The Argentine Re-Designation Drama: Notes From Two Frustrated Readers

By Mitu Gulati and Mark Weidemaier

In 2014, after much fanfare, a shiny new set of collective action clauses was released by ICMA (the International Capital Markets Association), with the endorsement of the IMF, the US Treasury, and others. The inspiration for these clauses? The fact that Argentina, after its 2001 default, got taken to the cleaners by hedge funds who found ways to exploit ambiguities (pari passu) and oddities (FRANs) in the terms of its debt contracts. The new ICMA CACs were supposed to protect against the risk of holdouts (by letting a super-majority of bondholders quash minority holdouts) while constraining opportunistic behavior by sovereigns (by limiting the sovereign’s ability to coerce creditors into supporting a restructuring). But for all of the good intentions behind these 2014 ICMA CACs, they are long, complicated, and leave gaps for clever parties to exploit. And Argentina’s 2020 restructuring proposal may just illustrate the problem.

Many creditors are irate about Argentina’s exchange offer, so much so that some of them say they no longer want the 2014 ICMA CACs. We have been struggling to understand why the offer got them so upset. Fortunately, Anna Szymanski of Reuters Breaking Views put out a piece titled “Argentina Gets Cheeky With its Creditors” earlier today that makes the basics of the drama clear (here). Cribbing from Anna’s research, here is how we understand what is going on and why creditors are irate.

For context, the ICMA CACs were supposed to be an improvement from the prior CAC model because they allowed a vote to be aggregated across multiple series of bonds. That is, if 75% or more of the holders in principal amount across all of the bonds having been selected for restructuring agreed, the dissenters would be forced to accept the same terms. The presumption is that, if a super-majority votes in favor, the deal is probably a fair one. The sovereign got to identify which bonds would be pooled together for the vote. But creditors were worried that pooled voting might allow for opportunistic behavior by sovereigns. One protection against this is a requirement that the sovereign designate the voting pools ex ante.

Section 11.3 of the 2016 Argentine Indenture (that draws from the standard ICMA language) states:

The Republic shall have the discretion to . . . designate which Series of Debt Securities will be included in the aggregated voting …; provided however that once the Republic . . . designates the Series of debt securities that [will be subject to the aggregated vote] those elections will be final for purposes of that vote or consent solicitation.

So, in the aggregated vote, Argentina must get the approval of holders of 75% of the principal amount. If it doesn’t the consent solicitation fails.

Argentina, however, seems to think that it retains the ability to “re-designate” some of the Series that voted against the restructuring to pretend that they hadn’t been included in the vote in the first place. To borrow Anna’s analogy, it’s like “a president holding elections and then afterward invaliding certain votes to make sure he wins.” (We can only imagine what might have prompted the analogy…)

Let’s do some rudimentary math. Assume there are 4 series of bonds A, B, C, and D, each with an outstanding principal of 100. The government pools them all together for a restructuring vote. It needs just over 75% approval, meaning holders of over 300 in principal amount. Alas, while just over 75% of each of series A and B vote in favor, only 60% of each of series C and D votes yes. Oops. No restructuring (75+75+60+60=270; 270/400 < 75%).  

But wait! Let’s re-designate the voting pool to include just series A and B. Now, we have just over 75% support in our re-designated pool and have successfully restructured those two series of bonds (into, say, series A* and B*). Series C and D escape the restructuring.

Here’s where the fun starts. Holders of series C and D are the proud owners of unrestructured bonds that the government has no intention of paying. Can they sue to enforce their claims? At least for the moment, yes. But let us imagine what might happen in stage 2.

Here, Argentina launches a second restructuring offer. Again, it proposes to aggregate all four series of bonds. It must offer the same terms to everyone in the voting pool. This is required by the “Uniformly Applicable” condition. But since series A* and B* have crappy terms—they were just restructured!—this is no problem. It proposes a restructuring on terms that are ever-so-slightly better than the ones in series A* and B*. Now, it expects to win the support of even more than 75% of those two series. For them, it’s free money. So let’s say Argentina gets 100% support from those investors. It is 2/3 of the way towards restructuring the entire pool. Whereas in stage 1, the 40% holdout rate in series C and D was more than enough to tank the deal, now the holdout rate will have to be 50% to keep the vote from exceeding the 75% threshold. Depending on how the numbers turn out, we can imagine additional rounds of re-designation and voting.

[Edit: A reader asks whether our math assumes the restructuring in round 1 involves only a reduction to coupon, rather than a principal reduction. It does, but this isn't necessarily required. It is true that, if the principal amount of the restructured bonds is reduced, their vote will contribute less to the favorable round 2 restructuring vote (which, again, requires the approval of 75% in aggregate principal amount of outstanding bonds). But even if there is a reduction to principal, the math works out as long as the increased support among holders of series A* and B* in the round 2 vote is large enough to compensate for the reduced principal of these bonds.]

It’s quite clever. Though it seems contrary to the purpose of requiring the government to designate the voting pool ex ante. We assume the purpose was to remove the incentive to make a low-ball restructuring offer. But if we understand Argentina’s argument correctly—and we confess we may not—that’s precisely what “re-designation” allows it to do. Of course, it also might be an invitation to litigation. (As Rodrigo Olivares-Caminal points out in his post on this matter, that litigation is likely to play out differently in England than New York, see here).

If there is a lawsuit, it might be that people involved in creating the 2014 ICMA CAC can shed light on the intent. We already have some clues from a blog post by Mark Sobel, the senior US Treasury official who is generally seen as the organizing force behind the 2014 ICMA CACs (here). In any event, this episode is perhaps an illustration of how hard it is—really, impossible—to draft bulletproof contracts.  

Comments

[Lucas Caldi]

Would you believe it is possible for Argentina to go along with the following strateg:

Following your expample, A and B bonds get restructured while C and D holdout.

Argentina files for a second restructuring attempt, but designates de pool of bonds to be restructured to include just A B and C. Therefore, the bloking power of C bondholders gets diluted even more.

Then it goes along to include a final restructuring proposal which now it includes the A B C and D bonds. Since C are already restructured and have lost their claims for the old bond terms, they would have incentive to approve the last restructured bonds (they me have improved terms and higher EY if Argentina wipes out holdouts). If A B C bondholders approve 100% the last restructuring terms, then D bonholder will inevitably get restructured too.

Thank you for you insight
Lucas