Imagine Riding the Ceteris Pari-bus into the Sunset ... in Argentina
Imagine sovereign debt without Argentina -- no Paris Club, no pari passu, no CACs, no SDRM ... even sovereign immunity might look totally different. History teaches that whatever happens in Argentina's imminent bond restructuring (revisiting, reprofiling, rejiggering, revamping --the difference is overblown) is likely to have consequences beyond the long-suffering Republic. The fact that Argentina has an actual government with authority over the economy and some capacity to execute a restructuring (unlike, say, Venezuela) justifies wading into the small print of its bond disclosure--as Mark and Mitu have done. Their able interventions free me to focus on two under-covered points. Methinks that (1) the single-minded focus on voting thresholds is misguided, and that (2) it helps to think of "uniformly applicable" as the latest incarnation of pari passu, which goes to show that inter-creditor equity remains a perennial problem in sovereign debt.
I. Basic Background - Limb Counting
- Argentina has two kinds of foreign bonds -- bonds issued in its 2005 and 2010 restructurings (the Kirchner bonds, or K-bonds) and bonds issued in and since the super-settlement of 2016 (the Macri bonds, or M-bonds).
- K-bonds have "two-limb" aggregated CACs. Any single bond series can be restructured with the consent of 75% of the bondholders, and that any two or more series can be restructured with the consent of 85% of the aggregated pool plus 66 2/3% of each series. Any series that does not clear the 2/3 vote drops out and is free to free ride, although its "yes" votes still count towards the 85%.
- M-bonds have a version of "single-limb" aggregated CACs, also known as ICMA CACs because ICMA played a vital role in designing and diffusing them, leading consultations with market participants before and after the publication of model clauses in 2013-2015 (scrolllll). "Single-limb" is a misnomer, because it describes only one of three voting procedures available to Argentina under ICMA CACs. Under this new single-limb procedure, Argentina can restructure two or more bonds with a single vote of 75% of all polled series. If that vote falls short of 75%, there is no restructuring; if it clears 75%, there are no holdouts. In theory, 100% of a series can object, and still get swept into the restructuring if three-quarters of the pool vote in favor.
- If Argentina does not want to hold a single-limb vote (more on why later), it could still use either of the two more established voting procedures, similar to the K-bonds, but with lower amendment thresholds for aggregation. Any one series of M-bonds can be amended with the consent of 75% of the bondholders, and any two or more series can be amended with 66 2/3 of all and 50% of each series polled.
- In K-bonds and M-bonds alike, the composition and number of voting pools, sub-pools, and series votes is entirely up to Argentina, as is the timing and sequencing of its offers to the creditors: "The Republic may select, in its discretion, any modification method for a reserve matter modification in accordance with the Indenture and to designate which series of debt securities will be included for approval in the aggregate of modifications affecting two or more series of debt securities. Any selection of a modification method or designation of series to be included will be final for the purpose of that vote or consent solicitation."
- M-bond CACs have another interesting feature: they let Argentina pool K-bonds and M-bonds in any aggregated vote of two or more series. Although K-bonds cannot be modified without clearing their respective thresholds (85% all, 66 2/3% each), their votes can count towards either two-limb or single-limb modification of the M-bonds, and help clear those lower amendment thresholds.
2. Pooling - Dive In!
The accepted wisdom seems to be -- because bond prices don't lie -- that higher aggregated voting thresholds in K-bonds would shield their holders from haircuts. All equal, this could be right. It might even be politically preferable for the new government to spare the bonds issued when the same party was in power, and inflict more pain on the opposition. But is ceteris paribus? (Admit it, you miss pari passu!)
- Imagine (1) a K-bond series with the face value of $900 million maturing in ten years, of which 60% is held by domestic banks and favorably disposed institutions, and (2) another with the face value of $100 million maturing in three months, of which 20% is held by a distressed debt fund and friends, determined to avoid the general restructuring. The distressed debt fund has more than enough votes to block a series vote, but not enough to block an aggregated vote, while Argentina starts with more than half the votes it needs.
- If the government felt certain of its prospects in this teeny two-bond wading pool, it could even drag along a $100 million M-bond maturing in a year, where aspiring free-riders hold 35%. The free-riders need more than 50% to force their series out of the deal under the terms of the M-bonds; the big K-bond would supply most of the necessary stock vote under two-limb aggregation, with some of the little M-bond supplying the balance.
- Even if the M-bond were entirely held by the free-riders, Argentina could try to amend it using single-limb aggregation, again leveraging the big K-bond vote. Recall it is certain of 60% of the voting pool, and needs only 15% more.
Put differently, the debtor can use a fragmented debt stock to its advantage by gerrymandering its voter pools. Before deciding whether 85% or 55% or any other number is enough, I would want to know how many series there are and who holds them (or at least where the major concentrations are). Who is in your life boat off the Titanic?
3. Equality, Revisited
As a rule, a government that launches a debt restructuring cannot afford a failed deal--the spillovers and dead weight losses are qualitatively different from anything observed in corporate restructurings. The success imperative is the ultimate safeguard against debtor abuses, and there is some evidence that debtors tend to ask for too little relief, rather than too much (see eg here). Nonetheless, single-limb aggregation combined with flexibility to configure voting pools gives the sovereign substantial bargaining power. Creditors have three layers of protection against sovereign abuse in single-limb aggregated CACs.
(i) The voting threshold is high -- 75% for the stock, same as for a single issue in a traditional series-by-series CAC. If you are trying to restructure a big stock of debt, 75% would take a fair amount of legwork on the part of the sovereign.
(ii) The new disclosure requirements are substantially more robust. In a partial or fragmented restructuring, the sovereign has to tell bondholders before soliciting their vote about its plans to restructure or exempt other debt (including official bilateral debt). In the much-discussed scenario where K-bonds are spared and M-bonds are shorn, M-bond holders must be told before they vote that K-bonds would be spared. Would they accept their fate quietly? Maybe, but probably not on a large scale.
(iii) Most famously, everyone in a single-limb voting pool must be offered "uniformly applicable" restructuring terms, defined as identical exit instruments or an identical menu of exit instruments. In theory, one could design a micro-targeted menu to appeal to different constituents holding different series, so that "uniformly applicable" ends up being variegated. Menu design is an art, and Argentina and others have practiced it in the past, but I suspect that it might be hard to get it just right without ascertaining more precisely who holds the debt and keeping the stock stable for some time. The risk of miscalculating is not trivial -- if the vote fails, no one restructures.
The uniformly applicable condition has been the subject of much creative speculation, some of which is recounted in the recent Weidemaier-Gulati post. The key difference between interpreting uniform applicability and the old pari passu covenant is that literally no one knew what the @#$^ pari passu meant when it came up in Peru, Nicaragua, Congo, and later Argentina, notwithstanding a massive research effort mining 19th century practice and a history of wacky court intervention, which resolved exactly nothing. The pari passu compromise reached in response to Argentina's pari passu saga was a redrafted clause that clarified what it did not mean, and judicial opinions that whittled the application of pari passu down to Voldemort.
"Uniformly applicable" was drafted by the same working group that redrafted pari passu, but this time starting from scratch and against the background of the pari passu interpretation drama (disclosure - I was part of the group, along with my co-authors here). Many of the topics in the chatter about Argentina today were specifically addressed in the drafting and in the outreach by ICMA, IIF, and others in the industry, as well as in public pronouncements by early adopters. For example, here is what ICMA's consultation paper had to say on the subject of using Net Present Value to measure equality:
The current proposal provides for the same offer on the same unit of par value; it should be noted, however, that it does not seek to benchmark or equalise a similar form of NPV loss across the different series of bonds being restructured. Considerable amount of work has been done to develop such a benchmark; a view has, however, been reached that, because of the number of assumptions that would need to be made in such a formula (for example, the discount rate to be applied), the level of complexity involved could lead to challenge where a party is minded to test whether the economics of the instruments being offered comply with the benchmark included. Similarly, such a formula or model would likely be so complicated and unwieldy that it might deter use of the single limb aggregation alternative altogether. The Uniformly Applicable concept, compliance with which can easily be ascertained, should therefore be preferable, such that both the debtor and the holders of affected series of bonds, as well as parties such as trustees or fiscal agents, can readily determine whether a proposal satisfies this important condition. The requirement to offer bonds of equal par value the same terms seeks to be consistent with principles of bankruptcy law, as well as to reflect the fact that in the event of non-payment, bondholders generally have the right to accelerate and demand full payment of the principal (although the Standard Aggregated CACs it should be noted may be used prior to a default). For zero coupon bonds, the accreted par value of their claim would need to be calculated.
Incentive payments and other side deals were also discussed and spelled out in the consultation papers, so that an offer "on the same terms" was intended to mean "the same offer on principal and the same offer on all accrued but unpaid interest (or other relevant
financial features of the bonds)..." -- although side payments are hard to detect in practice, especially when domestic creditors are involved.
Big caveat: Neither the drafting group nor ICMA are parties to Argentina's contracts, so that there is no guarantee that a court considering a challenge to "uniformly applicable" would take the consultation papers as proxy for the parties' intent at the time of contracting. Model clauses drafted by an industry group pose an interesting challenge for interpretation. In a slightly different context involving ISDA, one Professor Gulati wrote that courts should defer to industry meaning for standard terms as they do to legislative history for statutes. Still, recall that standard-form sovereign bond contracts are not in fact standard, and Argentina's contracts have enough idiosyncratic stuff in them that someone could persuade a judge that it meant to vary from ICMA's meaning even when it kept ICMA's words. On balance, my money is on ICMA.
Bottom line -- equality is a beautiful ideal constantly crashing into the messy reality of sovereign debt -- be it pari passu or "uniformly applicable." Sometimes you ride into the sunset, sometimes into the ditch.
- Feeds Categories: