Ghana's domestic debt exchange (DDE), which was launched on 5 December, is due to expire on Monday (19 December) – details of the offer can be found here. In this note, we set out some thoughts on the DDE and implications for the foreign bonds.
Its success (or otherwise) will be crucial in moving forward with the IMF programme, determining the debt sustainability outlook, and perhaps in shaping the terms on the foreign bonds. But the extent of participation in the DDE is unclear to us, amid suggestions that it has met with some resistance (labour unions for instance have called for pension funds to be exempt, according to reports). That said, there are likely to be punishments ("disincentives") for non-participation.
Indeed, the terms of the DDE seem harsh to us and we wonder if domestic debtholders will resist it or whether there will be sufficient participation for it to be effective. Put bluntly, it seems a bit rich for the government to have gone from saying just six months ago that they don't need an IMF programme to now demanding 50% debt relief.
The exchange terms
The DDE involves an exchange out of eligible bonds into a series of new bonds with a lower coupon profile and longer maturities (see here for the terms). There is no principal haircut. While, on the surface, no principal haircut may come as a relief, we think the implied PV reduction from the much lower coupons and lengthy maturity extension is sizeable.
Holders are being asked to exchange eligible bonds into a "strip" of four new amortising bonds maturing in 2027, 2029, 2032, and 2037, in proportions 17%, 17%, 25%, and 41%, respectively. This implies significant maturity extension for most bonds (although a shortening in maturity for holders of 2039 local bonds); that is holders of a shorter maturity aren't being asked to swap into a medium-dated bond, but rather 41% of their new holdings is in bond that matures in 2037!
The 2027 and 2029 bonds amortise in two equal annual instalments, the 2032 bond in three equal annual instalments, and the 2037 in five equal annual instalments. As such, a holder of the series sees amortisation begin in 2026 and continue thereafter annually until 2037 on a smooth and gradual declining basis.
For each series, there is no coupon payment in 2023, rising to 5% in 2024 and 10% thereafter. For context, the government's budget statement noted that the weighted average interest rate on domestic debt increased stood at c20% at end-September 2022. The first interest payment on each new bond will be made on 30 June 2024 (so that's no interest for 18-months).
The stock of eligible bonds is GHS137.3bn (equivalent to about US$9.8bn at the time the exchange was announced, at a GHS/US$ exchange rate of 14, although cUS$15.3bn now, after the cedi's sharp appreciation in the past few days to GHS/US$9, on Bloomberg prices). There are 69 individual bonds that are eligible in the exchange.
Government-issued bonds account for the majority of the eligible bonds (60 bonds and 92% of the outstanding). The government bonds have maturities ranging from 2023-2039. Some 30% of eligible bonds are due to mature in 2023, with another 12% in 2024 and again in 2025; that is, 54% of the bonds mature in the next three years.
There appears to be no minimum participation threshold at which point the government would reserve the right to cancel it (although one would think you'd want at least 50%, and ideally much more than that) and – unlike in foreign bond contracts – there are no collective action clauses that would bind a minority into accepting the decision of a super-majority.
Low participation, with the government seemingly rushing this through without much prior dialogue with stakeholders, would risk further damage to the government's economic credibility and delay IMF board approval (progress on debt restructuring is a prior action). But Ghana could extend the deadline, and perhaps sweeten the terms, or offer regulatory forbearance, to encourage participation.
Alternatively, it could take a more coercive approach. With the domestic debt under local law, it could just force bondholders into new instruments by changing the law; eg as we've seen elsewhere, it could change the law retrospectively to include collective action clauses in the local bonds ("retro-fit" collective action clauses). Or the authorities could introduce disincentives that reduce the value of the bonds not tendered (local holdouts) or simply rely on moral suasion. However coercion risks some reputational damage.
Such ploys have been seen before. As the IMF have noted in a recent policy paper, sovereign debtors have used the local law advantage as a legal device to restructure their domestic debt, although they point out they should carefully evaluate the potential adverse consequences of this option. They observe creditor participation in domestic debt restructurings can be influenced by carrots (sweeteners) and sticks (for instance, most notoriously, in Barbados and Greece, where high participation was supported by a retrofitted collective action mechanism that required parliamentary approval, or we've seen elsewhere the use of regulatory and tax disincentives, punitive risk weightings, limiting access to central bank emergency financing windows and their use as collateral, that can make old securities untradeable).
Indeed, the government has already said (via a recent FAQ) that holders who do not participate will continue to hold the old bonds but may face "additional regulatory measures" so they cannot benefit from their non-participation. It is not clear to us if the government intends to pay debt service on untendered bonds.
We make some other observations:
The government's investor presentation now puts government debt at just over 100% of GDP (around 103-105%). This compares with its budget estimate of 75.9% of GDP in September 2022, albeit with the admission that debt was now unsustainable.
It notes that Ghana's debt sustainability is assessed with regards to the IMF’s Low Income Country (LIC) thresholds for a country with "medium-debt carrying capacity". On this basis, it targets public debt below 55% of GDP in NPV terms by 2028. It also implies public external debt of 40% of GDP and 180% of exports in NPV terms, and external debt service of 15% of exports and 18% of revenue.
As such, this implies something like 48% debt reduction, on our calculations.
Which is close to what we think is the implied NPV relief in the domestic debt exchange (although the implied haircut is difficult to calculate). We value each of the new bonds at a constant 20% discount rate and average across the strip according to the percentage allocation (and assuming existing bonds trade at par). On this basis, our estimated recovery value for the new strip is GHS45 (per 100) – ie implying 55% PV relief.
Uncoincidentally, we calculated that a c50% PV haircut was implied by the deputy of finance minister's comments in the now infamous radio interview, when he referred to 30% principal haircut on foreign bonds and we assume a three-year grace period. We said at the time this sounded excessive and would expect bondholders to resist such onerous terms.
This seems harsh to us, for a country with debt/GDP of only 100%, with market access, and for a government which didn't admit that it had a problem six months ago.
For instance, we note that it would be at the high-end of recent restructurings; eg Greece (PV relief of 65-78%, on IMF estimates, but with public debt at 147% of GDP), and Barbados (PV relief of 30% on the foreign bonds, on our estimates, but with public debt at 159% of GDP). Conversely, Jamaica's second domestic debt exchange (NDX) in 2013 implied PV losses of 8.6% on IMF estimates (with public debt of 147% of GDP). And while Grenada's 2015 restructuring involved 50% nominal reduction (on its foreign bonds), with public debt of 103% of GDP, that was also motivated by climate vulnerabilities.
As such, Ghana is seeking more or similar relief to countries that have had much higher debt burdens. That is, we think it is difficult to justify 50% debt reduction in Ghana when the implied relief for Barbados and Jamaica was lower and they had much higher debt burdens.
Moreover, Ghana seems to be seeking debt relief as the solution to get from 100% to 55%, whereas some of this will (should) come from fiscal consolidation, and an increase in the denominator (GDP growth and real exchange rate appreciation, under a credible IMF-backed reform programme). Assuming instead that the burden is split evenly between fiscal consolidation and debt restructuring, that should imply PV relief on the debt of a more reasonable 24%.
Foreign bondholders may push back on this (as they have in Zambia) although this approach shouldn't come as a surprise – the LIC thresholds and medium debt carrying classification were used in the IMF's DSA in the 2021 Article IV. Rather, a 70% threshold for debt/GDP has being considered by the IMF for emerging markets and Ghana has been a country that has, until a year ago, enjoyed market access over the last 15 years (if not formally designated a market access country). We await the Fund's updated DSA after staff-level agreement.
Be careful what you wish for. The DDE could put foreign bondholders in a difficult spot. While foreign bondholders will have wanted domestic bondholders to be included in a restructuring (and rightly so in our view, given burden sharing and that domestic debt was a bigger part of the debt problem), if domestic bondholders accept a deal with 50% PV relief, it may make it harder (financially and optically) for foreign bondholders – who insist on domestic burden sharing and comparability of treatment – to resist it (although the comparison is imperfect, so there may be ways round it, and there may be ways to sweeten the terms). Moreover, a strong and well coordinated bondholder committee may be able to extract better terms.