Can Argentina Discriminate Against Bonds Issued Under Macri?

Mark Weidemaier and Mitu Gulati

We hope readers will forgive our trafficking in rumors, but this one is interesting and raises some fun and wonky questions about the relationship between Argentina’s different bonds. We talked about those differences in our last post. Basically, bonds issued 2016 or later are easier to restructure than bonds issued in the country’s 2005 and 2010 debt exchanges. This Bloomberg article explains the differences. Interestingly—and here’s the underlying driver of the rumor—the exchange bonds were issued during the presidencies of Cristina Kirchner and Nestor Kirchner, while Mr. Macri was in office when the 2016 and later bonds were issued. The rumor—relayed to us by some of our friends in the investor community—is that the new government has signaled that it might restructure the Macri bonds, or perhaps just default on them, while leaving the Kirchner bonds untouched.

We’re skeptical that the government really intends to do this, for two reasons. First, the plan sounds insane. That’s not exactly proof that the new Kirchner government won’t do it. But maybe some officials just believe that the government can improve its negotiating position if it seems willing to consider crazy stuff. That might not be sound negotiation theory or whatever, but maybe some in the new government take this view.

The second reason for our skepticism is that we’re not sure Argentina’s bond contracts give it a practical way to engage in this type of discrimination. But this question is actually quite complicated and highlights some ambiguities in Argentina’s bonds. Contractual ambiguities are our caviar and champagne, so that’s what we want to talk about here.

Could the government simply default on the Macri bonds while continuing to pay the Kirchner bonds? Sure, but doing so would eventually trigger the cross-default provisions of the Kirchner bonds. Here is a summary of the relevant provisions, which we extract from the 2010 prospectus. The discussion is simplified, but includes the key details:

The bonds define cross default as “any event or condition occurs that results in the acceleration of the maturity … of any of Argentina’s Performing Public External Indebtedness having an aggregate principal amount of U.S.$30,000,000 or more.” The definition also covers any payment default extending past the applicable grace period on bonds with an aggregate principal amount of at least $30 million.

“Performing Public External Indebtedness” includes most foreign currency debt issued after June 2, 2005. The intent here is fairly obvious. It is to prevent the cross-default provisions from being triggered by holdouts who were refusing to participate in the exchanges.

Given these provisions, a default on the Macri bonds would trigger the cross-default provisions of the Kirchner bonds, either once the Macri bondholders accelerated or after the 30-day grace period expired. This would then give holders of the Kirchner bonds the option to accelerate and sue. (Acceleration takes a vote of at least 25% in aggregate principal amount of each series.)

True, many holders of Kirchner bonds might be happy to collect their payments, rather than accelerate and demand full payment. But why would investors, who are already deeply skeptical about Argentina’s future economic prospects, not take the option to accelerate their payment schedule and walk away early? It only takes a subset of investors to get the ball rolling on acceleration. And if even one series of Kircher bonds accelerates, it will be hard for the others to stand pat.

What if Argentina were to restructure the Macri bonds, rather than simply default? Doing so would not seem to trigger the cross-default provisions in the Kirchner bonds. The reason is that a restructuring would not result in the acceleration of the restructured bonds, nor would it involve an uncured payment default on the restructured bonds.

The question, then, is whether Argentina can restructure the Macri bonds while leaving the Kirchner bonds untouched. We’ll assume the government tries to restructure the Macri bonds in a single-limb, aggregated vote. That would require the approval of holders of over 75% in aggregate principal amount of all affected debt but would not require the government to win the approval of each affected series. We make this assumption because holders of Macri bonds will be quite, um, unenthusiastic about receiving such discriminatory treatment. Any voting mechanism that requires a within-series vote will probably fail. Frankly, we’re skeptical that Argentina could get the 75% aggregate approval it would need for this strategy. (See point A about the plan being insane.) But we’ll indulge the possibility that a supermajority of holders of the Macri bonds would vote in favor.

Could the dissenting minority complain about receiving such discriminatory treatment relative to holders of Kirchner bonds? Perhaps. As we mentioned in our prior post, Argentina can conduct an aggregated, single-limb vote only if the restructuring satisfies the “uniformly applicable” standard. The standard requires (with our emphasis) that “holders of debt securities of any series affected by that modification are invited to exchange, convert or substitute their debt securities on the same terms for (x) the same new instruments or other consideration or (y) new instruments or other consideration from an identical menu of instruments or other consideration.” The provision continues by clarifying that the uniformly applicable standard is not satisfied when bondholders are “not offered the same amount of consideration per amount of principal … as that offered to each other exchanging, converting or substituting holder of debt securities of any series affected by that modification.”

On its face, the text seemingly allows a restructuring that does not touch the Kirchner bonds. The reason is that those bonds are not affected by the modification and so therefore seemingly are not included in the group of bonds that must receive uniform treatment. However, there’s also this provision, which we’ll quote at length and then summarize:

For so long as any series of debt securities issued [in the 2005 and 2010 exchanges] …  are outstanding, if the Republic certifies to the trustee and to the trustee under the 2005 indenture that a cross-series modification is being sought simultaneously with a “2005 indenture reserve matter modification”, the 2005 and 2010 debt securities affected by such 2005 indenture reserve matter modification shall be treated as “series affected by that proposed modification” as that phrase is used in the Indenture with respect to both cross-series modifications with single aggregated voting and cross-series modifications with two-tier voting; provided, that if the Republic seeks a cross-series modification with single aggregated voting, in determining whether such modification will be considered uniformly applicable, the holders of any series of 2005 and 2010 debt securities affected by the 2005 indenture reserve matter modification shall be deemed “holders of debt securities of all series affected by that modification,” for the purpose of the uniformly applicable definition. It is the intention that in the circumstances described in respect of any cross-series modification, the votes of the holders of the affected 2005 and 2010 debt securities be counted for purposes of the voting thresholds specified in the Indenture for the applicable cross-series modification as though those 2005 and 2010 debt securities had been affected by that cross-series modification although the effectiveness of any modification, as it relates to the 2005 and 2010 debt securities, shall be governed exclusively by the terms and conditions of those 2005 and 2010 debt securities and by the 2005 indenture; provided, however, that no such modification as to the debt securities will be effective unless such modification shall have also been adopted by the holders of the 2005 and 2010 debt securities pursuant to the amendment and modification provisions of such 2005 and 2010 debt securities.

Here's our summary of this provision, which we confess to find more than a bit confusing:

First, it applies when a proposed restructuring satisfies two conditions: (1) the Macri bonds are to be restructured using any aggregated voting method, whether or not the method requires a within-series vote, and (2) the restructuring also includes the Kirchner (i.e., exchange) bonds, which permit aggregated voting but always require a within-series vote. (Again, our prior post and the Bloomberg article linked above provide more details.)

Second, when a restructuring satisfies these conditions, the bonds are all lumped together for purposes of determining when the aggregated voting threshold has been met.

Third, if the Macri bonds are restructured in a single-limb aggregated vote, they and the Kirchner bonds must receive uniformly applicable treatment. Thus, the government could not, say, impose a 50% principal haircut to the Macri bonds but a 0.0000001% haircut to the Kircher bonds.

We used the last example to set up the obvious question: Can the government simply leave the Kircher bonds out of the restructuring, while benefitting from the single-limb voting provisions of the Macri bonds? The text of the uniformly applicable requirement, even when augmented by the provision quoted above, suggests that it can. This is consistent with the fact that the government is generally allowed to designate a subset of bonds for restructuring; it doesn’t have to choose between restructuring everything and restructuring nothing. 

At the same time, leaving out the Kirchner bonds produces functionally the same result as our last example (in which they were included but received essentially no haircut). And the provision quoted at length above doesn’t explicitly say what happens when the Kirchner bonds are excluded; it applies only when they are included.

Finally, the very last clause of the provision is worth emphasizing: “provided, however, that no such modification as to the debt securities will be effective unless such modification shall have also been adopted by the holders of the 2005 and 2010 debt securities pursuant to the amendment and modification provisions of such 2005 and 2010 debt securities.”

What this means depends on how one interprets “such modification as to the debt securities.” Remember, the provision begins by describing a circumstance in which the government proposes to modify both Macri bonds and Kirchner bonds. Perhaps “such modification” refers to that circumstance. So interpreted, this proviso simply means that, in a restructuring of both types of bond, the Macri bonds won’t be restructured without the Kirchner bonds, even if the aggregate voting threshold is satisfied.  Interpreted this way, the proviso does not address the situation where the government makes no effort to restructure the Kircher bonds. On the other hand, if one interprets “such modification” to refer to any restructuring of the Macri bonds, then this language could be interpreted to insist that all bonds be restructured together.

As a textual matter, this second reading is a bit strained, but it has a certain functional logic. After all, the proviso clearly seems designed to ensure equal treatment for the Kirchner and Macri bonds even if the uniformly applicable requirement is otherwise satisfied. To see the point, consider an identical restructuring proposal offered to all bondholders that wins the approval of over 75% in aggregate principal amount. Also assume that over one-third of holders of each series of Kirchner bonds votes against. This would exempt the Kirchner bonds, which require the approval of 66.67% of each series. In turn, it would also exempt the Macri bonds, even though they contain no requirement of within-series approval, even though 75% in aggregate principal amount voted in favor, and even though the uniformly applicable standard was satisfied. Given this repeated emphasis on ensuring equal treatment of Macri and Kirchner bonds, it seems like less of a stretch to conclude that the bonds must be restructured together.

Bottom line: Although the text supports one interpretation more than the other, the contract language here is not crystal clear. And ambiguity in contract language is a beautifully engraved invitation to the litigation minded. 

Finally, since we are talking about invitations to the litigation minded, how could we forget our old friend the pari passu clause. That topic probably merits a separate post because the law regarding the meaning of the clause has become entirely detached from the text of clause. We’re referring to two recent Second Circuit cases—BisonBee and Lucesco—that seem to have reduced the clause to a vague prohibition against “uniquely recalcitrant” behavior by a sovereign. Conceivably the clause would be interpreted to prohibit utterly irrational—in this case, a better term might be vindictive—differences in the treatment afforded to bondholders. Again, this is probably worth a separate post. But the lesson everyone should have learned from the last go around with litigation against Argentina is to be wary of contract terms that promise equal treatment. Leading experts are convinced that the pari passu theory that Elliott advanced in the NML case is now dead (see here). But, if the rumors we began this post with are right, pari passu might rise again from the grave.

Note that the prospectuses for Argentina’s modern bonds also purport to disclaim any interpretation that requires ratable payment. This language is untested, so it remains to be seen whether it effectively negates a remedy based on a violation of the pari passu clause. If a court decides, in light of these new Second Circuit pronouncements, that pari passu violations are all about bad debtor behavior (e.g., treating investors very unequally, out of spite) and not about the contract language (other than the promise of pari passu treatment), then the game will have begun again.