A “debt standstill” for the poorest countries: How much is at stake?

 A small but significant temporary relief

Why is a debt standstill necessary to complement other mechanisms to support health and economic activity? Recent trends in debt accumulation for low-income countries have created vulnerabilities predating the COVID-19 crisis. Three main changes occurred since the early 2010s for low‑income countries:

• An increase in debt levels due to higher deficits, financed by domestic and external borrowing, often due to weak debt management practices and loose fiscal policy;

• The diversification of external creditors, in particular the rising importance of Non-Paris Club official lenders and private creditors; and

• A shift towards less concessional terms.

As a result, debt sustainability for low-income economies was increasingly at risk before the COVID-19 crisis began (IMF, 2020[1]): out of 69 countries applying the Low-income Countries Debt Sustainability Analysis (LIC-DSA), in 2019 half are either already in “debt distress” or at high risk of debt distress, against 23% in 2013. Commodity exporters in particular have experienced high fiscal deficits leading to increasing debt. In addition, foreign currency debt increased rapidly (by about 10 percentage points of GDP between 2013 and 2018 for low income-countries), increasing risks linked to exchange rate movements. As a result of those trends, debt service is higher but also more difficult to restructure, as formal co-ordination institutions such as the Paris-club remain absent for those new actors.

The debt standstill covers a relatively narrow set of countries (see Annex for the full list). The criteria to select which countries would be covered by such a standstill has been the subject of debates. Some economists have argued that any emerging market should be included.3 Some others highlighted the risk to the African continent.4 A narrower set of countries such as least developed countries (LDCs) as defined by the United Nations, would focus on some of the most fragile and poorest countries. The 77 countries on which the G20 Finance Ministers converged are LDCs and those eligible to International Development Association (IDA) credits, or concessional lending window of the World Bank,5 including “blend” countries.6 This can be interpreted as not only to include the poorest economies, but also those at risk of witnessing parts of their population slide back to poverty. The debate on which countries should qualify highlights the blunt nature of a debt standstill: by definition, more debt service will be suspended to countries that have borrowed more in the past, and at more expensive rates. The current arrangement do not target, for instance, those which suffer most from the direct health impact of the virus, or its consequences on trade, tourism, or other poverty-alleviating activities. Those impacts remain uncertain. Other forms of support, complementary to this one, such as targeted loans by multilateral banks are playing this role.

The debt standstill provides temporary respite, not relief. The common term sheet adopted by the G20 agreement indicate that the standstill should be neutral in net present value (NPV), implying that the debt service not paid in 2020 will be fully repaid over 2022-2024 (with a grace period in 2021), with in addition interests to compensate for the delay (which could be significant for non-concessional loans). The possibility to maintain the standstill in 2021 will be reviewed in future G20 meetings. This could alleviate the uncertainty for borrowing countries, especially given that the economic shock and external financial pressure might very well be durable. It leaves open the question of whether countries that will request the standstill will be able to manage their financial commitments – which we focus on in the last section.

The limit on non-concessional borrowing will limit demand for the standstill. Under the agreement, countries will have to limit their use of non-concessional financing. While global financial markets have frozen in March and April, emerging sovereign issuers have managed to access finance since May. For “frontier markets,” which have already accessed international capital markets and aim at intend to use it for their gross financing needs in the short or medium run, this is likely to limit incentives for requests for a standstill, and lead to different approaches. Kenya, for example, has signalled that it would not request a standstill on that ground, but has approached official bilateral borrowers individually.7

The proposal would amount to a delay of around USD 16.5 billion in debt service to official bilateral creditors. This number is an upper-bound, if all eligible countries requested it and met the conditions, which is not certain. It also requires full participation of bilateral lenders including China, which will depend on implementation of the agreement. The World Bank’s International Debt Statistics forecast that collective debt service payments should have mounted to about USD 46 billion in 2020, quadrupling since 2012. Bilateral debt service is projected to represent USD 22 billion, or 48% of that total.8 Without information on the schedule of debt service intra-year, the assumption that will be maintained throughout is that 25% of the total was already paid in the first quarter. Overall, the proposed standstill for bilateral loans would be significant for recipient countries, representing about 3.7% of their government revenues, and 0.7% of their gross domestic product (GDP).9

While significant, this magnitude is not at the scale required by the current crisis, which is likely to be the largest in decades. The external shock is massive, sudden, and unexpected. Foreign direct investment (FDI) is expected to fall by at least 30%, and by even more in developing economies (OECD, 2020[2]). Africa’s GDP growth is expected to decline for the first time in decades, possibly down to -5% of growth in 2020, and it does not have the fiscal space to implement policies similar to those of advances economies (OECD, 2020[3]). Many countries in Latin America, already in the midst of a growth slowdown, are dependent to exports which have been vanishing this year: tourism and commodities. This is especially the case of Caribbean economies (OECD, 2020[4]). The challenges to policy responses are especially acute in low-income countries where the large informal sector prevents the kind of social protection or SME credit measures adopted in OECD countries. In addition, fiscal space – the ability to implement discretionary deficit – is also more constrained in developing countries (with considerable variations within this group): the estimated average stimulus related to COVID-19 is of 8% of GDP in advanced economies, compared to 0.8% in Africa.10 The UN Economic Commission for Africa (UNECA) called for a USD 100 billion stimulus. The official bilateral debt service standstill allows to partly lift this constraint, which can be allocated to health or general economic response.

 

Figure 1. Debt service for selected countries

USD billions

Note: The figures presented here are projected from debt stocks data collected by the World Bank at the end of 2018. For 2020, this is for the full year, so the amounts need to be reduced by 25% to obtain the estimated stand-still impact. See Annex for selected countries.

Source: (World Bank, 2020[5])International Debt Statistics, https://data.worldbank.org/products/ids

The G20 agreement also calls upon “private creditors (…) to participate in the initiative on comparable terms.” Involving private lenders can significantly raise the impact of the standstill, increasing revenues available to IDA and blend countries by USD 8.8 billion, an additional 0.3% of GDP. Again, this has to be perceived as an upper bound, as the request is voluntary, and incentives for the private sector to agree to the standstill are limited. The total liquidity provision would then amount to 1% of GDP, or 5.4% of government revenues. There are significant legal hurdles to constraining private lenders to postpone claiming debt service. For bonds, Collective Action Clauses (CACs) that usually allow for restructuring sovereign debt would be difficult to apply in this specific case.11 Relevant laws in the United Kingdom and the United States that govern a large share of sovereign bonds would have needed to be modified, a complex and potentially risky endeavour not suited to the current urgency. In addition, a significant share of the debt is actually in the form of bank loans, which are not governed by the same restructuring rules. The London Club, which co-ordinates commercial banks, is not designed for the emergency required by the standstill. Another potential model would be the Debt Reduction Facility (DRF) used under the HIPC initiative in 2005-2010, but it required extensive negotiations, and is likely not adapted for this specific purpose.12 Finally, any unilateral (and in some contracts, even negotiated) change in the debt repayment schedule could lead to a decline in the credit rating of the country, and reduce or even cut its market access. As a consequence, private lenders were called to voluntarily exert forbearance. The IIF letter, quoted above, indicates that this will be the case for a substantial number of lenders, but it could also leave a number of “hold-outs”. In mid-May, private creditors of African countries have recently formed a group which will be the main vehicle for negotiations.

Finally, the G20 agreement also asks multilateral development banks (MDBs) to further explore the options for the suspension of debt service payments over the suspension period, while maintaining their current rating and low cost of funding. Together, they hold about half of the debt stock of IDA and blend countries, but debt service tends to be lower as lending is largely under concessional terms. The difficulties are mostly operational, as they would need the support of their member countries: as indicated by the World Bank Group President David Malpass in a recent press conference, the view from MDBs is that this would require full compensation by shareholder contributions13. Indeed, lower debt service could lead to worsen their credit rating and thus their cost of funding, threatening their business model. If those difficulties could be overcome, projections for 2020 would indicate a maximum of USD 9.2 billion added to the standstill – a magnitude equivalent to the debt service due to the private sector. One approach adopted by the IMF is the use of the Catastrophe Containment and Recovery Trust (CCRT). This facility provides donor-funded grants to countries to cancel debt service on IMF loans, thus freeing immediate liquidity. Only 29 countries are eligible, which makes it very targeted. A recent proposed policy reform14 would help countries use this grant for the very short term in 2020. On 13 April, the IMF Board approved debt service relief on IMF loans for 25 countries, for USD 500 million, with support from the United Kingdom (USD 185 million), Japan (USD 100 million), and potentially others, including China15. While small in aggregate compared to the total debt service to multilateral institutions, this amounts to 0.2% of GDP of this specific group of 25 countries.

 African countries would benefit most from the debt standstill, and non-DAC lenders would bear most of the burden

In nominal terms, the proposed standstill would benefit primarily countries in sub-Saharan Africa. With 38 sub-Saharan African countries eligible for the debt service suspension, including large economies such as Nigeria and Kenya, most of the moratorium accrues to those economies. About USD 22 billion is due in 2020 by sub-Saharan African governments either to official bilateral creditors or the private sector. Deducting 25%, the suspension for 2020 would be of about USD 16.5 billion. In other regions, and in particular in the Pacific and in Latin America, many of the economies that would benefit from the standstill are small island developing states (SIDS). As a result, when compared to GDP, the effect of the suspension is sizable.

It also targets fragile countries heavily, as debt and fragility are tightly linked. Of the low-income countries assessed at high risk or already in debt distress before the COVID-19 crisis, all but three are either fragile contexts, small island developing states (SIDS), or both (OECD, forthcoming[7]). In addition, the economic crisis will hit more deeply commodity exporters, which have been accumulating debt faster, including commodity-backed debt (World Bank, 2020). These tend to have weaker institutions, in particular as relates to fiscal management.

Extending the standstill to 2021 would help reduce medium-term uncertainty, and provide USD 36 billion more in liquidity in 2021. The G20 agreement offers the possibility of maintaining the standstill up to 2021, which would be important as the possibility of a “second wave” of the pandemic might stifle the recovery. In total, the OECD estimates the debt service covered in 2021 at USD 36 billion, 44% higher than in 2020. Including multilateral debt service, the total would rise to USD 50 billion. A large part of this increase stems from the fact that only part of 2020 is covered16, while 2021 would cover the full year. Out of 69 countries, 8 have expected debt service in 2021 above 3% of GDP (12 when including multilateral debt service) and 20 above 1% of GDP17 (45 when including debt service to multilaterals). A few countries also face substantially higher debt service beyond the effect of timing: for example, Angola’s 2021 debt service is projected to go from USD 4.5 billion to USD 7 billion (5 to 8% of GDP), due to a jump in debt service to private lenders. Nigeria, Senegal, and Mongolia are also projected to face doubling of the debt service in 2021.

 

Figure 2. Debt service expected for 2021 by creditor type

Source: (World Bank, 2020[5])International Debt Statistics. https://data.worldbank.org/products/ids

Note: We show countries where external debt service (excluding multilaterals) represents over 1% of projected GDP for 2021.

Provided they implement it fully, official bilateral lenders outside of the OECD DAC are likely to contribute more to the debt standstill, as they have lent higher volumes, and at less concessional terms18. This is a direct consequence of the emergence of non-DAC members as major official bilateral creditors. China in particular has sharply increased its lending operations in Africa. It is also possibly under-reported, with some important gaps in data, as documented by numerous researchers and policy institutions.19 Saudi Arabia20 and the Russian Federation are also important bilateral lenders beyond the DAC. On aggregate, the stock of non-DAC members lending has increased to about 68% in 2017, the last available data point. In addition, on average, 25% of official bilateral loans are concessional, against 90% when considering loans from DAC members. As a result, it is likely that at least 70% of debt service due to official bilateral creditors is outside of DAC members. Given the lack of existing data on lending by non-DAC members, it is difficult to ascertain the relative shares of different countries, but existing academic sources indicate that China is likely the largest debt holder.

The implementation of the standstill from China remains uncertain. China is a signatory of the 15 April G20 agreement, but its implementation might be subject to different interpretations. For example, if China were to distinguish between loans extended by development banks (China Ex-Im Bank and China Development Bank) and official government loans, this would exclude at least two thirds of the debt stock from the standstill (according to (Horn, Reinhart and Trebesch, 2019[8]). In addition, there are uncertainties on the magnitude of Chinese lending: it is likely to be larger than what is indicated in borrower’s statistics recorded by the World Bank. The same authors estimate it at USD 150 billion. While debt service due in 2020 remains uncertain, it is likely to represent a sizeable share of the amounts covered, and thus magnitudes under standstill could be well below the USD 16.5 billion quoted above if Chinese participation was limited. Alternatively, other forms of restructuring could be extended by China, as has been the case in the past21, often under bilateral negotiations.

 

Figure 3. Evolution of the debt stock between DAC and non-DAC countries

Sources: (World Bank, 2020[5])International Debt Statistics (https://data.worldbank.org/products/ids) and (BIS, IMF, OECD, World Bank, 2020[9]) Joint External Debt Hub (JEDH): http://www.jedh.org/data.html.

 

Figure 4. Public and publicly guaranteed external debt stock composition

Source: (World Bank, 2020[5])International Debt Statistics (https://data.worldbank.org/products/ids).

Given these conditions and existing debt service data until 2018, the OECD estimates the debt standstill stemming from DAC members is between USD 1.5 billion and USD 3 billion. With total bilateral debt service suspension expected at USD 16.5 billion, and less than 30% of debt stock, the expectation would be less than USD 5 billion from DAC countries. However, taking into account that DAC loans are at least 3 times as likely to be concessional, the total debt service due to DAC members is likely closer to below USD 3 billion. In addition, in 2018, DAC members received about USD 1.8 billion in debt service from IDA and blend countries. Since 2010, debt service has fluctuated between USD 1.6 and USD 2 billion, with little clear trend, although it has slowly risen from USD 1.6 billion since 2015. With little additional available data, USD 1.5 billion is a robust lower bound. As such, the burden of the standstill is thus likely to bear mainly on China and other non-DAC members, estimated around USD 13.5 to USD 15 billion.

Within the DAC, debt service from official bilateral loans accrue to a small subset of official lenders. Development assistance or ODA from DAC members is delivered through grants and concessional loans, among other forms of support. In addition, official lenders/governments provide loans at non-concessional rates: those flows are called “other official flows”. Most countries focus their development co-operation budget on grants, and correspondingly receive little debt service as most of their aid is delivered via grants. However, a few donor countries, including the European Union, deliver aid through loans. This minority will bear most of the decline in debt service for 2020.

This debt moratorium will not impact the newly adopted headline "grant equivalent" official development assistance (ODA) of DAC countries, but will have an impact on net ODA flows. All forms of development finance should, to every extent possible, be protected and stepped up. Under the "grant equivalent" methodology, debt repayments are not recorded. As a result, the debt standstill will not impact the total ODA headline measure, but will have an impact on net ODA flow trends. Under the "net flow" methodology, if all other factors remain the same, the standstill could increase net flows in 2020 as debt repayments affected by the standstill would not be recorded, but the impact would be reversed in 2022 to 2024, which could result in a decrease in net ODA flows, if not offset by greater disbursements. In its recent statement on its response to Covid-19,22[1] the DAC agreed to “strive to protect ODA budgets, encourage other financial flows to support governments and communities in partner countries, and invite other development co-operation partners to do the same”.

 

Figure 5. Debt service to DAC members, in million dollars

Note: Some observations related to debt cancellations (in particular, Myanmar 2013/14) are removed

Source: (OECD, 2020[10])

 Conclusion: a temporary relief but leaving long‑term issues on the table

Overall, the adopted debt standstill is a positive step for international co-operation, which is more important than ever to respond to the global crisis, including through the G20 and multilateral institutions. While limited in its scope, the G20 agreement on a debt standstill will bring substantial liquidity to recipient countries, and represents constructive collective action in building and protecting global financial safety nets, notably for low-income countries. For counties with market access, the necessity to limit non-concessional borrowing during the suspension period might limit participation, but for others it could be an essential provision of liquidity.

The G20 agreement does not include language on future reassessment of debt sustainability, including restructuring and roll-over risk. The joint call of the World Bank and the IMF was clear: in addition to the urgency of a debt standstill, the G20 should also require a reassessment of debt sustainability, including to “prepare a proposal for comprehensive action by official bilateral creditors to address both the financing and debt relief needs of IDA countries”.23 If the global economic downturn proves to be long-lasting, several of those countries could indeed become insolvent. While the immediate need is to provide urgent resources, it would be necessary to consider deeper restructuring of the debt on a case-by-case basis.

The prospect of a future debate on a more comprehensive restructuring could lead to a lower participation in the standstill itself. Private lenders’ incentives are shaped by their expectations of future repayments. If they expect a restructuring in the future, they could decide to minimise their losses by claiming their debt service in 2020. Collective action issues add to those dynamics: if some actors expect others to not comply with the standstill, they might themselves do the same. It will be important to monitor compliance on the lenders’ side.

A second possible concern around debt restructuring is around how the resources made available by the standstill will be used. This was one motivation for the inclusion in the G20 agreement of monitoring by IFIs on the use of the reallocated debt service towards health or economic support, in order to avoid mitigate moral hazard. Another concern is more forward looking: the possibility that the debt standstill would somehow encourage less prudent fiscal behaviour in the future. However, in this case, the crisis is clearly exogenous, and not caused by any “imprudent behaviour”. The “conditionality” of the standstill is relatively light, but clearly aims to limit opaque practices, by requiring countries to disclose their financial commitments.

The issue of lack of transparency in lending practices is raised in the agreement, and should be strongly enforced to positively influence future resolutions. One of the conditions raised for applicants to the “debt service suspension initiative” is to “disclose all public sector financial commitments (debt)” with technical assistance from the IMF and the World Bank. While countries under the IMF programme are already required to disclose their commitment this could allow for significant progress.

The needs of vulnerable middle-income countries not covered by the existing standstill should be seriously considered. Countries dependent on oil exports or other commodities, as well as tourism, are experiencing a considerable economic shock, and could see a large share of their population slip back into poverty. In addition, the global recession could, depending on how long it lasts, push debt levels beyond what can be sustained. Several affected countries, such as Zambia (whose main exports, copper, experienced a 20% price decline), Ecuador (suffering from one of the worst outbreaks among developing countries), or Argentina (whose debt was already close to unsustainability before the COVID-19 shock) have already entered negotiations with lenders to restructure their debt. The international financial community has an important role in ensuring an orderly process for vulnerable countries.

Finally, the crisis also highlights new modalities for development co-operation. G20 responses in support of developing countries most hit by the COVID crisis should also look at mobilising all sources of finance, including the private sector, and highlighting principles of development co-operation (effectiveness, ownership), with modalities such as South-South and triangular co-operation. The DAC agreed to “strive to protect ODA budgets, encourage other financial flows to support governments and communities in partner countries, and invite other development co-operation partners to do the same”, and should use those scarce resources to leverage others and maximise their impact.