange rate pressures, which, combined with disruptions in trade and supply, could worsen a member’s external accounts and generate a balance of payments problem. Mitigating and reducing these BoP risks requires a consistent and deliberate approach, including setting macro-policy frameworks that can enable an environment where the necessary contingencies, safeguards, and investments can be promoted.

12. Addressing longer-term structural challenges typically requires frontloading measures—many of them costly—even as benefits may materialize only over a longer horizon. This presents lower-income and vulnerable middle-income countries with a difficult challenge: while lack of policy actions would have significant adverse macroeconomic consequences over the longer term, front-loaded action has an immediate fiscal cost and may thus lead to debt buildup and fiscal vulnerabilities over the short to medium term. Access to affordable long-term financing is thus critical to ensure that upfront action on climate change and pandemic preparedness can be taken, which then can involve significant long-term gains.

  • Climate change. Disruptions caused by natural hazards, as well as poor maintenance and mismanagement of infrastructure, cost households and firms at least $390 billion a year in low-and middle-income countries according to a 2019 World Bank study. The 2019 Climate Adaptation Report estimated that targeted US$1.8 trillion climate investment in five areas—early-warning systems, climate-resilient infrastructure, improved dryland agriculture crop production, global mangrove protection, and water security—by 2030 could generate US$7.1 trillion in net benefits globally, representing a return of almost 400 percent. This is because investment in resilience translates into higher growth, including from higher private investment due to the lower cost of capital, and increases the economy’s growth potential (Cantelmo et al., 2019). Higher growth contributes to stronger revenue growth and the enhanced resilience reduces recurrent spending needs (e.g., on reconstruction), thereby improving the fiscal outturn and the debt outlook relative to a business-as-usual scenario.

  • Pandemics. The health, economic and social impacts can be exorbitant, as demonstrated by the COVID-19 pandemic in which costs have recently been estimated to rise to US$13.8 trillion through 2024 (IMF, 2022). Disease outbreaks—that could become pandemics with a lack of preparedness and response—are set to rise further with advances in travel and trade, urbanization, civil unrest and war as well as global warming (International Working Group on Financing Preparedness, 2017). A 2021 report of the G20 High Level Independent Panel estimates that investing in pandemic preparedness and response now will have immense returns by reducing the risk of events whose costs to government budgets alone are 300 times as large as the total additional spending per year.

Balance of Payments Considerations

13. The IMF’s GRA and PRGT lending toolkit is focused on helping members resolve their balance of payments problems. A BoP problem was originally understood to be resolvable through policy measures adopted within three years or less (maximum SBA duration); and the “temporary use” of Fund resources (Article V, Section 3(a)) translated into maturities of up to five years. Subsequently, the EFF, introduced in 1974, recognized that policies to address BoP problems of a structural nature may take longer to implement and may resolve the BoP problem only over an extended period, which lengthened lending maturities up to ten years, and “grace periods” from 3¼ to 4½ years. In 1987 the Enhanced Structural Adjustment Facility (predecessor to the Poverty Reduction and Growth Facility and PRGT) went further in noting that low-income countries may face protracted BoP problems—i.e., financing could be provided when the member does not satisfy the traditional criteria of need applicable to GRA financing (i.e., BoP deficits or low reserves)9—and a single arrangement may not “fully resolve” the member’s BoP problem.

14. The RST adds to the lending toolkit by helping members address risks to prospective BoP stability stemming from select macro-critical longer-term structural challenges. While not necessarily posing imminent BoP problems, longer-term challenges such as climate change make countries more prone to severe BoP problems in the longer run by raising the likelihood and impact of future shocks and undermining growth prospects. Policy inaction—including on account of scarce financing—to address these challenges could increase these risks and jeopardize prospective BoP stability, as defined in H9. Helping member countries to address such risks through policy support and financing is consistent with the Fund’s mandate to support members’ BoP stability.

15. Longer-term structural challenges create a range of possible BoP needs. These needs that could be financed under the RST are typically multidimensional and can materialize over the short-, medium- or longer-term. In the case of climate change, potential sources of such needs— associated with adaptation, transition, and mitigation policies including energy security policies— include, inter alia10:

  • Costs of climate-related public and/or private investments, such as green energy generation, coastal protection infrastructure, energy-efficient retrofitting of existing building;

  • Costs associated with climate-focused reforms, such as transitioning to green technologies;

  • Offsetting the costs of policies typically required to enable a just transition, such as augmenting targeted social assistance in tandem with the unwinding of carbon subsidies;11 and,

  • Building up policy space and buffers necessary to mitigate risks to longer-term BoP stability, such as establishing and augmenting disaster funds, establishing and financing a multi-layered financial framework for disaster resilience, and augmenting international reserves to face financial stability implications of climate change.

Similarly, financing needs for pandemic preparedness can entail:

  • Cost of scaling-up basic health care infrastructure and human capital so that these can be readily deployed during pandemics;

  • Investing in early warning indicators and data collection and dissemination to inform pandemic policy response; and,

  • Creating policy buffers and financing contingency frameworks (akin to climate change) to provide a stronger and timely response during a pandemic.

16. RST financing is intended to enhance the overall longer-term policy space of member countries. As with all Fund lending, RST loans are provided as liquid and fungible BoP support and are not earmarked for specific projects. RST loans can be used to (i) cover shorter-term BoP needs stemming from the implementation of policy measures supported under the RST, (ii) increase medium-term policy space to continue implementing priority spending associated with addressing longer-term structural challenges, and (iii) augment buffers against risks associated with prospective BoP stability stemming from these challenges.

Lending Toolkit Considerations

17. Consistent with its purposes, RST financing will have significantly longer maturities than those under existing GRA and PRGT instruments. Reforms supported by the RST are expected to take much longer to yield economic benefits in terms of reducing risks to longer-term prospective BoP stability, relative to the policies and reforms supported under GRA and PRGT arrangements, which are expected to have a significant payoff before repayments start three to five years after disbursement. Moreover, there is typically considerable uncertainty about the period over which costs associated with RST-supported reform measures may materialize, bolstering the case for providing additional financial room for navigating the risks. The longer maturities, combined with envisaged modest interest rates, imply a high degree of concessionality of RST loans compared to the GRA, which is appropriate in light of the global public goods nature of RST support and the intended borrower group of LICs, more vulnerable MICs, and small states.12

18. Thus, the RST will be a third component of the IMF lending toolkit that complements existing facilities, albeit with a different focus than the GRA and PRGT. The RST is not intended to play the “lender of last resort” function—i.e., addressing short- to medium-term financing problems—associated with lending from the GRA and PRGT. Nor should it be viewed as part of the Global Financial Safety Net—where the Fund plays a central role as a financial backstop—since the RST does not aim to help members mitigate the risks of financial shocks. Instead, RST financing would help improve economic resilience and sustainability by addressing longer-term challenges that risk undermining prospective BoP stability. This objective underpins the distinct reform policies supported under the RST, as well the longer-term and more affordable terms of RST financing.

19. The RST would build on Fund support under an existing upper credit tranche (UCT)-quality instrument. Financing under the RST would require a concurrent Fund arrangement or instrument supporting a program with UCT-quality conditionality (henceforth called “UCT program”’13) to (i) provide adequate policy safeguards that contribute to mitigating the credit risk associated with RST financing and maintaining the reserve asset status of RST contributions, a key requirement for central banks of many potential contributors, (ii) mitigate the risk of facility shopping, whereby the RST’s longer and cheaper financing could be (mis)used to finance BoP problems normally addressed under GRA/PRGT programs, and (iii) support a stable macroeconomic environment in the borrowing country which is a necessary condition to pursue long-term reforms and support a catalytic role for the RST. From a BoP standpoint, the distinction between the RST financing and the UCT program is as follows:

  • The RST would focus on downside scenarios associated with select longer-term challenges. It would aim to lower the probability of such scenarios and/or reduce the severity of the BoP problems that would materialize should such a scenario come to pass. The recognition of these risks does not mean that the UCT program becomes inadequate in achieving its short- to medium-term goals.

  • A financing UCT program would focus on resolving an actual/protracted BoP problem under the baseline (for a disbursing arrangement), or insuring against a potential BoP problem under a likely downside scenario (for an arrangement treated as precautionary) over the medium-term. The UCT program is not designed to address all downside scenarios and potential shocks, as well as BoP needs over the longer-term.

  • Where a member has no need for IMF BoP financing, a signaling instrument like the PCI (accessible to all members) or PSI (accessible to qualifying PRGT-eligible members) would provide a suitable medium-term policy framework for the RST. The PCI/PSI would support the underlying macroeconomic framework, with RST financing helping address qualifying longer-term challenges.

C. Other Considerations

20. Multilateral context: Creating the RST requires broad participation of creditors and borrowers. A supportive multilateral context, a strong lender base, a diversified borrower base, the expected de facto preferred creditor status (PCS), and adequate credit risk mitigation, including important policy safeguards and financial buffers, are essential underpinnings to successfully operationalizing the RST.

21. Add it tonality: The design of the RST prevents facility shopping. The RST’s purpose—to improve prospective BoP stability—is distinct from the focus of GRA and PRGT financing. Thus, by construction, the RST provides additional financing and does not serve as a substitute to the GRA and PRGT. The RST’s proposed safeguards further mitigate the risk of substitution, including through access caps and concurrence of RST financing with an on-track UCT program.

22. Governance and financial architecture: As with other trusts established by the Fund under Article V, Section 2(b), the proposed decisions and the instrument establishing the RST will set out the purposes of the RST, the modalities of lending, the financial structure and clear rules for the Fund’s governance and decision-making in the administration of the RST, including regarding requests for financing from the RST, as further elaborated in Section IV.A The proposed RST’s financial structure is similar to that of the PRGT, with some necessary modifications to build up a large new trust from the ground up with appropriate safeguards (Section V).

23. Coordination with multilateral agencies: Given the massive scale of longer-term structural challenges, RST support should be seen as part of a broader concerted effort by the international community. It would complement policy and financing support from other international financial institutions and development banks to address these challenges. The financing needs of members to address these challenges are so large that RST financing is bound to be modest relative to the needs. Close coordination with the World Bank (as explained in Section IV.B) and other IFIs would be essential to provide coherent policy advice and to catalyze additional official and private financing where feasible. For instance, RST support for the development and implementation of overarching policy frameworks such as green public financial management would improve the integration of climate in policy formulation and enhance governance, thereby giving more comfort to other public and private lenders and donors to provide project financing and technical assistance. Moreover, potential lenders, including regional development banks and bilateral partners would benefit from the RST’s fiscal and debt sustainability framework that incorporates economic trade-offs. The underlying diagnostics—which would include those developed by the member country and/or by other international financial institutions, such as the World Bank—would be key inputs to help identify the main policy priorities that could be supported by different stakeholders, including the private sector.

24. Integration with Fund Capacity Development: Leveraging the synergies between the Fund’s surveillance, lending, and capacity development (CD) will help in the design and implementation of the RST reform measures. Fund CD with a medium-term programmatic approach can play an important role in supporting RST reforms. For example, diagnostic tools such as the Climate Public Investment Management Assessment (C-PIMA) and the Climate Macroeconomic Assessment Program (CMAP) help identify and assess capacity and policy gaps in areas that are likely to be prominent in RST operations.

Section III. Resilience and Sustainability Facility—Policy and Design Modalities

25. The proposed RST will provide loans to an eligible member through a financing instrument called the Resilience and Sustainability Facility (RSF), under an arrangement (“RSF arrangement”), with design features as elaborated below. Consistent with the considerations discussed in Section II.B:

The RST provides loans through the RSF with the goal of enhancing economic resilience and sustainability, thereby contributing to the member’s prospective BoP stability, by (i) supporting policy reforms that aim to reduce risks associated with qualifying longer-term structural challenges facing the member, and (ii) augmenting policy space and financial buffers to mitigate the risks arising from these longer-term structural challenges.

26. The RSF will be part of a policy and financing framework for structural transformation that is consistent with macroeconomic stability. Macroeconomic stability is a necessary condition for successful, longer-term structural reforms; without a solid macroeconomic foundation, reforms and investments aimed at prospective BoP stability would run the risk of stop-and-go cycles (as intermittent BoP problems take hold). Support under a qualifying UCT program (Section III.B) will help restore or preserve medium-term external balance and provide the macroeconomic foundation for an RSF arrangement to tackle longer-term challenges.

A. Uses of Resilience and Sustainability Facility Support

Qualifying Longer-Term Structural Challenges

27. The RSF helps address select qualifying longer-term structural challenges that create macro-critical risks to prospective BoP stability. The set of specific qualifying challenges calls for balancing several considerations: (i) macroeconomic criticality and consistency with the Fund’s mandate, (ii) demand from the membership, (iii) global public good nature of the challenges addressed by the RST; and (iv) operational capability given the Fund’s internal capacity and limits to collaboration with other agencies.

28. Coordination with other IFIs, particularly the World Bank, would be integral to the RST. The RSF’s ability to tackle qualifying challenges depends on strong diagnostics, ability to formulate policy priorities, and develop and monitor the appropriate conditionality to enhance the prospective BoP stability. Addressing these challenges requires close coordination with relevant multilateral and regional development banks, to best leverage each institutions’ expertise and experience in helping to identify reforms. A staff-level framework for RSF financing has been developed in coordination with the World Bank for climate change (see Annex III). An equivalent framework is being discussed with the World Bank and other stakeholders, including the WHO, for pandemic preparedness. In all such frameworks, the centrality of the Executive Board’s judgment as to whether relevant reform measures are appropriate and have been implemented would be ensured.

29. Based on feedback received in consultations with the Executive Directors, staff proposes that the RST supports reforms that address challenges associated with climate change and future pandemics at the outset.

  • Climate change could result in external vulnerabilities as climate-vulnerable countries are more at risk to the destruction of productive and export capacity, which create or exacerbate balance of payments problems. Countries’ reliance on carbon-intensive exports are at risk both from regulatory changes (such as the adoption of border carbon adjustment) as well as a decline in exports from the transition towards net zero. Hence, policies related to adaptation, transition, and mitigation – including policies that support energy security – are all important to enhance countries’ prospective BoP stability.

  • As the COVID-19 pandemic has demonstrated, failure to invest in better health policy and infrastructure to prevent/mitigate the risks from pandemics can create macroeconomic risks for countries and spillovers for the rest of the world. Measures and investments in pandemic preparedness can contribute to safeguard economic growth, reduce fiscal risks, mitigate losses to businesses and households, and protect balance sheets of different economic sectors, all of which have knock-on effects on external balance stability. The scope and timing of the Fund’s engagement on pandemic preparedness will be firmed up based on ongoing discussions with the World Bank and the WHO.

It is proposed that the Executive Board has authority to identify longer-term structural challenges qualifying to be financed using Trust resources beyond those identified at the establishment of the RST, subject to consent of contributors representing 70 percent of total commitments under borrowing agreements (see H73) and depending on operational considerations (see H75).

Use of Financing

30. RST loans would provide general BoP support to help meet financing needs associated with addressing qualifying longer-term structural challenges. RST financing would not be earmarked for specific projects but could be used to address a range of BoP needs, including: (i) covering any shorter-term BoP/fiscal needs directly associated with implementation of RST-supported reforms (1131); (ii) increasing policy space for fiscal spending14 and reforms associated with qualifying longer-term structural challenges, and (iii) augmenting longer-term buffers to strengthen the member’s ability to face shocks linked with qualifying structural challenges. RSF arrangements are expected to be drawing (i.e., not used on a precautionary basis); available financing is expected to be drawn promptly after the Board approves an RSF disbursement.15

31. In some instances, RSF-supported reforms may have short- to medium-term balance of payments implications, which would affect the interaction with the UCT program. In particular, RSF-supported reforms could:

  • Raise actual or prospective BoP needs. For example, expanding the social safety net to support reform of fuel subsidies or transitioning to green technology could increase BoP needs. Likewise, green public financial management (PFM) reforms may trigger higher imports for green investments. Accelerated decarbonization efforts—even when executed with carbon taxes—can also gradually increase BoP needs as private sector investment in green technologies surges.

  • Lower actual or potential BoP needs. For example, fuel subsidy reforms and carbon taxes may serve a dual purpose of improving prospective BoP stability and helping achieve the fiscal/BoP objectives of the accompanying UCT program. The resulting decarbonization may also lower the country’s sensitivity to energy price volatility over time, thus attenuating an important risk factor faced in many Fund-supported programs. Part of savings from such measures could be applied toward augmenting accompanying social assistance programs to cushion the impact of decarbonization on the poor and vulnerable.

32. While an RSF arrangement would accompany a UCT program, the latter must stand on its own. It means that the reforms and any financing under the UCT program must be assessed as sufficient to address the financing need and to ensure medium-term external viability under the baseline scenario, under the GRA, or make “significant progress”, under the Extended Credit Facility (ECF) of the PRGT. In Board documents, financing committed under an RSF arrangement would be recorded as separate from gap financing under the GRA and/or PRGT.

B. Eligibility and Qualification Eligibility

33. Scarce RST resources should be targeted at DCs, vulnerable MICs, and small states. As discussed, these members face higher risks to prospective BoP stability from longer-term structural challenges, and lack reliable access to affordable longer-term financing more suited to tackle them.

34. Staff propose to base initial RST eligibility on per capita income and population thresholds. Having analyzed a range of alternative methodologies for defining RST eligibility, staff consider that a simple income-based cutoff—with higher thresholds for small states—strikes an appropriate balance between targeting scarce resources, meeting countries’ needs, and ensuring predictability and evenhanded treatment. Using a combination of per capita income and population is an established and transparent method that also is used for determining PRGT eligibility. Specifically, it is proposed that for the initial list of eligible members, a member is assessed as RST-eligible if, based on qualifying data available as of October 2021:16

  • a. its per capita gross national income17 (GNI) in 2020 (or 2019, if 2020 data is not available) does not exceed ten times the 2021 International Development Association (IDA) operational cutoff ($1,205) or

  • b. it has a population below 1.5 million as of 2020, as reported by the World Bank, and its per capita GNI in 2020 (or 2019, if 2020 data is not available) does not exceed twenty-five times the 2021 IDA operational cutoff.

The per capita income cutoff for small states is set higher than for other countries given that most small islands with relatively high incomes face significant longer-term risks related to climate change but still lack steady access to affordable long -term market financing. Moreover, in order to not exclude low-income countries, in the few instances where the relevant GNI or population data are not available to assess the above criteria but the countries are PRGT-eligible, it is proposed that the countries be added to the RST eligibility list at the time of the establishment of the RST.18

35. Based on the proposed criteria, 143 countries will be RST-eligible at the time the RST is established (Annex I). Countries that do not plan to request RST financing (e.g., potential RST contributors) will have the option to opt-out of the RST eligibility list at any time by notifying the Fund in writing.

36. RST eligibility will be reviewed regularly. For the purpose of future eligibility reviews, a member will be added to the RST eligibility list if, based on the latest available qualifying data19, its annual per capita GNI is (a) less than ten times the IDA operational cutoff; or (b) less than twenty-five times the IDA operational cutoff if it has a population below 1.5 million, as reported by the World Bank. The first eligibility review will take place as part of the first review of the Trust. Subsequent eligibility reviews would be synchronized and combined with PRGT eligibility reviews, which are undertaken on a biannual schedule.20 Notwithstanding the regular review cycle, decisions on entry onto the RST eligibility list could also be adopted in the interim period between eligibility reviews. It would be expected that entry decisions would be adopted on a stand -alone basis where a member that is not on the list meets the entry criteria in the period between reviews, and where a delay until the next scheduled review of the RST eligibility list would disadvantage the member (in particular, if the member wishes to request support under the RSF). Specifically, future entrants to the PRGT would typically be added to the RST eligibility list at the same time, subject to an ad -hoc Executive Board decision. Re-entry on the list in the interim period between reviews would also be possible for any country that had previously opted out or been removed in the context of sanctions imposed for overdue obligations to the Fund (assuming the graduation criteria discussed below were not met).

37. Graduation rules will feature a buffer period. It is proposed that a country would graduate from the RST eligibility list if all of the following conditions are met: (a) its GNI has been above the applicable eligibility threshold (specified above) for at least the last five years (for which qualifying data are available), (b) its GNI has not been on a declining trend in the same period (comparing the first and the last relevant annual data) and (c) based on the latest qualifying data, its GNI is at least 10 percent above the applicable eligibility threshold (i.e., 27.5 times the IDA operational cutoff for small states and 11 times IDA operational cutoff for other countries).

38. It is proposed that changes in the RST eligibility list should not interfere with an existing RSF arrangement or ongoing discussions on new requests for an RSF arrangement. A member that has an existing RSF arrangement in place at the time of the Board decision approving its graduation from the RST would be permitted to continue its RSF arrangement, including any extensions or augmentations that may be possible thereunder. The graduation decision in respect of such a member would specify that its removal from RST eligibility would become effective only upon termination, expiry or cancellation of its ongoing RSF arrangement. Further, it is proposed that if a member is no longer RST eligible under a decision adopted by the Executive Board, there would be a transitional period of 5 months before such change would become effective. During the transitional period, a new RSF arrangement based on past or ongoing discussions with the member may be approved by the Executive Board. Members who become RST eligible would be able to access RST financing with immediate effect.

Qualification

39. To qualify for an RSF arrangement, eligible members would need to have: (i) a package of high-quality reform measures addressing qualifying longer-term structural challenges; (ii) a concurrent UCT program with at least 18 months remaining prior to its expiry; and (iii) sustainable debt and adequate capacity to repay. Specifically:

  • I. Relevant policy and reform package: Financing under an RSF arrangement would be based on a set of policies and reforms that are expected to help the member make significant progress toward strengthening the member’s prospective BoP stability by reducing macro-critical risks related to qualifying longer-term structural challenges. A guidance note would elaborate on how to design and assess the strength of the package of structural reforms, which would typically be developed by the Fund in consultation with other relevant institutions, including the World Bank.

  • II. Qualifying UCT program: An RSF arrangement would only be approved by the Executive Board when:

    • (a) The member is concurrently requesting approval of a qualifying IMF instrument supporting a program with UCT-quality conditionality or the completion of the review under such instrument, which, for the purpose of RSF qualification, would include SBA, EFF, PLL, FCL, SCF, ECF, PCI or PSI. The high ex-ante qualification bar for FCL arrangements is equivalent to the UCT-quality standard and is thus sufficient to provide the required policy safeguard (consistent with the FCL design/decision). SMPs—even those of UCT quality—would be excluded, as they are not Fund-supported (i.e., they are approved by Fund Management and not the Executive Board) and do not require the same level of financing assurances and safeguards. Emergency financing would not qualify as the RFI and RCF do not support a UCT-quality program and their outright purchase/disbursement nature is not compatible with RSF’s disbursement being tied to completion of reviews (see below). SLL arrangements would not qualify given their short duration (next bullet).

    • (b) There are at least 18 months remaining in the accompanying UCT program at the time of approval of the RSF arrangement. This would allow for adequate time to implement a set of high-quality structural reforms, while ensuring strong policy safeguards. As a transitional arrangement, during the first 6 months from the date on which the RST becomes operational later this year, ongoing and on-track UCT programs with as little as 12 months remaining would equally qualify.

  • III. Debt sustainability and capacity to repay: Like with all Fund lending, approval of an RSF arrangement and completion of reviews will require the members’ debt to be assessed as sustainable in the medium-term under the applicable debt sustainability framework (LIC DSF or MAC SRDSF). As with all Fund financing, requests for RSF arrangements and reviews would also require an assessment of the member’s capacity to repay the Fund.

C. Lending Terms

40. RSF lending terms are proposed to combine low interest rates with significantly longer maturities than in the GRA and the PRGT. RSF loans are meant to provide longer-term financing on terms that are considerably more affordable than available market financing. Under the terms proposed below, the cost of RST loans would be cheaper than GRA credit, and, at current interest rates, also more concessional than GRA-PRGT blend and regular PRGT terms (see Table 2 in Annex II). The relatively favorable lending terms are consistent with (i) the RST’s purpose to support reforms that tackle long-term structural issues that impact prospective BoP stability, (ii) reforms having a public goods nature with potential positive externalities, and (iii) the intended borrower group of LICs, vulnerable MICs, and small states.

41. Consistent with the RST’s purpose to contribute to longer-term prospective BoP stability, RSF loans are proposed to have a maximum maturity of 20 years. Given that RST resources are solely contributor resources, administered under Article V, Section 2(b), the requirements of temporary use of GRA resources under the Articles do not ap ply, and the proposed loan maturity is not unprecedented for Fund financing under Article V, Section 2(b), both as discussed above. It is proposed that principal repayments be made in 20 equal semi-annual instalments that begin 10!/2 years after the RSF disbursement. Together with the interest rate structure, the proposed maturity ensures that RST concessionality for the poorest and most vulnerable countries remains close to that in the PRGT (see below).

42. An RSF borrower would pay an interest rate with a modest margin over the 3-month SDR interest rate (SDRi). The margin above the RST’s funding cost (i.e., the SDRi paid on RST borrowing) contributes to the build-up of RST reserves and, together with the service charge on RSF disbursements for some country groups, covers the RST’s administrative costs, thus helping ensure the financial soundness and viability of the RST (Section V). RSF borrowing would not be subject to surcharges, and outstanding RST credit would not count towards the level-based threshold for GRA surcharges. A commitment fee is not currently contemplated.

43. Staff proposes a tiered interest rate structure that differentiates financing terms across groups of countries. The proposed margin structure has been calibrated to support adequate accumulation of the RST’s Reserve Account (Section V) while covering operational costs. The proposed margins would provide more favorable terms to lower-income countries and most small states (see Section VI and Annex V), with an expectation that future reviews would consider an increased degree of cross-subsidization of the poorest if the SDRi rises, as expected, to higher levels (see below). Three country groups would broadly mirror the differentiated financing terms across the GRA and the PRGT (See Annex V Table 2 for a list of countries under each group). The grouping of countries would be reviewed at each RST eligibility review; once allocated, a member would remain in that group until the next eligibility review. Irrespective of changes in a member’s blending status, income, or country grouping following an eligibility review during the lifetime of an RSF arrangement, the member would continue servicing credit outstanding arising from disbursements under that arrangement at the interest rate that corresponds to the group it belonged to when the arrangement was approved.

  • Group A countries are all PRGT-eligible countries that are not presumed blenders.21 22 This group currently comprises 51 countries and members in this group are proposed to pay a margin of 55 basis points above the SDRi and would be exempt from any service charges on RSF loan disbursements.

  • Group B countries are all presumed blenders under the Fund’s framework for blended access to GRA and PRGT resources and all non-PRGT eligible small states with per capita GNI below ten times the IDA operational income cutoff. This group is currently composed of 27 countries and is proposed to pay a margin of 75 basis points above SDRi and would be subject to an upfront one-time service charge of 25 basis points on each RSF loan disbursement.

  • Group C countries are all other RST-eligible countries. This group is currently composed of 65 countries and would pay a margin of 95 basis points above SDRi (just below the current GRA margin for the basic rate of charge) and would be subject to an upfront one-time service charge of 50 basis points for each RSF disbursement.

44. To protect the lowest income borrowers from rising market rates, an interest rate cap for Group A countries could be considered at future reviews (see below). Such an interest rate cap would imply that the margin paid by the poorest countries (Group A) would decline as interest rates rise above a certain level, and eventually would become negative once the SDRi rises above the specified level. The foregone interest income implied by such a cap would be financed through increased cross-subsidization of interest rates between borrowers, with the use of resources in the RST’s Reserve Account only possible if consistent with reserve-adequacy including reserve build-up (see Section V). A cap of 2¼ percent, including the margin, as assumed by staff for modeling purposes in this paper, would ensure that, together with the 10½-year grace period and 20-year maturity, the concessionality of PRGT and RSF loans would be at roughly the same level when SDRi rises above 2 percent (a grant element of around 30 percent at a 5-percent discount rate, see Annex II, Table 2). In the initial proposal for the level of margins, no interest rate cross-subsidization is envisaged at the start of the RST.

45. The margin structure and service charges will be subject to reviews, with adjustments as the Executive Board considers appropriate. The funding costs for RST loans is a given as the interest rates on these loans up to the SDRi will be hard-wired in the RST instrument and reflected in the borrowing agreements. Given the uncertainty about future demand, world interest rates, and RST finances, staff proposes that the Executive Board retain flexibility in setting the margin structure for RST loans, that balances the need to ensure the financial sustainability of the Trust and the need to protect the poorest members. At each interest rate review, the Board would consider a range of relevant factors to maintain or adjust the margins (as well as service charge structure), including the financial position of the RST (e.g., the projected pace of reserve buildup, demand across country groups, and the appropriateness of financing terms for borrowers). The review would also decide on any adjustments to the specification of country groupings.23 An interest rate review would take place during each regular RST policy review, and earlier as needed. Staff, for example, would expect that it would be triggered if the average SDRi were to rise above 1!/2 percent in any 12 -month period and if financial market indicators (i.e., SDRi projections) signal that this interest rate is not expected to decline below 1.5 percent within the coming quarters, which would give the Board the opportunity to discuss possible interest rate protection for the poorest borrowers through a reduced margin or an interest rate cap. An interest rate review could also be triggered by other unexpected events that have a bearing on the financial sustainability of the RST.

D. Access

46. The overall cumulative access cap for eligible members under the RSF is proposed to be set at the lower of 150 percent of quota or SDR 1 billion. Access considerations reconcile demand estimates (Section VI) with the anticipated size of the Trust and are an important tool to manage scarce Trust resources. Access caps may need to be revisited should there be excess or shortfall in contributions relative to existing fundraising targets (Section V). The RSF would have no annual access limits to facilitate flexibility in the timing of RSF disbursement but would limit the size of disbursements (see U58).

47. Access under RSF arrangements would be guided by norms and other criteria. The starting point of access determination would be an access norm of 75 percent of quota. The proposed use of norms reflects resource constraints associated with the limited size of the Trust and the difficulty of precisely assessing the BoP implications (both medium-term costs and long-term benefits) of RSF-supported reforms. However, the norm is neither a ceiling nor a floor on the level of access. It should help inform the assessment of appropriate access levels but should not be misconstrued as an access limit or entitlement. Deviations from the access norm may be considered depending on: (i) any direct short- to medium-term BoP needs associated with the implementation of corresponding RSF-supported reforms, (ii) strength and ambition of the package of RST-supported reforms; and (iii) capacity to repay the Fund, taking into account the member’s debt sustainability, debt carrying capacity, and composition of debt (in particular the prevalence of de facto senior obligations, especially in cases of elevated debt vulnerabilities).24

48. Policy Safeguards for High Combined Credit Exposure from the PRGT and GRA (the “PS-HCC Policy”), introduced in 2020, would not apply to financing under RSF arrangements. Features distinguishing the RST from other Fund financing—e.g., focus on prospective stability (as opposed to filling residual BoP gaps), longer maturities—underpin the rationale for treating RSF access separately from that extended under the GRA and the PRGT. The PS-HCC Policy, which applies the policy safeguards for HCC exposure under the GRA and the PRGT, would not be amended to include RST financing (which is for longer-term needs), or to take into account credit outstanding under an RSF arrangement in assessing access under PS-HCC. Likewise, the GRA and PRGT Exceptional Access Frameworks apply to respective GRA and PRGT credit and would not apply to RST financing. This approach also mitigates the risk that countries might seek to substitute GRA access with RSF access in light of the more favorable financing terms. Instead, the RSF design incorporates several alternative safeguards, which are more consistent with the terms of lending of the new facility:

  • Built-in safeguards: Policy safeguards for high access under the PS-HCC are designed to provide enhanced safeguards to GRA and PRGT resources, respectively. Under the RSF, a number of enhanced safeguards are built in directly through (i) debt sustainability requirements for RST qualification (1139) (ii) modest access caps (1146), (iii) access determination rules that explicitly account for debt vulnerabilities (1147), and (iv) enhanced analysis of debt structure, debt sustainability, and of capacity to repay (Section III.F).

  • Modest debt service schedule: The RSF’s much longer grace and repayment periods—resulting in modest effects on annual gross financing needs (up to 15 percent of quota per year)— positively distinguishes the impact of RST financing on the country’s debt risk profile from that of GRA/PGRT financing.

E. Lending Modalities

49. Request and Approval. An RSF arrangement requested by an RST-eligible member may only be approved by the Executive Board if that member either has an “on-track” UCT program with at least 18 months remaining until its expiration (12 months for RSF arrangements approved within the first six months of the RST’s operationalization), or is receiving approval for such a program concurrently with the consideration of the request for an RSF arrangement (see Section III. B on qualification). Approval of an RSF arrangement would normally take place either at the time of approval of the request for the UCT program or at a review under the UCT program. The authorities would need to make a request for an RSF arrangement in writing (normally as part of the LOI for the UCT program documentation). The Memorandum of Economic and Financial Policies (MEFP) — normally a single document covering both the UCT Program and the RSF arrangement—would then describe broad policy intentions in RST-relevant areas and specific structural reforms (“reform measures”, see 1151) that would be supported under the RSF arrangement.

50. Duration. The duration of the RSF arrangement would generally be expected to coincide with the duration of a new UCT program (when the two are requested together) or the remaining duration of an existing UCT program (when the RSF request occurs at a review of the UCT program). The minimum duration would be 18 months (12 months for RSF arrangements approved within the first six months of the RST’s operationalization), to allow adequate time for implementation of RSF reforms measures, including any necessary technical assistance (see ll(b) under 1139). The RSF arrangement would expire when all amounts available under the arrangement have been disbursed. Given the important safeguards provided by the accompanying UCT program, the RSF arrangement would also automatically terminate upon the termination, cancellation or expiry of the accompanying UCT program. The RSF arrangement could be extended at the time of a request of an extension of the qualifying UCT program if additional time is required to complete Board-approved reform measures (see discussion of conditionality below) or if new reform measures are identified for inclusion.

Conditionality

51. RSF conditionality shall take the form of Board reviews assessing reform measures. Similar to structural conditionality under UCT programs, a reform measure would be a single policy action or a set of very closely related actions constituting a single reform. Reform measures would be designed with the aim of being objectively monitorable. Reform measures will also need to be clearly linked to addressing qualifying longer-term structural challenges, and make a meaningful contribution toward strengthening the member’s prospective BoP stability (Box 1 provides illustrative examples for the case of climate change). Coordination with other IFIs/MDBs in the design of reform measures will be critical to complement their lending and to help catalyze additional financing, including from private sources.

52. RSF conditionality will follow the key principles underlying the Guidelines of Conditionality (GoC), with certain caveats. The following principles25 will apply to RSF conditionality: national ownership of reforms, tailoring to member’s circumstances, parsimony in the application of conditions, clarity in the specification of conditions, effective coordination with other multilateral institutions and the principle guiding the concept of a review by the Executive Board. There would be no cross-conditionality; as with all Fund lending, the views of other institutions, particularly in non-core areas for the Fund, would be an important input (for the RST, particularly in assessing completion of the measures). However, conditionality applicable to RST disbursements will consist of reviews assessing implementation of reform measures and none of the other types of conditionality provided under the GoC for lending in UCT programs—quantitative performance criteria and indicative targets, prior actions and structural benchmarks—will be used in RSF arrangements. Waivers of applicability and non-observance are not envisaged (see 1158 on completion of reviews). Further, the focus on helping the member resolve its BoP problems (GoC 116(a)) will not apply to RSF conditionality; it is instead focused on improving prospective BoP stability.

53. RSF conditionality will be typically separate from the conditionality of the concurrent UCT program, though select reform measures may be included in both. In keeping with the distinct but complementary purposes of the RST and its longer-term focus, reform measures must be compatible with the resolution (or prevention) of the BoP problem under the UCT program, but they are not required to directly contribute to it. For that reason, reform measures will be specified separately from the conditionality under the concurrent UCT program. This also means that reviews under a UCT program would not be impacted by delays in completion of reform measures (however, delays in completing reviews under the UCT program would delay RSF disbursements; see 1158). However, in those cases where the implementation of the reform measure is critical to the success of the concurrent UCT program (so-called “dual-purpose reforms”), the measure would also be included as part of the UCT program conditionality (typically as structural benchmarks; sometimes as prior actions), i.e., these conditions will be presented as conditionality for both the UCT program and the RSF arrangement. These dual-purpose measures would be assessed as per the policies applicable to conditionality for the purpose of completing reviews under the UCT program.

54. Reform measures would need to be implemented within the duration of the RSF arrangement, but with greater flexibility in their timing. To access RSF disbursements, the reform measure would need to be implemented over the duration of the concurrent UCT program. However, reform measure target dates would be indicative to take on board the difficulty to gauge the exact time needed for completion of reforms in relatively new policy areas. Flexibility in implementation timing also reflects that the reform measures will typically have no direct bearing on the resolution of the BoP problem addressed under the UCT program. However, if a reform measure is also a structural benchmark or prior action under the UCT program then it will also be monitored and assessed under the relevant framework applicable to the UCT program.

55. The macroeconomic framework and quantitative targets under the concurrent UCT program will include projections for RSF financing. RSF financing increases the overall financing envelope over the period of the UCT program. Based on the diagnosis and intended uses of RST financing, the macroeconomic framework under the UCT program will reflect how this financing is to be allocated between expanding the fiscal envelope, reserve accumulation and any other uses. Performance criteria and indicative targets under the UCT program will reflect this financing under the baseline, but also include adjusters linked to the expected timing of RSF disbursements (analogous to other exogenous sources of financing/revenue). Should the RSF financing not materialize (because the reform measure is not implemented or delayed), these adjusters fully automatically offset its impact on the medium-term macroeconomic framework. Since benefits of RSF-supported reforms are expected to materialize in the longer term, no other changes to the medium-term macroeconomic framework in response to their non-implementation would be generally expected, although this would need to be assessed on a case-by-case basis.

Reform Measures—Initial and Illustrative Considerations

The RSF would represent the first instance where IMF conditionality maybe exclusively focused on achieving longer-term objectives (as opposed to resolving short- to medium-term BoP problems). Staff intends to undertake a more detailed analysis and planning exercise—leveraging expertise inside and outside the institution—between now and the operationalization of the RST to define indicative areas and measures for RSF reform and conditionality. Beyond learning-by-doing, staff will also engage proactively with institutions, particularly the World Bank and regional development banks, to inform such reforms. To support country operations, staff will set up an interdepartmental working group to develop a knowledge hub and guide teams with best practices.

This box illustrates staffs preliminary and indicative thinking pertaining to climate-related reform measures which may be relevant depending on a member’s circumstances. Reforms for other qualifying longer-term challenges would be developed in parallel and in conjunction with the relevant IFI partners.

Climate-related reforms supported under the RSF would aim to integrate climate risks in economic decisions and support countries with adaptation, mitigation, and transition. Depending on country-specific priorities, RST reforms could support: (i) enhanced assessment of risks for climate-informed development strategies; (ii) improved adaptation that reduces costs of slow-moving climate change and climate-induced natural disasters; (iii) mitigation policies that reduce fiscal and balance of payments risks, while supporting the member to meet its international commitments; (iv) social protection strategies to mitigate the impact of climate shocks on the most vulnerable and also support a just transition; (v) development and implementation of a climate finance strategy aligned with countries’ NDCs; (vi) building buffers to cope with shocks; (vii) putting in place frameworks that support the integration of climate in public financial management and allow for increased transparency on climate reporting; and (viii) putting in place regulatory and supervisory frameworks to better assess climate-related financial sector risks.

The list below provides hypothetical examples in in the area of climate change. Specific measures under an RSF would have to account for country-specific circumstances, including the assessment of macro critical risks to BoP stability caused by climate change and the country’s starting positions in terms of climate policy and technical capacity. As noted in Annex III, policy priorities and conditionality will be developed in coordination with the World Bank; regional MDBs inputs will be sought as relevant for each country.

Climate Adaptation

  • Approve a National Adaptation Plan, which: (1) identifies climate vulnerabilities and adaptation gaps; (2) identifies adaptation solutions, including cost estimates for all major projects; (3) mainstreams adaptation into national planning and PFM; and (4) monitors and reports progress regularly.

  • Adopt measures to achieve at least full cost recovery in water utility pricing while ensuring access to adequate water consumption of the vulnerable to maintain affordability.

Climate Mitigation

  • Issue regulations on carbon pricing policy (e.g., removing exemptions, increasing carbon price levels, and expanding sector coverage).

  • Introduce/apply the standard VAT rate for electricity and fossil fuels.

  • Introduce/increase excises on coal, natural gas, and petroleum products.

  • Phase out agricultural subsidies that encourage emissions-intensive farming.

Climate Finance

Approve a disaster risk financing strategy for both low impact/frequent events and higher impact/low frequency events, using a risk layering approach covering both risk retention (budget reallocation/ augmentation) and risk transfer (climate risk insurance and climate-contingent loan rescheduling).

Public Investment Management

  • Define and publish climate-related elements among the criteria used by the government for the selection of all major public investment projects

  • Establish a standard methodology to conduct mandatory climate-related analysis for the ex-ante appraisal of all major public investment projects.

  • Produce centralized guidance and/or establish a central support unit to assist government agencies on the preparation and costing of climate-aware public investment strategies

Public Financial Management

  • Require that the climate implications of major new budget measures be systematically included in ex ante impact assessments and cost-benefit analyses and published in budget documents.

  • Develop a climate budget tagging system to improve allocation and monitoring of climate-related expenditure and publish report on climate-related expenditure alongside budget documents.

  • Develop a fiscal risk statement to include climate risks and natural disasters, and include narrative on risk management strategies, notably with respect to climate-related risks to public infrastructure assets.

Disbursements, Phasing and Reviews

56. Disbursements: A member’s total RSF access would be partitioned into several disbursements that are linked to the implementation of RSF reform measures. Each disbursement, defined as a percentage of quota, will be associated with the implementation of one reform measure. Recognizing the difficulty of “pricing” longer-term reforms, the presumption would be for equal access across disbursements; exceptions could be made for reform measures that are demonstrably more consequential to prospective BoP stability and/or create larger actual BoP needs.

57. Phasing will be specified at the time of the RSF request. Each disbursement will have an availability date linked to the expected time of completion of the associated reform measure, which will generally mirror the availability date of a particular purchase/disbursement under the UCT program;26 this “links” RSF disbursements to reviews under the accompanying UCT program.27 Multiple reform measures can be linked to the same availability date. Phasing should ensure that the cumulative access of all disbursements subject to the completion of a review under the RSF arrangement does not exceed 50 percent of quota. That said, if there are delays in the implementation of reform measures, disbursements exceeding 50 percent of quota that become available following the completion of a review are not precluded. Box 2 illustrates a range of scenarios. No disbursement would be available immediately at the time of approval of the request for an RSF arrangement. This ensures that an RSF arrangement only finances reforms that are additional to efforts the member has already undertaken.

58. Reviews: RSF disbursements would occur following the completion of a review by the Executive Board. RSF reviews will be used to monitor Implementation of reform measures, modify them, or establish new reform measures as part of an augmentation. Several features are of note:

  • Monitoring conditionality. RST loan disbursements are subject to the completion of an RSF review by the Executive Board. Completion of an RSF review28 will require a Board assessment, based on staffs recommendation included in the UCT program documentation, that the relevant reform measure was met. The Board may also, in rare cases, decide to complete the review and approve the associated RSF disbursement available where the reform measure was implemented with a minor deviation relative to its Board-approved design. In making an assessment of whether the deviation is minor, the Board will consider whether the member has substantially implemented the reform measure such that the objective of the reform measure has been met. The RST review will only make available the RST disbursement amounts corresponding to the reform measures that were assessed as implemented or implemented with minor deviation.

  • Modification and establishment of conditionality. In cases where the reform measure is assessed as not met, and the Board does not assess the deviation as minor, the review would provide an opportunity to adjust (or replace) the reform measure on a forward -looking basis if needed. Reviews can also be used to augment an RSF arrangement by increasing access of disbursements associated with existing reform measures with availability dates in the future or establishing new reform measures and associated disbursements.

  • Timing of reviews. Since the target dates for reform measures are indicative in nature (see 1154), regular reviews under an RSF arrangement are not warranted. Instead, reviews under an RSF arrangement will take place concurrently with reviews under the UCT program, once the expected date of completion of a reform measure has passed and management (i) recommends completion of a review based on staffs assessment that the reform measure was successfully implemented, or that any deviation was minor or (ii) a modification of reform measure(s) based on the authorities’ request for the same.29 To ensure that RST disbursements only occur when the UCT program is on track, delays in completing reviews under the UCT arrangement would also delay RSF reviews, and therefore delay the attendant disbursements, even when reform measures are implemented on time (Box 2, Scenario D).

  • Disbursement request. In keeping with the drawing nature of RSF arrangements (1130), members would be expected to request disbursement promptly upon completion of an RSF review. Reflecting the scarce nature of RST resources, the member would have 30 calendar days following completion of the review by the Executive Board to make a request (but no later than the end of the arrangement). If a member fails to make a request for disbursement within this period, the member may only request the disbursement in the 30-day period following the completion of a subsequent review under the RSF arrangement. This will also help minimize the risk that RSF disbursements occur when the UCT program is off-track.

59. Undisbursed amounts, cancellation, and expiration. As with all Fund arrangements, an RSF arrangement expires when all amounts under the RSF arrangement have been disbursed. An RSF arrangement may also be cancelled at any time by a member by notifying the Fund. The RSF arrangement would automatically terminate when the accompanying UCT program expires, is cancelled, or terminates (Box 2, Scenario E). Undisbursed financing after the expiration of an RSF arrangement—which may arise because reform measures were not implemented or the UCT program went off-track, expired or was cancelled—would not be available thereafter. However, the member may request financing under a successor RSF arrangement—with access up to the overall cumulative access caps for the RSF (in quota and SDR terms)—and an appropriate successor UCT arrangement.

Phasing and Disbursements Under Various Illustrative Scenarios

1. Phasing scenarios for a simple example

A hypothetical RSF arrangement approved together with an EFF has two reform measures phased for implementation ahead of the 2nd and 4th reviews. If both reform measures are implemented on schedule, ex-ante and ex-post RSF phasings are identical (Scenario A). If, however, implementation of the second reform measure is only completed ahead of the 5th review under the EFF, the corresponding review under the RSF arrangement and the associated disbursement would take place concurrently with the 5th EFF review (Scenario B). All reform measures must be implemented ahead of the last completed review under the EFF for the full amount under the RSF arrangement to be disbursed (Scenario C).

uA001fig01
Note: Scenarios assume simultaneous requests for the RSF and the UCT arrangement. RSF requests at the time of a scheduled review under the UCT arrangement would not substantively affect the scenarios.

2. Schedules of disbursement for a more complex example The following example presents disbursement tables for a hypothetical RSF arrangement accompanied by a three-year ECF, both approved on October 31, 2022. This example is separate from the phasing diagrams above, as it assumes more reform measures (four in total), with reform measures 2 and 3 expected to be completed by the time of the second ECF review (since all share the same availability date—April 1 5, 2024). A separate table would specify the content of reform measures 1 through 4.

Country: Proposed Schedule of Disbursements and Timing of Reviews Under ECF Arrangement

article image

Country: Proposed Schedule of Disbursements and Timing of Reviews Under RSF Arrangement

article image

F. Debt and Capacity to Repay Considerations

60. Additional debt and capacity to repay (CtR) safeguards are proposed for lending under the RST to account for the nature and terms of the lending, borrower’s debt profiles, and to mitigate excessive accumulation of de facto senior debt. Medium-term debt sustainability is a key qualification for accessing the RST (Section III.B). In addition, all Board documents for RST financing requests and augmentations will need to include analysis and discussion of (i) the composition of public debt, including the share of de facto senior debt, (ii) debt risk analysis over a longer time horizon (up to 20 years), and (iii) an extended CtR analysis that covers the RSF repayment period, taking into account all Fund borrowing by the member. These safeguards are warranted in part by the RSF’s longer repayment period and because RSF access does not fall under the PS-HCC Policy. Such additional safeguards are also intended to ensure that RST lending does not aggravate a borrower’s debt vulnerabilities, notwithstanding the additional exposure to de facto senior/non-restructurable debt (see discussion of the RST’s envisaged de facto Preferred Creditor Status, Section IV.B).

61. Debt composition. All relevant Board documents will need to discuss the structure of total public debt and its projected evolution overtime. In particular, analysis of external debt structures should take into account the amount and shares of debt owed to the Fund and other senior creditors, informed by tables showing two distinct breakdowns of public external debt: (i) de facto senior or difficult to restructure debt (debt to the IMF; debt to the World Bank and other international financial institutions; known collateralized debt) and other debt and (ii) multilateral versus official bilateral versus private debt.30

62. Long-term debt risk analysis. The LIC Debt Sustainability Framework (DSF) already requires a macroeconomic framework for a 20-year projection period, providing debt and debt service projections for the full repayment horizon of RST financing.31 The new Sovereign Risk and Debt Sustainability Framework for Market Access Countries (MAC SRDSF) will provide a 10-year horizon for debt and GFN projections. Both of these frameworks also have tailored stress tests for natural disasters to capture risks associated with one-off climate events over the medium term and the MAC SRDSF has a set of modules to analyze key longer-term risks: climate change, large amortizations, demographics, and natural resource scale-up or depletion.32 While these modules are optional for UCT lending in market access countries, they will be required for program documents with a concurrent RSF arrangement to inform on key long-term risk to debt, with only the natural resource scale-up/depletion module as optional. In particular, for the climate change module, countries will be expected to discuss the impact of climate change on key macro-fiscal variables (such as potential growth and spending) and, where feasible, draw out long-term implications for the debt, leveraging relevant analysis from the IMF’s Climate Macroeconomic Assessment Program (CMAP) documents, World Bank’s Country Climate and Development Reports (CCDR) or other country-specific work.

63. Capacity to repay. All relevant Board documents will need to present the evolution of the total IMF (i.e., PRGT+GRA+RST) projected debt and debt service to the Fund relative to key economic metrics over the course of the RST’s repayment period.

  • These long-term (20 years) projections, based on data provided by country teams, will be compared with available data from prior Fund lending, supported by a set of standardized charts.

  • In cases where the CtR indicators signal relatively high risks, the program documents underpinning the RSF request will discuss the severity of the implied risks and explain how the design of the RSF and concurrent UCT program—including access, phasing, and conditionality— seek to mitigate these risks.

  • The assessment will normally be based on the baseline macroeconomic framework of the RSF request, but may also need to reflect information on downside scenarios if these are also included in program documents (e.g., if the concurrent UCT arrangement is treated as precautionary).

G. Overdue Obligations

64. The Managing Director will not recommend an RSF arrangement for approval by the Executive Board if the member has overdue financial obligations to the Fund. This applies to overdue financial obligations in the General Resources Account, the Special Disbursement Account, the SDR Department, or to the Fund as Trustee (under the RST or under the PRGT). Similarly, if a member under an RSF arrangement incurs overdue financial obligations as described above, no further disbursements may be made under the RSF arrangement until such time as the overdue financial obligations are settled in full. Moreover, should a member incur overdue financial obligations to the RST, that member’s access to the GRA and PRGT would also be suspended.

65. The RST would follow the Fund’s strategy to prevent and remediate overdue financial obligations.33 This strategy consists of three elements: prevention, intensified collaboration, and remedial measures. As in the GRA and PRGT, in the RST, preventative measures provide a critical first line of defense to avoid the emergence of new arrears and would employ the Fund’s engagement with the member, including through surveillance, policy advice and safeguards, and technical assistance with the aim to proactively safeguard Fund resources. For members that cooperate with the Fund in resolving their arrears the strategy provides for intensified collaboration to assist the member in establishing a track record of policy and payments performance, mobilizing resources from other creditors and donors, and normalizing relations with the Fund. Once arrears have arisen, the strategy provides for the application of remedial measures of increasing intensity in order to encourage members to cooperate with the Fund in seeking a solution to their arrears.

66. The proposed cooperative strategy for resolving arrears under the RST would set remedial measures in cases of overdue financial obligations, broadly similar to those for the PRGT.34 The remedial measures would apply to member countries with overdue obligations (repayments of principal and interest) that do not actively cooperate with the Fund to resolve their arrears problems. An escalating timetable would guide the Executive Board in considering remedial measures of increasing intensity, although the application of each measure would take into account the specific circumstances of the member.35 Whenever a member would fail to settle a financial obligation to the RST on time, communication urging prompt payment would be sent to the member and the Executive Director concerned, and the member’s access to the GRA, PRGT, and RST will have been suspended. In all cases, overdue obligations (i.e., repayments of principal or payment of interest) to the RST would accrue interest at an interest rate equal to the applicable interest rate under the tier structure to compensate the Trust for the opportunity cost of arrears, subject to a minimum interest rate charge of the SDRi.36 In addition, in cases of protracted overdue obligations, remedial measures would also include the Executive Board’s consideration of removing the member from the list of RST eligible countries (6 months after emergence of arrears) and declaration of noncooperation with the Trust (12 months after emergence of arrears). Upon a declaration of noncooperation, the Fund could decide to suspend the provision of technical assistance. Nonetheless, as for the PRGT, these remedial measures would not extend as far as those for the GRA given that overdue obligations in the GRA are breaches under the Articles of Agreement whereas overdue obligations in respect of Trust loans are not.

67. The RST’s framework for overdue financial obligations would allow for and encourage the prioritization of GRA repurchases and PRGT repayments. The IMF’s lending under the GRA and PRGT has a unique “lender-of-last resort” function that is essential for the stable functioning of the international monetary system. Consistent with this role, and while recognizing borrowers’ freedom to attribute any payment to specific obligations, members would be encouraged to prioritize payments (arrears clearance or as appropriate periodic partial/”token” payments) to the GRA and PRGT in cases where a borrower has (or risks having) such overdue financial obligations. This approach, which would be built into the design of the RST, would be in line with the role that GRA and PRGT lending play in the Global Financial Safety Net and would help avoid any weakening in the Fund’s ability to perform its broader responsibilities in the International Monetary System.

H. Post-Financing Assessment and Ongoing Monitoring

68. Following the end of an RSF arrangement, monitoring could take place under the current Post-Financing Assessment policy. It is proposed that outstanding credit to the RST would count towards the quota threshold established under the Post Financing Assessment (PFA) policy, where PFA is expected for members that are not in a Fund arrangement and where combined GRA/PRGT/RST credit outstanding exceeds 200 percent of quota. In these cases, the evolution and impacts of reform measures and monitoring the member’s capacity to make repayments to the RST would be covered as part of the PFA Although in practice PFA is primarily triggered by credit levels relative to quota, the PFA policy also provides for absolute trigger of SDR 1.5 billion and SDR 0.38 billion for credit outstanding to the GRA and PRGT, respectively, which were calibrated relative to the loss-absorption capacity of the GRA and PRGT. Based on experience and the evolution of the RST’s Reserve Account, the calibration of an absolute level of outstanding RST credit as a trigger for PFA could be considered at the time of the first review of the RST policy. In the interim, the initial trigger for PFA in absolute terms is proposed to be set at the same level as the PRGT (i.e., SDR 0.38 billion). In addition, the PFA policy provides the Managing Director with the discretion to recommend the institution of PFA if there are developments that suggest the need for closer monitoring of the member’s capacity to repay, and particularly, where developments call into question the member’s progress toward external viability37 In the case of the RST, ongoing monitoring in the context of Fund surveillance would provide the opportunity to recommend PFA as needed over the longer period for which this credit remains outstanding.

69. Further monitoring will be covered as part of surveillance and lending. Following the end of an RSF arrangement, Article IV consultations will be the main vehicle to cover issues related to longer-term structural challenges addressed by the RST and follow up on the implementation of past IMF policy advice as long as they remain macro-critical. If residual BoP needs persist or a new BoP gap emerges, any follow-up UCT program would be expected to account for RSF repayments in debt sustainability and capacity to repay considerations.

I. Misreporting Framework and Other Lending Policies

70. Misreporting framework. A framework to address misreporting under the RST is proposed to be established. In particular, misreporting under the RST would occur either:

  • (i) when the member has received a non-complying disbursement as a result of inaccurate information being provided to the Fund on the implementation of a reform measure which leads to an incorrect assessment that (a) such measure has been fully implemented when it has in fact not been fully implemented or (b) that the deviation in implementation of the reform measure was minor when in fact it was not; or

  • (ii) there has been an Executive Board finding of misreporting under the accompanying UCT program, which was not assessed as “de minimis” or waived by the Executive Board. As the UCT program provides the short-term policy anchor and safeguards for disbursements under the RSF arrangement, a noncomplying disbursement under the accompanying UCT program, other than a “de minimis” misreporting or a misreporting that was waived, would also taint disbursements made at the same time under the RSF arrangement as noncompliant.

In cases of misreporting under (i) above, members are expected to make an early repayment of the non-complying disbursement(s), unless the Executive Board makes a determination that, based on the fact that the deviation in policy implementation is minor, the objectives of the reform measure have been achieved, notwithstanding the misreporting. With respect to cases of misreporting under (ii) above, the misreporting framework applicable to the UCT program would apply with no separate or additional steps required under the RST misreporting framework. Members would be expected to make an early repayment of the non-complying disbursement(s) under the RSF arrangement unless, pursuant to the misreporting framework applicable to the UCT program, the misreporting under the UCT program is de minimis or a waiver of non-observance is granted by the Executive Board. With the above exceptions, the misreporting framework of the RST is proposed to be procedurally analogous to that of the PRGT.

71. Other Fund policies. Given the close operational link between RSFs and UCT programs, key Fund policies applicable to UCT programs (e.g., lapse of time (LOT) procedures, side letter policy) will be applicable to RSF arrangements mutatis mutandis. Under the LOT procedures, a review under an RSF arrangement would be eligible for completion on a LOT basis where the review under the underlying UCT-quality arrangement or instrument meet the criteria for LOT completion of program reviews and the relevant reform measures have been met. In countries where governance issues are assessed to be critical to the resolution of the member’s BoP problem, these would be expected to be addressed under the UCT program in line with the 2018 Governance Policy. Staff would expect this aspect to be taken up based on experience at the proposed review of the RST. Other lending policies would be expected to guide RST policies as they apply to the accompanying UCT program as well. Any standalone RSF arrangement documents would be considered country papers within the scope of the Transparency Policy, with the same publication regime as applicable to documents on use of Fund resources. Similarly, any standalone RSF arrangement documents would be considered as documents discussing a member’s use of the Fund’s resources under the Open Archives Policy.

Section IV. Governance and Multilateral Considerations

A. Governance

72. Similar to the PRGT, the Fund would act as Trustee of the RST, and the Executive Board would have the authority to take decisions on issues pertaining to the RST as well as to approve individual RSF arrangements. This includes decisions related to eligibility, qualification, access, adjustments to certain financial parameters of the Trust (such as the interest rates), and the assessment of RSF conditionality. Decisions regarding the RST establishment, amendments and implementing decisions would be taken by a majority of the votes cast, as is the case with most decisions in the PRGT. The Executive Board would also have the power to terminate the RST.

73. Creditor consent would be required for changes to fundamental terms of the RST that affect creditor interests, including the purposes clause, lending facilities, and key financial terms and structure. As discussed in greater detail in the supplement to this paper, the RST Instrument specifies the provisions that could only be amended with creditor consent (“protected provisions”). These protections would assure contributors that no fundamental change would be made to the clause governing use of RST resources or in the key financial parameters for creditors without their consent or an opportunity to suspend their financial support for future trust operations in the absence of consent.38 If a Loan Account contributor does not consent to an amendment, further drawings under its borrowing agreement to fund Trust loans would be suspended. The contributor may further request the partial return of its Deposit Account and Reserve Account contributions, in an amount proportional to the remaining uncommitted portion of its commitments under its Loan Account borrowing agreement. The borrowing agreement, though, would remain open for encashment calls related to claims outstanding or committed prior to the effectiveness of the amendment, and outstanding claims under such borrowing agreement would remain outstanding and subject to repayment in accordance with the repayment schedule of the RST loans funded with drawings under the borrowing agreement. Where the consent of a stand-alone contributor to an amendment of the Reserve Account or the Deposit Account is required, the non-consenting contributor under such stand-alone contributions may request the repayment of its principal contribution, net of any losses or retained investment earnings.

74. In terms of the consent process for protected provisions, staff proposes that the consent of a creditor would be assumed if the contributor was given adequate notice and appropriate time to respond to a request for consent and no response is received. Specifically, the RST Instrument would authorize the Board to set a first deadline for explicit consent. This would be followed by a second deadline to those contributors that did not respond within the first deadline. The second formal notifications would indicate that if no response is received from the contributor on the proposed amendment, the contributor would be presumed to have consented to the amendment. It is proposed that after the amendment has become effective, any contributor that did not respond within the second deadline would have six months to notify the Fund of its objection to the amendment. If such notification is received, the contributor would have the same rights regarding suspension of future drawings and partial return of its share in the deposit and reserve accounts discussed in paragraph 73 above. The underlying rationale for this proposed consent process is that the purpose of the protected provisions is to give contributors enough time to consider whether they support an amendment and to take actions to protect their contributions if necessary, but not to provide them with a veto (including, e.g., through non-response). The proposal aims at balancing two competing interests, namely the need of giving contributors adequate time to consider an amendment and the ability to move forward with an amendment if a contributor does not respond. Staff is proposing a similar approach for the PRGT (1196).

75. Staff propose that the RST would be reviewed three years after it becomes operational, or sooner if warranted.39 This review could take stock of the experience with the RST, and consider if any modifications may be needed, including in respect to qualifying longer-term challenges, eligibility, margin and service charge structure, country groupings, and reserve coverage. The Board could also assess at that time whether RST financing should extend beyond the current drawdown period (November 30, 2030) or be sunset at a future date, in line with the goal of the SDR channeling to support more resilient and sustainable growth following the pandemic. The review would also decide the frequency at which RST policy reviews would be conducted going forward with the expectation that such reviews would generally cover the policy areas included in the first review. As set out above, depending on developments in the SDRi or unexpected events that have a bearing on the financial sustainability of the RST, an earlier review of the RST interest rate structure may be necessary. Such reviews would always also trigger a review of the country groupings.

B. Coordination with the World Bank

76. RST lending operations would greatly benefit from close coordination with the World Bank and other relevant agencies, including regional MDBs. IMF and World Bank staff have worked closely to develop broad principles for coordination on RST operations (see Annex III). At this stage, the Fund staff note concentrates on climate-related issues; they would need to be customized for other long-term challenges under the RSF to the specific features of those purposes. In line with the 1989 IMF-World Bank Concordat (SM/89/54, Rev. 1), coordination will be guided by principles that reflect each institutions’ respective mandate and expertise.

77. The five key areas for coordination are (i) diagnostics; (ii) policy priorities; (iii) conditionality; (iv) implementation supervision; and (v) program documentation (i.e., a vehicle to document the coordinated approach for RST lending).

  • RSF reform measures would be informed by country diagnostics developed in both institutions relevant to the RST’s purposes, as well as—where applicable—national authorities’ strategies. For example, on climate change, the Bank’s Country Climate and Development Reports (CCDR), if available, will be a critical input, complemented by other products such as the Fund’s Climate Change Policy Assessments (CCPAs) and its potential successor instrument, Climate Macroeconomic Assessments Programs (CMAPs), Climate Public Investment Management Assessment (C-PIMA), and Disaster Resilience Strategies (DRS). This would strengthen policy consistency and complementarity on policy recommendations, while leveraging the comparative expertise of both institutions for the best policy advice to the country. If neither a CCDR nor a CMAP are available, RSF policy design would be based on other available information and discussions with other IFIs/regional MDBs and the authorities.

  • Based on the diagnostics, Fund staff will identify policies that the member country would need to prioritize under the RST, in coordination with the WB. The idea is to promote complementarity with any active or planned WB development policy operation or climate program, while delineating areas of responsibility according to each institution’s mandate and role.

  • The IMF’s Executive Board will establish RSF conditionality and assess whether an RSF condition is met or not (consistent with the policy on not having cross-conditionality), drawing to the fullest extent deemed useful on the advice of WB staff in areas of the WB expertise. RST conditionality would be designed taking into account the principles underpinning the conditionality guidelines, including capacity to implement.

  • It is envisaged that Bank staff would provide an assessment letter on the authorities’ climate policies for information at the time of an RSF approval and for reviews where implementation of RSF reform measures is assessed. In addition, the staff report and MEFP will also document the coordination with the Bank, including a brief description of the Bank’s engagement in the sector and/or any Bank technical assistance related to RST-supported reforms. The staff report would also acknowledge Bank staff’s inputs into the RST in the cover note.

  • Broad expectations on modalities for engagement for both Fund and WB staff have also been agreed, with appropriate caveats for confidentiality of discussions with the authorities and information sharing.

78. A coordination framework will also be developed for pandemic preparedness by the start of RST operations. Based on a similar blueprint developed for coordination with the World Bank on climate change, staff are engaging with relevant stakeholders, including the World Bank and the World Health Organization, to develop broad principles for coordination with other institutions on RST lending for pandemic preparedness.

C. Preferred Creditor Status

79. The proposed lending framework presumes de facto PCS of RST loans, as is the case for Fund lending under the GRA and PRGT. The Fund’s de facto PCS is accepted by Fund members and the international creditor community (official and private)—by practice, not law—recognizing the public good nature of IMF financing, including the benefits to both the country and its creditors from Fund financial support under a Fund-supported program. Under this practice, creditors have excluded the Fund’s claims from sovereign debt restructuring by exercising voluntary forbearance in enforcing their claims.

80. To clarify intentions regarding the de facto PCS status for RST claims, the IMFC could call on all Fund members to recognize such status for RST claims, as it did for the Fund’s PCS in 1988. Major official bilateral creditors could recognize the de facto PCS of RST claims via the Paris Club or the G20, such as in creditor committees under the Common Framework of the G20 or Paris Club.

81. De facto PCS for the RST claims would be accompanied by additional safeguards. Appropriate access policies, and the debt sustainability and CtR safeguards discussed above would help minimize the risk that RST obligations could aggravate a member’s debt vulnerabilities as a result of the added exposure to de facto senior debt. In addition, to protect the unique lender-of-last resort function of the GRA and PRGT financing, the RST’s policy would encourage prioritizing payments to the GRA and PRGT over RST obligations when a borrower has overdue obligations to the Fund (see Section III.I—I) with no implication for the de facto PCS of the RST claims.

Section V. Financial Architecture

A. General Framework

82. The RST is proposed to be a loan-based trust administered by the IMF, with a governance and financial structure broadly similar to that of the PRGT. The roles of the Executive Board, management, and staff for RST lending and other decisions regarding the RST would be broadly the same as for PRGT and GRA lending. Similar to PRGT, RST resources would be mobilized based on voluntary contributions from members, including those wishing to channel SDRs for the benefit of LIC and more vulnerable middle-income members. The RST would in turn on-lend these resources to eligible borrowers. It is expected that the bulk of the contributions to the RST would be made in SDRs, although the provision of RST resources in freely usable currencies would also be possible. Contributions to the Reserve Account (RA) and Deposit Account (DA) could also be made in other currencies. All claims of contributors on the RST would be denominated in SDRs.

83. Establishing a large loan-based trust, and making it operational in a relatively short period of time poses challenges for the financial design of the RST relative to the PRGT, which had a funded Reserve Account from its establishment and has expanded incrementally since 1987. From the outset, the RST would require sufficient loan resources to meet potential demand from borrowers, as well as adequate financial buffers to help manage credit and liquidity risks. While lending and investment income can help build reserves overtime, there is a need for adequate upfront funding—this presents a major challenge since members’ budgets are constrained by the impact of the pandemic and parallel fundraising campaigns for other initiatives (e.g., PRGT, CCRT, IDA). This is the main difference over the PRGT which from its establishment had a reserve account funded with Special Disbursement Account (SDA) resources derived from gold sales profits.

84. The proposed financial architecture is designed to ensure the safety and liquidity of contributors’ claims on the Trust while minimizing the need for budgetary contributions. This is done through a combination of: (i) strong policy safeguards, including the link of RST lending to a UCT-quality program; (ii) a supportive multilateral context, including the expected de facto PCS for the RST; (iii) an adequate reserve buffer that is expected to grow over time through lending and investment income, with significantly higher reserves by the time RST loan repayments come due; (iv) a funding mechanism that combines modest upfront reserve account contributions with a much larger pool of loan resources and long-term deposits by contributors that create claims on the RST and retain their reserve asset character (through early repayment in case of balance of payments or reserve position need (“encashability”) and high asset quality); (v) a split of each contributor’s claims on the RST into resources for lending operations (drawn over time based on demand) and an upfront long-term contributions to bolster reserves (contributing to investment income and risk mitigation).

85. To operationalize these design features, the RST is proposed to have three financial pillars: a Loan Account, a Reserve Account, and a Deposit Account (see Figure 1 below and Annex IV for details).

Figure 1.
Figure 1.

RST Financial Framework

Citation: Policy Papers 2022, 013; 10.5089/9798400207730.007.A001

* Contributors and beneficiaries can provide/receive both SDRs and currency, facilitated through the VTA market when needed
  • Loan Account (LA). The LA is the RST’s conduit for lending operations, funded by voluntary loan commitments from contributors. As in the PRGT, RST loans would be made on a pass-through basis: the Trust extends loans to borrowers (in response to demand) by drawing on available LA resources committed by contributors (i.e., loan or note purchase agreements). Borrowing Agreements with contributors would be subject to uniform key substantive terms. In particular, all LA contributors are remunerated at interest rates up to the SDRi. Borrowers of RST loans will pay the SDRi plus a modest margin to cover administrative expenses and reserve build up, as well as a service charge in some cases (see H43). Repayments to LA contributors are on the same maturity schedule as the amortization of borrowers’ RST credit outstanding. LA contributors’ commitments will include an encashment buffer available for drawings in the event that another contributor requests early repayment of its LA claims by representing that its balance of payments or reserve position justify it (see Section V.B below).

  • Reserve Account (RA). The RA is the RST’s principal financial buffer to manage credit and liquidity risk, and also to cover administrative costs. It is funded through upfront contributions by contributors in proportion to their respective LA commitments. To limit the budgetary impact of the upfront RA principal contributions by LA contributors, the proposed structure starts off with a relatively modest RA balance that grows over time from lending margins, service charges, and investment income, net of administrative costs. The RA principal contributions are not remunerated but entitle each contributor to a share of the Trust’s net assets at the termination of the Trust. RST lending income from the margin is transferred to the RA quarterly, and any service charges by RST borrowers go directly into the RA. RA resources can be used to cover the administrative expenses of the Trust, and the margin and service charge income are expected to fully cover these costs. RA resources are distributed to contributors based on their share in the account at the time of the liquidation of the RAAn early partial distribution of RA resources may also take place based on an assessment of the adequacy of net reserves. Pending use, RA resources are invested in liquid, high-quality instruments in accordance with guidelines established by the Executive Board (see Section V.C below).

  • Deposit Account (DA). The DA, which adds to the RST’s gross reserves, would be funded upfront by long-term deposits from LA contributors in proportion to their LA commitments, remunerated at interest rates up to the SDRi. The purpose of the DA is to generate additional net investment income for the Trust that would contribute to building the RST’s net reserves, thereby minimizing any residual risks to contributors’ claims on the RST. Resources will be invested in liquid, high-quality assets, in accordance with investment guidelines adopted by the Executive Board (see Section V.C below), that would permit liquidation at short notice to allow for payment of a possible encashment in the event of a BoP/reserve need by a contributor.

86. Contributors’ claims on the LA and DA of the RST will have all the characteristics of reserve assets, supported by a multilayered risk management framework and the possibility to encash such claims in case of balance of payments or reserve position need. The quality and liquidity of claims on the RST is backed by:

  • Policy safeguards. RST lending would be approved alongside an arrangement supporting a UCT program allowing the RSF arrangement to leverage the UCT program’s financing, safeguards and monitoring framework (e.g., program design, debt sustainability analysis, capacity to repay assessment, PFA arrears policy). In addition, RST financing would have its own conditionality and additional policy requirements, such as an overall access cap, phasing rules, debt sustainability and CtR requirements, and charges and repayment structures (see Section III).

  • Multilateral context and de facto PCS. As with the PRGT, the RST would benefit from a supportive multilateral context. For instance, risk pooling and mitigation would be facilitated by broad member participation on the lending and borrowing side. The purpose and public good nature of the RST would motivate contributions to the Trust, and RSF-supported policies would possibly catalyze broader donor support on terms that moderate borrowers’ credit risk and promote sustained implementation of sound polices. Collaboration on the prevention and clearing of arrears and the agreement on remedial measures for overdue obligations would provide credit risk protection for the RST and, in turn, its contributors. Consensus by members that RST claims would have de facto PCS through an IMFC communique as discussed above will be an important factor in facilitating a wide pool of contributors to the RST.

  • Financial buffers. The proposed financial architecture is designed to provide strong buffers to safeguard against financial risk and to ensure the high quality and liquidity of contributors’ claims on the RST (i.e., LA and DA) from its establishment to underpin its readiness to meet potential demand.

87. The proposed financial structure envisages adequate financial buffers up-front, with significantly higher reserve levels expected by the time RST loan repayments begin. Gross reserves are initially funded through upfront contributions to the RA and long-term deposits (principal claims) in the DA Net reserves, defined as gross reserves net of contributors’ claims on the DA, are initially modest but are expected to grow steadily funded by lending margins and service charges income, excess investment income in the DA (net of SDRi paid to contributors), and investment income of the RA, net of administrative costs (see Section VI). The adequacy of reserves would be reviewed periodically together with the interest and charges’structure, with a view to ensuring an adequate, but not excessive, degree of reserve coverage.

88. Each LA contributor would sign a Contribution Package with three parts—the loan account contribution (borrowing agreement); an RA contribution; and a DA contribution.4041 Contributions to the RAand DA would be 2 percent and 20 percent, respectively, of a contributor’s commitment to the LA at a minimum. The DA contributions would have a uniform pre-specified maturity date, i.e., 20 years after the end of the agreed loan drawdown period when all LA claims are scheduled to have been repaid, irrespective of when final repayments under individual borrowing agreements take place. At the time of maturity of the DA, the principal value of a DA contribution is returned to the respective contributor. Any excess investment returns attributed to the DA contribution would be transferred to the RA and counted as a contribution of the contributor to that account. The RA contribution and DA net investment income would remain in the RST pending distribution to contributors. At liquidation of the RST, following repayment of all LA claims under borrowing agreements, and the repayment of the DA principal contributions and the associated transfer of excess investment income to the RA, each contributor would receive its calculated share in the total assets of the RA. Annex IV sets out in greater detail the principles and methodologies for DA and RA distributions. In the event of successive RST mobilization rounds, any additional RA and DA contributions would need to be determined at the time such extension of the RST is decided, taking into account the reserve adequacy of the Trust. As with the PRGT, staff would expect that contributions would be received from members or their institutions (packages or ad hoc contributions), and the framework is designed this way (in particular regarding use of SDR or encashment rights). If interest for contributions were received from other parties, the Managing Director would assess whether such requests could be accommodated under the RST framework (for example by excluding SDR transactions if the contributor is not a prescribed holder or limiting encashment rights), but would submit any such request for Executive Board consideration.

89. The Trust may also receive “standalone” contributions to the RA and/or the DA unrelated to LA contributions. Such contributions could come from contributors that are not able to make LA contributions but would want to contribute to the RA and/or the DA. Such standalone contributions to either or both accounts would be in the interest of the Trust, as they would bolster reserves at the outset and raise Trust income from investments. In contrast, standalone contributions to the LA (without RA and DA contributions) would increase the risk exposure of reserves and deposits from other contributors. Standalone contributions to the RA and DA would have a fixed maturity, guided by an investment duration consistent with reliably strengthening RST reserve accumulation. Staff considers that such contributions should normally have a maturity of 10 years. Standalone RA and DA contributions bear the same risks within the RST as RA/DA contributions from a Contribution Package.

B. Reserve Asset Character of Contributor Claims on the Loan and Deposit Accounts

90. Many potential contributors prefer (or require) that their claims on the Loan and Deposit Account of the RST qualify as international reserves.42 A key requirement for a reserve asset is that it must be readily available for use in case of a BoP or reserve need (i.e., be liquid) and be of generally high quality (i.e., with low credit risk). Preserving the reserve asset nature of lenders’ claims thus entails ensuring both the quality and liquidity of the underlying assets of the Trust, i.e., the quality of RST loans and investments. Under the proposed architecture, LA and DA claims would have all the necessary characteristics of reserves assets, while RA contributions would not.

91. LA claims. Contributors could count outstanding loan claims on the LA as international reserves, in view of the quality and liquidity of the underlying assets under the proposed credit risk management framework and encashment regime.

  • Asset quality of outstanding loan claims on the LA would be ensured by a multilayered credit risk management framework involving three pillars: (i) policy safeguards (including requiring an accompanying UCT-program, conditionality, and RST access policies) to ensure that borrowers have the capacity to service their debt; (ii) financial buffers in the RA and DA as described above, to protect contributors’ claims on the LA against potential arrears or credit losses; and (iii) a supportive multilateral context (e.g., a cooperative arrears framework, a diversified borrower base, and envisaged de facto PCS).

  • Liquidity of contributors’ claims on the LA related to outstanding RST loans would be ensured by an encashment regime stipulated in the RST instrument and incorporated into the loan agreements (all LA contributors would have to allow for encashment drawings), where contributors commit to provide access to their borrowing agreements for drawings in the event another contributor experiences a BoP need and requests early repayment its loan claim. Staff would seek to manage the level of undrawn commitments under borrowing agreements in a manner that preserves a buffer for possible encashment requests, parallel to the practice in the PRGT. A lender encashing its RST loan resources would commit to again finance calls for drawings under its loan agreement once it no longer has a BoP need. The encashment regime would require a sufficiently large pool of economically strong lenders.

92. DA claims. Lenders could count their claims on the DA as international reserve assets, on the basis of the proposed investment strategy and credit risk management framework.

  • Asset quality would rely on an appropriate investment strategy, with financial assets held in the DA invested in high-quality investment grade fixed-income instruments with limited risks to the principal value and in line with general practices for investing reserves. In general, the goal of generating investment returns above the SDRi will entail a certain degree of investment risk, especially over the short-term. While the strategy would aim to limit the probability and extent of investment losses over the medium-term, short-term transient losses stemming from mark-to-market volatility could arise but can be tolerated as the strategy would be expected to remain resilient over time and also because a risk exposure of this nature is commonly accepted for reserve assets. In addition, the RST’s comprehensive policy safeguards, an adequately funded RA (which would increase overtime based on lending margins and investment returns), and a supportive multilateral context, including de facto PCS of RST loans, would mitigate the risks to the value of claims on the DA. The value of contributors’ claims on the DA (i.e., deposit principal and attributed cumulative investment earnings net of SDRi remuneration) would be assessed and reported at regular intervals.4344

  • Liquidity of contributors’ claims on the DA would be ensured by investing DA deposit resources in high quality instruments, with sufficient liquidity of the asset pool to allow for encashment of individual contributors’ claims when they experience a BoP/reserve need. A contributor encashing their claim on the DA would commit to reconstituting its deposit once it no longer has a BoP need.

C. Investment Strategy

93. The investment strategy for the RST should support the purposes of the investment assets. Assets in both the DA and RA have a dual role of generating income for RST operations and providing security for RST loans. Maintaining high-quality investments and sufficient liquidity to meet potential withdrawal requirements would be important to support these purposes, including in counting DA claims as international reserves. Minimizing the risk of investment losses over a medium-term horizon would also be prudent in the event of unexpected needs (e.g., encashment or credit losses). To this end, the main component of RST investment assets should be invested in a short duration fixed-income (SDFI) strategy. In addition, a liquidity component which is expected to be relatively small, would be set aside in deposits to cover short-term operational cash requirements. 45 The proposed investment strategy described below is envisaged for Executive Board approval following the establishment of the RST. Similar to the PRGT, the RST Instrument would authorize the Executive Board to adopt guidelines for investment of RST assets in line with the proposed investment strategy described below.

94. An investment strategy similar to the one for the IMF’s own reserves in the Fixed-Income Subaccount (Fl) is envisaged for the SDFI component of RST investment assets.46 This strategy has recently been adopted for a portion of the PRGT portfolio and is well suited to meet the purposes described above. The strategy balances income generation with maintaining reserve asset like properties and is in line with general practices for investing reserves at peer institutions (e.g., central banks and multilateral development banks). The objective of the investment strategy for the SDFI component is to produce returns in excess of the three-month SDR interest rate (SDRi) by an average of around 50 basis points, while minimizing the frequency and extent of negative returns and underperformance over an investment horizon of three to four years. Eligible investments include a range of high-quality investment grade fixed-income instruments such as government and government agency obligations as well as credit instruments such as corporate bonds and asset backed securities. Under this strategy, the majority of the Fl portfolio has been actively managed in short duration diversified fixed-income instruments in which external asset managers can adjust risk exposure to duration and credit assets according to market conditions. Staff envisages that RST investments in the SDFI component would be similarly managed. The strategy for investing its reserves in the Fl has served the IMF well. It has a proven track record for generating consistent and meaningful returns in excess of the SDRi. The resilience of the strategy has been supported by the diversification of risk premia.

95. The proposed investment strategy for the SDFI component is well positioned to achieve a moderate margin above the SDRi over a three-to-four-year investment horizon. As with other investments of reserve assets, there will be fluctuations in the market value of the portfolio. The proposed strategy is appropriately conservative to limit excessive volatility and provide sufficient liquidity to meet withdrawal needs. The proposed strategy also has a relatively high probability of achieving a positive margin above the SDRi over a three-to-four-year horizon with an average expected margin of around 50 basis points over time. While the strategy has a relatively high probability of achieving a positive margin above the SDRi over shorter periods, this probability would be reduced under certain stress scenarios, for example, if interest rates rise faster than currently anticipated.47 However, under such a scenario, higher yield levels and an upward sloping curve would benefit returns over time allowing mark-to-market losses to be recovered.

96. Pooling of RST investment assets (RA and DA in one pool) with those of other trusts, in particular the PRGT, would facilitate timely implementation of the investment strategy and limit investment costs. Pooling of investment assets of accounts within a trust is already current practice. In addition, pooling of RST investment assets with the PRGT would allow the RST to benefit from the existing investment arrangements in place for the PRGT—limiting setup time and costs. Pooling of trust assets for investment purposes is a common practice followed by several multilateral development banks.48 If pooling was not possible, setting up separate investment arrangements for the RST could significantly delay implementation of the investment strategy. Pooled investments would also incur significantly lower investment costs over time than for separately managed accounts due to economies of scale. Pooling of assets would not be expected to have any material implications for PRGT contributors or PRGT investment operations. Under the current record keeping arrangements, detailed attribution is made of the share of each account in the investments, and this could be expanded for pooled investment across trusts.49

97. To enable pooling of RST assets with PRGT assets, an amendment to the PRGT Instrument is proposed, requiring the consent of subsidy account contributors. With respect to the PRGT, the provision on separation of trust assets under current PRGT Trust Instrument precludes pooling of investments across trusts. PRGT Instrument, Section VII, Paragraph 2 is among the so-called protected provisions of the PRGT Instrument. Protected provision can be amended but require the consent of affected contributors.50 Amending the PRGT to allow for the pooling of PRGT assets with other trusts for investment affects all current contributors to the subsidy accounts of the Fund whose contributions would be invested together with RST resources and possibly those of other trusts under such pooling arrangements. There is a total of 154 current PRGT subsidy account contributors whose consent would be required to make the proposed amendment effective.

98. Staff propose to seek consent from contributors to the PRGT subsidy accounts for the proposed amendment under a process that would allow for lapse of time consent for contributors that do not respond within the prescribed deadlines for consent (or an opportunity to be bought out). First, following approval of the proposed amendment by the Executive Board, a communication from the Fund would be sent to contributors to seek their explicit consent to the amendment within six weeks. Countries that do not wish to consent to the proposed amendment should notify the Fund in writing within the response windows. The objection by a contributor does not mean that the amendment could not become effective; rather, and as described in the MDRI context in 2005, the dissenting contributor could request back its subsidy contribution.51 52 After the expiration of the first deadline, staff would send a second formal communication to the contributor requesting a response within a further period of four weeks. The second communication would indicate that if no response is received by the second deadline, the contributor is assumed to have consented to the amendment. Fund staff would follow up bilaterally with authorities to ensure that a response is received within the second deadline. If no response is received within the second deadline, the consent is assumed for purposes of the amendment process. It is proposed that the contributor would be given six months after the amendment becomes effective to request the remaining share of its contribution to the PRGT subsidy accounts back if during this period it sent a communication to the Fund objecting to the amendment. Contributors would be given clear communications about this additional six-month period and its implications. In the event of any unexpected difficulties in securing the expected high level of contributor consents, management could return to the Executive Board prior to the expiration of the second deadline regarding the proposed amendments, including to extend the consent period for a further limited time to allow for engagement with PRGT subsidy contributors that have not yet responded, which is independent from the RST related decisions.

99. The proposed approach, while new in the context of PRGT amendments, is legally permissible as it appropriately protects the interests of affected PRGT contributors to have an opportunity to reject a proposed amendment while also taking into account the interests of other contributors and PRGT beneficiaries to have operations continue under an amended instrument. The PRGT Instrument is silent on the modalities for consent. While affirmative consent has been sought in the past, and will continue to be followed as the first step for the proposed amendment, adding a lapse of time element along with strong notice provisions and a significant period within which to request a buyout for creditors that do not respond provides appropriate safeguards to protect creditors’ interests. The underlying rationale of the amendment process for protected provisions is that a contributor can request its resources back if it does not consent to an amendment; the creditor does not have a veto over the amendment process. The lapse of time element for contributors that do not respond allows the Trustee to bring closure to the amendment process and avoids giving individual contributors a veto by not responding. Giving a contributor that does not respond within the deadlines an additional six months after the amendment has become effective to request its contribution back, offers a final opportunity for the contributor to express its objection and in this case request its contribution back.53

D. Making the RST Operational

100. To commence lending operations and therefore become operational, the RST would need a critical mass of resources from a sufficiently broad base of contributors with strong external positions. Once the RST is formally established, there will be an interim period before lending operations can start, with time needed to negotiate and make effective contribution agreements, set up new financial systems, and engage with members on UCT-quality and RST-supported policies. In this interim period, the Trust will be in a position to accept and hold RA and DA contributions and enter into loan agreements prior to the start of RST lending operations. Two general conditions would be particularly important for RST lending operations to commence: (i) sufficient loan resources in nominal terms to meet initial demand for RST resources in line with the envisaged access policy, and (ii) a sufficiently broad pool of participating RST contributors to ensure encashability of LA claims. It is proposed that lending operations commence once the Managing Director has notified the Executive Board that these two conditions are met and that sufficiently robust financial systems and processes are in place.

101. Fundraising would be initiated immediately after the establishment of the RST. The overall fundraising strategy would seek broad contributor participation and aim to secure the necessary resources for the operationalization of the Trust. As the bulk of contributions would likely be in SDRs, the initial fundraising will focus on a group of countries that have historically participated in other Fund financing initiatives (i.e., PRGT, NAB, BBAs) and participate in the VTA market (a total of 26 countries). Voluntary contributions from other VTA participants are also encouraged to help strengthen the RST financial capacity given the tight implementation timetable.

102. In addition, staff will implement in the interim period the necessary adjustments to the Fund’s financial systems and processes to ensure smooth RST operations, including appropriate internal controls and reporting. Staff will leverage existing Fund systems and expertise to implement the financial infrastructure for the RST as soon as possible. The implementation of changes needed to support a new Trust is, however, a significant and resource-intensive undertaking. Careful consideration will be given to ensure that system configurations will adequately safeguard the financial resources of the RST and its contributors. An assessment of system readiness will precede the commencement of RST lending operations.

Section VI. Demand Projections and Financing Scenarios

103. Demand for RST loans is estimated at around SDR 22 billion (US$31 billion) under a baseline scenario. The scenario is grounded on the proposed income-based eligibility (see Section III.B. Given the qualification criterion of countries requesting RST also having a concurrent on-track UCT-quality Fund-supported program, demand is calibrated using the sample of countries that had such a program in the past 10 years (70 countries of a total of 143 eligible countries). The baseline demand projection assumes that individual access would be 100 percent of quota for most countries,54 subject to a maximum nominal cap of SDR 1 billion.

104. A large number of DCs and small states are expected to request RST financing, reflecting their significant longer-term vulnerabilities and history of IMF program engagement. In staffs demand simulation, just under half of the 70 RST borrowers would be non-blend PRGT eligible countries (Group A see Section III.C on Lending Terms); about one-fifth would be presumed blenders and non-PRGT eligible small states with income below 10 times the IDA operational income cutoff (Group B), and over one-third would be non-PRGT eligible MICs and higher-income small states (Group C). In absolute terms, about SDR 10 billion of baseline demand would come from Groups A and B, and about SDR 12 billion from Group C (as MICs have on average larger quotas in absolute terms compared to LICs).

105. These baseline demand projections reflect a set of simplifying assumptions and are subject to significant uncertainty. Such uncertainty stems from the actual number of countries that would seek RST support, their economic/quota size, eligibility and qualification requirements (including having a UCT arrangement in place), and individual access levels, which in turn would depend on the strength of policies, debt vulnerabilities, and capacity to repay the Fund. Demand would also be affected by RST access policies, which may be subject to review over time. Given the high uncertainty, the financial model below also considers a high demand scenario at 25 percent above the baseline (SDR 27.4 billion) and a low demand scenario at 25 percent below the baseline (SDR 16.5 billion).

106. A recent survey of eligible members points to frontloaded near-term demand and suggests that the above projections are plausible for the medium term. Survey results suggest that there is significant interest in accessing the RST (Box 3). Based on 93 responses, 73 countries (of which 43 are PRGT eligible countries) have expressed some form of interest in accessing the RST with 15 being small states. Most countries are interested in accessing the RST to support reforms targeting climate change (69 countries) and pandemic preparedness (64 countries). Countries mostly favor accessing the RST in the context of a disbursing UCT program. Of the 73 interested respondents, 39 expressed interest in accessing the RST once it is operational (late 2022), with another 29 preferring “whenever the need for a traditional IMF program arises”, with 5 choosing 2023. Assuming access of 100 percent of quota for most countries55 and an SDR 1 billion cap, the total resources needed to meet the near-term demand of interested countries would be around SDR 21 billion, with around SDR 11.7 billion allocated to group C countries. Economic repercussions from the war in Ukraine could lead to additional requests for UCT programs in the near term which could have knock-on effects on demand for RSF arrangements.

RST Survey Response by Prospective RST-Eligible Members

uA001fig02
Notes: (i) Group A countries are all PRGT-eligible countries that are not presumed blenders. Group B includes all “presumed blenders” (of PRGT and GRA resources) and all small states with per capita GNI below ten times the I DA cutoff. Group C includes all other RST-eligible countries, (ii) As under baseline demand projections, access is assumed at 100 percent of quota. Countries whose quota is higher than 3 percent of their nominal GDP are assumed to have an access level of 50 percent of quota.

107. Staff has developed a financial model to analyze resource needs and the evolution of financial buffers, based on medium-term demand projections and the proposed financial structure (Annex V). In addition to demand, total upfront resources needed for the RST will depend on the desired size of the encashment buffer, as well as the RA and DA contributions. Consistent with the proposed financial structure above, staff assumes that loan resources would include an encashment buffer of 20 percent (as in the PRGT), implying the need to secure a total of SDR 27 billion in voluntary loan contributions. In addition, contributors would include an upfront reserve injection (in SDRs or currency) of 2 percent and an upfront long -term deposit of 20 percent relative to their loan commitments. Based on the estimate of total demand under the baseline, this implies that the RST would need about SDR 33 billion in total resources to be provided in its first five years. This would provide the RST with an initial RA balance of SDR 0.54 billion, and a starting DA balance of SDR 5.4 billion.

108. The financial structure supports a substantial build-up of reserves before RST loan repayments begin and adequate reserve coverage throughout the projection period. Gross and net reserves are initially modest, but adequate in light of the absence of any loan amortization over the first ten years of the Trust’s life and the very modest interest payments obligations. Over this time, gross reserves increase from SDR 6.0 billion to SDR 7.4 billion, and net reserves quadruple, from SDR 0.54 billion to SDR 2.0 billion. By the time loan amortization payments start coming due, gross reserves are expected to cover at least 35 percent of credit outstanding (between years 11 and 25), and at least 3 times annual debt service to loan resource contributors. Net reserves would cover at least 10 percent of credit outstanding and almost one year of debt service over the same period.

Figure 2.
Figure 2.

RST Reserves Accumulation: Combined Adverse Events

(In SDR million)

Citation: Policy Papers 2022, 013; 10.5089/9798400207730.007.A001

109. Cash flow stress tests show that reserves remain adequate under various adverse events. Staff conducted stress tests on the baseline financial model assuming lower investment returns, temporary arrears, permanent credit losses, encashment, and a combination of adverse events (see Annex V for the detailed results). Additional sensitivity analysis was also conducted to look at the effects of lower borrowing rates, defaults by large borrowers, and higher demand. The two worst scenarios tested come from the combined shock and simultaneous defaults by two of the largest borrowers. The analysis suggests that even under extreme stress scenarios it is highly unlikely that tapping the principal value of the DA would become necessary. Under the combination of adverse events, minimum gross reserves during the repayment period remain above 32 percent of credit outstanding, and over 2.5 times of annual debt service to loan contributors. The minimum level of net reserves during the repayment period would be SDR 1.5 billion, enough to cover 71 percent of annual debt service and 7 percent of credit outstanding.

Section Vii. Voluntary Trading Arrangements

110. A smooth functioning of the voluntary SDR trading market will be critical for effective SDR channeling through the RST. The liquidity of the SDR has for over thirty years been supported by voluntary trading arrangements (VTAs).56 VTAs are arrangements between the Fund and participants in the SDR Department and prescribed holders who have agreed to buy and sell SDRs within set trading limits.57 An increase in SDR channeling will likely lead to an increased volume of SDR transactions through the VTA market. Consistent with the current practice under the PRGT, there is a strong expectation that SDR contributors to the RST would have VTAs that could support the conversion of channeled SDRs into currencies in order to ensure broad distribution of SDR transactions among VTA participants. A well-functioning voluntary market would be key to ensuring that SDRs channeled to the RST can be converted into freely usable currencies as needed, and VTA members should therefore be prepared to participate more often in transactions and hold more SDR-denominated assets. Staff have recently engaged with current and prospective VTA participants with a view to expanding and strengthening the voluntary market (see Box 4)

111. The impact of SDR channeling via the RST on the VTAs will depend on several factors:

  • Total amount of SDRs channeled to the RST. Based on staffs initial assessment of projected demand for RST financing and implied contributions to cover Loan Resources and the Deposit Account (see Section VI), total SDR channeling to the RST could be about SDR 33 billion, assuming all projected contributions are made in SDRs. However, contributions to the RST can also be made in freely usable currencies, which would not impact the voluntary SDR market.

  • SDRs converted as a result of lending to borrowers. SDRs provided to the RST would be subsequently used to finance lending, of which a portion may be converted, either by the Trust or the borrower, into freely usable currencies to address borrowers’ BoP needs.58 Given the proposed pass-through nature of borrowing and lending in the RST and the link of RST lending to Fund-supported programs, the impact of SDR lending on the VTAs would be gradual over-time as the Trust draws on lenders' loan commitments to finance disbursements to RST borrowers. Staff’s baseline projections assume about SDR -5 billion in RST lending per year starting in late 2022, although the amounts and timing are subject to considerable uncertainty.

  • Channeling of SDR for investment purposes. Channeling of SDRs to the RST Deposit Account would have an early impact on the VTA market, because those SDRs would need to be converted into currencies for investment purposes. These amounts could be around SDR 5 billion during or shortly after the envisaged fundraising phase (similar to projected annual lending). As a result, an active period of VTA exchanges could be expected as the Deposit Account is initially funded in the RST.

Efforts to Expand and Strengthen the Voluntary Market

In advance of the General SDR Allocation in August 2021, Executive Directors emphasized that a strong voluntary SDR market was critical to support an expected increase in the volume of SDR transactions. Accordingly, in late-Spring 2021, staff initiated a two-pronged approach to strengthen and expand the voluntary market (see IMF Policy Paper No. 2021/049 Proposal for a General Allocation of Special Drawing Rights).

The approach for expanding and strengthening the voluntary market involved: (i) engagement with potential new VTA participants focusing primarily on members with strong external positions in the Financial Transactions Plan that did not have a VTA at that time (twenty-three members); and (ii) requesting existing VTA participants to provide additional operational flexibility under their VTAs. Specifically, staff sought VTA revisions in areas such as transactional limits, trading currencies, and notification periods.

Following extensive outreach with the Offices of the Executive Directors, the European Central Bank as a prescribed holder with a VTA, and member authorities, substantial progress has been made in expanding and strengthening the voluntary trading market:

  • As of mid-March 2022, six new VTAs have been agreed and become effective, increasing the number of VTA participants from thirty-two to thirty-eight. The new participants—Algeria, Brazil, Estonia (Republic of), Lithuania (Republic of), Oman, and Singapore, together with one other prospective participant (Luxembourg) in advanced stages of completing domestic procedures, would add about SDR 10 billion to the absorptive capacity of the voluntary market.

  • On operational flexibility, many participants scaled up their transactional limits and a few reduced notification period requirements. Some participants provided additional currency flexibility under their VTAs.

Staff will continue to engage with IMF members on possible steps to further expand and strengthen the voluntary market and calls for additional members to sign VTAs.

112. It is expected that SDRs channeled to the RST could be absorbed by a well-functioning voluntary market. The scaled up absorptive capacity of the voluntary market following the General Allocation stands at about SDR 242 billion as at end -February 2022 and is expected to be sufficient for future exchanges of SDRs, should the full amount of the estimated channeling of SDRs to the RST (SDR 33 billion) be converted through the VTAs.

Section VIII. Administrative Expenses

113. The implementation of the RST would result in new gross demands for staff resources that will require additional gross budgetary resources in the “steady state.” The RST would be a separate entity from the GRA and PRGT and additional expenses would be incurred to support Trust management, including Trust set up and ongoing Trust management. Set-up costs include the already ongoing development of the final design of the RST, as well as fundraising, IT system configuration, and the establishment of policies and guidance to govern lending under the Trust and its administration. The related costs have been funded in FY22 through reprioritization and extensive staff overtime and will be funded in FY 23 with a combination of temporary and permanent resources allocated to departments. Recurring “steady state” costs in the medium term would include:

  • Trust management: Execution of financial transactions under the Trust, reporting (including preparation of financial statements), external audit, resource mobilization and management, investment activity, financial and operational risk management, and system support. This also covers periodic and ad-hoc reviews of the RST lending policies and finances, and review and adjustments to policies related to financial terms and safeguards for the Trust’s resources, including its multilayered risk management framework.

  • Operations: Country program design and review, including economic policy analysis and supporting analytical work where needed in collaboration and coordination with the World Bank and other relevant organizations, together with policy discussions with the authorities, and the monitoring of program implementation including assessments of reform target completion. It also includes the aspects of periodic policy review beyond financial terms and safeguards. Costs in these areas could be frontloaded initially as area and functional departments establish new processes with other institutions, upskill (e.g., in the areas of climate and pandemic preparedness) and support possible pent-up demand for RST resources. The linkage of RST access to UCT-quality programs would allow for some cost synergies by drawing on the existing resource base in several areas and avoiding duplication (e.g., mission travel, program documentation, country review process, and Board approvals).

114. Total gross incremental costs are composed of direct costs (primarily labor) and overhead costs (such as additional office space requirements and additional overhead needs). Based on the demand outlook in Annex V, staff assumes that, in the “steady-state”, there could be 33 active RSF programs on average each year. On this basis, staff estimates that in the medium term, incremental direct costs for Operations would be in the range of about $9–10 million, or 27–29 FTEs annually. The cost of Trust Management, including all financial, policy, and related legal review needed to operate the Trust and safeguard its resources is estimated to be $5–6 million, or 16–18 FTEs annually. These cost estimates are subject to considerable uncertainty arising from a number of factors. In particular, actual costs would vary depending on the number of countries that will seek RST support. Also, expanding the range of long -term structural challenges that will qualify for RSF support going forward would require and/or draw on expertise across a range of new areas, in cooperation with partners, and would require continued review of cost impacts, especially for issues on the frontier of Fund work. As a preliminary estimation, using the direct cost estimates discussed above, and assuming that the RST would have broadly similar ratios of incremental overhead to direct costs as other expenditures, annual fully-loaded gross costs of the RST could be about $18 million.

115. In terms of budget impact, staff proposes a management fee for trust management activities, similar to that in place for CD-related trust funds. The resources would be categorized as receipts in the budget and therefore affect the gross administrative budget, but not the net administrative budget. The fee would be initially set at a fixed scale, linked to the estimated cost for trust management activities in relevant departments (FIN, LEG, and SPR) and drawn from the Reserve Account. Estimates for actual trust management-related costs would also be reported annually as part of the budget outturn report, and the scale and structure of the management fee reviewed at that time. Operational costs will be funded through the regular budget process through reprioritization, with the largest impact expected in area departments, ensuring that individual departments receive adequate incremental resources as needed. The overall budgetary impact of RST-related activities will be monitored with a review of relevant arrangements after sufficient experience is gained.

116. The proposed RST instrument also includes a provision that the GRA be reimbursed annually for the cost of administering the RST based on reasonable estimates of gross incremental costs with a view to ensuring cost recovery, consistent with the RST’s design as a self-sufficient trust. With the institution of a management fee, the scale of the reimbursement to income would be calculated to exclude costs covered by the fee. Total costs of administering the RST will be drawn from Reserve Account for: a) budget receipt covering trust management; and b) income reimbursement covering all other costs, i.e., operations. The annual income paper would report staff estimates of RST costs that would be reimbursed to income. It would be expected that decisions on annual reimbursements of these costs would be taken by the Board in the context of the Fund’s annual income paper starting from FY24, as is currently done for other reimbursable expenses.

Section Ix. Enterprise Risks

117. Potential enterprise risks span four key areas: business operations, finances, human resources, and the Fund’s reputation.

  • Business risks. These would include the possibility that the RST’s longer-term lending could duplicate other IFIs’ work unless it is coordinated and well-managed. Moreover, granting large and frontloaded access under the RST could delay reforms agreed under the UCT program. Additionally, greater focus on issues outside of the Fund’s core areas of responsibility could challenge the Fund’s ability to develop RSF programs and to monitor conditionality, and create mission creep which could distract from the Fund’s core lending operations.

  • Financial risks include credit, liquidity, and investment risks. Credit risk could stem from developments that erode repayment capacity, such as large non-restructurable debt or a sharp increase in the SDRi and global borrowing rates, which would worsen borrowers’ creditworthiness. Cross subsidization across borrower income groups would also require a diverse borrower group, which could be a challenge should demand be dominated by Group A countries. High demand and a high volume of transactions could put pressures on the VTA market. Additionally, a severe market shock could deplete RST Reserve Account investments. These risks have been carefully considered and are addressed in Sections V and VII.

  • Human resource risks stem from inadequate resourcing that could strain already overextended staff, including non-fungible experts. Planned skill upgrading in emerging areas may not be sufficiently rapid to meet RST demand.

  • Reputational risks arise from potential low demand amid high expectations, and an inability to mobilize sufficient resources to magnify the overall global impact of the SDR allocation. Alternatively, excess demand could challenge resources even if fundraising rounds meet expectations. Finally, country authorities could reverse their RSF-supported policies after receiving a disbursement, or implement measures aggravating risks from qualifying long -term challenges in areas that fall outside of scope of RSF conditionality. Although this risk exists in any program, it is more acute with an RSF given the long-time horizon over which the benefits of reforms are realized.

Policy and design modalities, the financial architecture, and other safeguards have been specified to mitigate these risks and will need to be continually assessed and modified as appropriate (Table 1).

Table 1.

RST Enterprise Risks and Mitigation

article image

Annex I. List Proposed of RST-Eligible Members

article image
article image
Notes: GNI data based on data current as of October 2021. Andorra, San Marino, and Venezuela excluded due to missing GNI data.

Annex II. Comparison of the Key Features of the Extended Fund Facility, Extended Credit Facility, World Bank Development Policy Financing, and Proposed Resilience and Sustainability Facility and Respective Concessionality Rates

Annex II. Table 1.

Key features of the Resilience Sustainability Facility and selected IMF/World Bank Financing Instruments

article image
Annex II. Table 2.

Concessionality of Resilience Sustainability Trust and selected IMF/World Bank Financing Instruments1

(In percent, unless otherwise indicated; as of February 22, 2022)

article image
Source: staff estimates.

A unified discount rate of 5 percent is used to calculate the concessionality of individual loans.

Current interest rates.

Concessionality calculations for IMF and IBRD lending are based on the assumed SDR and LIBOR rates remaining constant for the life of the IMF or IBRD loan, given the variable rate nature of the loans. According to the PRGT interest rate mechanism last approved by the IMF’s Executive Board in July 2021, for SDR rate < 2, ECF rate is zero; while for 2 < SDR rate < 5, ECF interest rate equals 0.25 percent. For EFF, interest rate is SDRi plus 100 basis points, surcharge is 200 basis points for outstanding credits above 187.5 percent of quota (zero for outstanding credits below 187.5 percent of quota) assuming timely repayment, commitment fee is zero assuming the amount is fully drawn, and service charge of 0.5 percent on amount drawn. For IBRD loans’ current rate is 6-month USD LIBORas of November4, 2021, 0.21325 percent, plus respective margins, for future rates assuming LIBOR equals SDRi and, therefore, interest rates equal SDRi plus respective margins. For IDA loans, IDA fixed SDR rates, 0.75 for regular loans and loans to small economies and 2 percent for blended loans, have remained unchanged since their introduction; IDA’S financial planning is premised on holding these rates unchanged.

EFF for outstanding credit < 187.5 percent of quota.

Concessionality is a weighted average of grant elements: 1/3 ECF and 2/3 EFF.

For RST Group A, interest rate SDRi plus 55 basis points is capped at 2.25 percent, i.e., at SDRi = 2 percent and SDRi = 3 percent, the interest rate remains at 2.25 percent; service charge is zero percent. For RST Group B and C, interest rate is SDRi plus 75 basis points and 95 basis points and service charges 0.25 percent and 0.50 percent, respectively.

IDA loans comprise three groups of lending with different terms: (i) regular loans as illustrated in the table, (ii) loans for small economies, and (iii) loans for blending countries, also illustrated, with concessionality of 53.26 percent, 60.62 percent, and 35.18 percent, respectively. IDA also offers grants to some of its client countries.

There a re four IBRD country groups (A, B, C, and D) with different margins added to 6-month USD SOFR repayment terms a re flexible and must fall within the policy limits of 35 years maximum final maturity (including the grace period) and 20 years maximum average repayment maturity. For illustrative purposes, table shows estimates for loans to IBRD Group B with 20 years maturity including 10 years of grace period and two equal repayments per annum; the front-end fee is 0.25 percent, while commitment fee is assumed zero. Margin for Group B is 1.35 percent, while for Groups A, C, and D the respective margins are 1.15 percent, 1.55 percent, and 1.80 percent.

Annex III. Broad Principles for Bank-Fund Coordination in RST Operations

  • 1. The proposed new IMF-administered Resilience and Sustainability Trust (RST) focuses on supporting policy reforms to help build resilience and sustainability by addressing long-term structural challenges that pose risks to eligible members’ prospective BoP stability (e.g., climate change and pandemic preparedness) and that may also affect the members’ actual, potential, or protracted BoP problems.

  • 2. RST lending operations would greatly benefit from close coordination with the World Bank, and potentially also with other multilateral development banks (MDBs), given these institutions’ expertise and experience in the areas to be supported by RST lending. This note summarizes guiding principles and a proposed practical approach for such coordination. These principles are based on the experience with coordination on PRGT/GRA-supported programs and Bank programs and are informed by discussions between the staff of the two institutions. The principles and key steps in this note are specific to Bank-Fund coordination on RST-related issue.1

  • 3. This note focuses on climate change, as there is ongoing work in the two institutions to further mainstream climate strategy. Modalities of engagement discussed in this note may be carried over to other purposes of the RST that may be added, if these purposes are under Bank expertise, although customization to the specific features of those purposes maybe needed.

Guiding Principles

4. In line with the Concordat, coordination between Bank and Fund staff will be guided by principles that reflect each institutions’ respective mandate and expertise. The guiding principles below build on a long experience of close coordination on Fund and Bank-supported programs at the country level.

  • RST measures would be informed and expected to be consistent with country diagnostics developed in both institutions relevant to the RST’s purposes. On climate change, the Bank’s Country Climate and Development Reports (CCDR), if available, will be a critical input, complemented with other products such as the Fund’s Climate Change Policy Assessments (CCPAs) and its potential successor instrument, Climate Macroeconomic Assessments Programs (CMAPs). In practice, Bank and Fund staff will coordinate the production of CCDRs and CMAPs to complement and ensure consistent advice between the two products for member countries, in line with the agreed coordination between Bank and Fund staff on CCDRs and CMAPs. Fund staff are expected to discuss with their Bank counterparts areas of the CCDR or other diagnostics that they intend to include in the RST program to ensure complementarity. In instances where countries already have an advanced climate change framework, Fund staff could use these inputs flexibly as part of the analytics informing the RST program.

  • RST operations will focus on areas of Fund expertise, complemented by Bank expertise as needed on a case-by-case basis. The intention of the coordination is to seek strong complementarity with the Bank’s policy and development financing and to best leverage the expertise in both institutions for the best policy advice to the member countries. At the same time, the governance of the RST is one that is fully Fund-administered, requiring all RST conditionality measures to be designed, monitored, and assessed by the Fund.

  • The purpose of RST lending would be to support structural macro policy reforms that enhance long-term resilience and sustainability, and thereby improve prospective BOP stability. The Bank will continue to take the lead on development prospects and related policies and investments.

Possible Key Steps

5. The key areas identified for coordination are: (i) Diagnostics; (ii) Policy priorities; (iii) Conditionality; (iv) Implementation supervision; and (v) Program documentation (i.e., a vehicle to document the collaborative approach for RST lending).

I. Diagnostics: Fund staff will draw on country diagnostics to assess the need for an RST-supported program in line with the Fund’s mandate and to identify policies that could be supported by an RST.

Specifically on climate, Fund staff should, if possible, draw on CCDRs, a new product that is being rolled out in over 30 countries in FY22 and will be further rolled out to other countries. The CCDR— the Bank’s new core diagnostic tool to help countries align climate action and development efforts — synthesizes country authorities’ and other climate-related work, deepens the analytics, and provides policy recommendations. The CCDR will be a critical tool for diagnosing issues and identifying policy priorities in climate-related RST-supported programs. As per the agreement between Fund and Bank staffs on CCDR/CMAP coordination, CCDR teams are expected to engage with the respective Fund country teams, and CMAP teams are expected to engage with the respective Bank country teams and relevant Bank global practices/themes, including sharing early drafts of preparatory notes and reports.

The Fund’s capacity development (CD) in climate-related work—including on CCPAs/CMAPs, green public financial management, the new climate change module of the Public Investment Management Assessment, carbon taxation, subsidy reform, etc.—would also be important diagnostic inputs into the team’s assessments. CMAPs are expected to gradually become available for more countries, though many fewer than the CCDR. Fund surveillance products, such as Article IV and Selected Issues Papers as well as other Fund analytical work, would also inform RST policy design, as would national climate plans or similar strategies/documents (e.g., Disaster Resilience Strategies).2

II. Policy priorities: Based on the diagnostics, Fund staff will identify policies that the member country would need to prioritize under the RST. Fund RST teams will coordinate with their Bank counterparts the policy actions to be included in the RST program to promote complementarity with Bank-supported programs. Not all policy priorities will become RST conditionality; some of the priorities could become MEFP commitments by the authorities in their concurrent UCT-quality program. In countries where there is an active or planned Bank DPO or climate program, staff will ensure overall policy advice consistency and complementarity between the RST and DPO on climate policies and delineate areas of responsibility according to each institution’s mandate and role. Fund staff will seek to ensure that overall policy reform requirements do not overload implementation capacity constraints.

III. Conditionality: RST “reform measures” would satisfy requirements under the Fund’s Guidelines on Conditionality. In program design, IMF staff will focus conditionality in areas of Fund expertise (orjoint Fund-Bank expertise), such as fiscal, monetary, and financial issues, as well as statistical and modeling capacity. However, there may be some cases where other macro-critical areas arising from areas of Bank expertise may also become subject to RST conditionality (similar to a social safety net reform rising to being a macro-critical structural benchmark in a PRGT/GRA-supported program). In these cases, IMF staff would draw the conditionality from existing Bank analysis/advice, in consultation with the Bank. The Fund’s Executive Board will establish RST conditionality and assess whether an RST condition is met or not (consistent with the existing policy of not having cross-conditionality), drawing to the fullest extent possible on the advice of Bank staff in areas of the Bank’s expertise.3

IV. Board documentation: Fund staff will prepare RST documentation based on their assessments of the RST-supported programs (including justification of the need and appropriateness of an RST-supported program; diagnostics of key priorities and needed reforms; current climate strategy/policies; climate policies going forward, including any RST-supported reform measures; and, for reviews, assessment of policy implementation and commitments). The staff report and MEFP will document the coordination with the Bank, including a description of the Bank’s engagement in the sector and/or any Bank technical assistance related to RST-supported reforms. Bank staff will provide an assessment letter on the authorities’ climate-related policies. Timely provision of assessment letters will be important, as with Fund staff assessment letters provided for Bank operations.4 The staff report would also acknowledge Bank staff’s inputs into the RST in the cover note.

V. Modalities/process of engagement: In developing RST-supported programs, Fund staff are expected to coordinate with Bank staff (and/or other MDB staff as relevant, in line with existing modalities) in programs that include RST-supported reforms. Such coordination would be subject to appropriate adjustments to handle confidentiality and/or operational considerations at both the Bank and the Fund.

Specifically, to promote coordination, Fund staff are expected to:

  • Share Fund diagnostic or analytical work on RST-related issues (e.g., CMAP, Selected Issues papers) with Bank staff.

  • Engage early in the process with Bank staff on RST reforms by, for example, discussing/brainstorming on RST policy measures.

  • Incorporate the outcome of this engagement into the RST/UCT arrangement Policy Note.

  • Draw significantly on the CCDR, if available, and any Fund CD or other diagnostics for climate-related RST reform measures.

  • Periodically share with Bank staff an advance calendar of expected RST operations and tentative document circulation timelines.

  • Circulate—with a 5-day turn around—all RST-related background and conditionality sections of the Policy Note, Staff Report, and Letter of Intent/Memorandum of Economic and Financial Policies to Bank staff for comments.

  • Subject to adjustments needed for confidentiality or other operational reasons, invite Bank staff to comment on the Policy Note, or relevant parts thereof, and/or attend either a unified UCT arrangement/RST Policy Consultation Meeting (PCM) or a separate RST-only PCM.

  • When necessary and feasible, participate in each other’s mission meetings on RST-related issues.

  • Support ongoing engagement for each review to ensure continued “note sharing” across institutions.

As per agreement with the Bank, Fund staff can expect Bank staff to:

  • Engage Fund counterparts in the preparation of CCDRs, particularly in macro-fiscal areas.

  • Engage with Fund staff in early discussion/brainstorming on the design of RST-supported policies.

  • Provide comments on draft documents prepared by Fund staff and join Fund mission meetings as appropriate and feasible.

  • Prepare, upon request by the Fund Mission Chief to the Bank’s Country Director, an Assessment Letter, in line with agreed timelines.

  • Support ongoing engagement at all steps (CCDR preparation; DPO related to climate change; RST preparation; RST program period) to ensure continued “note sharing” across institution.

Annex IV. Financial Architecture of the RST

This Annex builds on the discussion in the main body of the paper and provides further detail on the key features of the various accounts and the underlying principles.

A. Guiding Principles

1. The RST is a loan-based trust administered by the IMF, with a financial structure broadly similar to that of the PRGT. In particular, and similar to the PRGT, RST resources would be mobilized based on voluntary contributions from members, including those wishing to channel SDRs for the benefit of low-income and vulnerable middle-income members. The RST would in turn on-lend these resources to eligible borrowers via the Loan Account (LA) at a moderate interest rate, covering funding and administrative costs and contributing to reserve build-up. The RST would have a Reserve Account (RA), along with a Deposit Account (DA) to build reserves overtime and backstop the RA in unforeseen extreme tail risk events. The overall size of the Trust is initially targeted at SDR 33 billion, with a pre-specified, uniform drawdown period for loan disbursements through November 30, 2030 and a maximum maturity of 20 years for loan repayments, which start IOV2 years after a disbursement.1

2. The proposed financial structure is designed to build and maintain adequate reserve buffers for this financing envelope and maturity horizon. Gross reserves are initially funded through modest upfront contributions to the RA and long-term deposits to the DA. Net reserves, defined as gross reserves net of contributors’ long-term deposit claims on the DA are expected to grow over time through lending and investment income, net of RST administrative costs, such that reserves are expected to be significantly higher by the time RST loan repayments come due. The level of reserves would be reviewed periodically together with the structure of interest rates and service charges, with a view to ensuring an adequate degree of reserve coverage.

3. A broad-based initial loan mobilization round would aim to secure resource commitments sufficient to meet expected demand until 2030. While the current architecture assumes a one-time resource mobilization, resources could be mobilized periodically, based on the decision of the Board, the availability of Trust resources, and expected demand. Contributors would be free to decide on their participation in any loan mobilization round, including on the size of their commitment. In cases where new loan resources would be mobilized, the Fund would review the adequacy of the RST financial structure, including reserve adequacy to support future lending, and in that context modifications of the modalities set out below could be adopted, subject to the provisions on amending the RST Instrument.

B. Contributions -General Considerations

4. The RST Instrument would set out the key financing terms and conditions for contributions to the RST, including the key rights and obligations of contributors. The individual contribution agreements with each contributor would reflect these terms and specify only a few additional aspects, in particular the contribution amounts. While some tailoring of individual agreements will be possible, in line with the practice for the PRGT, for example regarding media of payment, the key terms would be uniform across contributors. Consistent with the approach under the PRGT, the Managing Director would be authorized to enter into individual contribution agreements. Once finalized, the agreements would be circulated to the Executive Board for information—as is the case for the PRGT.

5. Each LA contributor would sign a “Contribution Package” with three parts—an LA contribution (borrowing agreement); an RA contribution; and a DA contribution.2 Contributions to the RA and DA would be a minimum of 2 percent and 20 percent, respectively, of a contributor’s commitment to the LA. Resources committed under a borrowing agreement with the LA are drawn as needed during the pre-specified uniform drawdown period. In contrast, DA and RA contributions are provided to the RST upfront.

6. The Trust may also receive “standalone” contributions to the RA and the DA unrelated to LA contributions. Such contributions could come from contributors that are not in a position to make LA contributions but that would want to contribute to either or both of the RA and DA. These contributions would be in the interest of the Trust as they would bolster reserves at the outset and raise Trust income from investments which adds to net reserves overtime. Standalone contributions would have a fixed maturity, at an investment duration consistent with reliably strengthening Trust reserve accumulation, where a maturity of 10 years would be preferable.3

7. The SDR would be the unit of account for RST financial reporting, and all commitment of resources to the Loan Account of the RST would be denominated in SDRs, as would be any contributor claims to the LA, RA, and DA. The SDR denomination of RST claims under borrowing agreements eliminates exchange rate risk as RST loans are SDR denominated. The SDR denomination of commitments to the LA ensures the application of uniform upfront contributions to the RA and DA that are expressed as a fixed percentage of the LA commitment. The SDR as the unit of account aligns RST financial reporting with the GRA and other accounts of the Fund.

C. Loan Account and Resources

8. The LA is the RST’s conduit for lending operations, funded by voluntary commitments from contributors. Borrowing agreements with contributors would be subject to uniform key financial terms: in particular, all claims under borrowing agreements would be remunerated at the SDRi, or lower with the agreement of the contributor, and agreements would provide for the same maturity structure as RST lending. Loans from the RST to borrowers would be made at the SDRi plus a modest margin (which varies by country grouping), and some borrowers also pay a service charge (see below).

9. Key financial terms of LA borrowing agreements and related conditions and modalities include:

  • Pass-through principle. As in the PRGT, it is proposed that lending to the RST take place on a pass-through basis, i.e., the RST would draw on borrowing agreements as needed to finance disbursements to borrowing members and would repay lenders according to the RSF repayment schedule. At the time of an RST loan disbursement, the LA makes a drawing on lender(s)’ borrowing agreement(s) which is(are) credited to the LA as RST borrowing, and the LA then disburses these funds as RST loans to borrowing members. In the same way, all repayments of principal and payments of interest flow through the LA

  • Borrowing agreements. Contributors would commit resources under agreements that would specify the amount of committed LA resources for specified drawdown periods, with the drawdown period discussed below. Similar to loans made to the PRGT, loans to the RST would not be revolving.4

  • Loans or notes. Resources contributed to the LA can be provided under loan or note purchase agreements, depending on the contributor’s preference, and the terms of loan and note purchase agreements would not differ materially. As in the PRGT, both loan claims and notes could be made transferable to other RST contributors, Fund members, and prescribed holders of SDRs.5

  • SDR denomination of RST commitments under borrowing agreements and contributor claims on the RST. As mentioned above, RST loan commitments under borrowing agreements, and claims resulting from drawings under RST borrowing agreements would be SDR denominated. In contrast to the PRGT where the maximum commitment under a borrowing agreement to the loan accounts can be expressed as the SDR-equivalent of a currency amount, such commitments would not be possible for the RST owing to the parallel contributions to the RA and DA that provide for contributions as a fixed percentage of a contributor’s LA contribution. To ensure the minimum percentages of 2 and 20 percent, respectively, the LA commitment also has to be expressed in SDRs in all agreements, otherwise the SDR value of the maximum commitment in currency would fluctuate with the fluctuation of SDR basket exchange rates.

  • Uniform maximum drawdown period. The borrowing agreements would provide for a pre-specified uniform drawdown period, i.e., through November 30, 2030. On an exceptional basis, the Managing Director could agree to shorter drawdown periods where domestic constraints prevent a member to commit to the November 30, 2030 drawdown period. All contribution agreements, however, would provide for drawings for encashment purposes for as long as RST loan claims remain outstanding. Any extension of the drawdown period, or a new drawdown period under a new resource mobilization round, would require an amendment of the RST Instrument (and the agreement of individual LA contributors as regards drawings under their borrowing agreements).

  • Drawings for the benefit of all RST eligible members. Drawings can be made for the benefit of all RST eligible members. LA contributors would have no bilateral relationships with the borrowing member; rather the contributor would lend to the RST, and the Fund as Trustee of the RST, would provide loans to the borrowing member. As in other circumstances of joint donor funding (e.g., contributions to multi donor accounts), contributors could apply notional earmarking in cases of domestic limitations regarding support of specific countries.6

  • Proportionate Drawings. Drawings by the Trustee under loan contribution agreements would take place with the view of having broadly balanced positions of loan resources among contributors overtime.

  • Media of transactions. Borrowing agreements would provide for drawings in SDR or freely usable currency, with flexibility under individual agreements as to what media are used for drawings. Repayment of principal would be in SDR or a freely usable currency, as specified in individual agreements. Payment of interest would normally be made in SDRs to the SDR account of the relevant member in the SDR Department, provided that the Trustee and the contributor can agree on other media of payment.

  • Maturity and grace period. RST loans to borrowing members are proposed to have a maximum 20-year maturity, with repayment of each disbursement in 20 semiannual instalments starting 10 ½ years after the disbursement.7 The maturity schedule of claims under LA borrowing agreements would mirror the scheduled maturity of RST loans. As in the PRGT, borrowing agreements can provide for shorter, renewable maturity dates, provided that the maturities can be renewed unilaterally by the Trustee for maximum maturities in line with the repayment schedule of RST loans.

  • Encashability. Borrowing agreements will allow for the early repayment of outstanding claims in the event a contributor represents that its balance of payments and reserve position justify the early repayment, and the Trustee, having given this representation the overwhelming benefit of any doubt, agrees.8 The encashing contributor would have to reopen its agreement for renewed drawings as soon as its balance of payments has improved sufficiently, as evidenced by the inclusion of the currency of the member into the Financial Transactions Plan for transfers. The encashment regime is elaborated below.

  • Interest rates. Lenders’ LA claims on the RST would be remunerated at the SDRi or lower with the agreement of the contributor. RST borrowers pay an interest rate of SDRi plus a margin differentiated by different borrower groups. The margin would be subject to periodic reviews. As discussed in greater detail in paragraph43 of the main body of the paper, the initial margin is envisaged as follows: 55 basis points for Group A countries (i.e., PRGT-eligible members that are not presumed blenders under the Fund’s framework for blended access to PRGT and GRA resources), 75 basis points for Group B countries (presumed blenders and non-PRGT eligible small states with per capita GNI below ten times the IDA operational cutoff), and 95 basis points for Group C countries (non-PRGT eligible members).

  • Interest rate on overdue financial obligations: Overdue principal or interest obligations on RST loans will be charged interest at the interest rate applicable to the loan claims on which obligations are overdue from the time they become overdue to compensate the Trust for the opportunity cost of arrears.9 The interest rate charge on overdue obligations will be subject to a minimum of the SDRi, which is also the interest rate applied to overdue obligations in the PRGT.

  • Service charge. As discussed in greater detail in Paragraph 41 of the main body of the paper, it is proposed that a service charge be levied on RST disbursements. Similar to the margin, staff proposes a tiered system that differentiates among the same three groups of RST eligible members. Staff proposes that initially RST borrowers in Group B and C pay a service charge at 25 and 50 basis points respectively, while Group A countries would be exempt from a service charge. Service charges would be reviewed periodically, as part of the periodic reviews of the operations of the RST or earlier as warranted by circumstances.

  • Net interest income transfers and service charges. The LA would receive interest income from RST borrowers (at SDRi plus the applicable margin) and make interest payments to RST LA contributors (at SDRi) whose borrowing agreements have been drawn upon. Margin income and income from service charges are intended to cover the RST’s administrative costs, with margin income expected to contribute to the build-up of RST reserves. Margin income will be promptly transferred from the LA to the RA following the end of each financial quarter, and service charges will be credited directly to the RA when paid by borrowers on the disbursement date of each loan.

10. Interest rate reviews would be undertaken as part of the periodic reviews of the RST, or earlier if circumstances warrant. The initial margins and service charges would be established at the time the RST Instrument is approved. The level of the margins would be reviewed at the time of the periodic reviews of the RST (i.e., normally at three-year intervals) or earlier if circumstances warrant. In the view of staff, such earlier review should be triggered if the average SDRi were to rise above 1.5 percent in any 12-month period and if financial market indicators signal that the SDRi is not expected to decline below 1.5 percent within the coming quarters. An interest rate review could also be triggered by other unexpected events that have a bearing on the financial sustainability of the RST. Given the role of margins and of service charges in covering administrative costs and the role of the margin in building reserves, the considerations guiding reviews would include adequacy of RST reserves, coverage of administrative costs, and appropriateness of the financial terms to the borrowers.

11. As discussed in greater detail in H44 of the main body of the paper, an interest rate cap could be considered at a future review of the RST. A cap for LICs in that group implies that a negative margin would apply once the SDRi rises above a certain level, with the foregone interest income (from the shortfall to the SDRi plus the margin on Group A loans) to be funded from the RA. Such an interest rate cap would only be established, however, in the context of an assessment of the adequacy of RST net reserves over the lifecycle of the RST and would require prior consultation with LA contributors.

12. All borrowing agreements with LA contributors would authorize the Trustee to make drawings to fund the early repayment of claims under borrowing agreements of other contributors in case of encashment requests. The right of lenders to early repayment (“encashment”) of their claims under LA borrowing agreements in the event of balance of payments or reserves need is a key pre-condition to ensure the reserve asset status of LA claims.10 Many potential contributors to the RST have indicated the need for their claims on the LA to have reserve asset status. Accordingly, the RST instrument requires that all borrowing agreements would authorize drawings by the Trustee to fund a request for encashment by other contributor(s), irrespective of whether or not a contributor is interested in the encashability of its own claims.11 This requirement ensures the broadest contributor base to support the reserve asset nature of LA claims. When drawing under borrowing agreements to finance RST lending, the Trust would leave a buffer (similar to the practice under the PRGT) for possible encashment requests and take the need for this buffer into account when mobilizing LA resources. It should be noted that, as under the PRGT encashment framework, borrowing agreements would remain open for encashment calls for as long as RST loan claims remain outstanding (i.e., through November 30, 2050, assuming a drawdown period through November 30, 2030 and a maximum 20-year maturity for RST loans).

13. An encashing contributor would be required to reopen its agreement for drawings as soon as its balance of payments and/or reserve position has improved sufficiently. If the balance of payments/reserves position of the contributor becomes sufficiently strong, as evidenced by the member’s inclusion in the Financial Transactions Plan for use of the Fund’s holdings of its currency in GRA purchase transactions, any undrawn portion of its borrowing agreement would again become available for drawings, including encashment calls by other lenders for as long as RST loan claims remain outstanding. Such drawings would be made with the objective of having broadly balanced position among contributors over time.

14. In the event of overdue obligations by RST borrowers, claims under LA contribution agreements would be repaid with RA resources according to the original maturity schedule for interest and principal payments. Only if an extreme tail risk event materializes that depletes the RA would DA resources be used as a backstop (see section D below). Upon clearance of arrears, the resulting payments to the RST would be directed to the DA and RA with any such repayments first allocated to the DA for any amounts of DA principal used from that account to repay LA contributors. As with the PRGT, no rescheduling of RST loans by the Trustee will be permitted. Interest at the SDRi plus the relevant margin would be charged on any overdue interest or principal from the date the obligation becomes overdue (see above).

15. Similar to the PRGT, the RST’s accounting policies will have a framework for assessing the credit quality of outstanding RST loans, which will determine the provisions for credit losses, if any. The credit quality assessment of outstanding loans would be grounded in the context of the nature of the RST’s lending and the unique institutional status of the IMF, as Trustee. This includes the credit risk management practices discussed above, interest on overdue repayments and interest, and the envisaged de facto PCS of the RST. Taken together, these factors are expected to significantly reduce the likelihood of the RST recognizing a provision for credit losses in its financial statements.12

D. Reserve Account and Deposit Account

Reserve Account

16. The RA is the RST’s principal financial buffer to manage financial risks and protect contributors’ claims on the RST, and to cover the RST’s administrative costs. If needed, the RA would be available to fund temporary interest and principal arrears, and to absorb provisions for credit losses and potential ultimate credit losses (write-offs). The RA would also cover the costs of administering the RST through reimbursement of the GRA. The key features of the RA and the contributors’ RA claims include:

  • Resources. The resources of the RA consist of (i) upfront unremunerated contributions by contributors in proportion (a minimum of two percent) to their respective LA commitments13, (ii) any additional standalone unremunerated contributions to the RA unrelated to an LA contribution (see next bullet), (iii) transfers of net interest income from the LA, (iv) service charge income paid by borrowers of RST loans; (v) investment earnings on RA balances; (vi) transfers of excess investment earnings from the DA; and (vii) transfers of payments of overdue principal or interest from the LA The minimum RA principal contribution is 2 percent of total LA commitments, provided that contributors would be able to provide higher contributions in excess of the 2 percent minimum.

  • Standalone contributions. The RST Instrument would authorize standalone contributions to the RA This authority is intended for contributors that are generally not in a position to contribute to the LA but would like to support the RST for the purposes of reserve build up and credit protection. Standalone contributions would have a fixed maturity with a preferred maturity of 10 years in view of the RST investment strategy. They are subject to the same risk of principal loss as any other RA contribution for as long as the standalone RA contribution is outstanding.

  • No reserve asset status. Reserve Account contributions would not be encashable and would not count as international reserve assets of contributors. RA contributors do not have the right to request early repayment of their contributions in the event of a balance of payments need.

  • Principal RST buffer. RA resources are the principal financial buffer in the case of arrears on Trust loans, including for provisioning. In particular, RA resources (by way of a transfer of resources to the LA) would be used first in the case of overdue obligations on RST loans (see above) to ensure continued payments of interest and principal under LA borrowing agreements in accordance with the repayment obligation schedule of RST loans. If RA resources are used for payment of interest or principal to LA contributors, the payments by RST borrowers to clear any overdue obligations funded by the RA would be directed to the RA provided that if payments had also been made from the DA principal because RA resources had been exhausted (if an unforeseen extreme tail risk events were to materialize), any such payments from arrears clearance would first be directed to the DA until any principal amounts used from the DA for that purpose have been replenished. Any provisioning for accounting purposes (for example, as a requirement under IFRS 9) would also be first charged against the LA and corresponding amounts would be transferred from the RA.

  • Administrative expenses. RA resources would be used to cover administrative expenses. This is not expected to affect the reserve accumulation from RA contributions as margin and service charge income are expected to more than cover administrative expenses, (see Section X of main body of the paper).

  • Investment of RA resources. The RST Instrument would provide for the authority of the Managing Director of the Trust to invest RA assets, pending their use, in accordance with guidelines approved by the Executive Board.

  • Distribution mechanism and contributor share. The resources of the RA are to be distributed at the time of liquidation of the RST, or at such earlier time as the Trustee decides (see below). Each contributor has a calculated share in the RA that is based on (i) its original principal contribution, (ii) attributed net investment earnings/losses on calculated share, (iii) any transfers of excess investment returns from the DA attributed to its contribution following the repayment of the DA principal contribution, (iv) amounts related to repayment of overdue obligations where the payments to LA contributors were funded with DA investment earnings, with such resources to be attributed among RA contributors based on the share of a contributor in the DA, and (v) use of RA resources to make payments on overdue obligations to LA contributors, and payment of administrative expenses, proportional to each contributor’s overall share.14

  • Terminal distribution. At liquidation of the RST, following repayment of all LA claims under borrowing agreements, and the maturity of the DA and the associated transfer of excess investment income to the RA, each contributor would receive its calculated share in the total assets of the RA. Contributors would have flexibility to direct their share in the terminal distribution to other IMF-administered Trusts.

  • Early distribution. It is proposed that the Trustee may decide on an early partial distribution of RA resources to all contributors, in proportion to their share in the account, if it determines that the coverage from net reserves of the remaining RST loans justifies an early distribution. Such decision would be based on an assessment of both RA resources and excess investment returns in the DA, which constitutes the net reserves of the RST. Early distribution from the RA could become relevant in the final years of the RST when outstanding loan claims are being largely repaid and when reserves have grown as expected (see below on sequencing of RA and DA distributions if the Trustee decided to reduce RST reserves). Contributors would have flexibility to direct their share in the early distribution to other IMF-administered Trusts or their own accounts.

Deposit Account

17. The purpose of the DA is to generate additional reserves for the Trust and to minimize any residual risks to contributors’ claims on the RST. Specifically, the DA (i) bolsters gross reserves upfront, including in the early years when RA balances are still modest; (ii) invests its assets to generate investment earnings above the SDRi over time with a view to building sizeable additional net reserves by the time borrowers start repaying their RST loans; and (iii) in unforeseen extreme tail risk events, acts as a backstop to the RA by providing an orderly mechanism for pooling any residual risks to contributors’ total claims on the RST. Its key features include:

  • Funding. The DA would be funded upfront by long-term deposits that are denominated in SDR from LA contributors in proportion to their LA commitments. Each LA contributor would be expected to make a DA deposit equivalent to at least 20 percent of its LA commitment. DA contributions at higher percentages would be permitted.

  • Standalone contributions. The RST Instrument would authorize standalone DA contributions. Such contributions are generally intended for members that are not LA contributors. Standalone contributions would have a fixed maturity with a preferred maturity of 10 years based on the investment horizon of DA assets, and the remuneration could not exceed the SDRi.

  • Remuneration. As with LA claims, lenders’ claims on the DA would be remunerated at the SDRi (or lower) with interest paid following the end of each quarter.

  • Investment of DA resources. DA resources would be invested in liquid, high-quality assets that aim at generating investment income in excess of the SDRi to build additional net reserves over time while also allowing for the possibility of encashment. The RST Instrument would provide for the authority of the Managing Director of the Trust to invest DA assets under guidelines approved by the Executive Board.

  • Backstop to the RA. Only in an extreme tail risk scenario that causes large financial losses where the resources available in the RA are insufficient would the DA resources be used for credit protection. This backstop function is as follows: if RA resources were exhausted, DA resources could be relied upon to make payments due to contributors, using all excess investment returns before using DA principal. When RST borrowers clear any overdue obligations, they would first be allocated to the DA. Once DA principal is fully replenished, clearance of overdue obligations is directed to the RA. To limit the potential for a reduction in DA principal for any contributor, the use of excess returns would be attributed to each DA contributor according to its share in the total excess returns of the DA and a corresponding addition would be made to its RA contribution. If necessary, any use of principal for a contributor would be made according to its share in total DA principal claims, with replenishment of the DA allocated on the same basis. In cases of provisioning (reflecting an increase in the probability of an ultimate loss), non-cash transfers from the DA to the LA would be made to ensure the LA is made whole. As permitted under IFRS 9, provisions can be reversed when credit quality improves (i.e., the probability of protracted arrears and/or an ultimate loss decreases), resulting in the reversal of the non-cash transfers between the DA and LA. In attributing losses for encashment of DA principal, any such provisions would be allocated first to excess investment earnings, and then to DA principal, according to the same approach as outlined above. This DA backstop function further reduces any residual risks to contributors’ total claims on the RST by providing an orderly burden-sharing mechanism for risk pooling in extreme tail risk events (see Box 1). Given the RST’s multilayered credit risk management framework, including the envisaged de facto PCS, the likelihood of the residual backstop function being applied is expected to be extremely low (see Section VI on stress simulations).

  • Encashability. DA contributors would have the right to request early repayment of their DA principal prior to maturity, net of any attributed losses or excess investment returns if the contributor represents that its balance of payments and reserve position justify the early repayment, and the Trustee, having given this representation the overwhelming benefit of any doubt, agrees.15 Any excess investment income attributed to the encashing contributor would remain in the DA pending DA maturity or the completion of early distribution of the DA principal or the activation of the DA backstop function to cover payments to LA contributors, at which time it would be transferred to the RA as the RA contribution of that contributor.

  • Encashment operations. The encashment request would be funded by a liquidation of DA assets equivalent to the principal amount of the deposit, net of any attributed losses.16 Investment earnings attributed to the deposit and retained in the DA would not be returned in the event of an encashment call; rather they would be retained in the DA and accounted for as the remaining share of the contributor. Investment returns on such undisbursed excess investment returns and other DA receipts (whether from the reversal of provisioning or repayment of overdue obligations where payments to LA contributors were funded with DA principal contributions) would continue to accrue notwithstanding the encashment. While encashment can reduce the amount of a contributor in the account, it does not eliminate the status of a DA contributor.

Reconstitution of DA resources following encashment. If the balance of payments and reserves position of the relevant member that has encashed its DA contribution becomes sufficiently strong again (as evidenced by the member’s inclusion in the Financial Transaction Plan for use of the Fund’s holdings of its currency in GRA purchase transactions), the contributor would have to reconstitute its deposit in a timely manner for an amount equivalent to the SDR value of the amount received under the encashment.

  • Reserve asset status. Lenders’ claims on the DA would count as their international reserves based on the RST’s multilayered risk management framework, investment in high-quality (investment grade) assets, and the encashability of DA claims.17

  • Maturity. The maturity of DA contributions would be based on the maximum final maturity of loans funded under the RST resource mobilization round. With a proposed drawdown period through end-November 2030 of the first round and a maximum loan maturity of 20 years, the uniform maturity date of the DA contributions in the initial round of contributions would be end-November 2050. This uniform date would also apply in exceptional cases where the drawdown period under an individual LA contribution agreement would be shorter than end —November 2030 due to domestic constraints. Having a uniform maturity date ensures that the risks resulting from RST lending are shared equitably among all contributors, independent of the timing of when loans were drawn.

  • Distribution of DA principal at maturity. At the time of maturity, the principal amount of the original deposit is returned to a creditor, net of any attributed losses. Any investment returns attributed to the deposit would be transferred to the RA and counted as a contribution of the contributor to that account. The maturity of the deposit would normally coincide with the liquidation of the RST unless the Trustee decides on a future loan mobilization round that extends the life of the RST beyond end-November 2050. In that context, it would need to be decided whether the DA is maintained and whether contributors could extend their deposit beyond its original maturity. This would be done through an amendment of the RST.

  • Early distribution of DA principal. Prior to maturity of the DA deposits, the Trustee could decide on an early distribution of part or all of DA principal contributions if such distribution is justified based on the net reserve coverage of RST loans on a prudent forward -looking basis. As the DA serves as a backstop to the RA, any reserve distribution would be first done from the DA and attributed to the principal amount of deposits. If DA principal is fully repaid, excess investment earnings attributed to each DA contributor would be transferred to the RA and attributed to the RA share of the contributor. An early RA distribution would be made to all contributors in proportion to their calculated shares in the account as discussed above.

  • Termination. In the event of liquidation of the Trust, the DA contributors would receive their principal net of any attributed losses. All remaining accumulated DA excess investment earnings would be transferred to RA and be distributed as discussed above.

Enhanced Risk Management through the Deposit Account

The DA is an integral element in the proposed financial architecture of the RST. The purpose of the DA is to endow the RST upfront with sufficient reserves, generate investment earnings to build the Trust’ net reserves overtime, and to ensure an orderly pooling of residual risk across contributors in unforeseen extreme tail events. Taken together, these features significantly reduce any residual risks to contributors’ total claims on the RST. Contributions to the DA will preserve their reserve asset status given their investment in high-quality assets and contributors’ ability to encash their contributions upon representation of a balance of payments or reserves need.

Ensuring adequate reserve coverage from Day One. Compared with DA resources, RA resources are scarcer and more time-consuming to mobilize in large quantities, including because contributors cannot count their RA claims as reserve assets. By endowing the RST with ample reserves right from its inception, the DA facilitates speedy implementation of the RST.

Generating additional net reserves over time. The DA resources would be invested in high-quality assets, while maintaining sufficient liquidity to meet potential encashment to ensure the reserve asset nature of contributors’ DA claims. The excess returns, i.e., investment income exceeding the SDRi paid on contributors’ DA principal, would build up additional net reserves (defined as the RA balance plus cumulative DA excess returns) over time. As excess returns would be retained in the DA for further investments, they would benefit from a compounding impact. They would also protect the contributors’ DA principal against potential impairment from credit events and investment losses.

Backstop to the Reserve Account. The RA is the RST’s principal financial buffer to manage credit and liquidity risks, and cumulative excess DA investment returns provide an additional layer of protection. The DA’s backstop function to the RA further minimizes creditors’ risk to their total claims on the RST by setting aside a share of contributors’ total loaned resources for an orderly risk pooling mechanism.

  • In a hypothetical scenario with unforeseen extreme tail risk events that cause large financial losses exceeding the RA balances, the book value of contributors’ claims on the DA would be temporarily reduced, first using the pool of DA investment returns (in proportion to the attributed share of each contributor in those returns) and then by using the DA principal. This minimizes the chances of DA principal value being reduced even in an already remote hypothetical scenario. In the absence of this backstop function, such extreme tail events could imply an uneven impact on contributors because at any given time, contributors’ loan claims may not be fully harmonized relative to LA commitment shares. As a result, in such a tail risks scenario, the total claims on the RST of some contributors could be reduced by more than would be the case with the burden-shared DA backstop function. Given the RST’s multilayered credit risk management framework, including the envisaged de facto preferred creditor status (PCS), net reserves are expected to be more than sufficient to absorb a broad range of severe adverse events, including a combination of reduced investment income, encashment, and sizable arrears and credit losses (see Section VI), making the application of the backstop function extremely unlikely.

  • The backstop is designed to protect the principal in the Deposit Account. Before tapping DA principal, the excess investment returns in the DA would be fully utilized, providing a significant buffer for that principal. Moreover, as soon as there is a reversal of financial losses (e.g., arrears repayments by any borrower), any reduction in the DA principal would be replenished first.

Annex V. Demand Projections and Financing Scenarios

Demand Projections

Total baseline demand for RST financing is estimated at SDR 22 billion based on the stylized assumptions below:

  • a. A subset of eligible countries with past IMF programs would request RST financing (Annex V Table 1 below). Of the 143 RST-eligible countries (i.e., countries with 2020 GNI per capita below 10 times the IDA operational income cutoff and countries that have a population size of less than 1.5 million and GNI per capita below 25 times the IDA operational cutoff), it is assumed that all of those that had a UCT-quality program (as defined above) over the past 10 years would come forward for RST financing. Applying this assumption leads to 70 countries seeking RST financing. These are reflected in the illustrative demand scenario.

  • b. Individual country access is assumed at 100 percent of quota for most countries, subject to a nominal cap on access of SDR1 billion. Countries with quotas greater than 3 percent of their respective nominal GDP, are assumed to receive access of 50 percent of quota. Countries with Flexible Credit Lines in the last decade that meet the income eligibility criteria are not assumed to request access under the baseline.

Countries pay interest following a tiered interest rate structure (SDRi plus margin) that groups eligible countries into three categories. The SDRi is assumed to rise over time, stabilizing at 3 percent by year 9 of the projection period.

  • Group A countries are all PRGT-eligible countries that are not presumed blenders.1 This group pays a margin of 55 basis points above the 3-month SDRi and would be exempt from any service charges on RST disbursements.

  • Group B countries would include all “presumed blenders” (of PRGT and GRA resources) and all small states (below 1.5 million inhabitants) with per capita GNI below ten times the IDA operational income cutoff. This group would pay a margin of 75 basis points above SDRi and would be subject to an upfront one-time service charge of 25 basis points levied on each RST disbursement.

  • Group C countries would include all other RST-eligible countries, i.e., all eligible (non-small) MICs and all small states with income above ten times the IDA operational income cutoff. This group would pay a margin of 95 basis points above SDRi and would be subject to an upfront one-time service charge of 50 basis points on each RST disbursement.

Annex V. Figure 1.
Annex V. Figure 1.

Interest Rates Paid by Resilience and Sustainability Trust Borrowers Under Tiered Rates1

Citation: Policy Papers 2022, 013; 10.5089/9798400207730.007.A001

1 The horizontal axis indicates years of RST operations. The SDRi interest rate is assumed to rise then stabilize at 3% from year 2030. Group A includes PRGT-eligible countries that are not presumed blenders; group B includes presumed blenders and small states with GNI per capita below 10 times the IDA threshold; group C includes all other eligible members that are not included in groups A or B.

Financial Model

Starting off from the above demand estimate and interest rate structure, staff has developed an illustrative financial model with the following features:

  • Demand for RST financing. SDR 22 billion, disbursed evenly over the first five years of the Trust.

  • Loan terms. 20-year final maturity; 10½-year grace period; borrowers pay the tiered interest rates and service charges depending on the country group they belong to.

  • Contributions. Contributors sign contribution agreements with three parts—a loan contribution; a reserve account contribution; and a deposit contribution. It is expected that the bulk of the contributions would be in SDRs, but the provision of loan resources in freely usable currencies would also be possible.

  • Loan Account (LA). Total loan commitments are assumed to cover the projected demand and a 20 percent encashment buffer (SDR 27 billion); the SDRi rate is paid to lenders on drawn loan commitments.

  • Reserve Account (RA). Initial funding through upfront reserve injections equivalent to 2 percent of contributors’ respective loan commitments. Reserve balances increase over time from RST lending margins and investment returns, net of administrative costs.2 Balances in the DA are invested and earn returns at SDRi + 45 basis points.

  • Deposit Account (DA). Funded upfront by long-term deposits provided by contributors, remunerated at the SDRi. The initial balance of DA would be equivalent to 20 percent of each contributor’s loan commitment. Balances in the DA are invested and earn returns at SDRi + 45 basis points.

  • Gross and net reserves. The RST’s gross reserves consist of the total balances in the RA and the DA. Net reserves are defined here as gross reserves net of the DA principal liability, which is equivalent to the RA plus any cumulative excess DA investment returns (i.e., returns in excess of SDRi paid to contributors as remuneration).

Financial Stocks and Flows—Baseline

The baseline scenario tracks the evolution of financial stocks and flows over time using the above demand projections and financial design features.

  • Lending operations. Disbursements are assumed to be made at equal installments of SDR4.4 billion annually for a period of five years; credit peaks at SDR 22 billion; amortization starts in year 11; annual debt service due to lenders peaks at SDR 2.7 billion; all loans are fully repaid after 25 years.

  • Reserve Account. RA balances start at SDR 544 million and grow over time from lending margins and investment returns, reaching SDR 1.8 bn in year 11 when amortization of RST loans starts. The terminal value after 25 years (excluding cumulative DA excess investment returns) would total SDR 3.4 billion.

  • Deposit Account. Lenders deposit upfront SDR 5.4 billion in the DA; balances are invested in high-quality, liquid assets, and grow to SDR 5.8 billion by the time amortization of RST loans starts and to SDR 6.4 billion after 25 years, based on excess investment returns above remuneration of lenders at the SDRi.

  • Gross reserve coverage. Gross RST reserves (RA + DA) cover at least 35 percent of credit outstanding in the repayment period (years 11–25), and more than 3 times debt service.

  • Net reserve coverage. Net reserves (RA balances plus cumulative excess DA investment returns, i.e., gross reserves net of DA principal), are at least SDR 2.2 billion during the repayment period (years 11–25), providing a buffer before the DA principal would be affected by potential arrears. Net reserves cover at least 10 percent of total credit outstanding and one year’s worth of total debt service.

  • Terminal value. The “terminal” value of the Trust (i.e., net reserves after 25 years when all loans have been repaid), would be SDR4.3 billion, an eight-fold increase from inception. However, this is based on the stylized assumption that there is no further RST lending after the first 5 years, while net reserves continue to build over that time. This is unlikely to materialize in practice because (i) the RST might see further rounds of borrowing and lending and/or (ii) margins would be reduced once adequate reserve coverage is reached.

Annex V. Figure 2.
Annex V. Figure 2.

Resilience and Sustainability Trust: Stylized Lending Example

(In SDR million)

Citation: Policy Papers 2022, 013; 10.5089/9798400207730.007.A001

Annex V. Figure 3.
Annex V. Figure 3.

Resilience and Sustainability Trust Reserve Accumulation and the Components of Net Reserves1

Citation: Policy Papers 2022, 013; 10.5089/9798400207730.007.A001

1 Balance (stock) of net reserves versus flow of its components (net lending income, net return on Reserve Account, net return on Deposit Account, and administrative expenses).

Adverse Events

To stress-test reserves under a range of adverse events, the baseline assumptions were modified in a few illustrative hypothetical examples (see Annex V Table 3):

  • Lower returns. If investment returns for the RA and the DA are 20 basis points below the baseline assumptions, it will not significantly affect minimum reserve coverage ratios during the repayment period. The minimum net reserve cushion in the repayment period would fall from SDR 2.2. billion under the Baseline to SDR 2.0 billion.

  • Temporary arrears. If 10 percent of debt service (principal and interest) is not paid for five years (years 8–12), and arrears are cleared thereafter, the minimum total reserve coverage ratio over credit outstanding would decline from 35 percent to 34 percent. A net reserve cushion of at least SDR 1.8 billion would remain available during the repayment period to cover any additional arrears before DA principal would be affected. Debt service coverage ratios would not be significantly affected.

  • Permanent credit losses. If 10 percent of principal due in the first five years of the repayment period (years 11 -15) is never repaid, it would not significantly affect minimum reserve coverage ratios. The minimum net reserve cushion in the repayment period would decline modestly to SDR 2.1 billion.

  • Encashment by a lender. If one contributor encashes its entire loan and DA principal in one year, it would not significantly affect minimum reserve coverage ratios of the Trust.

  • Combined adverse events. If all the above adverse events were to materialize, it would have the most negative impact on the trust relative to the other scenarios. Despite this, the minimum total and net reserve coverage ratios to credit outstanding would decline from 35 and 10 percent to 32 and 7 percent, respectively. A net reserve cushion of at least SDR 1.5 billion would remain available during the repayment period to cover any additional losses before DA principal would be affected.

Additional Sensitivity Analysis

Additional illustrative examples provide ranges of plausible developments under alternative assumptions:

  • Uniform margin on lending. If the interest margin paid by borrowers is uniformly 75 basis points, minimum total reserve coverage relative to credit outstanding improves modestly. A net reserve cushion of at least SDR 2.3 billion would remain available during the repayment period.

  • Lower administrative costs. If annual administrative costs are SDR 18 million rather than the hypothetical SDR 25 million assumed in the baseline, it would not significantly affect minimum reserve coverage ratios.

  • Default by largest borrower. A permanent default by two of the largest borrowers on interest and principal due during the repayment period (SDR 2 billion) would have a moderate impact on minimum reserve coverage ratios relative to credit outstanding (as credit outstanding is assumed to decline in the outer years), but the minimum reserve coverage ratios relative to debt service would decline more substantially. The minimum net reserve cushion to cover any additional arrears would decline to only SDR 0.8 billion vs. SDR 2.2 billion in the baseline.

  • Higher demand by borrowers. If demand by borrowers is 25 percent greater than assumed in the baseline, reserve coverage ratios would marginally improve. The net reserve cushion during the repayment period would be SDR 2.8 billion (compared to SDR 2.2 billion in the baseline).

Lower demand by borrowers. If demand is 25 percent lower than in the baseline, reserve coverage ratios would marginally decline. The net reserve cushion during the repayment period would be at least SDR 1.5 billion (compared to SDR 2.2 billion in the baseline).

In addition to the robustness of reserves vis-a-vis cash flow risks, it is also expected that projected resources in the RA would comfortably absorb any required provisions for credit losses under baseline and stress simulations for calculating expected credit losses (ECL) under accounting standards. Using the Fund’s existing ECL calculation model as a proxy for the RST, a number of high-level simulations were run to determine the extent of provisions for credit losses. The simulations were run at a point in time in the projected baseline for loan demand when credit outstanding is at its peak at SDR 22 billion. For each simulation and according to the model, the RST loan portfolio was allocated into three buckets; (i) unchanged credit quality; (ii) deteriorated credit quality (i.e., significant increase in credit risk); and (iii) default (i.e., protracted arrears). Probability-weighted losses for each bucket were calculated for three scenarios; (i) baseline (i.e., on-time payments); (ii) protracted arrears cleared after 7 years; and (iii) ultimate failure to pay, to arrive at total ECL3 In the baseline and stress simulations (which reflect conservative assumptions of portfolio credit quality relative to the Fund’s experience of arrears cases) estimates of total ECL range from 0.1 percent to 22 percent of the projected minimum RA balance (SDR 0.54 billion), indicating that resources in the RA as estimated in the illustrative example would be more than adequate to absorb ECL under impairment assessments, implying that the DA principal would not be affected by credit losses even in such severe stress scenarios.

Annex V. Table 1.

Country Groups and Estimated Demand

article image
Source: Finance Department

Group A consists of PRGT-eligible countries, excluding presumed blenders. Group B includes presumed blenders and small states with GNI per capita below 10 times IDA cutoff. Group C includes eligible countries that are not in

Based on the eligibility criteria effective upon Board approval of the RST, including Syria, which is expected to become PRGT eligible on March 18, 2022.

To estimate demand, the 143 eligible countries are screened for those that have had a UCT program in the last 10 years. Eligible FCL countries are

Baseline demand estimates assume an eligible country would borrow at 100 percent of quota on average with a nominal cap of SDR 1 billion. However, an eligible country, whose quota is higher than 3 percent of its nominal GDP, is assumed to borrow at 50 percent of the quota.

Annex V. Table 2.

Countries by Group1

article image

This reflects the grouping of countries based on the eligibility criteria effective upon Board approval of the RST.