2021 Special Drawing Rights Allocation—Ex-Post Assessment Report
Author:
International Monetary Fund. Strategy, Policy, & Review Department
Search for other papers by International Monetary Fund. Strategy, Policy, & Review Department in
Current site
Google Scholar Close
and
International Monetary Fund. Finance Dept.
Search for other papers by International Monetary Fund. Finance Dept. in
Current site
Google Scholar Close
Publication Date:
29 Aug 2023
eISBN:
9798400254574
Language:
English
Keywords:
SDR obligation; SDR conversion; SDR channeling option; SDR interest rate projection; SDR assets
Download PDF (1.2 MB)

This report follows up on the impact of the historic $650 billion 2021 SDR allocation on the global economy, documenting IMF members' use of the allocation and assessing its economic effects. The report finds that the allocation was beneficial for the global economy, helping meet the long-term global need for reserves and supporting market confidence. Members used the allocation mostly to increase international reserve buffers, with some emerging market and developing countries also using it to meet fiscal and external financing needs. While SDR interest costs have increased, members’ capacity to service SDR obligations remains generally adequate. Members’ use of the allocation was mostly in line with Fund advice, and the transparency and accountability of SDR holdings and use has been broadly appropriate, although some gaps remain. Voluntary SDR channeling from economically stronger to more vulnerable members has helped amplify the benefits of the allocation.

Abstract

This report follows up on the impact of the historic $650 billion 2021 SDR allocation on the global economy, documenting IMF members' use of the allocation and assessing its economic effects. The report finds that the allocation was beneficial for the global economy, helping meet the long-term global need for reserves and supporting market confidence. Members used the allocation mostly to increase international reserve buffers, with some emerging market and developing countries also using it to meet fiscal and external financing needs. While SDR interest costs have increased, members’ capacity to service SDR obligations remains generally adequate. Members’ use of the allocation was mostly in line with Fund advice, and the transparency and accountability of SDR holdings and use has been broadly appropriate, although some gaps remain. Voluntary SDR channeling from economically stronger to more vulnerable members has helped amplify the benefits of the allocation.

Introduction

1. In August 2021, the Board of Governors (BoG) of the IMF approved a general allocation of Special Drawing Rights (SDRs) equivalent to USD 650 billion. At the time, the COVID-19 crisis had taken a catastrophic humanitarian toll, causing considerable stress in the global economy and threatening the stability and smooth functioning of the international monetary system (IMS). This SDR allocation, which is the largest in the IMF’s history, became effective on August 23, 2021 and was aimed at addressing the long-term global need for reserves and boosting confidence through a powerful multilateral response to the COVID-19 crisis (Box 1).1

2. At the time of the approval of the allocation, staff committed to prepare an ex-post report two years after its implementation. The report was to “review the allocation against the broad macroeconomic context and policy priorities following the COVID-19 pandemic” and to “discuss broad patterns of holdings and exchange of SDRs into freely usable currencies; the use of SDRs for transactions with the Fund or ex-post voluntary channeling; and a broad characterization of public spending and macroeconomic trends (e.g., reserve buffers, inflation, and growth) […] [as well as] examine the potential effects of the allocation on the stability and resilience of global financial markets (e.g., sovereign risk)” (IMF 2021a). The report was among several measures proposed by staff and endorsed by the Executive Board in response to the call by the International Monetary and Financial Committee to enhance transparency and accountability in the reporting and use of SDRs.2

3. In line with the above commitment, this paper analyzes the implications of the allocation from multiple perspectives. It first discusses the impact of the allocation on global reserves, confidence, and borrowing costs, as well as on global inflation. The next section documents members’ use of the allocation—which can be either retained as international reserves or used to cover fiscal or external financing needs—and assesses the economic implications. Section IV provides staff’s assessment on whether the use of the allocation was in line with Fund policy advice and transparency and accountability principles. This is followed by looking at the voluntary channeling of SDRs from economically strong members to the most vulnerable members and a summary of conclusions. The findings of this paper are supported by the IMF’s publicly available SDR Tracker, novel empirical and theoretical analyses, surveys of members and IMF mission chiefs, and recent safeguard assessment reports.

Special Drawing Rights (SDRs)—A Primer1

The SDR is an international and unconditional reserve asset created by the IMF in 1969 to supplement its members’ official reserves. It is a potential claim on the freely usable currencies of IMF members. As such, SDRs can provide IMF members with liquidity. The value of the SDR is based on a basket of five currencies: the U.S. dollar, the euro, the Chinese renminbi, the Japanese yen, and the British pound sterling.

Only participants in the SDR department (currently all IMF members), prescribed holders, and the IMF itself, can hold, buy, and sell SDRs. Currently, there are 20 organizations approved as prescribed holders. Only participants in the SDR department can receive allocations of SDRs.

uA001fig01

SDR Allocations

(SDRs Billions)

Citation: Policy Papers 2023, 035; 10.5089/9798400254574.007.A001

Source: IMF Finance Department

To date, a total of SDR 660.7 billion (equivalent to about USD 943 billion) have been allocated. SDRs are allocated to countries proportionally to their paid quota share in the IMF. So far, four general SDR allocations have been approved, including during the global financial crisis in 2009 (SDR 161.2 billion) and following the COVID-19 pandemic in 2021 (SDR 456.5 billion).2

Members can use their SDR holdings unconditionally. Upon allocation, IMF members receive SDRs as a foreign-currency asset (the SDR holding), with a corresponding liability to the Fund (the SDR allocation). Some members hold SDRs to boost their reserves. Others may buy or sell SDRs for currencies or use them for other financial operations authorized by the IMF Executive Board. The decision about using the allocated SDRs rests with each member.3

A participant or prescribed holder in the SDR Department may exchange SDRs for currencies bilaterally or through the Voluntary Trading Arrangements (VTA) market. VTAs are arrangements between the Fund and participants and prescribed holders who have agreed to buy and sell SDRs against freely usable currencies. The role of the IMF in this market is to act as an intermediary, matching willing buyers and sellers of SDRs. The VTA market is backstopped by the designation mechanism under Article XIX of the Fund’s Articles of Agreement. Under this mechanism, if a participant with a balance of payments need requests to exchange SDRs but no willing counterparty can be identified through the VTA or bilaterally, the Fund may designate participants whose balance of payments and reserve positions are deemed sufficiently strong to provide freely usable currencies to the requesting participant in exchange for its SDRs. The Fund has not had to activate this designation mechanism functions since 1987.4

A participant earns interest on SDR holdings and pays charges on cumulative allocations, both at the SDR interest rate. When a participant’s SDR holdings fall below its cumulative allocations, it would incur a net interest expense. The SDR interest rate is determined weekly, based on the weighted average of the interest rates of short-term financial instruments of SDR basket currencies.

SDR Department participants with strong external positions have historically used some of their SDR holdings to help countries in need. Specifically, several countries have used part of their SDR holdings to scale up the IMF’s PRGT’s loan resources and contribute to the RST.

1 For additional background on SDRs, see Questions and Answers on Special Drawing Rights. 2 A special one-time allocation of SDR 21.5 billion took effect on September 9, 2009 to correct for the fact that members that had joined the IMF after 1981 had never received an allocation (the Fourth Amendment special allocation). 3 Members do not have to meet any requirements to receive their share of SDRs in a general allocation and they have the right to use SDRs in transactions (i.e., to exchange their SDRs to obtain currencies) or in operations authorized by the Fund (including, payments of financial obligations, loans, pledges donations, swaps, and forward transactions) with other SDR Department participants or prescribed holders. 4 Prescribed holders cannot request an exchange of SDRs in transactions through designation.

Impact of the Allocation on the Global Economy

This section examines the impact of the allocation on global reserves, confidence and borrowing costs, and global inflation. It also presents the members’ views on the timeliness and usefulness of the allocation for the global economy.

A. International Reserves, Confidence, and Global Inflation

4. The 2021 SDR allocation aimed at meeting the long-term global need to supplement existing reserves. As described in the 2021 Proposal for a General Allocation of Special Drawing Rights, in line with Article XVIII of the Fund’s Articles of Agreement, the Managing Director can make a recommendation for an SDR allocation “to meet the long-term global need…to supplement existing reserve assets in such a manner as will promote the attainment of its purposes and will avoid economic stagnation and deflation as well as excess demand and inflation in the world.” At the time of the allocation, staff estimated the long-term global need for reserves assets to be around USD 1.1–1.9 trillion.3 After considering non-SDR funding sources, and in line with the methodology used in 2009, the 2021 allocation covered a significant share of the estimated residual long-term global reserve need.4

5. The allocation boosted reserves at a time of unusual uncertainty and stress in the global economy.5 Members’ gross, and in most cases net, international reserves increased once the allocation became effective.6 Although the bulk of the allocation went to AEs, in line with members’ respective quota shares, the allocation as a share of GDP and reserves was larger in LICs than EMs and AEs (Figure 1).

Figure 1.
Figure 1.

Impact of the 2021 SDR Allocation1

Citation: Policy Papers 2023, 035; 10.5089/9798400254574.007.A001

Source: IMF data and staff analysis1 Allocation corresponds to the increase in SDR holdings.

6. The allocation helped bolster confidence and lower borrowing costs, particularly for credit-constrained EMDCs. The SDR allocation was less costly than borrowing from markets at the time of implementation and provided rapid, unconditional liquidity to all members, including non-market access LICs, without contributing to global imbalances or posing immediate rollover risk. Staff analysis indicates that the allocation helped reduce borrowing costs for EMDCs, improving overall confidence (Box 2). Specifically, EM’s sovereign risk premia (i.e., credit-default swaps, CDS) declined by around 3–4 percent in the week following the allocation. The decline was larger in countries with higher sovereign spreads prior to the allocation.

7. The allocation does not appear to have been a material contributor to increased global inflation. The allocation represented only a small fraction (less than 0.5 percent) of total global broad money in 2021 and was implemented when the global output gap was large and negative (IMF 2021a). Moreover, only about 5 percent of the total allocation was exchanged into freely usable currencies (see next section), and monthly SDR sales via VTAs (on average USD 850 million) were much smaller than AE central bank balance sheet expansion (about USD 210 billion per month) during March 2020–April 2023 (Figure 2).7,8 While both demand and supply factors contributed to the global inflation surge since 2020 (IMF 2022d), there is no evidence that SDR sales by EMDCs (as AEs did not exchange their allocated SDRs) were a key factor. This is further confirmed by the survey results in which authorities and mission chiefs indicate that the allocation did not materially impact inflation (see background paper).

Figure 2.
Figure 2.

SDR Exchanges and Changes in Major Central Bank Asset Holdings

(in billions USD, monthly change)

Citation: Policy Papers 2023, 035; 10.5089/9798400254574.007.A001

Source: IMF WEO, Finance Department data, and staff analysis.

Impact of the SDR Allocation on Sovereign Spreads

Staff conducted an event study analysis on the impact of the allocation on CDS spreads for 49 AE and EM members.1 Daily data from August 23 through August 31, 2021, indicate that CDS spreads declined by an average of 2.97 basis points or 2.88 percent a week after the SDR allocation was made effective (Box Figure a).

The analysis was refined to control for global and country-specific factors using regression analysis. The analysis is based on the following specification:

In(CDSi,t)=α+βSDRi,tDummy+γ′In(Xi,t)+Ii+Qt+ϵi,t

where CDSi,t represents country i’s CDS spreads on day t. The SDR dummy variable (SDRi,tDummy) is the variable of interest, taking a value of 1 after the allocation and a value of 0 before. Xi,t represents control variables: (i) the bilateral spot exchange rate vis-à-vis the USD and (ii) a domestic equity market index. The regression includes country fixed effects (Ii) and time fixed effects (Qt). The sample period is August 2–31.

uA001fig02

(a) Decline in CDS Spreads following the 2021 SDR Allocation

Citation: Policy Papers 2023, 035; 10.5089/9798400254574.007.A001

Source: IMF staff analysis.

Results suggest that sovereign risk premia declined after the SDR allocation (box table). The β coefficient, which captures the effect of the allocation, is between 0.03–0.04 and statistically significant in the linear specifications (table columns 1–3). This suggests that on average CDS spreads are about 3–4 percent (around 3–4 basis points) lower in the post-SDR allocation period once country and time fixed effects are considered. Such estimates may be viewed as a lower bound of the overall effects, as the spreads might have already reacted to plans for the allocation prior to its approval in August 23—for instance, when the IMFC called for the allocation (April 8) or when it was approved by the IMF Executive Board (June 25).

uA001fig03

(b) Decline in CDS Spreads Conditional on Countries’ Pre-existing Vulnerabilities1

Citation: Policy Papers 2023, 035; 10.5089/9798400254574.007.A001

1 Pre-existing vulnerabilities measured by the CDS levels prevailing prior to the allocation, where countries are grouped into four categories according to their quartile ranking in the relative values of their CDS levels.Source: IMF staff analysis.

More vulnerable countries benefitted most from the SDR allocation, as they saw larger relative declines in their CDS spreads. Staff conducted additional analyses to examine how pre-existing financial conditions affected the impact of the allocation, by differentiating country groupings based on the distribution of their CDS spread levels that prevailed prior to the allocation. Results suggest that countries in the top quartile of CDS levels prior to the allocation saw a more significant reduction in their CDS spreads relative to countries in the other quartiles. Specifically, the top quartile of countries based on CDS spreads saw a decrease of 6.7 basis points, compared to 1.8 basis points for other (i.e., relatively more credit-worthy) quartiles. This finding is robust to controlling for global and country-specific factors through the following non-linear regression:

In(CDSi,t)=α+β1SDRi,tDummy+β2SDRi,tDummy×Groupi+γ′In(Xi,t)+Ii+Qt+ϵi,t

where Groupi represents a categorical variable for country i’s quartile ranking in the distribution of their CDS spreads, and the coefficient β2 captures the interaction effect between the SDR allocation and country’s original CDS’ spreads levels (results reported in the “Non-linear Specification” column).2,3

article image
Note: All specifications include country and time fixed effects. ***, **, and * stand for 1, 5, and 10 percent significance, respectively.
1 The sample does not include LICs and some EMs as high frequency data were not available. 2 After controlling for global and country-specific factors, the regression results (non-linear specification in the Box table) indicate that the allocation reduced CDS spreads by an additional 0.6 percent for countries in the 2nd quartile; 1.2 percent more for countries in the 3rd quartile, and 1.8 percent more for countries in the top quartile, when compared to countries whose pre-allocation CDS spreads fall into the bottom quartile. 3 Staff also found that the reduction in spreads was greater for EMs relative to AEs in a statistically and economically significant manner, thereby confirming the finding that the SDR allocation benefited the more vulnerable countries.

B. The Membership’s Views

8. Members generally perceived the SDR allocation as helpful, necessary, and timely for the global economy (Figure 3). The survey of member authorities indicates that all AE and EM respondents and about 95 percent of LIC respondents agreed that the SDR allocation was helpful for the global economy.9 Moreover, a large majority of members across all country income groups agreed that the allocation was timely and necessary for the global economy. The majority of surveyed member authorities also considered that the allocation’s size was sufficient for the global economy. However, 35 and 50 percent of EM and LIC respondents, respectively, considered it insufficient to meet the membership’s financing needs, while recognizing that the allocation was an important part of the global response. Some respondents further noted that allocating a greater proportion to EMs and LICs rather than to AEs would have been preferable.

Figure 3.
Figure 3.

Impact of the SDR Allocation on the Global Economy: The Membership’s View

(In percent of total respondents)

Citation: Policy Papers 2023, 035; 10.5089/9798400254574.007.A001

Source: Country authorities’ survey; staff analysis.

Economic Implications at the Country Level

This section focuses on how members utilized the allocation, from the immediate impact on their SDR holdings, to conversions and the ultimate economic use of SDRs. It examines patterns of SDR holdings and conversions and the macroeconomic characteristics of SDR converters. It discusses how members used SDRs, examining country groupings, specific uses and impact on IMF programs. It also assesses the implications of higher interest rates for SDR converters and concludes with members’ and staff’s views and assessment of whether its use has been in line with Fund policy.,

A. Changes in SDR Holdings and Conversions after the Allocation

9. Members’ SDR holdings generally increased following the 2021 allocation, reflecting the main utilization as a reserve buffer, although some countries used their holdings for repayments to the Fund and other purposes (Table 1, Figures 4 and 5). SDR holdings can be either kept as international reserves, directly used in transactions with the Fund (GRA repurchases and charges; PRGT repayments; contributions to concessional trusts), converted into freely useable currencies through VTAs, or used for a range of other financial operations (see background paper). Currencies received from the conversion of SDRs can, in turn, be retained as reserves buffers or used for other fiscal or balance of payments financing purposes. AEs generally held on to their SDR allocations and VTA participants increased their holdings further to allow conversions by EMDCs, with the aggregate holdings-to-allocation ratio rising to about 104 percent at end-June 2023 from 98 percent in July 2021.10,11 The aggregate SDR holdings-to-allocation ratio of EMs and LICs was 88 and 54 percent respectively, as of end-June 2023, compared to 74 and 70 percent right before the allocation. Despite the sizeable allocation, a considerable number of EMDCs have already nearly exhausted their SDR holdings: 33 EMDCs (of which 19 LICs) have SDR holdings-to-allocation ratios below 5 percent and 9 EMDCs (of which 4 LICs) have ratios in the range of 5–20 percent.

Table 1.

SDR Allocations and Holdings at end-June 20231

(SDR billions)

article image

When the IMF allocates SDRs, participants in the SDR Department receive unconditional liquidity represented by an interest-bearing reserve asset (SDR holding) and a corresponding long-term liability to the SDR Department (SDR allocation).

Source: IMF data.
Figure 4.
Figure 4.

EMDCs’ SDR Uses since the 2021 Allocation

(In percent of total inflows from the 2021 allocation through the indicated date)

Citation: Policy Papers 2023, 035; 10.5089/9798400254574.007.A001

Source: IMF data
Figure 5.
Figure 5.

Changes in the Distribution of SDR Holdings-to-Allocation Ratio1

(Number of countries indicated on each bar)

Citation: Policy Papers 2023, 035; 10.5089/9798400254574.007.A001

1 The 2021 allocation bar corresponds to observations at end-day of August 23, 2021.Source: IMF data.

10. LICs exchanged a sizeable portion of their SDRs for freely usable currencies through the VTA market (“converters”) (Figure 4). During August 2021–June 2023, SDR sales by IMF members through VTAs amounted to SDR 26 billion (5.7 percent of the total 2021 SDR allocation) and involved 139 sales requests by 64 members (about one third of the Fund’s membership).12,13 LICs accounted for one third of the total sales volume (SDR 8.3 billion, equivalent to 56 percent of total SDRs allocated to them in 2021). After a sharp increase of SDR sales in September 2021 (SDR 4.3 billion), the sales volume declined to a monthly average of about SDR 1 billion more recently (Figure 6). While the number of VTA sales increased after the start of Russia’s war in Ukraine, the volumes did not, reflecting conversions by smaller countries.14 Meanwhile, SDR transactions with prescribed holders remained modest.15 In tandem with the allocation, the trading capacity of VTAs and number of participants was expanded.16 As of end-June 2023, the buying and selling capacities of VTAs stood at about SDR 210 billion and SDR 163 billion, respectively. This trading capacity— which was expanded and strengthened after the 2021 allocation—together with the broad regional representation of the VTAs, is expected to be able to meet future demand for SDR-to-currency exchanges and continue supporting the smooth functioning of the SDR market.17

Figure 6.
Figure 6.

VTA Sales by IMF Members, 2006–2023

Citation: Policy Papers 2023, 035; 10.5089/9798400254574.007.A001

Source: IMF data.

11. A large number of LICs and several EMs used their SDR allocation to settle Fund obligations. The SDR allocation helped 42 countries to make PRGT repayments, totaling SDR 1.2 billion, providing welcome financing space for LICs facing a broader funding squeeze. For instance, countries in Sub-Saharan Africa have been confronted with a substantial decline in aid budgets, from about 4 percent of regional GDP in 2006 to about 2.5 percent in 2022, as well as declining Eurobond issuances amid the tightening global financial conditions.18 On the GRA side, 13 members met their obligations in SDRs, with total GRA repurchases in SDRs totaling SDR 26.3 billion.19

12. SDR converters tended to exhibit higher vulnerabilities (Box 3). Staff analysis suggests that both LIC and EM converters tend to have lower reserves and larger current account deficits than non-converters. EM converters also tend to have lower GDP growth and higher public and external debt levels. Finally, countries with small SDR holdings (relative to their cumulative allocation) before the 2021 allocation tended to convert the newly allocated SDRs more than others (Figure 5).

13. Countries with Fund-supported programs converted a larger share of their allocated SDRs into freely usable currencies, compared with those without a program (Figure 7). According to SDR sales data, 33 countries with IMF-supported programs (“program countries”)20 at the time of SDR allocation converted, on average, 48 percent of their allocated SDRs to freely usable currencies, compared to an average of 25 percent in the 157 countries without IMF programs (“non-program countries”).21 Including program countries under through end-2022 reveals an even starker difference, with 49 program countries converting on average 54 percent of the allocated SDRs compared to 21 percent in non-program countries. This difference likely reflects higher vulnerabilities and financing needs in program countries, necessitating larger conversions of SDRs.

Figure 7.
Figure 7.

Conversion of SDRs: Program vs. Non-Program Countries

(In percent of total SDR allocation)

Citation: Policy Papers 2023, 035; 10.5089/9798400254574.007.A001

Sources: IMF staff analysis.

Analyzing Characteristics of SDR Converters

Staff conducted a correlation analysis to analyze the characteristics of members that converted SDR holdings after the allocation. Specifically, pairwise correlation coefficients of selected economic indicators and a binary variable indicating whether a country converted its SDR holdings since the 2021 allocation, were computed.1 This coefficient measures the association between country decisions to convert their SDR holdings and their macroeconomic characteristics.

Several macroeconomic characteristics were considered. These include real GDP growth, international reserves, the current account balance, public debt, external debt, and sovereign spreads.2 For sovereign spreads, averages of daily data points for the J.P. Morgan Emerging Market Bond Index (EMBI) between July 20 and August 20, 2021 were used. The sample included 69 LICs and 94 EMs, although the final sample varied across different macroeconomic characteristics depending on data availability. The correlation coefficients for both LICs and EMs are presented below. Overall, the estimates suggest that SDR converters tend to exhibit relatively higher vulnerabilities.

Correlations between SDR Conversion and Selected Macroeconomic Characteristics

article image
Note: green (yellow) shaded cells correspond to correlation coefficient greater than 0.1 (less than -0.1).
1 The conversion dummy takes a value of 1 if the country had at least one SDR sale transaction in the VTA market between August 23, 2021 and end of June 2023. 2 For all variables except sovereign spreads, the data used to capture these characteristics were based on the July 2021 vintage of the IMF’s WEO database, to capture the characteristics of countries just before the allocation. As a robustness check, staff conducted the same exercises using averages over April 2021–July 2022, since several countries converted their SDR holdings at a later date after the allocation; the results were similar.

B. Economic Uses of the SDR Allocation

14. The central policy question for members is whether to retain SDRs as reserves or convert them to finance fiscal or external needs. Welfare analysis based on DSGE model simulations suggests that the optimal policy response to the SDR allocation—namely maintaining SDR holdings to boost reserves, converting SDRs to support the economy, or a combination of both—depends on country-specific characteristics, including policymakers’ preferences and the sensitivity of critical macro parameters (Box 4). Specifically, the more sensitive a country’s borrowing costs are to changes in reserve coverage, the larger the share of SDRs kept as reserves under the optimal policy response. In addition, it may be beneficial to retain SDRs to preserve future policy space as long as there is no pressing need to convert them.

Modeling the Policy Implications and Tradeoffs of the SDR Allocation1

Staff developed a small open economy DSGE model to capture countries’ trade-offs and quantify the costs and benefits of using SDRs. In this novel setup, the central bank’s balance sheet is explicitly modeled, where the allocated SDRs constitute both an asset and a liability. The country’s risk premium (Φ) is an endogenous function of international reserves (Res):

Φt=φae−φrRe⁡st+ϵt

where φa is the average risk premium, φr > 0 the sensitivity to reserves, and ε an exogenous, autoregressive component. In this framework, holding the allocation as additional reserves boosts the country’s liquidity buffers, contributing to lower sovereign risk premia. Alternatively, the allocated SDRs can be converted into domestic or freely usable currencies and used to purchase domestic or foreign bonds to smooth consumption in the presence of adverse shocks, gain policy space, and avoid contractionary and distortionary policies. The central bank is assumed to maximize welfare by adjusting the size and composition of its balance sheet, according to the following identity:

Mt+S¯t=Bt+Ft+St

where M is base money, S¯ the exogenous allocation of SDR, and B, F, and S are the amounts of domestic bonds, foreign currency assets, and SDRs the central bank chooses to hold on the asset side (with Res = F + S). Since SDR access is limited to the size of the allocation (0≤S≤S¯), converting the allocated SDRs today reduces SDR holdings, and therefore the ability to rely on them in following periods.

Policymakers face important trade-offs regarding the use of the allocated SDRs. First, each type of asset entails different benefits (in terms of reducing borrowing cost versus allowing for consumption smoothing). This drives the intra-temporal decision on whether to convert or retain the allocated SDRs as reserves. Second, there is an inter-temporal trade-off between converting now (to increase either domestic or foreign assets on the central bank balance sheet) or later.

uA001fig04

(a) Utility under different conversion patterns following an SDR allocation (relative to steady state)

Citation: Policy Papers 2023, 035; 10.5089/9798400254574.007.A001

Source: Bloomberg and IMF Staff Calculations
uA001fig05

allocation (for high and low sensitivity φr)

Citation: Policy Papers 2023, 035; 10.5089/9798400254574.007.A001

The optimal response depends on country-specific characteristics. To illustrate this, staff calibrated the model for two types of economies and carried out stochastic simulations assuming two distinct SDR conversion patterns: “full conversion,” where the allocated SDRs are fully converted, and “full retention,” where the allocated SDRs are fully kept as reserves.

  • Figure (a) shows the impulse response of utility following an SDR allocation and illustrates the inter-temporal tradeoff in the decision to convert versus retain the allocated SDRs. Conversion entails a higher positive change in utility initially (as it boosts domestic demand), while utility is higher under retention in the outer years.

  • Another important characteristic is the sensitivity of the country’s risk premium to changes in reserve coverage. Figure (b) plots the discounted sum of utility after an SDR allocation under two alternative values for φr. If markets are sensitive to changes in a country’s reserve levels, it is optimal for the central bank to retain SDRs as reserves to lower borrowing costs. By contrast, in a country with less sensitive spreads, it is optimal to convert the allocated SDRs and support the economy. To the extent that EMs are more highly integrated with international financial markets than LICs, and their borrowing costs more sensitive to fluctuations in global conditions, the model predicts that EMs should maintain a larger share of their allocated SDRs as reserves than LICs.

1 Based on N. End, J. Koosakul, K. Tanyeri, and C. Tovar (forthcoming).

15. Most members used the SDR allocation to increase foreign reserves and/or finance fiscal needs, broadly in line with staff’s theoretical analysis (Box 4). According to the SDR Tracker (Box 5), the most prevalent use of the allocation was holding them as reserves, with 115 countries (out of 142) at least partly doing so.22 Fiscal support is the second most reported use of the allocation, with 46 countries at least partly using the allocation for this purpose.23 There are differences across country groups:

  • AEs have used the allocation to increase foreign exchange reserves: All 17 AEs covered by the Tracker increased their foreign reserves using the SDR allocation. One country (Greece) used part of the SDR allocation to finalize the repayment of obligations to the IMF early. No AE reported using SDRs for fiscal support.

  • Most EMs used the allocation to boost their reserves: 61 EMs (out of 72 covered by the Tracker) used some or all the SDR allocation to boost their foreign exchange reserves. In addition, 13 EMs used at least part of the SDR allocation for fiscal support, six for debt repayment, three for external financing needs, and two for clearing arrears.

  • LICs had more varied use of the SDR allocation, including for pandemic-related needs: 33 LICs (out of 53 covered by the Tracker) at least partly used the allocation for fiscal support, including explicitly for pandemic-related spending (e.g., Gambia, Guinea Bissau, São Tomé and Principe, Senegal, and Zimbabwe). 38 LICs held (at least part of) the allocation in reserves, nine used the allocation for debt repayment,24 six used it to cover external financing needs and three to clear arrears.

Uses of the Allocation based on the SDR Tracker

The IMF launched the SDR Tracker to help understand and document how countries used the 2021 SDR Allocation, while promoting transparency and accountability. Per the 2021 SDR Guidance Note, country teams were expected to cover members’ treatment and use of the SDR allocation in IMF Staff Reports. The SDR Tracker compiles excerpts from IMF Article IV Staff Reports on the use of SDRs. It was launched in April 2022, with bi-monthly updates thereafter.

A vast majority of the membership was covered by the Tracker. In total, 142 out of 190 members (or around 75 percent of the IMF’s membership) were covered. Coverage was at 77 percent for both EMs and LICs, but only 63 percent for AES, possibly given the lack of systemic importance of their allocation. Uses of the allocation included increasing reserves, fiscal support, debt repayment, other external financing needs, channeling, and clearing arrears. EMDCs, especially LICs, used the allocation for non-reserve purposes (see charts).

uA001fig06

Uses of the SDR Allocation Based on the SDR Tracker

Citation: Policy Papers 2023, 035; 10.5089/9798400254574.007.A001

Source: SDR Tracker and IMF Staff Reports

The allocation helped support EMDCs’ response to the COVID-19 pandemic. For instance, the Gambia used 25 percent of the allocation to mitigate the impact of the pandemic, including for a facility that could help bolster vaccination rates. Guinea-Bissau used part of the allocation to fund vaccination and improvements in health services. São Tomé and Príncipe used about half of the allocation to finance hospital renovations and acquisitions of medicines and medical equipment, while Senegal used part of its allocation on the health sector and domestic vaccine production. Among EMs, North Macedonia used SDRs for pandemic-related subsidies to vulnerable households, vaccines, and other health-related expenditures. Bosnia and Herzegovina used the entire allocation for pandemic-related spending, while Paraguay financed expenditures related to the COVID-19 emergency plan.

The SDR allocation also facilitated debt management operations. Several EMs used their SDRs to repay debt, including Sri Lanka and Ukraine. Colombia’s central bank sold foreign exchange to the Finance Ministry (an amount equivalent to the SDR allocation) in exchange for local currency treasury bonds at market prices, boosting its liquidity without debt issuance. Several LICs used the allocation to reduce debt vulnerabilities and costs—for instance, Liberia used the SDR allocation to retire T-bills, while Papua New Guinea used it as a source of inexpensive financing. Gabon used its entire 2009 and 15 percent of its 2021 SDR allocation to repay domestic debt and improve the composition of domestic financing.

C. Implications for Fund-Supported Programs

16. Program countries used a substantially larger share of their SDRs for fiscal support (Figure 8). Data from the SDR Tracker indicates 54 percent of LICs and 22 percent of EMs with a Fund program used SDRs for fiscal support, compared to 31 and 11 percent of non-program LICs and EMs, respectively. In contrast, 75 percent of non-program EMs and 43 percent of LICs maintained the allocated SDRs as foreign reserves, compared with 58 percent and 35 percent, respectively, with Fund-supported programs.

Figure 8.
Figure 8.

Program vs. Non-Program Countries

(Use of SDRs based on Staff Reports, in percent of countries in the SDR Tracker)

Citation: Policy Papers 2023, 035; 10.5089/9798400254574.007.A001

Sources: the authorities’ survey, SDR tracker, and IMF staff calculation.

17. Program targets were adjusted in line with program objectives and country-specific circumstances (Table 2). According to the IMF MONA database, among 23 programs at the time of SDR allocation,25 11 had fiscal targets relaxed (primary balance target lowered or Net Domestic Assets (NDA)/Net Domestic Financing (NDF) targets raised) to accommodate additional fiscal spending.26 Fiscal targets were unchanged in 4 programs and tightened in 3 programs, reflecting a stronger need for fiscal consolidation in countries with higher risk of debt distress. At the same time, reserve targets (NIR, GIR, or NFA) were increased in 10 programs to rebuild external buffers. In 3 program countries, however, reserves targets stayed unchanged, and in one they were lowered, reflecting program definitions27 or unexpected adverse shocks. In several cases, changes to program targets included both relaxing the fiscal stance and increasing reserves.

Table 2.

Change in Program Targets following the 2021 SDR Allocation1

(Number of programs)

article image

Comparison between program approval/reviews before and immediately after the SDR allocation.

“N/A” indicates cases where relevant program targets (QPCs) were not specified, or data are unavailable.

Sources: MONA database, IMF staff calculations.

18. Post-allocation adjustments to program targets allowed countries to smooth fiscal adjustment and/or enhance external buffers. Comparing World Economic Outlook (WEO) projections in April 2021 (pre-allocation) and October 2021 (post-allocation) can provide some insight into the degree to which fiscal and reserves paths were modified following the SDR allocation.28 That said, the data should be interpreted with caution, as projections may reflect other changes that occurred during this period. For countries that relaxed program fiscal targets, the projected PB after the allocation envisaged, on average, a modest deterioration of around 0.2 percent of GDP per year in 2021–23. For countries that raised their program reserves targets, the post-allocation data indicate an average increase in reserves coverage (in months of imports) of about 0.4 in 2021, declining to 0.3 and 0.1 in 2022 and 2023, respectively (Figure 9). In contrast, projections for non-program converters show a tighter fiscal balance post-allocation, indicating that these countries used the allocation for other purposes. Upward adjustments in reserves paths for these countries were relatively higher than for program countries.

Figure 9.
Figure 9.

Changes in Staff Projections (April 2021 vs. October WEO)

Citation: Policy Papers 2023, 035; 10.5089/9798400254574.007.A001

1 Average data for countries under a program at the time of SDR allocation. The left panel covers 11 countries that relaxed PB or NDA/NDF targets. The right panel covers 10 countries that raised NIR or GIR targets.2 For non-program countries that converted SDRs, there were sizable data revisions for 2020 between the two WEO vintages.Sources: MONA database, WEO database, IMF staff calculations.

19. The SDR allocation appears to have temporarily reduced demand for Fund financing as countries benefitted from stronger reserve buffers (Figure 10). The initial surge in emergency lending likely explains the slower pace of new upper-credit tranche (UCT) program requests following the pandemic compared to that observed during the global financial crisis. While the pace of emergency lending slowed, approvals of new UCT arrangements plateaued from July 2021—to November 2021, when SDR sales by members reached record highs. While a full-fledged empirical analysis is beyond the scope of this report, the data for non-program countries with relatively modest reserves coverage suggests that the additional substantial external buffers created by the SDR allocation may have reduced or slowed demand for Fund financing during that period (Figure 11).

Figure 10.
Figure 10.

Number of New UCT Arrangements

(Monthly cumulative, from crisis onsets)

Citation: Policy Papers 2023, 035; 10.5089/9798400254574.007.A001

Sources: IMF staff calculations.
Figure 11.
Figure 11.

Effects of SDR Allocation on Non-Program Countries with Low Reserve Coverage

(In months of imports)1

Citation: Policy Papers 2023, 035; 10.5089/9798400254574.007.A001

1 Countries with a reserve coverage below three months of imports before the SDR allocation that did not have a program during August 2021 to end 2022. Excluding AEs. Sources: IMF staff calculations.

D. Implications of Higher Interest Rates for Members with Negative Net SDR Positions

20. The expected path of the SDR interest rate has shifted significantly upward since the allocation (Figure 12). Since August 2021, the SDR interest rate has risen by over 390 basis points, leading to higher interest payments for members with negative net SDR positions (i.e., SDR holdings below their cumulative allocation).29 To assess the expected interest costs for these members, staff constructed market based SDR forward curves (Box 6). These curves indicate that the expected path of SDR interest rates has shifted upward by about 230 basis points on average from August 2021 to June 2023 across the forward curve, with the long-run market-implied SDR interest rate increasing from about 1.4 percent in August 2021 to about 3.2 percent in June 2023.30

Figure 12.
Figure 12.

SDR Interest Rates Since the Allocation

Citation: Policy Papers 2023, 035; 10.5089/9798400254574.007.A001

1 The grant element numbers for SDRs are calculated based on end-June 2023 interest rates. The numbers for ECF and IDA are based on illustrative estimates discussed in International Monetary Fund (2023c) based on January 2023 interest rates.Source: IMF staff calculations.

21. The higher SDR interest rate implies that the financing terms of SDRs are just below the concessionality threshold. By convention, the terms of a financing instrument are considered concessional if the grant element—the percentage difference of the face value of a loan relative to the net present value of service payments associated to it—is greater than 35 percent.31 Given current interest rates and baseline projections, staff estimates the grant element of SDR conversions to be 34 percent,32 just below the consessionality threshold and the grant element of the IMF’s Extended Credit Facility (ECF) under the PRGT (about 36 percent) (Figure 12d).33 The grant element of SDR use would decline futher if interest rates were to increase further, if a discount rate lower than 5 percent is applied, or if the SDR allocation were to be cancelled after 30 years or sooner.

22. The expected nominal cost of servicing SDR obligations has increased significantly (Figure 13). One-year ahead payments of net SDR charges for the median country increased from USD 0.1 million to USD 7.8 million during August 2021—June 2023. The (infinite horizon) expected cost based on the present value (PV) of net payments to the SDR Department has increased almost threefold (from USD 37 million to USD 134 million) for the median country. For the top quartile country, the one-year-ahead and the PV cost measures are estimated at USD 19 million and USD 333 million, respectively; for the member with the largest negative SDR Department position in nominal terms, this cost is estimated at USD 259 million (one-year-ahead) and USD 4.5 billion (PV), respectively.

Figure 13.
Figure 13.

Estimated One-year-ahead Payments and PV of Members’ SDR Department Obligations for SDR converters1

(By forward curve vintages, USD million)

Citation: Policy Papers 2023, 035; 10.5089/9798400254574.007.A001

1 Outlier observations (with values exceeding 1.5 times the interquartile range) are not shown.Source: IMF staff calculations.

23. Despite the higher expected cost, most members’ capacity to service SDR obligations remains broadly adequate, with some exceptions. The SDR Department’s annual payments represent less than 0.1 percent of GDP for 90 percent of members with negative net SDR positions, and less than 1 percent of reserves for 90 percent of that group. Meanwhile, the PV of SDR obligations amounts to less than 1 percent of GDP in 69 percent of this group of members and between 1–2 percent of GDP in 23 percent of members. However, in 8 percent of EMDCs , the PV of SDR obligations amounts to 2–9 percent of GDP (Figure 14). Among LICs, São Tomé and Principe,Yemen, and Zimbabwe stand out as having a PV of SDR obligations greater than 2 percent of GDP and external debt greater than 40 percent. Three EMs with PV of SDR Department exposures greater than 2 percent also have external debt levels greater than 60 percent of GDP (Jamaica, Suriname, and Venezuela). Furthermore, for several members, SDR Department obligations are a significant part of their external debt stock.34 These members will need to take SDR Department obligations into account when determining their macroeconomic and financial policy settings to ensure an adequate capacity to service them over time. Moreover, for members with low levels of SDR holdings, staff will need to monitor SDR holdings closely and encourage early purchases of SDRs to meet payments of charges to the SDR Department and the GRA. This will help reduce the risk of technical arrears but may increase operational costs of the SDR Department. Staff also stands ready to assist members through policy advice and technical assistance on broader external debt management issues.

Figure 14.
Figure 14.

SDR Department Obligations and External Debt Levels

Citation: Policy Papers 2023, 035; 10.5089/9798400254574.007.A001

Source: LIC-DSF database, staff calculations. The PV of SDR Department obligations is measured in percent of 2023 GDP using SDR interest rate projections of June 2023.

Projecting the Expected Cost of Reducing SDR Holdings1

The expected cost of reducing SDR holdings is the present value (PV) of all future net interest payments to the SDR Department. The expected cost is assessed via the following equation:

ExpectedCostt=PVt=Σq=0n((SDRaq−SDRhq)*Et(SDRiq)/4)(1+r)q/4+(SDRaq*−SDRhq*)(1+r)q/4

The PV of interest payments in quarter q are determined by the prevailing SDR interest rate of this quarter (SDRiq), based on expected future SDR interest rates at that quarter at a given time (Et(SDRiq)), the difference between the member’s SDR allocation and holdings in the same quarter (SDRaq — SDRhq), and the discount rate (r). The second term in the right-hand side of equation (1) represents the “face value” redemption of the member’s net open position in the SDR Department (captured by the term (SDRaq* — SDRhq*), discounted at the quarter of redemption (q*).

Several additional assumptions were made. Basket weights of SDR basket currencies are fixed at the initial 2015 weights through July 2022, after which they follow the August 2022 basket weights for the rest of the projection period. Basket instruments are fixed. We use an infinite projection horizon (no cancellation assumed) and a discount rate of 5 percent, in line with convention. These assumptions yield the below equation:

PVt=(SDRaq−SDRhq¯)*[Σq=1120Et(SDRiq/4)(1+5%¯)q4+Σq=121∞Et(SDRι120¯/4)(1+5%¯)q4]

Staff estimated the expected SDR interest rate in each future quarter based on available market data. Specifically, the expected SDR interest rate (Et(SDRiq/4)) was calculated by first creating zero-coupon bond yield curves for the SDR basket currencies and then deriving 3-month forward rates using no arbitrage conditions.

1 Based on “The Financial Cost of Using Special Drawing Rights: Implications of Higher Interest Rates,” by Neil Shenai, Nicolas End, Jakree Koosakul, and Ayah Said. IMF Working Papers (forthcoming).

24. SDR obligations would remain relatively contained in a stress scenario for most members (Figure 15). To further assess the interest rate risk for countries with negative SDR positions, a stress scenario was developed with the long-term SDR interest rate increased by one standard deviation relative to the baseline.35 Even under the stress scenario, most members’ expected debt service metrics appear serviceable, with an average PV of SDR Department obligations of 1.60 and 0.89 percent of GDP for all LICs and EMs, respectively. However, 21 members (18 percent of the total) would have a PV of SDR Department obligations greater than 2 percent of GDP compared to 9 members under the baseline. And 17 countries (16 percent of the total) would have a PV of SDR Department obligations greater than 20 percent of total reserves, compared to 9 countries under the baseline.36

Figure 15.
Figure 15.

Distribution of the Expected Cost of SDR Conversion Under Baseline and Stress Scenarios 1/

Citation: Policy Papers 2023, 035; 10.5089/9798400254574.007.A001

1/ Authors’ calculations.Source: World Economic Outlook, authors’ calculations

25. For the majority of countries, the allocation has not led to changes in debt sustainability risk ratings, although a few LICs in debt distress have high SDR Department obligations (Figure 16).37 Out of 39 LICs with negative SDR positions, the majority (31, or 80 percent) experienced no change in their LIC-DSF risk rating during July 2021–June 2023 (almost all other LICs also saw no change), three (8 percent) observed an improvement, and six (16 percent) experienced a deterioration. Of the latter: four (Djibouti, Lao PDR, Malawi, and Zambia) entered debt distress due to a combination of factors not related to the SDR allocation, although Malawi stands out as having relatively higher SDR Department payments; Comoros and Tanzania saw an increase in their risk rating reflecting country-specific vulnerabilities. Some members were already experiencing significant debt vulnerabilities or outright debt distress prior to the allocation, and thus did not experience a change in their ratings following conversion of their holdings. São Tomé and Principe and Suriname are two such examples, which also stand out for their relatively higher debt servicing costs to the SDR Department.

Figure 16.
Figure 16.

Change in LIC-DSF Risk Rating1

(Converters vs. non-converters, percent of countries in each category)

Citation: Policy Papers 2023, 035; 10.5089/9798400254574.007.A001

1 The change in LIC-DSF risk rating is determined by comparing the risk rating of LICs in July 2021 (just before the allocation) and in April 2023. The “converters” sample includes members that exchanged SDR holdings via VTAs.Source: LIC-DSA database and staff analysis.

E. Members’ and Staff’s Views

26. IMF members have a generally positive view about the impact of the SDR allocation on their respective countries (Figure 17). Almost all EM and LIC authorities agreed that the allocation was helpful given their respective country circumstances. About 95 percent of LIC respondents agreed that the allocation was both necessary and timely for their country, while affirmative responses for EMs were somewhat lower at 72 and 85 percent, perhaps reflecting that many EMs had adequate reserves prior to the allocation (Figure 1). In contrast, while about 80 percent of EMs agreed that the allocation was sufficient for their country, only 47 percent of LICs agreed, likely reflecting larger financing needs the latter faced due to the pandemic and subsequent shocks, as well as reduced access to market and concessional financing sources, as noted earlier.

Figure 17.
Figure 17.

Views from Recipient Countries on the Allocation for their Individual Economies

(In percent of respondents)

Citation: Policy Papers 2023, 035; 10.5089/9798400254574.007.A001

Sources: Country authority survey; staff analysis.

27. IMF mission chiefs consider that countries’ policies were broadly in line with Fund advice (Figure 18). Mission chiefs assessed that member policies related to the allocation were broadly consistent with the principles set forth in the Guidance Note on the Treatment and Use of SDR Allocations. For 24 members where staff had limited contact with the country authorities either due to recognition issues or the member’s unwillingness to engage with the Fund, mission chiefs were not able to assess the extent to which the member’s policy choices were consistent with Fund advice.

Figure 18.
Figure 18.

Impact of the SDR Allocation on Members’ Economic Policy

(Percent of Surveyed IMF Mission Chiefs)

Citation: Policy Papers 2023, 035; 10.5089/9798400254574.007.A001

Sources: Mission chief survey; staff analysis.

28. For some EMDCs, particularly LICs, mission chiefs believe that the allocation may have led to delays in needed adjustment and reforms (Figure 18). For 20 percent of LICs and 13 percent of EM countries, mission chiefs considered that the SDR allocation led to material delays in needed policy adjustment and reforms. This perception is corroborated by evidence on the implementation of structural benchmarks (Figure 19), which deteriorated between 2021 and 2022 for program countries where mission chiefs noted delays, while in other program countries the implementation of structural benchmarks improved slightly (by 2 percentage points).38 Moreover, in 12 cases (7 EMs and 5 LICs), mission chiefs think that the SDR allocation may have delayed a needed debt restructuring. Finally, for 3 EMs and 9 LICs, mission chiefs reported that the allocation exacerbated fiscal dominance, and in 4 EMs and 8 LICs the allocation appears to have adversely impacted central bank independence (see next section). This underscores the importance of continued tailored and granular staff advice to the membership to ensure SDRs are used to pursue sustainable policies.

Figure 19.
Figure 19.

Implementation of Structural Benchmarks in IMF-Supported Programs

(Percent of structural benchmarks met, by years of test dates)

Citation: Policy Papers 2023, 035; 10.5089/9798400254574.007.A001

1/ Delays in adjustment/reforms is based on survey of IMF Mission Chiefs.Sources: MONA and staff estimates.

Transparency and Accountability in the Holding and Use of SDRS

This section examines whether countries adhered to transparency and accountability principles in their reporting and use of SDRs. First, it provides an overview of domestic holding structures and possible mechanisms to use SDRs for fiscal purposes. Second, it discusses transparency and accountability of domestic holding structures and challenges associated with the fiscal use of SDRs. The analysis is based on a review of 94 countries subject to ongoing monitoring under the IMF Safeguards Assessment Policy. Third, it assesses the statistical treatment and recording of the allocation in national statistics. Finally, it concludes a discussion of how staff stepped up publication of data in line with the commitments made by staff at the time of the allocation.

A. Domestic Holding Structures

29. IMF members typically hold SDRs at their central banks.For most members, domestic legislation considers SDRs as an international reserve asset and determines that the SDRs are held and controlled by the central bank, which is consequently also financially responsible for the corresponding liability vis-à-vis the IMF. In such cases, the SDR holdings and allocation are booked on the central bank balance sheet. There are, however, members where the SDRs are held instead by the government, in some cases but not always through a specific holding structure (e.g., an “exchange stabilization fund”) under the auspices of the Ministry of Finance (MoF). There are a few countries that hold their SDRs through both the central bank and the MoF (Figure 20).

Figure 20.
Figure 20.

Who Holds the SDRs?

Citation: Policy Papers 2023, 035; 10.5089/9798400254574.007.A001

Source: IMF data

30. To fund budgetary outlays, some members thus had to transfer their SDRs to the government, which proved challenging and often required complex mechanisms. In the few countries where the fiscal authorities had control of SDRs under the domestic legal frameworks, such fiscal use was relatively straightforward. If, however, the SDRs were initially under the control of the central bank, any envisaged fiscal use required the central banks to transfer these SDRs to the fiscal authorities. In some countries, the transfer required legislative amendments, as existing legislation protects the financial autonomy of the central bank. Where it was legally possible, members transferred SDRs or the foreign exchange proceeds from SDR conversion through a variety of financial-legal mechanisms (Box 7). The implementation of these mechanisms typically required the adoption of legal instruments such as loan agreements or memoranda of understanding (MoU), depending on the organic law of the central bank and the mechanism for transferring SDRs to the government.

31. The transfer modalities are key to safeguard the financial autonomy of central banks but were not always respected. Where central bank acts as paying agent for SDR charges toward the Fund, adopting an MoU to govern and clarify the financial responsibility of the government stemming from the transfer of SDRs for fiscal use has proven necessary. The purpose of such an MoU is to avoid the central bank carrying the cost of financial liabilities of the government vis-à-vis the IMF. However, in several instances, the transfer of SDRs from the central bank to the government has resulted in financial structures that are not conducive to central bank financial autonomy.39

32. Currency unions utilized specific modalities to use the policy space provided by the SDR allocation, which may challenge the financial autonomy of the regional central banks. For instance, regional central banks in Sub-Saharan Africa provided local currency loans (monetary financing) in proportion to, but unlinked to, the SDR allocation received by member countries. In fact, higher SDR interest rates would translate into an interest rate risk for these central banks if the repayment terms do not compensate the regional central bank for interest obligations to the SDR Department.

Transferring SDRs from Central Banks to Governments

Members have used a broad variety of mechanisms to transfer SDRs that were originally held by their central banks to the government:

  • (1) Transfer of assets and liabilities. The SDR holdings (asset) and SDR allocation (liability) are transferred from the central bank to the government pursuant to a legal instrument (e.g., law, agreement). By consequence, the central bank derecognizes the SDR-related asset and liability from its balance sheet and the government acquires control over the SDR holdings and assumes the ensuing financial liability. Such an operation requires careful discussions with the central bank’s external auditors on the financial reporting implications.

  • (2) On-lending of SDRs or FX equivalent. If allowed by legislation, this takes place through a loan agreement between the central bank and the government. Ideally, the terms of the loan should mirror the features of the central bank’s SDR-related obligations to make the central bank financially safeguarded (SDR conversions trigger quarterly interest obligations based on the floating SDR rate; section V).

  • (3) Sale of SDRs or FX equivalent. The government can purchase from the central bank SDRs (or FX if SDRs were converted), against consideration in the form of local currency or bonds. Such transactions should be at “arm’s length.” Exchanging SDRs or FX for government bonds raises complex financial and accounting issues that require careful consideration, including the central bank’s external auditors who need to validate the appropriateness of the exchange. The application of IFRS may also raise issues, such as valuation requirements.

  • (4) Cession-Retrocession. This specific form of sale consists of a transfer (the cession) of the SDR holdings from the central bank to the government against consideration in the form of an assumption by the government of the central bank’s corresponding liability to the IMF (SDR allocation). To fund budgetary expenditures, the government will have to subsequently sell the SDRs back to the central bank (retrocession).

B. Transparency and Accountability in the Holding and Use of SDRs

33. SDRs held through central banks are subject to broadly robust transparency and accountability mechanisms. Almost all SDR-holding central banks publish annual financial statements that are independently audited by external firms or State Audit bodies, thus providing external independent assurances on reporting. Most central banks have appropriate disclosures and accounting treatment for their SDR holdings, even though staff encountered cases where the central banks were not specific on the accounting treatment of SDR holdings or made errors in the initial recording and accounting treatment of SDRs.

34. SDRs held by governments generally have a lower level of transparency and accountability.40 Of the 94 reviewed countries, only a minority (two) of the small group (five) that hold their SDRs through the MoF issue annual financial statements, and only one of these statements is externally audited by the State Audit body. This should be seen in the broader context of challenges pertaining to balance sheet financial reporting by sovereigns.41

35. The fiscal use of SDRs originally held by central banks is seldom reported transparently. Information on the occurrence and modalities of transfer from the central bank to the fiscal authorities was available in only one third (26) of the cases where the central banks originally held the SDRs (78). Moreover, SDR-related disclosures, including on fiscal use, are very scarce (almost virtually inexistent) in most cases once the shift from the central bank to the fiscal authorities is completed. In engaging with member authorities on using SDRs for fiscal purposes, staff has repeatedly stressed the importance of developing in the MoF appropriate structures to hold the SDRs and manage the exposure stemming from their fiscal use, as well as its interaction with the broader legal framework for public financial management, even though traction has been limited. Staff will continue to help members strengthen fiscal SDR holding structures and more generally, fiscal balance sheet reporting.

C. Quality of Statistical Recording

36. Since the allocation, the IMF Statistics Department (STA) have guided the authorities and IMF country teams on the treatment of the SDR allocation in macroeconomic statistics. STA prepared a compendium of frequently asked questions (FAQs) on the statistical treatment of SDR allocations (in line with IMF 2009c; IMF 2016), as well as a guidance note on the recording in government finance statistics (GFS) in line with IMF (2014a).42 Upon request from authorities and IMF country teams, STA has also provided country specific guidance in over 30 instances.43

37. External sector statistics reported to the IMF by most member countries on the 2021 SDR allocation have been broadly in line with standards. As of June 2023, 164 out of 190 Fund member countries have already reported their 2021 balance of payments to STA. The SDR allocation transaction (liability) is missing in 21 cases.44 Following pre-established procedures, STA used the IMF Finance Department information to disseminate country’s SDR transactions and positions data in its statistical publications―balance of payments and international investment position (IIP), International Financial Statistics (IFS) and the Balance of Payments Statistical Yearbook (BOPSY), correcting any discrepancies with the actual figures.

38. Most member countries have recorded their SDR position in the central bank balance sheet; some other countries of those where SDRs are held by the government, do not report SDR assets and liabilities through GFS. External sector statistics do not indicate which domestic agency includes the SDR allocations and holdings on their balance sheet. This information is available from the monetary and financial statistics (MFS) and government finance statistics (GFS). Of the 173 economies reporting monthly MFS in SRF 1SR to STA as of June 2023, the central bank balance sheets for 146 reporters reflect full or partial SDR allocation. However, regarding GFS, only twelve of the 121 countries that have reported 2021 data to the IMF’s GFS database report data on government holdings of SDRs (see background paper).45

D. Publication of SDR Operations by the IMF

39. The Fund has increased the publication of information about SDR holdings and transactions, further strengthening the smooth functioning of the SDR Department. In October 2021, the Fund published its annual update on SDR trading operations.46 In addition, on December 15, 2021, the Fund published, for the first time, a new quarterly financial report that provides net changes in member’s SDR holdings split in two broad categories: (i) those related to IMF operations; and (ii) SDR trading and other uses.47 These measures, together with the monthly reporting on the IMF Finances webpage on SDR holdings and allocations, have increased the publicly available information on SDR transactions, and have been positively received by the membership. Finally, at the time of the 2021 allocation, staff also conducted significant outreach with the membership on the treatment and use of SDR allocations, including on transparency and accountability.

Voluntary Channeling of SDRS

This section discusses voluntary channeling of SDRs to support vulnerable countries since the 2021 allocation.

40. In October 2021, the G20 pledged to channel USD100 billion, or 20 percent of their SDR allocations, on a voluntary basis, for the benefit of vulnerable countries.48 At the time of the allocation, the Executive Board discussed three options to channel SDRs for: (i) scaling up the PRGT; (ii) contributing to the new IMF-administered RST; and (iii) providing SDRs to prescribed holders, including multilateral development banks (MDBs). The financial architectures of the PRGT and the RST explicitly allow contribution of SDRs or freely usable currencies and ensure the safety and liquidity of contributors’ claims on their respective loan and deposit accounts so that these claims can have all the characteristics of reserve assets (IMF 2022c).

41. There were operational challenges and legal considerations for members to contribute SDRs. Contributions were at times delayed or could not occur when parliamentary approval for budgetary outlays was required, when members had to first join the VTA market,49 or when domestic legal frameworks precluded SDR contributions. In the eurozone, the ECB had to confirm the reserve asset status of contributions to the RST and more broadly their consistency with the requirements under EU legislation so that the region’s central banks could channel SDRs. In a few cases, the determination by the OECD on whether SDR contributions classified as overseas development aid was a consideration given some contributors’ ODA limits.

42. The smooth functioning of the VTA market continues to be critical for effective SDR channeling. As envisioned at the time of the SDR allocation, SDR channeling has increased the number of SDR transactions and the total SDR amounts exchanged for currencies in the VTA market. This outcome reflects the strong expectation that participants channeling SDRs stand ready to convert SDRs into currencies and ensures that there is broad distribution of SDR transactions among VTA participants. Going forward, as more and more pledges are realized, VTA participants who channel SDRs should be prepared to participate more often in SDR transactions as their SDR holdings-to-allocation ratio drop and also hold more SDR denominated assets.

43. The G20 channeling ambition of USD 100 billion has been reached in terms of pledges. According to the G20 International Financial Architecture Working Group Co-Chair Note, pledges by both G20 and non-G20 countries reached the target as of June 23, 2023.50 Pledges were made by 29 countries and amount to USD 103.4 billion (SDR 73 billion), of which the proportionally largest came from France and Japan, channelling 40 percent of their allocations, followed by Australia (38 percent), China (34 percent), and Saudi Arabia (29 percent).

44. Pledges toward IMF fundraising efforts, both in SDRs and equivalent currencies, have also progressed but have not yet fully reached the Fund’s goals:

  • Poverty Reduction and Growth Trust (PRGT). SDR channelling was instrumental in helping finance the pandemic-related July 2021 increase in PRGT access (IMF 2021d). As of end-May, pledges for loan resources are still falling short of the first stage fundraising targets by about SDR 1 billion, and those for subsidies by about SDR 0.9 billion. Ensuring a self-sustaining PRGT that is fully able to meet the needs of the IMF’s poorest and most vulnerable members over the longer term requires additional resources, which could be facilitated by further SDR channeling.51

  • Resilience and Sustainability Trust (RST). When the RST was established, it was foreseen that the bulk of the fundraising of the SDR 33 billion total RST resources, including loan resources, would be made via SDR channeling (IMF, 2022c). As of end May 2023, pledges to the RST, including loan resources, totaled SDR 25.5 billion (of which SDR 22 billion were delivered), leaving a loan resource gap of SDR 6.5 billion. Additional pledges could help meet the strong, frontloaded member demand for additional RSF arrangements.52 In addition, given strong demand from members, the Fund has an ambition to raise its funding target for the RST.

45. Staff has continued to support interested multilateral development banks (MDBs) in exploring SDR channeling options. While the channeling of SDRs to MDBs is ultimately a matter for discussion between MDBs and potential contributors, staff has provided technical advice to interested MDBs (e.g., the African Development Bank and the Inter-American Development Bank), including sharing the experience of channeling SDRs to the Fund’s concessional trusts. The legal procedures by the IMF to implement MDB channeling are still under discussion, as are assessments of the possible implications for the SDR market.

Conclusions

46. This ex-post report discusses the use of SDRs and the implications of the SDR allocation from multiple dimensions. It provides an assessment of the impact of the allocation on the global economy and its implications at the country level. The report also reviews whether the use of the allocation was in line with Fund policy advice and transparency and accountability principles and documents efforts to voluntarily channel SDRs for the benefit of vulnerable countries.

47. The 2021 SDR allocation had significant benefits for the global economy. It provided needed liquidity to the membership during a period of exceptionally high uncertainty, helping meet the long-term global need for reserves and supporting market confidence. For many EMDCs, especially LICs, the SDR allocation represented a significant share of GDP that supported reserve adequacy and alleviated external and fiscal financing constraints, thus contributing to global financial stability. There is no evidence that allocation materially contributed to global inflation.

48. The allocation has been largely used by members to increase international reserve buffers, with some EMDCs also using it to meet fiscal and external needs. While most members retained at least part of their SDR holdings as international reserves, some countries used SDRs to meet obligations with the Fund or converted them into freely usable currencies to, among others, increase fiscal spending and amortize more costly debt. Notably, about half of LICs used at least part of the SDR allocation for fiscal support, including to address immediate needs due to the COVID-19 pandemic. In many Fund-supported programs, fiscal targets were relaxed to accommodate the use of SDRs for fiscal spending, while in many others reserve targets were increased to boost reserve buffers.

49. While interest costs have increased for members with negative net SDR positions, they are assessed as manageable for most members. SDRs are not cost-free, as a reduction in holdings below the cumulative allocation generates a net interest expense. With rising interest rates, expected interest costs (in net present value terms) are estimated to have more than tripled. Nonetheless, most members’ capacity to service SDR obligations remains broadly adequate, with some needing to manage it carefully. Staff stands ready to assist these members through policy advice and technical assistance on broader external debt management issues.

50. Members’ use of the allocation has been broadly in line with Fund advice, with some exceptions. IMF mission chiefs generally note that the allocation did not delay needed macroeconomic adjustment and reforms or debt restructurings and did not exacerbate fiscal dominance or impact central bank independence. However, for some EMDCs, particularly LICs, mission chiefs believe that the allocation may have led to delays. This underscores the importance of close monitoring and continued tailored and granular staff advice to the membership to ensure SDRs are used to pursue sustainable policies.

51. The transparency and accountability of SDR holding and use has been generally adequate, although there are some gaps. Central bank holdings of SDRs are generally subject to a high degree of transparency and accountability, while transparent reporting of SDRs transferred to governments faced some challenges, especially where SDRs are held by the fiscal authority. Staff will continue to support the authorities to improve transparency and accountability, including by enhancing fiscal reporting.

52. The G20 target of re-channeling USD 100 billion from economically strong members to vulnerable members has been reached in terms of pledges. This has helped scale up IMF lending to the membership, both through the PRGT and the more recently established RST. Efforts now focus on translating the pledges into contributions and increasing the ambition of voluntary channeling for the benefit of vulnerable countries.

References

1

The legal basis of SDR allocations is “to supplement existing reserve assets in such manner as will promote the attainment of its purposes and will avoid economic stagnation and deflation as well as excess demand and inflation in the world” (Articles of Agreement, Article XV).

2

See for example the April 2021 Communiquéof the International Monetary and Financial Committee (IMFC) and The Chair’s Summing Up in the Proposal for a General Allocation of Special Drawing Rights (June 2021).

3

This estimate was based on projections of standard indicators of reserve adequacy for the five years following the allocation (see IMF, 2021a).

4

The methodology used to decide on the size of the 2009 SDR allocation, which covered 30–60 percent of the estimated long-term global reserve need, was also used in determining the size of the 2021 allocation. Using the same percentage translated into a suggested allocation range of US$0.3–1.2 trillion (about SDR 0.2–0.8 trillion) in 2021, see IMF (2021a).

5

In this paper, low-income countries (LICs) are those eligible to use the PRGT; emerging markets (EMs) are the rest of the “Emerging Markets and Developing Countries (EMDC)” grouping used for the IMF quota classification (i.e., this category includes all EMs and LICs); and advanced economies (AEs) are the remaining countries.

6

Provided a member does not subtract long-term reserve-related liabilities from net international reserves (NIR), which is the practice in most countries, NIR increases in proportion to the increase of the member’s SDR holdings.

7

The August 2021 SDR allocation was worth less than 3 percent of total AE central bank balance sheets at the time of its effectiveness.

8

Even where SDR holdings were converted to freely usable currencies, these actions often reflected reserve management operations (with no injection of liquidity) or may have been offset through sterilization, which could have further muted the impact of the allocation on inflation.

9

For the purpose of reporting survey results, “strongly disagree” and “disagree” responses, as well as “strongly agree” and “agree” responses, are aggregated. In total, staff received 101 responses of 190 members surveyed, of which 26, 53, and 22 were from AEs, EMs, and LICs, respectively. For more details on the survey, see background paper.

10

The holdings-to-allocation ratio in this section refers to the cumulative SDR holdings and cumulative SDR allocation including the SDR allocation 2021, unless otherwise specified.

12

The number of transactions (212) is higher than the number of sale requests since multiple counterparties and/or currencies might be needed to fulfill one sale request.

13

A survey of IMF mission chiefs in early May 2021 projected potential demand for freely usable currencies following the allocation of about SDR 20 to 26 billion (IMF 2021).

14

Earlier spikes in VTA exchange volumes in 2016 to 2019 were primarily related to GRA lending activities following the payment of the quota increases in SDRs from the 14th General Review of Quotas.

15

As of end-June 2023, only five of the twenty prescribed holders held SDRs.

16

The trading capacity of the VTAs is determined by considering the minimum and maximum SDR amounts that a VTA member is willing to hold vis-à-vis the actual SDR holdings of each VTA member.

17

Since the 2021 allocation, the number of VTA members has increased from 32 to 40 and VTA members have provided significant additional operational flexibilities (see background paper section II).

18

See IMF (2023d), especially Figure 3, on the evolution of the sources of financing for Sub-Saharan Africa since 2000.

19

This amount represents total repurchases made in SDRs, which may have been funded from holdings at the time of the allocation, SDRs received subsequently under Fund programs, and other inflows.

20

This group excludes countries with non-financing Fund-supported programs (i.e., Policy Coordination Instruments).

21

In countries that converted more than 100 percent of the 2021 allocation, the conversion ratio is capped at 100 percent.

22

Several countries have used their allocation for more than one purpose.

23

Tracking the specific fiscal use of the allocation is difficult, however, given that the proceeds of SDR exchanges are fungible.

24

For instance: in Guinea-Bissau, the authorities used the SDR 27.2 million allocation (about USD 38.4 million) to prepay the debt service on the West African Development Bank (BOAD) non-concessional loans due up to 2022, in addition to covering COVID-related expenditures. In Rwanda, 30 percent of the allocation (0.6 percent of GDP) was to be used to retire the remaining 2013 Eurobond amount maturing in 2023.

25

For the purpose of the analysis, precautionary arrangements and programs without completed reviews after the SDR allocation were excluded from the 33 programs at the time of SDR allocation.

26

Changes in program targets indicate changes in Quantitative Performance Criteria (QPCs) for December 2021 and June 2022. Data from MONA were complemented by those in staff reports.

27

In a few cases, adjustment in reserves target was not necessary, as the target was set on the NFA, or the definition of the NIR netted all reserve-related liabilities (including long-term liabilities). In these cases, the SDR allocation increased both reserve-related assets and liabilities by equal amounts and had no impact on the reserves target.

28

Staff did not include the SDR allocation in their WEO macroeconomic projections before July 27, 2021. Inclusion of the SDR allocation in macroeconomic projections started with staff reports going to the Board after August 23, 2021.

29

As explained in Box 1, a net interest expense at the SDR interest rate is generated when a member’s SDR holdings fall below cumulative allocations, while members with SDR holdings above cumulative allocations will receive a net interest income. Holdings may fall below cumulative allocations when members exchange SDRs for currencies through the VTA or bilaterally, or use SDRs to meet Fund obligations.

30

This result is robust to using alternative methods of SDR interest rate projections from WEO estimates.

31

More formally, the grant element of the SDR as a financing instrument is determined by the following equation: Grant element = (FV – NPV|SDRi)/FV.

32

This assumes a standard 5 percent discount rate and no cancellation of the allocation. This discount rate is also used in the Debt Sustainability Framework for Low-Income Countries (LIC-DSF) and in the IMF’s Debt Limits Policy for calculating the grant element of a financing instrument. Since the creation of the SDR, the BoG has never cancelled SDRs. Should SDRs be cancelled, then members with negative SDR balances (cumulative SDR allocations greater than holdings) would need to repay this balance “promptly,” per the Fund’s Articles of Agreement.

33

IMF (2023c). The financing terms of SDRs entail a higher grant element than financing under the Resilience and Sustainability Fund (RSF), which has a grant element of 29 percent (for Group A-countries, with the interest rate capped at 2.25 percent).

34

For instance, among LICs, Burundi, Central African Republic, Chad, Democratic Republic of Congo, Haiti, Liberia, São Tomé and Principe, Somalia, and Yemen have a PV of SDR Department obligations greater than 5 percent of total external debt as of end-June 2023. Among EMs, Equatorial Guinea, Eswatini, and Guyana, exceed this 5 percent threshold. This exposure might be particularly relevant in debt restructuring cases where senior debt (including SDR obligations) cannot be restructured.

35

More specifically, the forward rates from year 30 onward are assumed to be higher by 2.9 percentage points. From years 0 to 30, the SDR forward rates are assumed to rise linearly from the baseline projection in quarter 0 (that is, zero increase in the baseline projection at quarter 0) to the new higher steady state interest rate in year 30 onward.

36

Due to the assumptions underlying the stress test, near-term interest rate forecasts do not change materially, and thus the one-year-ahead debt service indicators are almost the same in both baseline and stress scenarios in Figure 15.

37

Per the 2021 SDR Staff Guidance Note, for members where SDRs are recorded on the central bank’s balance sheet, which is usually outside of the DSA perimeter, there is no direct SDR-related input into the DSA. In cases where SDRs are recorded on the government balance sheet and the net SDR position is negative, then the shortfall in SDR holdings relative to allocations are included in the stock of nominal debt, while the net SDR interest obligation must be reflected in future interest expenses of the government in both the LIC-DSF and Sovereign Risk and Debt Sustainability Framework for Market Access Countries (SRDSF). In the LIC-DSF, the PV of the net SDR interest expenditures is incorporated as part of the PV of debt as well.

38

While both groups experienced a decline in implementation during the pandemic, they exhibit different trends during 2022. It is too early to assess whether these delays have been made up.

39

For example, the foreign exchange in which the SDRs were converted was “on-lent” by the central bank to the government, albeit without concordance between the SDR charges and the interest charged on the central bank loan—effectively subsidizing the State, with adverse consequences for the central bank’s financial performance.

40

Notable exceptions are countries with well-developed frameworks for “exchange stabilization funds” or similar structures.

42

IMF (2022b), aiming to foster public-sector transparency and accountability in the use of SDRs, provided granular, step-by-step guidance on the recording of SDRs in GFS, the government’s statement of operations and balance sheet, gross and net debt, and debt measurement for debt sustainability analysis (DSA) purposes.

43

Among other topics, guidance focused on the accounting and statistical treatment in macroeconomic statistics of on-lending of proceeds from the conversion of SDRs, the transfers and/or cession-retrocession of SDRs, and the impact of such transactions on official reserve assets.

44

Some of these countries are not BPM6 reporters and others, while being classified as BPM6 reporters, currently deviate from the BPM6 recommendation on the recognition of the SDR allocations as a long-term external debt liability of the country. The remaining 26 members had not yet reported their 2021 balance of payments statistics to STA as of June 2023.

45

GFS compiled by countries cover only government entities and do not include central banks and are reported annually to the IMF on a voluntary basis. As of February 2023, member countries that had reported GFS data for the general allocation of SDRs in 2021 were: Australia, Bosnia Herzegovina, Greece, Japan, Latvia, Republic of Moldova, New Zealand, San Marino, South Africa, Ukraine (based on 2020 reporting of previous SDR general allocation in 2009), United Kingdom, and United States of America.

46

The 2022 update was published in October 2022.

47

The Fund has continued to publish this report on a quarterly basis.

49

See IMF (2022c), para. 110: “…a smooth functioning of the voluntary SDR trading market will be critical for effective SDR channeling through the RST[...]Consistent with the current practice under the PRGT, there is 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.”

51

In the April 2023 review of PRGT resource adequacy, staff noted that in addition to the stage 1 subsidy needs, higher lending commitments and the net impact of the faster than expected rise in the SDR interest rate meant that subsidy costs to maintain the PRGT’s self-sustained annual lending capacity of SDR 1.65 billion had risen by an additional SDR 2.3 billion. The IMF Board called for a multi-pronged strategy to address the elevated near-term needs of the PRGT while also advancing efforts required to ensure the PRGT’s longer-term sustainability (IMF 2023c). Such a strategy will be further developed as part of the upcoming PRGT review.

52

A further SDR 5.1 billion in standalone contributions support RST reserves.

Other IMF Content

Other Publishers

  • Collapse
  • Expand
Cover Policy Papers
2021 Special Drawing Rights Allocation—Ex-Post Assessment Report
Author:
International Monetary Fund. Strategy, Policy, & Review Department
and
International Monetary Fund. Finance Dept.
Volume/Issue:
Volume 2023: Issue 035
Publisher:
International Monetary Fund
ISBN:
9798400254574
ISSN:
2663-3493
Pages:
44
DOI:
https://doi.org/10.5089/9798400254574.007
  • View in gallery

    SDR Allocations

    (SDRs Billions)

  • <sup><xref rid="RA001fig01" ref-type="fig">Figure 1.</xref></sup>
    View in gallery
    Figure 1.

    Impact of the 2021 SDR Allocation1

  • <sup><xref rid="RA001fig02" ref-type="fig">Figure 2.</xref></sup>
    View in gallery
    Figure 2.

    SDR Exchanges and Changes in Major Central Bank Asset Holdings

    (in billions USD, monthly change)

  • View in gallery

    (a) Decline in CDS Spreads following the 2021 SDR Allocation

  • View in gallery

    (b) Decline in CDS Spreads Conditional on Countries’ Pre-existing Vulnerabilities1

  • <sup><xref rid="RA001fig03" ref-type="fig">Figure 3.</xref></sup>
    View in gallery
    Figure 3.

    Impact of the SDR Allocation on the Global Economy: The Membership’s View

    (In percent of total respondents)

  • <sup><xref rid="RA001fig04" ref-type="fig">Figure 4.</xref></sup>
    View in gallery
    Figure 4.

    EMDCs’ SDR Uses since the 2021 Allocation

    (In percent of total inflows from the 2021 allocation through the indicated date)

  • <sup><xref rid="RA001fig05" ref-type="fig">Figure 5.</xref></sup>
    View in gallery
    Figure 5.

    Changes in the Distribution of SDR Holdings-to-Allocation Ratio1

    (Number of countries indicated on each bar)

  • <sup><xref rid="RA001fig06" ref-type="fig">Figure 6.</xref></sup>
    View in gallery
    Figure 6.

    VTA Sales by IMF Members, 2006–2023

  • <sup><xref rid="RA001fig07" ref-type="fig">Figure 7.</xref></sup>
    View in gallery
    Figure 7.

    Conversion of SDRs: Program vs. Non-Program Countries

    (In percent of total SDR allocation)

  • View in gallery

    (a) Utility under different conversion patterns following an SDR allocation (relative to steady state)

  • View in gallery

    allocation (for high and low sensitivity φr)

  • View in gallery

    Uses of the SDR Allocation Based on the SDR Tracker

  • <sup><xref rid="RA001fig08" ref-type="fig">Figure 8.</xref></sup>
    View in gallery
    Figure 8.

    Program vs. Non-Program Countries

    (Use of SDRs based on Staff Reports, in percent of countries in the SDR Tracker)

  • <sup><xref rid="RA001fig09" ref-type="fig">Figure 9.</xref></sup>
    View in gallery
    Figure 9.

    Changes in Staff Projections (April 2021 vs. October WEO)

  • <sup><xref rid="RA001fig10" ref-type="fig">Figure 10.</xref></sup>
    View in gallery
    Figure 10.

    Number of New UCT Arrangements

    (Monthly cumulative, from crisis onsets)

  • <sup><xref rid="RA001fig11" ref-type="fig">Figure 11.</xref></sup>
    View in gallery
    Figure 11.

    Effects of SDR Allocation on Non-Program Countries with Low Reserve Coverage

    (In months of imports)1

  • <sup><xref rid="RA001fig12" ref-type="fig">Figure 12.</xref></sup>
    View in gallery
    Figure 12.

    SDR Interest Rates Since the Allocation

  • <sup><xref rid="RA001fig13" ref-type="fig">Figure 13.</xref></sup>
    View in gallery
    Figure 13.

    Estimated One-year-ahead Payments and PV of Members’ SDR Department Obligations for SDR converters1

    (By forward curve vintages, USD million)

  • <sup><xref rid="RA001fig14" ref-type="fig">Figure 14.</xref></sup>
    View in gallery
    Figure 14.

    SDR Department Obligations and External Debt Levels

  • <sup><xref rid="RA001fig15" ref-type="fig">Figure 15.</xref></sup>
    View in gallery
    Figure 15.

    Distribution of the Expected Cost of SDR Conversion Under Baseline and Stress Scenarios 1/

  • <sup><xref rid="RA001fig16" ref-type="fig">Figure 16.</xref></sup>
    View in gallery
    Figure 16.

    Change in LIC-DSF Risk Rating1

    (Converters vs. non-converters, percent of countries in each category)

  • <sup><xref rid="RA001fig17" ref-type="fig">Figure 17.</xref></sup>
    View in gallery
    Figure 17.

    Views from Recipient Countries on the Allocation for their Individual Economies

    (In percent of respondents)

  • <sup><xref rid="RA001fig18" ref-type="fig">Figure 18.</xref></sup>
    View in gallery
    Figure 18.

    Impact of the SDR Allocation on Members’ Economic Policy

    (Percent of Surveyed IMF Mission Chiefs)

  • <sup><xref rid="RA001fig19" ref-type="fig">Figure 19.</xref></sup>
    View in gallery
    Figure 19.

    Implementation of Structural Benchmarks in IMF-Supported Programs

    (Percent of structural benchmarks met, by years of test dates)

  • <sup><xref rid="RA001fig20" ref-type="fig">Figure 20.</xref></sup>
    View in gallery
    Figure 20.

    Who Holds the SDRs?

SDR Allocations

(SDRs Billions)
Figure 1.

Impact of the 2021 SDR Allocation1
Figure 2.

SDR Exchanges and Changes in Major Central Bank Asset Holdings

(in billions USD, monthly change)

(a) Decline in CDS Spreads following the 2021 SDR Allocation

(b) Decline in CDS Spreads Conditional on Countries’ Pre-existing Vulnerabilities1
Figure 3.

Impact of the SDR Allocation on the Global Economy: The Membership’s View

(In percent of total respondents)
Figure 4.

EMDCs’ SDR Uses since the 2021 Allocation

(In percent of total inflows from the 2021 allocation through the indicated date)
Figure 5.

Changes in the Distribution of SDR Holdings-to-Allocation Ratio1

(Number of countries indicated on each bar)
Figure 6.

VTA Sales by IMF Members, 2006–2023
Figure 7.

Conversion of SDRs: Program vs. Non-Program Countries

(In percent of total SDR allocation)

(a) Utility under different conversion patterns following an SDR allocation (relative to steady state)

allocation (for high and low sensitivity φr)

Uses of the SDR Allocation Based on the SDR Tracker
Figure 8.

Program vs. Non-Program Countries

(Use of SDRs based on Staff Reports, in percent of countries in the SDR Tracker)
Figure 9.

Changes in Staff Projections (April 2021 vs. October WEO)
Figure 10.

Number of New UCT Arrangements

(Monthly cumulative, from crisis onsets)
Figure 11.

Effects of SDR Allocation on Non-Program Countries with Low Reserve Coverage

(In months of imports)1
Figure 12.

SDR Interest Rates Since the Allocation
Figure 13.

Estimated One-year-ahead Payments and PV of Members’ SDR Department Obligations for SDR converters1

(By forward curve vintages, USD million)
Figure 14.

SDR Department Obligations and External Debt Levels
Figure 15.

Distribution of the Expected Cost of SDR Conversion Under Baseline and Stress Scenarios 1/
Figure 16.

Change in LIC-DSF Risk Rating1

(Converters vs. non-converters, percent of countries in each category)
Figure 17.

Views from Recipient Countries on the Allocation for their Individual Economies

(In percent of respondents)
Figure 18.

Impact of the SDR Allocation on Members’ Economic Policy

(Percent of Surveyed IMF Mission Chiefs)
Figure 19.

Implementation of Structural Benchmarks in IMF-Supported Programs

(Percent of structural benchmarks met, by years of test dates)
Figure 20.

Who Holds the SDRs?