Macro Analysis /

IMF/WB call for debt standstills: further thoughts and implications for bonds

  • There is little more to go on in terms of how the recent call for debt standstills by bilateral creditors would work

  • If implemented, it would be the biggest international debt relief effort since HIPC and the Brady plan

  • Possible bailing-in of bondholders is also unclear; but if necessary, established principles should be followed

IMF/WB call for debt standstills: further thoughts and implications for bonds
Stuart Culverhouse
Stuart Culverhouse

Chief Economist & Head of Fixed Income Research

Tellimer Research
7 April 2020
Published byTellimer Research

In this note, we set out some further thoughts on the recently announced joint appeal by the IMF/World Bank for official bilateral creditors to provide a debt standstill to some of the world’s poorest countries to help them cope with the impact of Covid-19. The poorest countries targeted are the 76 countries that are IDA-eligible as defined by the World Bank. Of these, 22 have international bonds and these could get caught up in this initiative too. 

We should stress that there is still little more to go on at this stage in terms of how this would work since the announcement on 25 March, although recent IMF comments suggest that the standstill on debt service to official bilateral creditors should last for a period of one year. The IMF/WB hope this proposal will be endorsed by the G20 at the forthcoming Spring meetings over 16-17 April. We may hear more about the policy and operational details after that. But, despite the urgency, we do question how quickly this can be implemented, and for the first countries to benefit from it, given the time it might take to coordinate the official sector response, particularly to undertake debt reconciliation and ensure the effective participation of non Paris Club creditors. 

We also stress that the possible involvement (bailing-in) of bondholders, while likely in our view, albeit on a case-by-case basis, is still unclear at this stage but nonetheless should follow established principles where it is deemed necessary; we may hear more on this in coming weeks too as the market digests the news.

Scope of debt relief 

If implemented, we think this initiative (perhaps it might be called the Coronavirus Debt Relief Initiative or CDRI?) could be – on some measures – the biggest international debt relief effort since the HIPC/MDRI debt relief initiative and the Brady plan, at least by number of countries it covers if not the size of debt covered or amount of debt treated. According to our analysis, it would be the biggest ever in terms of number of countries covered (potentially up to 76 IDA-eligible countries vs 39 and 18, respectively). The amount of debt covered is also large. We think this initiative could cover debt amounting to some US$165bn of bilateral debt owed by IDA-eligible countries on our calculations, and potentially up to US$220bn if their bonded debt is also included (which we estimate at US$58bn), and applied to all eligible countries, let alone the addition of other commercial debt too, such as loans. By contrast, however, we think Brady's covered more debt. Brady bonds were issued in an aggregate amount of over US$160bn and, given the typical face value reduction in Brady deals, this might imply a debt stock before treatment in excess of US$250bn, according to our back of the envelope calculations (even more in today’s money). 

But it would certainly not be the biggest by amount of debt actually treated, which here is intended to be much smaller than the other two initiatives. This initiative involves flow relief (the suspension of debt service due for a year, which is presumably to be repaid at a future date). There is no debt reduction, or debt forgiveness as such (as far as we are aware), nor is it applied over multiple years (at least yet). One year’s debt service may not be that large (even on a large stock of debt, given the likely long-term and concessional nature of much of it, albeit some may be more expensive ECA-related debt), although it can be significant for each individual country. We calculate that the amount of debt (debt service) to be treated here could potentially amount up to some US$21bn (based on recent historical numbers – projections are not available). Of this, cUS$16bn arises from debt service payments (principal and interest) by all IDA-eligible countries to bilateral creditors, based on World Bank IDS data for annual debt service payments, and a further US$4.8bn arises from interest payments on international bonds issued by these IDA-eligible countries, according to our own calculations (there would be only a small amount of principal due over the next year), if all the countries deemed eligible were to use it. For comparison, the total cost of providing assistance under HIPC is estimated at US$76bn, according to the IMF (as of end-2017 in NPV terms).

We estimate total public and publicly guaranteed (PPG) external debt owed to bilateral creditors by (all) IDA-eligible countries is cUS$165bn, based on World Bank IDS data (latest figures for 2018, see Table 1). Based on separate Paris Club data, we calculate the amount outstanding from IDA-eligible countries to Paris Club creditors is US$55bn. This might suggest some US$110bn is owed to non-traditional bilaterals. However, this is only an illustrative comparison given the different data sources (a bit like comparing apples and oranges). But we think these figures demonstrate the importance of getting wider buy-in across all bilateral creditors (Paris Club and non-Paris Club) in order to maximise its effectiveness, as well as reducing the risk of free riding.

Of the 22 IDA-eligible countries with international bonds outstanding, we calculate the stock of PPG external debt owed to bilateral creditors is US$93bn, while that owed to Paris Club creditors is US$25bn. This suggests some US$68bn is owed to non-traditional bilaterals. We reckon the stock of bonds for these 22 IDA-eligible countries is of similar order, about US$58bn (see here).

In terms of debt service, and hence a guide to how much might be subject to standstill (treatment), we calculate that total debt service owed to bilateral creditors by IDA countries was cUS$16bn in 2019, comprising US$11bn in amortisation and US$5bn in interest. Of the subset of 22 IDA countries with bonds, we calculate that total debt service owed to bilateral creditors was cUS$10.5bn in 2019, comprising US$7.4bn in amortisation and US$3.1bn in interest. However, these figures are quite lumpy; in terms of debt service, Pakistan accounts for most of it (nearly a third), followed by Kenya, Ghana and Ethiopia (collectively another third). We estimate, for this subset of 22 issuers, debt interest on their bonds is of similar order, cUS$4bn (see here).

We make some more observations

  1. Practicalities. There are few further details available at the moment on how this official sector debt relief/standstill will operate. We note, however, that it is envisaged the standstill on debt service (principal and interest) owed to official creditors is for a period of one year, according to remarks by IMF Managing Director Kristalina Georgieva at the joint IMF-WHO press briefing on 3 April. Georgieva noted that when the economy is standing still, debt service obligations should be in a standstill. This may raise questions – especially if it applies to bondholders – over what happens if individual countries take longer to recover (and who decides), and if there will be sunset clauses after which time the initiative expires. One of our observations from the global financial crisis is that sovereign defaults may occur with a lag; there were few sovereign defaults at the time (Ecuador and Seychelles). However, this initiative may be more efficient in terms of supporting an immediate response rather than procrastination, which would merely delay the inevitable and require even deeper action in future. 
  2. Debt reconciliation. There may need to be a debt reconciliation exercise by bilateral creditors with IDA-eligible debtor countries to agree what they are owed, by whom, and on what terms. This will need to be reported to the IMF/World Bank to inform their assessment of liquidity needs, debt sustainability and debt relief requirements. We suspect this could take time. 
  3. Role of China. There may be some concern about China’s attitude to this initiative and if, and how, it might participate. In the aforementioned IMF-WHO press briefing, we note that Georgieva stated China is engaged constructively on this issue and has framed “a set of principles it would be interested to see being integrated in that process”. It is not clear what this means, although appears to suggest China has its own ideas on conditionality, but at least shows it is aware of the initiative and the tacit expectation that it does its bit. We note that China has little experience in sovereign debt restructuring in a multilateral context, preferring bilateral agreements instead, although Beijing may have learned some valuable lessons in the Republic of Congo.
  4. Non traditional bilateral creditors. While agreement among G20 leaders will send an important signal, it might be a worry that there seems to be limited ability to enforce any agreement on non-Paris Club bilateral creditors. It is instructive that, to this day, non–Paris Club bilateral creditors as a whole have delivered only half their share of HIPC Initiative debt relief, and one-third of these creditors have not delivered any relief at all, according to the IMF.

Implications for bondholders

The implications for bondholders (and other commercial debt) are still unclear at this stage. We find it difficult to believe that this initiative will not apply to bondholders, and therefore implying a standstill too on upcoming debt service on bonds, either through application of the principle of comparability of treatment or moral suasion, but nor do we think most bondholders would want to resist it anyway. Indeed, we think it would be difficult for any fund with ESG principles to argue otherwise. 

Yet, as we noted in our earlier research, we think a standstill on bond payments will be operated on a case by case basis. Some countries might seek such forbearance, some may not (although whether the decision is entirely up to the country is moot). 

However, bondholders – whilst sympathetic – would still want to ensure that any requests for payment standstills and subsequent negotiations followed established best practice in terms of the IIF/G20 principles of sovereign debt restructuring (whether for these eligible countries or anyone else). This would include, for example, principles of goodfaith, transparency, creditor engagement, and fair treatment. 

We also highlight some practical considerations, although of course, with sufficient commitment and goodwill from all sides, nothing would be unsurmountable: 

  1. Debt perimeter. While for bonds the amount outstanding and financial terms is public and shouldn’t be controversial, there may be issues around the possible inclusion of non-bonded debt (bank debt and loans), where the expectation (subject to individual legal terms) might be that these should receive equal treatment to bondholders too. However, lack of transparency over non-public debt could be a factor. 
  2. How will a standstill on bonds be effected? The purpose will be to reach an agreement with bondholders to suspend payments temporarily without triggering an actual default (what happens to ratings and CDS may be another issue). Can this be done informally and voluntarily, or does it – more likely – require a formal voting process? We see two possibilities. First, voluntary agreement not to make certain payments. It doesn't normally require a vote or consent to miss a payment, but might require something to prevent this triggering other actions. This might be a tacit (or explicit) commitment from bondholders not to declare default, although something more formal may be needed. We’ve seen examples of other countries missing payments and going into default, but without bondholders declaring default, or voting to accelerate/calling cross default. However, that might be possible in isolated cases, less so as the number of defaults grows. Second, a formal bondholder vote to change payment terms (either for existing or future coupons) or to waive an event of default. Agreeing certain actions may be made easier in situations where bonds contain collective action clauses (CACs), whereby a super majority (typically 75%) binds everyone, but could be harder if they don’t have CACs, so that unanimity is required. Most modern international bond issues – which will include those issued by IDA-eligible countries under this scheme – contain CACs, although some may not. 
  3. How – and when – will interest that has been suspended be repaid after the standstill is over, and what happens if the standstill needs to be longer than envisaged ex ante? Interest could be capitalised, repaid when coupons resume either in one go or under a repayment plan (a la Cote d’Ivoire in 2012), or embedded in a new PDI instrument (akin to Barbados recently), although, depending on their size, this would be small and illiquid.
  4. Implications for debt sustainability and will more need to be done? While the initiative is framed as a suspension of debt service for a year, which we think bondholders can tolerate (we think the PV hit would be relatively mild in most cases), it may be that some countries need forbearance for even longer. Moreover, given (a) the adverse economic impact from the pandemic and (b) that many of these countries came into this crisis with weak positions and debt sustainability concerns, bondholders may be concerned that the official sector will try to force a bigger treatment on them (even debt forgiveness). We observe that 8 out of the 22 IDA-eligible countries with outstanding bonds were classified as being at high risk of (external) debt distress or with debt in distress (Table 1). A further six were at moderate risk (and it’s not difficult to imagine some of these slipping into the high risk category) while the five for which we don’t have classifications may also slip (or already have slipped) into high risk too. Only three were classified as low risk (Rwanda, Uzbekistan, and Honduras). Having said that, Republic of Congo, Mozambique and Grenada have all recently completed restructurings, so there might be little additional benefit to them of a standstill on their bonds (although Republic of Congo didn’t restructure its bonds, so we wonder if it is at greater risk if China, for instance, seeks to evoke comparability of treatment here). 
  5. Bondholder coordination. Each country, and within them, each bond, will have different holders so committees (assuming there needs to be one) will be different in each case which may prove to be inefficient and cumbersome. A single global committee, which represents all bondholders, irrespective of country/bond, may however not be practical (given diversity of bondholders). Still, we think bondholders can probably move more quickly than the official bilateral sector. 
  6. The official sector (and poor countries alike) may be concerned about the threat of litigation. This may be overstated, in these circumstances, but solutions may include use of exit consents, (temporary) changes in law governing bonds in key jurisdictions (eg New York and London) that make it harder, if not prohibit, legal action being taken against debtors, or extra-territorial legislation (eg a UN resolution) – for instance, see Lee Buchheit and Sean Hagan. Indeed, borrowing from the HIPC experience, we recall the UK Government barred creditors from taking legal action against HIPC countries in UK Courts through The Debt Relief (Developing Countries) Act 2010.