The Challenge of Reallocating SDRs: A Primer

Introduction

The approval of a new allocation of Special Drawing Rights (SDRs) by the International Monetary Fund (IMF) is the first effort to deal with the financial impact of the COVID-19 crisis on a global level. The extra $650 billion of reserve assets to be injected onto the balance sheets of the world’s central banks is an important response to the financial squeeze felt by many countries, especially lower- and middle-income countries (LMICs) that do not have the space for the kind of fiscal and monetary expansion that many developed countries have undertaken.

While the allocation will provide an important level of flexibility to these reserve-constrained countries, the majority of SDRs will be allocated to developed countries whose external reserves position is not constrained and already have the fiscal and monetary tools to react to the pandemic-induced economic downturn. Thus, there has been a call to reallocate (or recycle) some portion of developed countries’ SDRs to LMICs who could make use of more financing. ^[1],[2]

In principle the reallocation is easy to accept; in a time of global crisis, no resources should sit idle. But, in practice, an international agreement on how to reallocate the SDRs and for what purpose has been elusive. This obstacle is two-fold: there is not yet an international consensus on what the most pressing need is for these “new resources,” and, perhaps more importantly, there is a misunderstanding of what exactly SDRs are and how they function, making the technical decisions on reallocation difficult.

The purpose of this note is to reframe the concept of SDRs and then to outline in broad the types of proposals that have been mooted as a basis for more detailed work over the coming months.

We want to see what proposals are in line with international mandates for SDR use and what political considerations come into play, but most importantly outline the technical work that needs to be done so that the SDRs can be “put to work” when they become available in August 2021.

First, we will give some of the economic context for and background on the SDR allocation.

The economic context for SDRs

The economic strain put on the world economy by the COVID-19 pandemic is enormous. The global economic loss is estimated to be around $4 trillion in 2020, and while economic growth is expected to be positive in 2021 and 2022, the prospects for emerging and developing countries have dimmed, given slow vaccine rollout and the emergence of new strains of the virus. The risks to the outlook are firmly on the downside.

Faced with an unparalleled economic crisis in 2020, developed countries spent huge amounts of money (and gave considerable tax breaks) to support their populations. Advanced economies spent about 20 percent of their GDP (or about $10 trillion) in 2020 to support their people, and offered about 10 percent of GDP in loans, guarantees, and equity. To do this, they flexed their financial muscle through their central banks to borrow and to “print money” to support the added spending.

Emerging markets and low-income developing countries (LICs) did not have the economic muscle or flexibility to respond in a like manner. For emerging markets, the 2020 fiscal stimulus was about 5 percent of GDP, while in LICs it amounted to only 2 percent of GDP. The result was severe economic strain and deprivation leading to increased poverty throughout the developing world, with as many as 115 million people being pushed into extreme poverty (less than $1.90 per day) by 2021, bringing the global total to 150 million.

While one might expect developed countries to respond to an acute crisis in the developing world by stepping up their foreign aid substantially, the response was quite tepid. Official Development Assistance in 2020 increased by only 3.5 percent (in real terms) to $161.2 billion. Put another way, developed countries spent an extra $10 trillion on their own people in 2020, but only spent $6 billion more in aid—and many developed countries spent less. This compares to a loss of $2 trillion in economic output among emerging and low-income developing countries.

Background on the allocation^[3]

While the International Monetary Fund (IMF) and the World Bank stepped up their support during 2020, the global financial community coalesced around the idea of an allocation of SDRs as a way of giving all central banks around the world a quick and unconditional boost to their foreign currency reserves, which in turn would allow governments increased latitude in managing their internal and external finances. While some argued that an SDR allocation of as much as $3 trillion was needed, given the scale of the financing challenging emerging markets across the world, the figure of $650 billion was the most that could be approved without the assent of the United States Congress.^[4] The Trump administration did not support the allocation, and thus it was stalled until the Biden administration took office, allowing it to go forward. The allocation will now take place on August 23, 2021.

By IMF rules, the new SDRs will be allocated proportional to countries’ quotas at the IMF^[5]. Low-income developing countries will receive about $21 billion worth of SDRs, with middle-income emerging countries (excluding China) receiving a $212 billion boost to their reserves. The remaining $433 billion of SDRs will go to advanced economies, whose need for increased reserves is substantially less. Thus, many have argued that all or some part of the advanced countries’ share of the new SDRs should be reallocated to needier countries.

In the rest of this note, we first take a step back to understand in a bit more detail what SDRs are and are not. And then look at four types of reallocations that are being proposed and their technical feasibility.

SDRs are promises, not money

SDRs are a reserve sharing mechanism. They are not money in the traditional sense of the word as they cannot be used to purchase anything real. They are, however, a financial asset that can be used among sovereign nations as a medium of exchange; for example, a country can pay its debts to another country (or to certain international institution) in SDRs. Their value rests on a promise by all members of the IMF that, when a country needs so-called hard currency (e.g., dollars, euros, yens, sterling pounds, renminbi), it can exchange its SDRs for hard currency, which is in turn provided by another member of the IMF out of its reserves. The providing country then “holds” the SDRs as a promise that, if and when it needs more hard currency, it can exercise its right to ask yet another member to do a similar exchange.

An example might help. A (fictional) country Nambia, has seen a drop in the value of its main export and thus is short on the US dollars it needs to pay for vaccine imports, which are needed immediately. Nambia has 20 million SDRs in its account at the IMF and it approaches the IMF to exchange its SDRs for dollars. The IMF approaches a rich (fictional) country, Fritaly, which has plenty of dollars at its central bank. Since each SDR is worth $1.40, Fritaly agrees to give Nambia 28 million in US dollars in exchange for Nambia’s 20 million SDRs. Nambia then has enough dollars to buy its needed vaccines. This exchange gives Fritaly the power to purchase more hard currency in the future when needed in the amount equivalent to 20 million SDRs. Note that because of this exchange, neither Nambia’s nor Fritaly’s total foreign reserves have changed—all that has happened is that Nambia has exchanged SDRs, which are unusable for vaccine purchase for dollars for which Fritaly had no immediate use. Nambia is however now able to buy vaccines.

The willingness to increase the world’s SDRs signals countries’ willingness to share more of their hard currency reserves. The proposed increase is substantial, more than tripling the total SDRs from about $300 billion to just under $950 billion. Put another way, IMF member countries have agreed as of August 23 to make up to 7 percent of the world’s $12.7 trillion of reserves available to be shared as opposed to a ceiling of 2.4 percent now.

Whether the new shareable reserves are a lot or a little depends on one’s point of view. Central bankers tend to be extremely cautious in using reserves for purposes other than financial system defense, while others see such reserves as idle resources that could be put to some use, especially in times of emergency. This tension is not a new one, particularly in countries where resources are scarce, but it has been elevated by the economic crisis caused by the COVID-19 pandemic, which has raised fundamental questions about the adequacy and use of global reserves.

There does seem to be a global political consensus that wealthy countries, and some middle-income countries with lots of reserves, do not need the extra SDRs right now. They would be unlikely to switch their SDRs into hard currencies and thus the global sharing agreement would be more limited in scope, unless the advanced countries lent or donated their reserves to countries in need. But the advanced country financial custodians of these new reserves will be cautious. After all, the potential claims on their hard currency reserves have tripled. They want to be sure that in opening the global sharing agreement represented by SDRs, they have not opened their central bank purses to untransparent or wasteful spending. While there is no current threat to rich countries reserves, central bankers will want to avoid any future threat as well as the reputation or political risks that some uses of SDRs may entail.

Whatever one’s point of view, the question the world now confronts is how to mobilize these shareable reserves. The recent communique of the G20 finance ministers put forward clearly three requirements for any use of new SDRs (emphasis added):

We also invite the IMF to present proposals to enhance transparency and accountability in the use of the SDRs while preserving the reserve asset characteristic of the SDRs. In parallel, we ask the IMF to explore options for members to channel SDRs on a voluntary basis to the benefit of vulnerable countries, without delaying the process for a new allocation.

These three demands are of the IMF, but they represent the best consensus (as of April 2021 at least) of the international community as to how they are willing to share their collective reserves.

Two questions arise:

• can all three of the demands be met, and

• if not, where might there be a political consensus to compromise along one or more of the three dimensions?

The first is largely a technical question that involves the jurisprudence of using SDRs and central bank reserves, including understanding what exactly “the reserve asset characteristics entail,” while the second is a political one. But to the extent that technical solutions can be found to help meet these three criteria, the demand for political compromise will be lessened and the time needed to get SDRs “to work” will be shortened. Thus, examining the technical challenges surrounding SDR reallocation is important work to be done sooner rather than later.

A taxonomy of proposals for using SDRs

In the rest of this note, we will look at some of the proposals now being entertained by the IMF and proposed by other groups, grouped into five categories. We only intend to present the landscape here, with deeper technical analysis to be done later. While some of these proposals are already quite advanced in the technical and political dialogue, others deserve serious consideration in the months to come.

Before going into some of the specific proposals, there are three important technical constraints that inform all the choices to be made.

First, SDRs belong to individual countries, not to the IMF. Thus, any decision to reallocate or transfer SDRs must be made country-by-country. Groups such as the G7 or G20 might reach a consensus about what they want to do, but it is up to each country to act. The IMF plays no direct role other than accounting for any reallocations that take place. The only way the IMF might play an implementation role is in crafting facilities for its own lending, for which SDRs can provide sources of funds (see section 1 below).^[6]

Second, there is a cost to each country in reallocating its SDRs. Every IMF member country has been allocated a certain amount of SDRs, which are held at the IMF in the country’s SDR account. When it uses those SDRs—either by converting them to hard currency or transferring them to another SDR holder—its SDR holdings fall below its SDR allocation. When holdings are below the country’s allocation, it pays interest on the difference—currently almost zero (0.05 percent per year). (Similarly, if its holdings are above its allocation, it receives interest at the same rate). Thus, rich countries that donate any of their new SDR allocation by any of the schemes outlined in this note will pay interest, which will have to be budgeted according to each country’s procedures. While the interest rate is currently extremely low, it is updated daily, based on global interest rates and could increase rapidly.^[7]

Third, there is a distinction between lending and donating SDRs. If countries lend their SDRs, they remain the asset of the country. As with any lending, SDRs may earn interest, have some prospect of being returned after the life of the loan, but also incur some risk of non-repayment^[8]. This risk, along with the liquidity of the loan, is critical in evaluating whether the SDRs maintain their “reserve asset characteristic” when loaned. Countries may donate their SDRs to another country, making it a permanent transfer. In this case, they are no longer an asset of the donating country. For most countries, this amounts to a budgetary transfer, requiring approval by parliament. Furthermore, there is a perpetual annual cost to the donating country; it must pay the SDR charges forever (unless it comes by replacement SDRs in some other manner). Donated SDRs no longer constitute reserve assets for the donor country (but may for receiving country). ^[9]

Keeping these three technical issues in mind, we can look at five types of proposal being put forward to reallocate (or recycle) SDRs.

1. Proposals that keep the SDRs within the IMF

Various proposals have been made to repurpose the new SDRs to increase the financial support of the IMF to countries in need. All of these proposals involve loaning SDRs to facilities within the IMF.

The advantages of keeping the SDRs inside the IMF are two-fold: first, the financial structures within the IMF that receive the SDRs are often remunerative, have some risk guarantee and can be called back when needed (encashability) – the three reserve asset characteristics of the SDR. Second, the IMF criteria for on-lending to vulnerable countries are transparent; they also ensure appropriate safeguards to the use of the financing by vulnerable countries, overseen by the IMF’s Executive Board.

Those who are skeptical of the within-IMF proposals see some disadvantages: IMF on-lending limits the use of the SDRs in vulnerable countries to balance of payments problems, as defined by the IMF, and often requires countries to establish an IMF-supported program with associated conditions, which some view as either too high a bar or too slow to implement, inflexible or non-transparent. In some cases, the eligible countries are limited and the ability to access financing may be slow, given the IMF’s stringent requirements for lending. From a more political standpoint, IMF lending conditions are also seen in some quarters to serve the IMF and/or its more powerful shareholders, rather than the needs of its vulnerable member countries. Some countries fear a stigma attached to receiving loans from the IMF, that indicates a failure of their economic policies.^[10]

Current options being considered for inside-the-IMF lending include:

Augmenting the Poverty Reduction and Growth Trust (PRGT)

The PRGT is a trust fund within the IMF designed to support lending to a specified group of low-income countries (LICs). SDRs have been used in the past to support the PRGT and technically it is the quickest solution within current regulatory structures. We have discussed elsewhere the technical advantages and disadvantages of going this route. Its major limitations are its restriction to serving only LICs, limits on the amount of money that could be made available quickly, with differing views on the efficacy of the conditions needed to disburse the money.

Establishing new facilities with the IMF for short-term COVID-related needs

Many groups have noted the need for immediate financing to confront challenges arising from the pandemic, particularly vaccine financing.^[11] This seems like a legitimate reason to dip into the world’s central bank reserves through the SDR allocation, as there will be long lasting damage if the world is not fully vaccinated. If correctly designed, keeping such a facility within the IMF could preserve the reserve asset nature of the SDRs. The technical questions here revolve around getting the money “out the door quickly” and ensuring that it is available when needed and is put to appropriate use.^[12] There is a broader question whether the IMF is the appropriate institution to manage such a facility, given its predominance expertise of macroeconomics and international finance. Also, it is unclear if such a facility is needed right now, as simpler solutions exist (such as a window within existing IMF’s rapid financing facilities. Also, other facilities for vaccine financing are going unused and financing appears to be less of a problem than the logistics of distribution and administration. Moreover, the establishment of such a dedicated facility runs the risk that the resources are set aside for an event like the current pandemic but are then not used for an indeterminant amount of time.^[13]

Establishing facilities within the IMF for longer-term needs, including sustainability

For example, supporting health preparedness needs, including vaccinations and more broadly to support a green and equitable recovery in the most affected and vulnerable countries.^[14] The IMF is considering establishing what they are calling a Resilience and Sustainability Trust (RST). The COVID-19 pandemic has demonstrated the economic vulnerability that results from the integrated global economy and the inequitable impact of the crisis. It has reinforced the need for transition to a more sustainable and equitable economic structure. However, like any investment, the immediate costs of the transition will need to be financed so that the more-than-offsetting long-term benefits can be realized. Use of a small part of the world’s excess reserves to finance such a transition seems a reasonable proposition to many, and SDRs provide a mean to do so. The question again is whether this can be accomplished, in part, within the confines of the IMF to maintain the reserve aspect of the SDRs and the transparency in use that comes with IMF-supported economic reform programs. The IMF would need to take a new view of sustainability and the objectives and time horizon of balance of payments support. The Fund has done so in the past when poverty reduction and economic growth became more central to its financial support of low-income countries.^[15] This type of proposal raises a series of technical questions on how to design such a facility to evaluate the longer-term financial support needed country-by-country and appropriate financial programmatic structures to administer it, as well as whether longer-term support can be structured to maintain the reserve asset characteristics of the SDRs.

2. Proposals that use prescribed holders of SDRs as financing partners

SDRs are a reserve sharing mechanism for IMF member countries and cannot be used for any financial transactions other than with a limited number of financial institutions. They can be used in financial transactions with so-called “prescribed holders of SDRs,” 15 financial institutions outside the IMF that can hold SDRs as an asset. These include four central banks, three intergovernmental monetary institutions and eight multilateral development banks (MDBs), including the World Bank (both IDA and IBRD).^[16] These prescribed holders can, for example, receive SDRs as repayments for loans or for any other appropriate transfer from their own member countries. SDRs are an asset on their balance sheets but because they typically acquire SDRs in transactions as a convenience for their client countries, their SDR holdings are relatively small.

The advantage of using these prescribed holders is that the SDRs do not have to be converted to hard currency to augment the financial holdings of the prescribed holders, and thus they do not drain the hard currency reserve coffers of IMF member countries who are ceding their SDRs^[17].

Most of the proposals currently being mooted in this vein are to supplement the resources of multilateral development banks (MDBs) and regional development institutions. There are two types of proposals:

• Lend SDRs to MDBs so that the MDBs can in turn lend them to countries in need. The MDBs could then support financing in a particular area, where they are better placed to lend to developing countries than is the IMF (e.g., vaccines, climate mitigation/adaptation, agricultural support).

• Use the SDRs to increase strengthen the balance sheet of MDBs. SDRs would thus be leveraged to raise more lending resources in capital markets.

The key policy issue here is whether using SDRs at the MDBs could be done in a way that it would preserve the SDRs’ reserve asset characteristics.

• For loans to MDBs, the structure of the PRGT at the IMF could be mimicked: (i) the MDB compensates the SDR lender, either from the proceeds of the loans it makes to developing countries or from another pot of money designed for such compensation and (ii) the MDB guarantees that the SDR lender can retrieve its SDRs on demand.

• For use of SDRs to strengthen the capital base, preserving the reserve asset characteristic of the SDR appears to be much more difficult. First, it is unclear from the regulations governing the use of SDRs whether a country can use its SDRs to take an equity position in another entity. Second, it is likely such a capital contribution would be remunerated and whether conversion on demand could be supported. Further technical exploration is needed to understand whether there is a way to maintain the reserve asset characteristic or whether this demand would have to be abandoned if SDRs were used in this way.

(Again, an example may help. Suppose our fictional rich country, Fritaly, wanted to strengthen the financial position of a regional MDB (RMDB). If Fritaly lent SDR 10 billion to RMDB, its overall assets would not decrease (with the loan replacing the SDRs as the asset) and it may earn interest from its loan to RMDB to offset the interest paid to the IMF. But the loan could not count as a reserve asset unless RMDB could ensure that the claim remains liquid and that the risk, however small, of non-repayment is effectively mitigated.^[18] If instead Fritaly donated SDR 10 billion to RMDB. Fritaly’s reserves would decline by SDR 10 billion and FLARs assets would increase by the same amount. RMDB could then use those assets as the basis for increased lending. However, Fritaly would pay interest to the IMF on its deficit position in its SDR account.)

Another question arises as to why countries should donate their SDRs rather than other reserve currencies? While donating SDRs will keep hard currency reserves unchanged, on the balance sheet it will amount to spending reserves and will, thus be governed by the donating country’s rules regarding expenditure of reserves, and these vary considerably country to country. And the donation will result in an ongoing interest payment to the IMF. However, hard currency reserves are often invested and earn interest, so their use involves some risk and opportunity cost as well.

In essence, a decision to lend or donate or give SDRs to these prescribed-holder institutions would be a decision to let them use some portion of world’s currency reserves to strengthen their own balance sheets. That is a political question, but the technical question is whether SDRs are the best way of doing so if the political consensus exists.

3. Proposals that use SDRs for financial purposes outside of the prescribed holders of SDRs

There is at least one prominent proposal^[19] where SDRs would be lent to an outside-the-IMF fund designed to underpin commercial borrowing by developing countries that relatively recently accessed international capital markets and are thus particularly vulnerable during this crisis. These countries are currently paying higher interest rates than similarly situated countries with longer and deeper market exposure. This results from high perceived risk due to unfamiliarity, a relative scarcity of lenders, and the lack of a repo market to allow rapid exit of creditors if needed^[20]. Proponents make the case that a relatively small amount of assets from rich country central banks could serve as a backstop for a possible repo facility and thus overcome the imperfections due to an immature market. And they have proposed that the “excess SDRs” resulting from the allocation might provide the source of these funds.

The use of SDRs would directly benefit those countries that already had access to international capital markets, but whose market access was jeopardized by the pandemic-induced economic crisis and thus squarely fit within the mandate of the G20 that SDRs be used for vulnerable countries.

Such proposals have the same advantage as those for prescribed holders of SDRs, as they do not necessarily require the conversion of SDRs into hard reserve currencies as they would only provide a financial underpinning for the repo facility and could earn interest. However, this begs the technical questions as to where the SDRs would be held, as such as facility would not have standing as a prescribed holder. Possibilities would include at the IMF in some sort of administered trust or at one of the central banks that can hold SDRs.

Various other technical issues remain to be worked through. This use would not prima facie preserve the reserve characteristic of the newly allocated SDRs. Some technical fixes could be considered as with lending to the prescribed holders, but it would be a harder case to make given that such a facility would have no financial track record and thus be seen as riskier. The transparency of use of the SDRs would be clear, until such time as they were called on to cover any actual future default—why did default occur and what activities is the underlying capital covering? So technical work underpinning the structure of the fund would have to be completed. And like the use of SDRs for prescribed holders, the question arises: why use SDRs rather than hard currency reserves, given that it would take central bank acquiescence to establish such a fund.

4. Proposals to use SDRs as a short-term financing mechanism outside of the prescribed holders of SDRs

The short-term budgetary financing needs of developing countries are acute in the aftermath of the initial waves of the pandemic and could become more acute in the next 6-12 months if the rate of infection does not recede. Most vulnerable countries do not have the fiscal or monetary space to engage in the kind of unbridled spending those rich countries used in 2020 to prop up their economies and ensure the well-being of their citizenry. So, they need to look elsewhere for short-term financing.

In addition to increasing short-term financing from the IMF or the MDBs, there have been proposals to use SDRs to underpin short-term financing through other existing institutions special-purpose funds (education, health, agriculture). Many existing sector-specific funds would like to expand their support to needy countries and see the SDRs as a pool of money to be tapped to this end. And some see need for new funds.

These expenditure proposals are meritorious, and many would address the acute needs of vulnerable countries. The financial structure of these funds would have to be appropriate for using SDR either through donations or by leveraging SDR loans as leverageable capital.^[21] In essence, these proposals would use global central bank reserves for fiscal spending, and this makes the political hurdle high. Use of reserves through the guise of SDRs may appear to be a way to avoid direct budgetary transfers from rich countries to these various funds. But in fact, in most developed countries, SDR reallocation to such uses would be scored in the government’s budget or at least require legislative approval.

Setting the political problems aside, the technical questions would be roughly the same as those in the section immediately above, with one important addition. The SDRs would need to be converted to hard currency before they are transferred to the recipient funds, as such fund cannot be holders of the SDRs – SDRs are of no use in paying teachers or buying medical supplies. One solution would be to locate these funds within the MDBs and use MDBs hard currency reserves to make disbursements, but that brings us back to our second schema above. The other would be to argue that the new allocation has increased central bank reserves beyond what is needed and thus the central bank should spend its existing hard currency on these good causes, either as directly spending or as leverageable capital.

5. Proposals to use SDRs as long-term financing of country needs outside of the prescribed holders of SDRs

It is not just short-term financing to tide over vulnerable countries, but, as noted above, longer term financing is needed to begin the transition to a green and equitable future. As with short-term financing, some see a role for others than the IMF and MDBs in assuring this financing.

The challenges here are the same as those for short-term financing, but, if in the form of loans, these uses would tie up SDRs/central bank reserves for a much longer period. This increases the political stakes, but also raises two other technical issues worth consideration.

First, over time ensuring transparency and effective use of the SDRS would be more difficult to prove. Although MDBs, DFIs and other institutions/funds have established methods for tracking results, including through results-based financing, their effectiveness wanes with time and may give little solace to the world’s central bankers whose reserves are being encumbered. Straight up disbursement of the funds at the outset might be even harder to justify and would certainly sacrifice any notion of “reserve asset character.”

The second is the interest paid on using SDRs. As noted above, rich countries would incur an interest cost for any deficit in SDR holdings resulting from lending or donating SDRs. With the interest rate currently very low, this might be a tolerable expense for the short-term. But longer-term loans of SDRs or outright donations will be more expensive just because of duration. And they will be riskier as that the prospect for global interest developments is unknown.^[22]

Some concluding remarks

The impression some readers had after reading earlier versions of this paper was that the technical hurdles of reallocating SDRs, especially outside the IMF, are so great that there is no point trying. This impression is certainly true if there is no political will to think differently about what SDRs are and how they might be used. The technicians defending the sanctity of central bank reserve assets will find a host of technical and legal ways to block the use of SDRs in all but the most conservative ways.

But the world’s leaders can decide that they want to leverage (or even spend) some portion of the world’s $12.5 trillion in foreign reserves to confront the twin global crises of health and climate. While this could be done directly with the actual foreign exchange reserves, SDRs provide a mechanism to pool their resources, act collectively and thereby share and minimize the risk to any one country or to the entire financial system. Given the enormity of the challenges facing the globe, leveraging SDRs to provide the financial wherewithal to confront the challenges of Covid and climate seems a sensible thing to do.

Ongoing work at CGD will look in more depth at the technical challenges of the four alternative uses of SDRs exposited in this to provide policy makers. We hope to present a clearer picture of the pro and cons of the various options and the political hurdles that will need to be overcome, as a necessary ingredient to policymakers’ decisions over the next few months.