Mark Weidemaier, Ugo Panizza and Mitu Gulati teach international finance and law at the University of North Carolina, the Geneva Institute, and the University of Virginia, respectively. Mark and Mitu co-host the podcast “Clauses and Controversies” where Ugo is a frequent guest.
An investor friend taught us a new acronym a few days ago: FUD. He used the term — short for Fear, Uncertainty and Doubt — to describe his feelings about a funky Ghana bond maturing in 2030.
Ghana is on the brink of default after a borrowing spree over the past decade, including a spate of recent kick-the-can issuances. It owes about $50bn in total. Of this, roughly $14bn is denominated in foreign currency, including a $1bn bond due in 2030.
FUD is usually used around deliberate misinformation (or real information that someone is trying to discredit) and anyway, we didn’t initially understand why anyone would have FUDDY feelings about the Ghana 2030? Unlike other Ghanaian eurobonds, it is backed by a World Bank guarantee of 40 per cent of face value.
But upon closer reading, multiple sources of legitimate FUD reveal themselves.
Issued in 2015, the Ghana 2030 bond has the latest next-generation aggregated ‘Collective Action Clause’, which have been gradually introduced since 2014 to ensure orderly debt restructuring. What distinguished these new provisions was an “aggregation” mechanism that was absent from prior generations of government bonds. Simplifying a little, the aggregated CACs let countries restructure multiple bond series in a single vote across all series as long as 75 per cent of creditors vote yes and everyone is offered the same deal (“uniformly applicable” treatment).
Now, for the FUD. The fancy new CAC in the Ghana 2030 bond permits aggregation with other series of bonds that constitute “Debt Securities Capable of Aggregation.” That term is defined to mean:
. . . those debt securities which include or incorporate by reference this Condition 13 and Condition 14 (Aggregation Agent; Aggregation Procedures) or provisions substantially in these terms which provide for the debt securities which include such provisions to be capable of being aggregated for voting purposes with other series of debt securities.
In short, the Ghana 2030 can be pooled for voting purposes with other foreign currency bonds that have the same aggregated CAC, or “provisions substantially in these terms.” Then, if 75 per cent of the voting pool are in favour of a restructuring, the 2030 holders will be forced to take the same deal. And, because all bondholders must be offered the same deal, any restructuring proposal will almost certainly not benefit from a guarantee. Bye bye World Bank guarantee!
(Adding the local currency bonds into the aggregation adds yet more FUD, but we put that aside.)
In principle, the bond documents do not have to be interpreted to allow this result. Ghana’s other foreign-law bonds with aggregated CACs — the ones without guarantees — do not have exactly the same aggregated CAC as the 2030 bond. The reason is that CACs list a dozen or so key matters (such as payment terms) that can be modified only by supermajority vote. These are called “Reserved Matters.” The Ghana 2030 has one more Reserved Matter than the others: the guarantee. But remember, the 2030 bond can be aggregated with any other bond that has an aggregated CAC that is “substantially in these terms” (ie, the terms found in the Ghana 2030). And it is hardly obvious that this difference makes the other bonds not “substantially” the same as the 2030.
It might seem unfair to give holders of the Ghana 2030 bond the same restructuring deal as holders of non-guaranteed bonds. But the aggregated CAC does not require the bond issuer to respect differences in the economic value of different bonds. In fact, the “uniformly applicable” requirement seems to require the contrary. To simplify, it requires the issuer to offer each bondholder the same terms (or the same new instrument). The logic underlying this requirement was that investors holding different securities would find that their interests converged in a crisis. Except, well, they don’t.
Could all this happen? Ukraine recently concluded a ‘debt reprofiling’, with foreign creditors agreeing to a two-year freeze on payments. It has multiple bond series outstanding, some of which had aggregation provisions like those in Ghana’s post-2014 eurobonds. Ukraine’s reprofiling encompassed regular sovereign bonds and two corporate bonds that had a state guarantee. Ukraine did not hesitate to aggregate these bonds together, despite the fact that the guaranteed bonds were arguably superior credits. We are willing to wager that the argument for aggregating guaranteed corporate bonds with regular sovereign bonds was that all counted as “Debt Securities Capable of Aggregation.”
Perhaps the World Bank will step in to protect the 2030 issue. We imagine it would like to protect the value of its guarantees. If the Bank does want to step in, it might be possible to cash out the 40 per cent guaranteed portion of the Ghana 2030, leaving the rest to be restructured like an ordinary Eurobond. Something like this was done when Ecuador restructured its collateralised Brady Bonds in 2000, but that took a lot of fancy lawyering.
But even this path involves FUD. Ghana is broke. If the World Bank pays on the guarantee, then Ghana immediately owes the World Bank. And the World Bank, like the IMF, has priority status — it gets paid first and in full.
In theory, that means that any restructuring attempts come to a standstill until the arrears to the Bank is cleared. Maybe there is a way around this roadblock. But it is another wild card, and hardly the type of uncertainty that the holder of a guaranteed bond wants to contemplate.
So how does the market see all of this?
Back in May 2022, the market thought the FUD bond was going to pay out in full, whereas the other bonds were going to get haircut. The FUD was trading par, with a yield to maturity of 10.8 per cent (the coupon on that bond is 10.75 per cent). By contrast, two Ghanaian Eurobonds bonds due in 2029 and 2032 were trading at discounts to par value of 40 per cent and 60 per cent respectively (for yields to maturity of about 16.5 per cent).
Today, investors seem, well, a bit more fearful, uncertain and doubting. The FUD bonds trades at a 30 per cent discount (19 per cent yield-to-maturity), while the other two bonds have yield-to-maturities which are well above 30 per cent and trade at a 68 per cent and 78 per cent discount, respectively.
Is the market predicting (praying for?) an Ecuador type solution?
