I recently wrote a twitter thread about sovereign bonds from Imperial China, and while it was intended to be humorous, a number of people were more interested in the serious aspects of sovereign debt. I’m hesitant to write about sovereign debt, as I’m not really an expert. Anyone looking for an in-depth understanding of the topic should read as many papers as possible written by Mitu Gulati, a professor at Duke. There are also a number of journalists who do a great job covering sovereign debt for the public, such as Felix Salmon and Joseph Cotterill. On top of that, there are practitioners both in law and finance who undoubtedly know this stuff better than I do.
Without stepping on anyone’s toes, I want to use this post to explain what I think is useful to know about sovereign debt restructuring. Better introductions to the topic probably exist, but if I’m going to point someone to an introduction, it’s going to be an introduction that covers everything I think is important. So in this post, I’m going to talk about a few notable examples of sovereign debt restructuring, and introduce various clauses and restructuring techniques.
There is no sovereign bankruptcy regime; a country can’t simply file for bankruptcy and receive protection from creditors while an orderly payment schedule is worked out. If a country wants to impose a haircut on creditors, i.e. reduce the value of their claims, it has to turn to contract law. The country is going to need to look at the terms of its bond prospectuses, and ask what rights creditors have and how a creditor might enforce those rights.
More specifically, a country needs to determine whether the prospectuses have any clauses that allow bondholders to cram down changes on other bondholders. If so, what clauses can be changed? What is the governing law of the bond, and in what courts can a dispute be heard? What constitutes a default, and who can bring a claim? Has the country waived sovereign immunity in that country, and in relation to what issues has sovereign immunity been waived? Unrelated to the prospectus itself, the country is going to want to determine what assets it owns overseas, and how easily a creditor could attach those assets.
The first country I want to talk about is Ecuador, which in 2000 decided to restructure its external debt. Ecuador’s bonds were governed by New York law, and could be modified by a majority vote of the bondholders (or 66.6%, I forget in this case). However, unanimous consent was required to modify payment terms, including the timing of payments, the amount of principal and interest, and the currency.
In order to force a haircut on all bondholders, Ecuador utilized “exit consents”. Ecuador created new bonds with reduced payment terms, and offered bondholders the right to exchange their old bonds for these new bonds. However, in order to accept the exchange offer, the bondholders had to vote to change the non-payment terms of the old bonds. In particular, they had to vote to remove the cross-default clause (which says that a default on any debt obligation constitutes a default on the bond), the negative pledge clause (which says that Ecuador cannot pledge its assets to another credit if that would harm the bondholder’s security), and the listing requirement (requiring that the bonds be listed on a stock exchange).
The aim was to leave the old bonds “gutted”, such that the value of the old bonds would be considerably diminished. In turn, this induced bondholders to accept the exchange offer. 98% of bondholders exchanged their bonds, and in doing so agreed to modify the nonpayment terms of the bonds they were exiting, affecting the rights of the remaining 2%.
This was the first time exit consents were used to effect a sovereign debt restructuring, and this served as an example to other countries. In 2003, Uruguay decided to restructure its debt using this approach. In the case of Uruguay, modifying the payment terms of the old bonds was not a requirement of the exchange offer, but rather was made optional. Nonetheless, it was pretty much a given that the exiting bondholders would vote for the changes. And in many ways, Uruguay’s use of exit consents was actually more aggressive. In addition to the clauses that were modified in the case of Ecuador, Uruguay also removed the waiver of sovereign immunity in an attempt to render the bonds unenforceable.
It should be evident that using exist consents isn’t a great solution to sovereign debt restructuring, and as such countries and other public actors began to search for a solution. The result of that process was to include collective action clauses (CACs) in New York law bonds. CACs come in various forms, but most of the time when someone mentions CACs they are referring to a clause that allows a bond’s payment terms to be modified with the approval of a certain threshold percentage of bondholders. The threshold percentage is usually higher than the percentage of bondholders required to approve a change to non-payment terms. This type of CAC is therefore a contractual solution to force a haircut on holdout bondholders. If you want to learn more about the drafting of CACs, read this paper by Mitu Gulati.
There are other types of CACs as well, namely collective acceleration and reverse acceleration clauses. A modern example might read:
The idea is basically that bondholders as a group might be made worse off if an individual bondholder could claim the full principal and accrued interest as soon as a default occurs. A collective acceleration clause requires some threshold percentage of bondholders to agree in order to accelerate the payment of principal and interest, and a reverse acceleration clause allows that to be reversed upon the vote of some other (often larger) percentage of bondholders. This is seen as a useful protection for bondholders as a whole against opportunistic actors such as vulture funds.
There is more complexity and variation in the wording of these clauses, especially when you start to consider the way the voting is calculated and the aggregation of votes across different bonds. However, once you understand the basic problems these clauses were created to solve, it is easy to figure out the mechanics of any given bond by reading the prospectus.
A CAC is a contractual solution, and ordinarily must be included in the contract to begin with in order for it to be an effective tool in restructuring sovereign debt. However, there is sometimes a workaround, as seen in the case of Greece.
Greece had a number of English law bonds containing CACs, and when Greece needed to restructure its debt in 2012 these bonds were not a problem. However, the majority of Greek debt was governed by Greek law and did not contain CACs. However, Greece could modify the terms of the bonds simply by changing Greek law. Greek law bonds don’t have CACs? No problem, Greece can simply change the law to impose CACs on the bonds. This sort of retrofit CAC is exactly what Greece chose to do, passing the Greek Bondholder Act 2012.
Of course, Greece could have just changed the payment terms directly, but I think the perception was that imposing CACs was a less coercive measure, and might avoid potential challenges under the Greek Constitution and the European Convention on Human Rights. Only a 50% majority was required to change the payment terms, and this majority applied across all Greek law bonds as a whole, rather than requiring 50% of each bond issue.
The idea of changing domestic law to alter the terms of bonds isn’t exactly a novel or interesting idea, and I highly doubt it surprised anyone. I only raise the point here because it explains something more fundamental: just like how debt denominated in a local currency can always be repaid, as the government can simply create more money, local law debt can always be restructured, as the government makes the rules.
There are a number of countries that decided in the last decade or so to add CACs to their bond prospectuses. Some of these countries, however, have bonds from the late 90s that are still outstanding, and these bonds often lack CACs and are governed by New York law. This results in a somewhat tricky situation, whereby maybe only 85% of a country’s bonds can be restructured using CACs. This happens to be the situation Venezuela finds itself in, as I noted about a year and a half ago. I also know this is the true of Jamaica as well, or at least it was true as of about two years ago.
Interestingly, one could take the view that Venezuela’s bonds that do not include CACs are easier to restructure, as exit consents work just fine and require a lower voting threshold than is required under the CACs. I have my doubts, however. In particular, I am interested to see how a court would view an aggressive use of exit consents alongside bonds that contain CACs. I also just wonder whether it’s worth it.
That is more or less everything I want to say. There are a number of topics that I completely left off, and a lot of detail that I glossed over. Not only is there a lot more to be said about the content of sovereign debt contracts and various restructuring techniques, but sovereign debt restructuring also touches on other fascinating topics like sovereign immunity, recognition of foreign judgments, and enforcement of judgments. I spent my third year at the University of Hong Kong, and wrote an entire research paper on sovereign debt restructuring in investor-state arbitration. However, despite the never-ending depth to this subject, I think it all makes a lot more sense once you understand the basics. Hopefully this introduction is helpful in that regard.