The Puzzling Pricing of Venezuelan Sovereign Bonds

by Mark Weidemaier & Mitu Gulati

Venezuela’s sovereign bonds differ in ways that should, in theory, be reflected in market prices. For example, depending on the bond, the vote required to modify payment terms through the collective action clause (CACs) varies from 100% (requiring each holder to assent), to 85%, to 75%. Bonds with higher voting thresholds are harder to restructure and one would think prices would reflect this. Two bonds issued by state oil company PDVSA also have legal features that one might expect to have pricing implications. One bond benefits from a pledge of collateral (the PDVSA 2020) and, in consequence, should be priced higher than otherwise-comparable bonds. A second was issued at a particularly large original issue discount (OID); this is a potential legal defect that should lower its price. This is the so-called “Hunger bond” (PDVSA 2022 —see here, here and here for more)).

Although these differences seem like they should matter, reports from the European markets (where the bonds can still be traded) indicate that bid prices for Venezuelan sovereign bonds range from around 13.0 to 13.5 cents on the dollar, while ask prices range from about 14.5 to 15.5. Moreover, prices on the bonds with different voting thresholds are identical. That is, the bonds that cannot be restructured except with each creditor’s assent are trading the same as bonds that allow a creditor majority of 85% or 75% to force restructuring terms on dissenters. But why? Venezuela is in full-fledged default, when legal protections should matter the most.  Shouldn’t these non-US investors (US investors can’t buy, given OFAC sanctions) be offering higher prices for bonds with better terms?

One possibility is that the non-US market lacks specialist holdouts like Elliott or Aurelius, who are willing to hold a recalcitrant sovereign’s feet to the fire. Strong contract terms, after all, don’t matter unless someone is willing to enforce them.

Another possibility is that these differences in voting thresholds don’t matter all that much in this context. To be a real pain in the sovereign’s neck, a creditor needs to do more than hold out. It also needs to activate other bond provisions—such as the acceleration clause—that require a stake of 25% or more. An investor with a stake large enough to accelerate would necessarily be able to block modification, no matter which CAC was in the bond. (For discussion, see here).

A third possibility is that the investors who are willing to litigate, and who therefore value differentials in contract rights, have not entered the market yet. We’ve heard that such investors often enter when the price drops to around 10 cents on the dollar.

Plausible conjectures, we think. But then, why do PDVSA bond prices seem to vary in ways that suggest contract rights matter? As of about a month ago, these bonds were trading in the 12 (bid) to 14 (ask) cents on the dollar range. That is slightly lower than the Venezuelan sovereign bonds, even though both sets of bonds essentially reflect Venezuelan sovereign credit.

When we look to the PDVSA 2020s (the collateral bond) and the PDVSA 2022 (the Hunger Bond)—the two with unique legal attributes—we see variance in pricing. The PDVSA 2020s, which benefit from a pledge of collateral, are priced significantly higher than other bonds (in the 30 to 40 cents range). The Hunger Bonds, with the potential OID problem, are priced lower than the other bonds (in the 9 to 12 cents range). 

So then, why are legal aspects of PDVSA bonds priced, while legal aspects of sovereign bonds are not? Perhaps this lends some credence to the explanation (above) that it is really the threshold for acceleration that matters to sovereign bond pricing. That is, perhaps the variance we see in the CAC term (effectively requiring a stake of 25% or less to hold out) doesn’t matter, since prospective holdouts will need a 25% stake to accelerate. Or maybe the particular attributes of PDVSA’s bonds (collateral protection for the PDVSA 2020s, a heightened risk of repudiation for the Hunger bonds) differ in kind from differences in voting thresholds. 

Anyway, these pricing differences are puzzling and may become more so as prices drop. Rumor is that some trades for Venezuelan sovereign bonds over the past few days have been priced under 10 cents on the dollar. If so, perhaps this signals the arrival of a new investor type—one interested in litigation—and we will soon see some pricing variation.

Finally, the collateralized PDVSA 2020 bond has seen an intriguing price drop. Yes, it remains at a significant premium to other PDVSA bonds. And it is backed by meaningful collateral in the form of a controlling interest in CITGO. Early in the crisis though, it traded in the 80-95 cent range, when other PDVSA and Venezuelan sovereign bonds were in the 30 cent range. Why have its prices dropped so much? One possibility is that these bonds were not properly authorized by the National Assembly in 2016, which might provoke a challenge to the collateral pledge. We plan to examine that possibility in a separate post in the next few days. But we’d love to hear insights from folks who trade and actually determine these prices.

UPDATE - Oct 8, 2019

One reader reminds us that Canadian mining company Crystallex is inching closer to holding an execution sale on PDVSA's interest in PDV-Holding, the parent company of CITGO Holding. As a reminder, PDVSA owns PDV-H, which owns CITGO-H, which owns CITGO. The collateral for the PDVSA 2020s consists of 50.1% of the CITGO-H shares. It is possible that Crystallex will wind up owning PDV-H. This is an interesting point, and conceivably it might help explain the drop in PDVSA 2020 bond prices. We wonder, though: Why would a potential ownership change for PDV-H have a large impact on PDVSA 2020 pricing? If the PDVSA 2020 bondholders foreclose, they will have 50.1% of CITGO-H and a minority partner of some sort (either PDVSA or Crystallex). Why would the partner's identity be such a big deal? 

There is also the matter of whether the Guaido legal team's announcement that it planned to offer essentially the same terms to the holders of all the bonds (except the legally dodgy ones) has caused the prices of the 100%, 85% and 75% bonds to merge.  That is a possibility. But the fact that all of these bonds would be given the same initial offer is standard practice in sovereign restructurings.  It is what happens next that matters. Some holders will reject that offer (assuming they have the votes to protect against a cram down) and a subset of them will move to litigation, hoping that they can induce a better offer. It is what happens then that is key in determining the ultimate recovery.

Comments

[Carlos de Sousa]

Dear professor Weidemaier and Gulati,

Just one clarification. The opposition's restructuring guidelines clearly suggest that PDVSA 2020 bondholders will be given preferential treatment in an eventual restructuring.

“Claims that benefit from a valid first priority security interest in property of the Venezuelan state or its public sector entities may be given separate treatment in the renegotiation.”

All bonds collapsed when JP Morgan cut the weight of Venezuela in the EMBI to 0% this summer. The asset freeze imposed on August 8 also weighed on the PDVSA 2020, as Guaidó's team said they could now default on the bond without losing the collateral. We know that's not true if General License 5 (GL5) is not amended, but it raised doubts among investors. Will OFAC amend GL 5 before October 27 to prevent bondholders from seizing Citgo? GL 5 was issued to prevent Maduro from defaulting without facing consequences, so they should update it. But OFAC has stated that it doesn't want to intervene with the normal operation of the market...

I hope my comments contribute to this incredibly interesting discussion.

Best wishes,
Carlos de Sousa.
Lead Economist at Oxford Economics
Venezuelan

[Roland Pettersson]

Two main factors explaining the 2020 price collapse would be: (i) JP Morgan's decision to drop Venezuelan debt from its emerging markets index and (ii) the uncertainty on the actual stance of the U.S. Executive on shielding CITGO from Veny's creditors.

Regarding Prof. De Sousa's comments on the need of amending GL 5 to protect CITGO from holders of the collateralized 2020 bonds, I can only say that it shouldn't be a decisión taken lightly, and potential market repercussions should be accounted for.

Creditors like to see thorough planning and are usually willing to play along with their distressed counterparties, provided that doing so reflects positively on overall perspectives for a turnaround. Isolated actions precluding them for enforcing their credits and affecting the value of their investments is more often than not met with strong resistance.

At the end of the day this is all about trust. Trust in a plan. Trust in the decision makers. And last but not least: Trust in the decision makers' ability to carry out such plans to the end-line.