Suriname's sovereign restructuring on the move
Bond restructuring in the books, all eyes now on the IMF?
In May 2023, Suriname reached a restructuring agreement with holders of its two outstanding Eurobonds, contingent on obtaining a Staff Level Agreement with the IMF for the next program review. The country was placed in a tough spot, with the defaulted bonds capitalizing at a high rate while the default could have hampered FDI badly needed to kickstart oil exploration and lift up economic growth.
The IMF should soon unveil an updated macro-fiscal framework, but the restructuring deal appears to fall short of the hypotheses underlying the 2021 program. Since then, the government had to cope with a sharp deterioration of the macro environment, while trying to implement required reforms. There is a risk that a remaining gap in an updated IMF program would need to be bridged through additional fiscal consolidation instead of debt restructuring, right as the government faces significant economic headwinds.
The next step will be for Suriname to get the IMF program back on track, putting China into the limelight as textbook financing assurances – never formally obtained in the first place – will probably be required for disbursements to resume.
Deep-dive into the terms of the deal
The main terms of the agreement in-principle have been laid out in a press release – the blog is based solely on this information which might be complemented when the deal closes. Interestingly, the deal is contingent on Suriname reaching a Staff-Level Agreement with the IMF by June 15.
In short, the two outstanding Eurobonds will be bundled into one fixed income instrument with a face value of $650m – that is a 25% principal haircut on original face value and accumulated past due interest (PDI) – a 10-year maturity, and a coupon of 7.95%. During the first two years, only 4.95% will be paid in cash with the remaining 3% capitalized.
A first observation is that PDI has been capitalized at the hefty prevailing contractual rate for the past 3 years, that is 12.875% and 9.25% for the two existing bonds. In fact, the 2023 bond originally bears a 9.875% coupon but features a provision (“Additional Interest” in the prospectus) which states that the interest rate will increase by 300 basis points after any missed payment by the issuer.
In the end, the PDI significantly increased the size of the claims: after the haircut the total face value is only reduced from $675m to $650m, that is a c. 4% reduction.
Using the Sturzenegger and Tresbesch formula for haircuts (1-PV_new/PV_old) I find that the NPV reduction is c. 16% with a 5% discount rate (the one used by the IMF). Looking at market-relevant discount rates, we get a c. 29% NPV reduction at 10% exit yield, and c. 40% reduction at 15%. These are ballpark estimates though Daniel Munevar finds similar figures. I made the simple (and conservative) assumption that missed coupons are settled in year 1 for the old bonds which gives a slightly higher haircut.
The fixed income instrument described above will be complemented with a value recovery instrument (VRI) paying out only if the country manages to exploit its yet-untested oil reserves and receives royalties. After an initial $100m of oil royalties goes to the government, creditors will be able to claim 30% of the yearly oil royalties from block 58 until 2050. The initial amount of the VRI is set at the value of the initial haircut multiplied by a factor of 1.2, meaning if it pays out bondholders would not only recover their losses but also make decent money in the process. The maximum amount payable under this instrument is set at $689m, coming on top of the fixed income instrument, as the initial amount of the VRI increases over time at a 9% capitalizing interest rate.
The VRI features a call option enabling the government to prepay the instrument without penalty or premium. As soon as royalties start coming in, the government will face a difficult trade off: if it expects the 30% of royalties from block 58 will exceed the outstanding amount of the VRI before 2050, it could be tempted to retire the VRI right away to limit the payments to creditors. However, this would severely drain public finances and there is still a risk that royalties would not materialize in the end.
These VRIs will surely be difficult to price: not only investors need to factor in oil price forecasts, they also need to price the actual volume of oil in Suriname’s block 58, estimate the ability of the government to exploit these reserves, the volume of royalties they would derive from it, and the probability that the government calls the instrument. However, with the concentrated investor base (5 investors control c. 75% of the outstanding bonds according to the authorities), a secondary market for these instruments is unlikely to exist and they will probably sit on the books of the participants in the exchange until expiration.
Finally, the press release indicates that “the full amount of Suriname's royalty payments from Block 58 shall be deposited” in an offshore payment account, over which the investors have some sort of control. More precisely, upon the occurence of certain events (not specified for now) investors will be able to exercise a put option and sell the VRI back to Suriname using the funds docked in the offshore account.
The information available for now begs serious questions about sovereignty when it comes to the benefits of the country’s natural resources, especially since the amounts held in the offshore account will be higher than just the share of royalties investors are owed. When the final deal is inked the precise contours of the so-called put events will be a matter for scrutiny.
More broadly, the use of quasi-collateral like offshore accounts is rather unusual for Eurobonds (no precedent immediately springs to my mind), but quite common for Chinese loans as showed by Anna Gelpern and colleagues. Coincidently, the IMF discovered that China Exim had drawed funds from an escrow account in Suriname between the approval of the program and the first review. The use by bondholders of similar structures could prove useful to China when it comes to pushing back on “debt trap” accusations by western governments, answering that commercial creditors located in the main financial jurisdiction share the same practices.
These mechanisms can come back to bite: they make it difficult for issuers to restructure bonds or even run arrears, complicating the negotiations of IMF programs since the Fund can ask further assurances from such creditors. For now, it will be interesting to see what the IMF has to say about the offshore payment account in its staff report when the program makes it to the Board.
A deal at odds with the IMF program?
The Suriname IMF program is currently off-track and the IMF recently indicated work is ongoing with the authorities and a mission is expected soon. The new program will likely feature an updated macro-fiscal framework and debt sustainability analysis which is not yet public.
In the meantime, we can compare the characteristics of the last deal to the restructuring parameters that formed the bedrock of the IMF program approved in 2021 – quotes in this section are from the 2021 staff report. Disclaimer: any relief calculation for bondholders is somewhat of a lower bound since it does not take the VRI into account – if the security pays out it is likely that bondholders will not grant any relief.
First, about nominal haircuts:
Under this scenario, the face value of external commercial debt, including arrears—bonds and non-ECA backed loans—is reduced by 40 percent at end-2022, and the amortization of remaining debt outstanding would be paused for 3 years.
The IMF hence expected a bigger face value reduction than the 25% featured in the deal. The delay in reaching the deal and significant capitalized PDI, make the actual reduction of the face value of the claim even smaller.
About coupon reduction, the IMF expected an average coupon of 3.4% for non-ECA-backed commercial debt:
In addition to the nominal haircut and extension of maturity, the restructuring scenario also assumes interest payments starting in 2023 with reduced average coupon rates of 3.4 percent for non ECA-backed commercial debt and Eurobonds, and around 1.1 percent for official and ECA-backed commercial debt.
Non-ECA-backed commercial claims other than Eurobonds represented 1% of GDP at end-2021 compared to 31% of GDP for bonds. Since the new bond will bear a 7.95% coupon, it is unlikely that non-bonded debt instruments can have a low enough coupon to get the average down to the IMF requirements.
Then comes the NPV relief:
This restructuring scenario results in NPV reduction of around 36 percent for official bilateral and 45 percent for external commercial creditors at a 5 percent discount rate, and 62 percent for official bilateral and 58 percent for external commercial creditors at a 10 percent discount rate.
As discussed in the first section, the latest deal with bondholders seems to result in a NPV reduction of c. 16% at a 5% discount rate and 29% at 10%, significantly below the IMF’s assumptions. This could also be a breach of the Paris Club’s comparability of treatment principle, as the NPV relief appears significantly lower than the one granted by official creditors – the final call on this will come from the two other measures the Club uses to assess CoT in a holistic manner (nominal flow relief and duration extension).
Overall, the restructuring appears to fall short of the IMF expectations. Problem is, since the approval of the program at end-2021 the macro and fiscal backdrop have arguably deteriorated with lasting scars from the pandemic, spillovers from Russia’s invasion of Ukraine, and inflationary pressures. The deterioration was the reason for which the program went off-track in the first place, as the authorities considered that reforms required by the IMF had become unpalatable.
A deeper restructuring would therefore have been expected to stay in line with the updated IMF program parameters. As things stand, either the program will have loose assumptions – we have the Sri Lanka precedent in that regard – or the fiscal adjustment will be much stricter – something that the government can ill afford in the current economic context.
A glimpse of negotiation dynamics in recent months
Another way to understand how the negotiations unfolded in the last stretch is to compare the final terms to the ones that were on the table in July 2022 from both the authorities and bondholders, with significant gaps at that time. I outline the key differences between the two parties below, and indicate on which side the final deal ended up. Disclaimer, bondholders “stated that the alternative restructuring scenario they presented should not be regarded as a proposal”, but it still brings valuable information on the direction of travel.
For the fixed income instrument:
Government proposed a 33% face value haircut, bondholders 20% – the 25% value is closer to the bondholders
Government wanted PDI to accrue at 8% instead of contractual interest rate after 2021, bondholders wanted contractual interest rate until the restructuring date – in the end PDI accrues at contractual interest rate up to the closing date
Government wanted a 6% average interest rate, bondholders 8.75% – the 7.95% value is closer to the bondholders
Government wanted a step-up coupon, bondholders wanted a straight coupon – in the end bondholders do get a straight coupon, with part of it capitalized in the first two years
For the VRI:
Government wanted a one-off floor of $500m in royalties before payments to creditor kick off, bondholders wanted a $50m floor – the $100m value in the final deal is closer to the bondholders
Government wanted the interest rate on the VRI to be set as the average coupon of the fixed income instrument, bondholders wanted a step-down from 12% to 7% – the 9% in the deal is somewhat in line with what bondholders wanted, and higher than the coupon on the new bond
Government wanted the VRI to expire in 2035, bondholders did not provide a date – in the end the expiration date in 2050 is significantly later than the government’s proposal
Government wanted a hard cap on the VRI payments, bondholders did not want a hard cap – in the end there will be a hard cap at 2.5 times the value of the nominal haircut initially accepted by bondholders
Government wanted to allocate 10% of royalties to bondholders, bondholders wanted 30% – the final deal features a 30% allocation ratio
Bondholders wanted a security interest included in the structure of the offshore escrow account, which the government did not propose – the final deal does feature the security interest
Overall, it appears that the needle moved in the direction of bondholders for most indicators except the inclusion of a hard cap in the VRI.
A new normal for bond restructuring dynamics?
It is interesting to try and understand the incentives and leverage of all stakeholders that might have led to the final deal, before drawing takeaways.
First, time was not on the side of the government for several reasons. PDI capitalizing at the prevailing contractual rate was rapidly increasing the size of the claims, by more than $60m per year. As bondholders apparently refused to have a cut-off date after which the accrual rate would be lower (see the previous section), it was urgent for Suriname to get a deal at least in order to stop the bleeding and lower the interest rate.
Additionally, it would have been harder for the government to attract the FDI needed to finance the upcoming exploration of oil while in default to bondholders. Empirical literature tends to show that sovereign defaults have a significant negative impact on FDI attraction. For what it’s worth, the national oil company announced the signature of profit sharing contracts for two offshore blocks with international oil companies a few days after the restructuring deal.
In parallel, we can assume that bondholders had significant leverage: these are small bonds, outside of the main indices, and the investor base was particularly concentrated with 5 investors holding 75% of the outstanding. This is a point that, in my view, warrants more attention in current policy debates: smaller issuers have to cope with a constrained investor base, less liquid secondary markets, facilitating the obtention of blocking stakes giving a few holders the upper hand in negotiations and reducing the impact of majority voting provisions. It is especially the case for small island developing states, which often undergo restructurings when climate shocks occur.
Therefore any takeaway from the Suriname deal for other restructurings is to be taken with a pinch of salt: the behavior and leverage of bondholders is probably different that in, say, Zambia or Sri Lanka – that’s even before throwing the oil exploration story into the mix.
Still, the Suriname deal aligns with a trend of increasing proactiveness from bondholders in restructurings as I alluded to in my takeaways from the IMF meetings. It started earlier this year when investors made a spontaneous restructuring offer to Ethiopia for a restructuring of its outstanding Eurobond, before the bilateral restructuring under the Common Framework and even before the IMF came out with its updated DSA.
Therefore, I still expect more movement from bondholders in coming months, trying to circumvent IMF constraints and the logjam among bilateral official creditors. Indeed, even when the bonds accrue PDI at a high interest rate in theory, investors still earn nothing on the position for several years, making these instruments a tough sell when US t-bills are paying 5%.
This dynamic can put governments in a difficult position: while they are surely willing to accomodate the demand of bondholders for faster deals, they also need to comply with contradictory injunctions. As part of Paris Club or G20 Common Framework agreements, countries have to respect the comparability of treatment principle, meaning earlier bond restructurings deals can constrain a debtor in its subsequent negotiations with bilateral creditors. Then again, if bondholders strike deals before the conclusion of IMF programs, countries might be forced to implement additional fiscal adjustment in order to comply with program assumptions later on.
What’s next? China back in the limelight
In May 2022, the IMF reached a Staff-Level Agreement on the second review of the program, which never made it to the Board as the program went off-track. The next step for the country will be to get the program back on track, as the Fund hinted at an upcoming mission during the spring meetings in April 2023.
We could expect a Staff-Level Agreement on a new review in coming weeks, especially since the IMF signalled its flexibility to adapt the program, but China could become a roadblock once-again towards the approval of the review by the Board and the resumption of disbursements. This, as an aside, might also be why bondholders made the deal contingent on a Staff Level Agreement and not an IMF Board approval.
As I discussed at length in previous blogs, the IMF used its arrears policies in 2021 to override the need for financing assurances from China and approve the program. However, with the evolution of policy discussions in recent months it is unlikely that the same trick could be used again without triggering reactions from major shareholders.
During the spring meetings, the acting director of the IMF’s Western Hemisphere Department touched exactly upon this issue:
And in the program previously, we had basically moved forward on the program on the basis that they were essentially not paying the debt to China and would be eventually restructured. So, I think we are looking for some progress in terms of that restructuring of the Chinese debt. The Surinamese already reached agreement with its Paris Club creditors and it's close to reaching an agreement with India, which is the other big creditor. And so, I think having some more progress on the debt restructuring talks with China would really help, both help both with the program and help the country.
Like in Zambia, the IMF does not say precisely what kind of progress would be necessary to get the program back on track. I would expect that the IMF is asking China for a letter similar to the one provided to Sri Lanka, which most importantly included a commitment to provide a debt treatment in line with IMF program parameters.
An interesting question is to what extent the bond deal could impact China’s behavior. In a podcast, Jeromin Zettelmeyer said that China was reluctant to restructure in Suriname in part because it wanted to make sure bondholders would also restructure. When the dust settles, if it appears the deal struck by bondholders is more attractive than the Paris Club deal, China could demand to align with these latest terms.
Then if the IMF program resumes, it will be interesting to see whether China turns these financing assurances into a concrete restructuring deal, something that has proven difficult so far in Zambia and Sri Lanka. One risk in that regard is that the Paris Club has – for the first time in its history to the best of my knowledge – made its restructuring agreement contingent on the respect by Suriname of the comparability of treatment principle. If China does not restructure its claims, Paris Club claims due after 2025 would not be restructured and the debt amortization would further steepen.
Suriname finds itself in a tough spot again, and the road ahead is bumpy at best. There are a lot of moving pieces which will be interesting to watch:
The update of the IMF macro-fiscal framework will give a first indication on the feasability of the fiscal adjustment to restore sustainability given current restructuring parameters.
Getting China on board is set to be the main challenge for the government in coming months, including the obtention of financing assurances and a subsequent restructuring deal.
The light debt treatment for Eurobonds might result in a very high exit yield, hampering market access and leading to the need for subsequent restructurings.
The Paris Club will see its assessment of comparability of treatment tested by the terms accepted by bondholders and China, which might be an occasion to increase the transparency of its assessment, right as there are policy discussions about a common formula across creditors to assess CoT.
I am grateful to Teal Emery for valuable comments.
If you want to discuss this further, please reach out: firstname.lastname@example.org
You can also follow me on twitter @TheoMaret
Views expressed here are my own, no my employer’s.