Euro Area Sovereign Bonds: CACs or no-CACs?

Mitu Gulati and Mark Weidemaier

Beginning January 1, 2013, Euro Area authorities required member countries to include “collective action clauses,” or “CACs,” in sovereign bonds with a maturity over one year. CACs are a voting mechanism by which a bondholder supermajority (e.g., 66.67% or 75%) can restructure bond terms in a vote that binds dissenters. Before 2013, the vast majority of sovereign bonds issued by Euro area countries not only lacked CACs; they essentially said nothing about restructuring. For much more on CACs, European and otherwise, see here, here and here.

Because of this policy change in 2013, almost every Euro Area sovereign has two sets of bonds outstanding: CAC bonds and no-CAC bonds. Is either type of bond safer for investors to hold in the event of a restructuring?

Italy just slipped into recession and has a gargantuan debt stock. Lorenzo Codogno, former chief economist at the Italian finance ministry, sees a crisis around the corner. To quote him (via the Guardian):

“All the leading indicators suggest the first quarter of the year will be as bad as the last, and the second quarter will be flat. It’s likely things will pick up from there, but even then, it will mean the economy finishes the year in a weak position”

If Italy or another Euro Area country needs to restructure, the difference between CAC and no-CAC bonds will become important. As an initial matter, it is tempting to think that the no-CAC bonds protect investors from restructuring, because these bonds implicitly give each bondholder the right to veto a restructuring of the bond. To be clear, the no-CAC bonds don’t actually say that investors have this right--the bonds say nothing at all--but this isn’t an unreasonable interpretation. But even if so interpreted, it isn’t clear that the no-CAC bonds really offer more protection. The reason is that almost all of the bonds—CAC and no-CAC—are governed by the local law of the issuing sovereign. And the sovereign can change that law to facilitate a restructuring. (For discussions of the local-law advantage, see here and here.)

The crucial question, then, is whether CAC and no-CAC bonds differ in terms of the protection they offer investors against changes to local law. It seems to us that this question depends on the answer to a number of other, subsidiary questions:

First, when the bond is governed by the issuing sovereign’s law, to what extent do CACs protect investors against adverse changes to this law? We previously wrote about this question when considering the risk of currency redenomination. Our question then was whether CAC bonds, which also require a super-majority vote to change the currency of payment (see the definition of “reserved matter”), protect against redenomination risk. This question may prove important in the event that Italy’s financial situation worsens. To broaden our focus beyond redenomination, perhaps CAC bonds give investors some meaningful protection--though hardly complete immunity--from changes to the governing law.

Second, when a bond includes a CAC, must the issuing government use it? Or can the government restructure via some other method (such as imposing a restructuring by fiat via a change to local law)? If the former, CACs would offer investors more protection. We’d expect CAC bonds to trade at higher prices (lower yields) than no-CAC bonds, because they are safer.

 Third, do no-CAC bonds in fact give each investor the right to veto a restructuring that would affect their bond? As noted above, we think this is a reasonable interpretation, but there is no clear answer as yet. We suspect this is how many investors understand the bonds, and it corresponds to what many Euro Area policy makers thought about the mandate. (For reports on interviews, see here.)

Euro Area authorities have said very little about how the new CACs will work in a restructuring. Perhaps they do not want to give any indication that another restructuring is in the works. (The official line is mostly that there will never be another debt restructuring in the Euro Area and any question that implies otherwise is terribly stupid.)

Several research papers recently examine the CAC vs. no-CAC question in the context of local-law Euro Area bonds. The papers are by a group of researchers at the ESM, (Picarelli, Erce & Jiang - here), a group from the ECB and the Central Bank of France (Steffen, Grund & Schumacher -- here), and by a group of academics (Carletti, Colla, Ongena & Gulati - here). The papers all use somewhat different datasets and empirical strategies. But the bottom line is the same every time. CAC bonds are viewed by the market as less risky.

These studies, however, don’t give a definitive answer as to whether CAC bonds are safer, and they certainly don’t explain why they would be safer. Leaving aside suspicions about whether markets price legal risk efficiently, we are still looking for a coherent legal theory to explain why local-law CAC bonds offer more protection than local-law no-CAC bonds. If there is an Italian debt restructuring, this question will play a central role.

Comments

[Charles]

Question Three gets at why CAC bonds appear safer: investor interpretation. As an investor, you see three possible outcomes for your investment: paid at par, paid below par, and debtor default. One reason a sovereign may be more likely to default than to pay below par is the inability to get investors on board with a restructuring. The investor in possession of a tool that facilitates restructuring may see the probability of being paid below par rise compared to debtor default, and thus purchases CAC bonds accordingly. CACs may not guarantee anything legally, but they raise the expected value.

[Ashley Shan]

The mandatory CAC itself is an indication that there might be a restructuring down the road. CAC-bonds are trading higher probably because the CAC introduces more certainty into the restructuring process (so investors know what would happen). Similarly, local-law CAC bonds might offer more protection in that, compared to local-law non-CAC bonds, at least they *address* the possibility and process of restructuring, whereas the silence re: investor holdout situation in local-law non-CAC bonds potentially expose investors to greater risks of local law change.

[Nick Buchta]

The first thing that I think is important to look at in terms of a potential restructuring is that Italy's growth prospects constrain options for making the debt stock sustainable moving forward. Codogno's comments reflect the view, often proven wrong, that the Italian economy will be able to rebound. Really since 2010-11, projections have relied on decent (if not good) growth and those projections have not been born out. I would worry about the economy's ability to start back up even by Q3 2019.

Beyond that, however, I find the third question to be the most interesting. Although the majority of Italy's debt stock is post-2013, any restructuring (or even, likely, reprofiling) is going to have to come up with some plan to get around those bondholders holding up action. This approach could mean offering some kind of bond swap to sidestep the problem but no matter what it will present a sizable challenge for Italian officials.

[Hailey Klabo]

Italy presents an interesting case for how CAC bonds will work out, given that around 2/3 of Italy's debt is held domestically. This means that the Italian banks and retail investors hold quite a bit of power with respect to what the Italian government wants to do--they have the margins if they choose to stand together, even if it was just the banks, who hold around 37% of Italian debt. Would Italy have to divide and conquer? On the flip side, if the Italian government gets the domestic bondholders on board with a proposed restructuring, foreign bondholders will almost be powerless to stop it (under the two-thirds standard).

Given that Italian law already allows for unilateral restructuring, would Italy need to pass new laws or just rely on Article 3 of the Consolidated Act of the regulations and legislative provisions regarding public debt?

[Federico Espinosa Fernández]

I believe that the First question gives the key issue. Thus, if we assume that an issuer is enabled to modify the terms and conditions contained in the issuance documents through passing new laws (or using another internal method), bonds should become in the most expensive source of financing given the enormous uncertainty that they bring. In other words, how is possible to assure to the investors that in the future, using the local law, the issuer will not convert a CAC bond for a Non-CAC bond? Then, any potential answers to the Second and Third question lose relevance.

[Zach]

I have trouble seeing how restructuring could occur as outlined in Question Two. The ESM model CAC states that a 2/3 majority of bondholders in writing, or a 75% vote, is required for changes to a "reserved matter," as defined in (1)(h). The list of "reserved matters" is very broad - it covers options such as changes to coupon and maturity dates, haircuts, guarantees, seniority, and changes of jurisdiction and immunity.

Clause (1)(h)(vi) states that a 2/3 bondholder majority in writing is required to "impose any condition or otherwise modify the Issuer's obligation to make payments on the Bonds," while (2.5)(a) states that any non-reserved matter may be modified with the Issuing state's consent and "the affirmative vote of holders of more than 50% of the aggregate principal amount" or "a written resolution signed by or on behalf of holders of more than 50% of the aggregate principal amount." This modification requirement also applies to cross-series modifications, under clauses (2.2), (2.3, and (2.4).

Given the high bar for imposing changes of "any agreement governing the issuance or administration of the Bonds," it seems to me like a government's restructuring by fiat would not be a valid modification.

[Charlie S]

It seems that the threat of a government changing their local law to force a restructuring is present in both CAC and no-CAC bonds. In fact, this is what Greece did in order to facilitate a restructuring in 2012, by changing local law to minimize the amount of holdout creditors. In this case, perhaps the "safety" offered by local-law COC bonds is merely a perception resulting from a cognitive bias stemming the ability to affirmatively change a contract term via a vote.

[Natalie P]

I think Charles makes a good point that an investor interpretation matters a lot here. However, my concern is that if CAC bonds do not offer as much protection as investors currently think, investors will quickly wise up and look to alternative protections. For example, with the unilateral changes Greece made, a lot of investors thought Greece was overreaching and were angered by the actions. This makes it much harder for Italy to potentially use this action.

[Rob H]

The Euro authorities' obstinance on this issue could come back to haunt them. Because so much ambiguity permeates the CAC-inclusive Italian bonds — comprising about 75% of the debt stock — it would be far preferable for authorities to go on the record with regard to a future restructuring. As Ashley noted, the very presence of the CACs betrays Area leaders' current attitudes.

Given that so many of Italy's sovereign bondholders are domestic, I could foresee the government arguing that general public opinion operates as an enforceable proxy for CAC provisions and thus overrides them. Because of doom loop concerns, even non-bondholders should have a voice, the argument would go.

And to the extent this regime can manipulate public opinion, as a CAC-inclusive bondholder I wouldn't make too many assumptions about my protection.

[Emma]

To Hailey's point about the importance of the banks here. It does seem like the case that a restructuring of Italian debt would almost require the agreement of the Italian banking industry. I wonder if they would likely participate given how a restructuring of the debt might hurt stability. If there is a restructuring like there was in Greece this might push them towards failure, depending on how much of their capital is in Italia bonds. If the banks, or even some of them were to fail, they would be tossed into the single resolution mechanism for resolving bank failure, which was adopted in 2014 and has not been widely tested.

[Brett T]

I guess my question is, if the gov't can change local law to facilitate whatever form of restructuring they choose (my understanding is that for the most part this is the case), then no-CAC or CAC does not really matter. If Italy decides to restructure via the CAC's then it will change the law for the no-CAC's' to make them CAC's. If it decides bypass the CAC's and restructure via some other method, then it will simply change local law to equalize all the bonds for whatever purpose they need. Unless the fact that the commercial contracts had CAC's in them prevents the gov't from removing or altering the CAC's (which I don't think the gov't would be prevented from doing so), then it doesn't really matter. If the Greece restructuring taught us anything, it's that foreign and international European Courts (FCJ, ICSID, ECtHR, ECJ) are not going to interfere with the sovereign's right to change local law bond contracts to avoid imminent insolvency and disaster. It is possible that some of these foreign or international courts will see the retroactive removal of a CAC as different from the retroactive insertion of a CAC, but I just don't think it's likely.

[Isabelle S.]

I do think it helps with investor expectations to know how the voting mechanisms work to amend the bonds, so the CAC bonds may be “safer” with regards to knowing how things work – but that’s if the sovereign choices to amend those bonds. But for all we know, the sovereign can say “yea, we don’t like those old bonds. You have to exchange your bonds to these new bonds (with terms we like), or you’re not getting paid and you can’t do anything about that because we’re a sovereign,” and perhaps they pull an Argentina and declare by law that holdouts will not be paid (which as a side note may have a better chance of working because if there’s no contract terms in the non-CAC, there’s no pari passu…..), or simply threaten that holdouts will not be paid.

Question 1: There appears to be two conflicting contractual terms: the CACs and governing law as local. On one hand, investors can argue to the court that they expected to have a say via CAC voting on bond amendments, so the enacted law goes against this term. But to be honest, I don’t find much sympathy for this argument: we see in the Gold Clauses and Greece (2012 – i.e. before these CAC bonds) how far we allow sovereigns to go with changing the law to suit their bond needs, so as sophisticated parties, these investors should have made more explicit demands for more protections against a change to governing law. But perhaps I am too used to thinking how a New York court may view this situation and Europe may sympathize with the investors more. But also, I think the answer to this question largely depends on what the law is doing. Like is the law directly going against a contract term? For instance, Argentina enacted a law that went against a contractual term, and SDNY was mad about that.

Question 2: We need to consider whether the bond explicitly say the government must use the CAC? Unlikely/Not to my knowledge. Instead sovereign could do an exchange offer saying “exchange or you’re SOL”. Perhaps if the sovereign wants to employ a holdout deterrent it utilizes the bond’s voting rules through an exit consent. But considering the government could just enact a law saying they won’t pay holdouts or simply threaten that, the exit consents are not necessary.

Question 3: Honestly, learning that a very very very expensive “contract” has no contractual terms is just shocking…. Like billions of dollars can be held up because 1 bondholder with a nominal stake says “no thank you.” Whether this is allowed may depend largely on the jurisdiction a potential case goes to. Like does the court care more about “rightness” or not. For instance, Mamatos Court cared a lot about the ultimate “public interest aim” and allowed an allegedly “coercive” restructuring go through. So in this type of court, I think a court may be more willing to side on the sovereign’s side and say investors should not have the ability to halt an entire restructuring when the domestic/international economy is at stake. Whereas a New York court is probably less prone to care about “rightness,” but instead look at what the contract says and declare “you’re sophisticated and did not contract for it, so oh well” (see Sharon Steel). But here is the problem for non-CAC bonds: nothing was contracted for… On one hand, the investors didn’t contract for this right, but perhaps the investors expectation can play a role here that they thought (and the market concurred) that they had this right. So, the NY Court may be more likely to “punish” the sovereign for being silly enough to enact such a ridiculous bond with no contract terms, and teach them to do better next time.

[Victoria Dodev]

For question one, I do not see any compelling reason why the CACs would prevent Italy from changing its domestic law in a way that affects bond terms, and the EFC, which drafter the model-CACs, agrees with this in its official explanatory note. Take Italy for example, which adopted the model-CACs through decree, found here: http://www.dt.mef.gov.it/modules/documenti_en/debito_pubblico/normativa_spalla_destra/CACs_decree_of_7.12.2012_no._96717.pdf. They indeed include a reserved matter for "change the law governing the Bonds." Firstly, this language is itself ambiguous (does this mean change from the governing law of one sovereign's to another's (arguably the more common interpretation) or changes in the substantive law affecting bond terms). Regardless, note ft 5, which says this reserved matter is "[t]o be included if the Bonds are governed by a foreign law." 98% of Italy's issuances are domestic law governed decree bonds, which would not include this reserved matter.

The EFC's explanatory note adds additional color on the addition of this footnote: https://europa.eu/efc/sites/efc/files/docs/pages/explanatory_note_draft_on_the_model_cac_-_26_july.pdf. "[T]he standardised CAC treats a change in the law governing a bond as a reserved matter only if a bond is governed by a law other than the law of the issuer. The Committee believes there is no need to treat a change from the issuer’s own domestic law as a reserved matter because the issuer already has the power, at least in theory, to adopt any desired modification by means of domestic legislation without changing the law governing its bonds. In so treating a change in governing law, the Committee does not mean to suggest that euro area government issuers will in fact exercise their sovereignty in relation to any series of bonds, or that modifying the terms of a bond by means of domestic legislation does not itself raise difficult legal issues."

Thus, the question one can be answered with "to no extent." Whether bondholders would challenge this change in a domestic or international court is a separate matter, but the CACs themselves to not prohibit an adverse change in domestic law.

For the second question, I would also answer no. The ESM treaty was written with the idea of restructurings should somehow be standardized across eurozone sovereigns. Art 12 of the treaty says that CACs shall be included "in a way that ensures their legal impact is identical," but the treaty fails to ensure identical legal impact in any way.

Bonds which include the model-CACs are overwhelmingly governed by domestic law. Thus, the interpretation and the implementation of the CACs is left up to domestic government, its legal and judicial system. Application and interpretation is not standardized across sovereigns. Further, the model CACs are not even part of the text of the treaty itself, but merely "common terms of reference" (basically, informal standard language that is up on the Commission's website) which are not legally binding. Lastly, nothing in the treaty says "we will uphold these CACs in all events and use them in a restructuring," but merely that CACs will be "included" in issuances starting in 2013. Sovereigns are not prohibited from anything in the treaty from removing CACs conducting a restructuring through other methods. Again, if they do so, they risk being challenged through bondholder litigation.

While I do not think that the CACs present any legal obstacle to a sovereign's power to change domestic law, amend or remove CACs, or conduct a restructuring as it please, they may offer some protection to the extent that it would be controversial for a government to completely disregard them. Caselaw that is a legacy of the Greek saga (see Mamatas) also cautions against a measure that so extremely goes against investor expectations. But, as the ECtHR has shown, desperate times call for desperate measure that are sometime upheld.