Reallocating SDRs into an IMF Global Resilience Trust


The August 23 allocation of SDRs has given low- and middle-income countries (LMICs) breathing space on their balance sheets to confront the monetary and fiscal challenges of the economic crisis induced by the COVID-19 pandemic. In the next few weeks, at the UN General Assembly meetings and the annual meetings of the IMF and World Bank, the international community will discuss the possibility of reallocating (or recycling or channeling) developed countries´ SDR allocation, largely unneeded by them, to LMICs. 

In an earlier note, we outlined a set of mechanisms that could be envisaged for SDR reallocation and the considerable challenges of doing so. In a follow-up to that note, CGD convened a group of experts to consider the technical challenges of SDR reallocation in more detail. As a result of those discussions, we will publish a series of technical notes examining the nitty gritty of what needs to be done to reallocate SDRs, highlighting areas where policy and technical decisions must be made by the international community.

Each note will first describe the basic design elements of one of the reallocation modalities.

The note will then consider a set of technical evaluation criteria:

  1. Does the reallocation modality maintain the reserve asset nature of the SDR?
  2. Does it meet legal constraints of the IMF and SDR donors on use?
  3. Does it ensure that the use of the funds is transparent?

It will then look at a set of policy criteria that might help the international community decide whether to pursue the modality:

  1. Does it address priority policy areas?
  2. Does it cover targeted group of countries?
  3. Does it link well with other elements of the international cooperation architecture?
  4. Does it leverage or catalyze other resources?
  5. Does it ensure the funds will be used?

Each note will conclude with a summary of the major challenges in implementing the particular reallocation modality being considered.  

Once the notes are complete, we expect to publish a synthesis paper, but given the pace of international discussion, we will publish the technical notes on each allocation modality as they are ready. What follows is the first of those notes on an IMF resilience fund. 


The reallocation of SDRs into the Poverty Reduction and Growth Trust (PRGT) is a channel that has already been used by several countries (see here and here), and the IMF has called on advanced countries to use some of their new SDRs to buttress the PRGT further.

But many have pointed out that loans from the PRGT can only be used to support a prescribed list of low-income countries for a limited set of purposes. Thus, there have been calls for another IMF-based fund that can support a larger number of countries, particularly vulnerable middle-income countries (MICs), and can be used to provide financial flexibility as LMICs confront recovery from the pandemic and the transition to a resilient, sustainable, and equitable future global economy.

This note considers the technical challenges of establishing such a fund, which we dub a Global Resilience Trust (GRT).[1]

Basic design elements

The IMF would put in place a new trust fund like the PRGT, with a loan account and reserve account, the latter of which would provide an additional buffer against credit risks.[2] If the IMF wanted to consider making subsidized loans out of this trust fund, it would also need to establish a subsidy account with donations in hard currency.[3]

Advanced countries would agree to lend SDRs (or hard currency) to the GRT, with built-in encashment schemes to counter liquidity risk.

The GRT will then on-lend to LMICs, with the loans approved by the IMF Executive Board. The purpose, terms, country coverage, qualification criteria, and conditionality frameworks for the loans would need to be decided by the IMF Executive Board (see technical criteria below). This could be done without any revision to the IMF’s Articles of Agreement (the IMF charter document).

IMF staff would then establish the framework for negotiation of the loans, presentation to the Executive Board, and links to other lending by the IMF, or perhaps other multilateral institutions or bilateral creditors, allowing the possibility of leveraging IMF lending.

Technical criteria

1. Would an IMF-based GRT maintain the reserve asset nature of SDRs?

Like the PRGT, a GRT could be structured to address liquidity risk and credit risk. However, this would likely require additional hard-currency resources for a reserve account to guard against credit risk and for a subsidy account if the intention is to provide concessional loans. If the GRT were meant to be self-sustaining (with the investment income offsetting running subsidy or administrative costs), the reserve and subsidy accounts would need sufficient supporting resources.  

2. Does it meet legal constraints on use?

From the IMF’s perspective? There is no legal impediment to the IMF establishing a GRT that would on-lend SDRs. However, in putting together the details of the fund’s purpose and use, it must meet other requirements that ensure it is consistent with the purposes of the IMF in helping countries adjust to immediate or prospective balance-of-payments needs. These include:

  • Articulating a set of policy objectives—such as addressing pandemic vulnerability, climate change, biodiversity sustainability, or broader SDG goals—that helps the country deal with an immediate or prospective balance-of-payments need, which cannot be addressed through other established IMF facilities, including the General Resources Account and the PRGT.
  • Establishing unambiguous country eligibility criteria.
  • Establishing policies that define when a country qualifies for financial assistance.
  • And given a particular balance-of-payments need, disbursements would have to support a program that would work to address or resolve that problem over the time of the loan, and with an appropriate conditionality framework (for example on policies that address climate change and biodiversity).

From the creditor countries’ perspective? Every creditor country has its own set of rules governing the use of SDRs and the conditions under which they can be lent to another party. As most countries view SDRs as reserve assets, lending them would require remuneration to offset fiscal costs, risk mitigation to ensure the assets are of reserve quality, and some ability to liquidate the loan and get the SDRs back if needed (see section 1). Often these constraints would also entail a maximum length of the loan to recipient countries. For the PRGT, the loan term is 10 years, but a longer term has been proposed for a GRT.[4] Whether this is possible would be determined by the willingness of advanced countries to go along with longer periods and their perception of the associated risk, which in turn would depend on the conditions of lending.

3. Does it ensure that the use of the funds is transparent?

IMF governance together with policies on transparency and safeguards are well established and would provide a good basis for lending from the GRT to LMICs. Whether the existing policies are sufficient would depend on whether the IMF has the expertise and policies in place for the new scope of activities. For example, if the GRT were used to support climate-favorable regulatory change or investment, could the IMF frame and monitor these uses? While partnership with other institutions may help fill any expertise gaps, the IMF would ultimately have to attest to the sound usage of lending from the GRT.

Policy criteria

4. Does it address priority policy areas?

The purpose and policy focus would need to be established in the language of the instrument that establishes the GRT. In principle, there is nothing to prevent the purpose embracing resilience to climate change (through policies aimed at mitigation, adaptation, and structural transition of economy), sustaining biodiversity, guarding against digital crises, as well as other SDG-related purposes. However, as noted above, some link to a particular balance-of-payments need would have to be established for the GRT to be consistent with the IMF Articles of Agreement. In the case of the PRGT, the policies to reduce poverty and increase growth were assumed to create or exacerbate the balance-of-payments need; and similar—or extended—arguments would have to be developed for the new trust and be distinct from policies that other IMF facilities address.

5. Does it cover a targeted group of countries?

Country eligibility to draw on an IMF-administered trust could reflect whatever criteria the Executive Board decides. Possibilities include: the same as for the PRGT, or a much larger coverage to encompass MICs, or even all, of the membership. . (A similar facility established using IMF general resources, rather than advanced countries’ SDRs, would automatically grant eligibility to all members meeting the conditions for financial support.)  

6. Does it link well with/enhance other aspects of the international cooperation architecture?

There is a well-established framework to link loan disbursements from an IMF trust for specific purposes to the assurance of macroeconomic stability and fiscal sustainability—and safeguards for repayment. These are important signals domestically as well to international creditors, for example through providing a framework for disbursements of bilateral aid or through setting formal requirements for multilateral debt relief mechanisms. The purposes and conditions could also be designed to play directly to meeting the global public goods objectives consistent with the Paris agreement and established under the UNFCCC. Partnership with the World Bank would be essential, which could provide needed expertise and complementary financing. Also, given its global reach, the IMF lending role would complement its broader surveillance function and its associated responsibility to speak truth to power.

7. Does it leverage/catalyze other resources? 

A starting point would be to design the lending instruments so that the loans from the GRT would be additional, both to those available under existing facilities at the IMF, as they need to serve different purposes, and to other internationally mobilized resources. Beyond that, an IMF signal of assurance of macroeconomic stability and prospects for inclusive growth, while tackling the “resilience” purposes, should permit or encourage financing from other institutions or countries and, importantly, from private sector investors, who will need to provide the bulk of the financing for the transition to a green and equitable economic structure. While the GRT would provide by its nature general program financing, it could catalyze project financing from multilateral development banks and the private sector and encourage innovation in the provision of other kinds of green finance. 

8. Does it ensure funds will be made available and used?

Availability of funds. The design and implementation of the GRT would need to be sufficiently attractive to creditor  countries to ensure that sufficient SDR reallocation commitments are, in fact, forthcoming. As noted above, these elements include interest compensation, risk mitigation, the ability to liquidate SDR loans, and transparency and accountability in use, including appropriate conditionality to prevent misuse.

Frequency of need. As opposed to setting up a trust specifically for a pandemic response—which may be needed only periodically (and at uncertain intervals)—a GRT set up to deal with the longer-term economic transition would have potentially continuous demand for a generation.

Stigma. Countries tend to be averse to borrow from the IMF. In fact, the IMF’s lending capacity in the General Resources Account far exceeds demand. As the IMF designs a Resiliency Trust, it will need to confront the reasons for potential borrowing countries’ reticence. The difficulty is that the design elements that will attract creditor  countries to lend their resources to the GRT (such as conditionality which also mitigates credit risks) are the same that will make borrowing countries hesitate.

Applying a “Goldilocks” amount of conditionality. Following the above logic, there should be not too much conditionality, nor too little. A minimum would include justification of balance-of-payments need and a safeguards assessment. But in developing the conditionality framework, can an acceptable balance be achieved that would convince creditors that country objectives and the purposes of the Trust would be met, without discouraging countries from accessing this source of finance?


The PRGT has paved the technical way for a GRT to be established at the IMF. Four thorny challenges will confront the design and implementation of such a trust:

  • The GRT will have to be secure enough to garner support for creditor  countries to lend their SDRs to it. This will require some amount of complementary “hard currency” resources to mitigate liquidity and credit risk and provide any desired subsidies.
  • The loans from the GRT will need to support policies to deal with a balance-of-payments need not met by other IMF facilities, and the IMF will have to have the capacity to judge whether the loans are being appropriately used.
  • The loan conditions will have to be strict enough to assure SDR contributors that the funds will be used effectively to mitigate the balance-of-payments need and thus be repaid at the end of the loan. But these same conditions must not be so onerous as to discourage countries from borrowing.
  • Even with very broad support, the GRT would be able to provide only a fraction of the climate-related financing needs of LICs and MICs. The emerging catalytic role of the new Trust would need to be closely monitored.

We thank the CGD Technical Working Group on SDRs for their insightful contributions to this piece and to the CGD work program on SDRs in general. Any errors remain our own.

[1] The IMF is working on a similar trust which they are calling the Resilience and Sustainability Trust (RST). At this point the GRT and RST monikers could be used interchangeably.

[2] See here for a description of how the loan, reserve, and subsidy accounts with the PRGT function. And note that if MICs were eligible to draw from the GRT, the credit risk would be substantially larger than under the PRGT, thus requiring a heftier reserve fund.

[3] There is some possibility that SDRs could be donated or lent to the subsidy account to support subsidies, but these add another layer of technical complexity. See this article for an elaboration. Alternatively, creditors could lend SDRs at the same highly concessional interest rate that applies when they are on-lent to LICs. In this case, the subsidy cost is absorbed directly by the creditor.

[4] Liberal rollover policies can de facto result in longer terms, as has happened with many PRGT-eligible countries, but an explicit lengthening of the term would be consistent with the duration needed for the policy reform being supported.

Rights & Permissions

You may use and disseminate CGD’s publications under these conditions.