Sovereign CDS – Hellenic Republic CDS
Reports came out on Friday that Greece was now negotiating directly with creditors regarding some form of restructuring. We believe that this increases the risk of a Credit Event occurring, but ultimately it will still have to be a Failure to Pay Credit Event because any agreement with the bondholders would still be considered “voluntary” as bondholders have the legal right to say “No” and would not be “bound” by any other bondholders’ decision.
If we are going to look at CDS contracts then two sets of documents are important: the ISDA Credit Derivative Definitions (and Supplements) and the indentures under which the relevant bonds were issued. We can talk about haircuts, defaults, restructurings, voluntary, what is fair, etc., but the documentation has to be the guide as it is ultimately what any court of law (and ISDA Credit Event Determination Committee) would base their decision on. We will attempt to be clear here and ensure that any defined term from the CDS documentation is capitalized.
Would a 50% change in notional be a restructuring or a default? Certainly! But would it be a Restructuring Event, or a Credit Event? That is the question that needs to be answered!
What would constitute a Credit Event? It is worth noting that it is not called a “Default Event” and that there is no such term as a “Default Event”, there is only “Credit Event”. CDS should really have been called Credit Event Swaps since final payment is based solely on a Credit Event being triggered. This does not have any impact on what the rating agencies call it, or the ECB calls it, or how banks have to treat the debt, this is solely an analysis of the CDS contracts. Any court of law would also take into account the “intention” of the document. So as we look at the specific language we will try to examine the intent. This would play a secondary role in any decision but it is important nonetheless.
Over the years, we were very involved in the development of the documentation and in the development of the CDS products, and have traded close to a trillion of CDS in our careers. Depending on your view of CDS that may be good or bad, but it does qualify us to try and walk you through the details of the product. The devil is in the details, and the situation in Europe as a whole, and Greece in particular, takes us into the bowels of the documentation. In the U.S., the standard for corporate CDS is NoR, which means that the only Credit Events are “Bankruptcy” and “Failure to Pay”. Restructuring was dropped because it inevitably led to confusion and legal wrangling. European corporate CDS still trades with MMR (modified modified restructuring) which limits the Deliverables if the Credit Event is triggered by the Buyer of protection. Sovereign CDS still trades with old R (restructuring), because the reality is that sovereigns tend to restructure more than they actually fail to pay.
With Greece now directly involved in the negotiations, the risk that they trigger a Repudiation/Moratorium Event has increased. With some concern about whether Greece will receive the next tranche of IMF money, there is a potential that they have a Failure to Pay on the bonds maturing on December 19, 2011, which would take us into the Grace Period Extension section of the document. Greece being directly involved in the negotiations does NOT increase the likelihood of a Restructuring Credit Event. We believe that it will be even more difficult to reach agreement among the banks and that increases the probability of a Failure to Pay Credit Event. This is confusing to even some people who trade the product (which is one thing that is wrong with Wall Street) so bear with us as we try and walk through how this all works. If you are bored or confused or annoyed at this stage, then there is no point in reading further, but if you want to understand how the CDS is going to work, then you need to suffer through the rest.
So although Greece being directly involved in the negotiations does NOT increase the risk of a Restructuring Event, we believe that the risks of a Credit Event have increased as it will be far more difficult for Greece to achieve a high participation rate without forcing the issue by actually defaulting on payments that are due. Greece’s direct involvement also brings some maturity extension provisions into play, as they would be triggered as Greece potentially tries some hardball strategies with the bondholders.
A Restructuring Event Remains Highly Unlikely For Greece
From the ISDA documentation (copied here without their express written consent)
Section 4.7. Restructuring.
(a) “Restructuring” means that, with respect to one or more Obligations and in relation to an aggregate amount of not less than the Default Requirement, any one or more of the following events occurs in a form that binds all holders of such Obligation, is agreed between the Reference Entity or a Governmental Authority and a sufficient number of holders of such Obligation to bind all holders of the Obligation or is announced (or otherwise decreed) by a Reference Entity or a Governmental Authority in a form that binds all holders of such Obligation, and such event is not expressly provided for under the terms of such Obligation in effect as of the later of (i) the Credit Event Backstop Date and (ii) the date as of which such Obligation is issued or incurred:
(i) a reduction in the rate or amount of interest payable or the amount of scheduled interest accruals;
(ii) a reduction in the amount of principal or premium payable at maturity or at scheduled redemption dates;
(iii) a postponement or other deferral of a date or dates for either (A) the payment or accrual of interest or (B) the payment of principal or premium;
(iv) a change in the ranking in priority of payment of any Obligation, causing the Subordination of such Obligation to any other Obligation; or
(v) any change in the currency or composition of any payment of interest or principal to any currency which is not a Permitted Currency.
It is quite clear that the negotiations with Greece and the IIF of bondholders would likely satisfy at least clauses ii and iii, while possibly also satisfying clause i. This is the easy part of the analysis. The tricky part of the definition is the first part of the paragraph. Notice that nowhere is “voluntary” used. In fact, nowhere in the entire 106 pages of CDS definitions is the word “voluntary” used.
The basic intention is to ensure that a basis package should be protected. You should be able to hold a bond and buy CDS to the bond maturity and over the life of the trade, either the bond will mature and the CDS will expire, or that the bond will experience a default and the CDS would trigger (this will come up again in the Grace Period Extension analysis). These documents were drafted with corporate credit in mind. Some corporate bonds have voting provisions that if X% of bondholders agree to something with the issuer, all bondholders would be “bound” to that agreement. This “cramdown” is the scenario that the language is trying to pick up. Say there was a provision that said if 90% of holders agreed to a maturity extension the bond would have its maturity extended. If 95% voted in agreement, then 100% of the bondholders would have to accept the new terms. That would be a Restructuring Credit Event because 5% voted no, but still had the terms of their bonds changed. The idea was to protect a minority of investors from having bond terms changed against their will without being able to trigger CDS.
With that as the intention, you can make more sense of the terms used.
“occurs in a form that binds all holders of such Obligation, is agreed between the Reference Entity or a Governmental Authority and a sufficient number of holders of such Obligation to bind all holders of the Obligation or …”
The language talks about binding all holders and that a sufficient number agrees such that all holders (even those that didn’t agree) would be bound by the agreement. That is consistent with what we believe the intention was, but it is a bit sloppy to the extent that 100% of bondholders agree (as discussed below).
You can also see that it doesn’t make a difference whether it is a third party or the government negotiating with the bond holders. So long as no bondholder can be bound by another bondholder’s decision, it would NOT be a credit event.
We have only reviewed a couple of the Hellenic Republic Bond offering circulars. We did not see any evidence of a “cramdown” provision, so basically any debt that doesn’t agree to a restructuring should remain outstanding with its original terms. This is the key, each bondholder has a legal right to say “No” to a restructuring, and if they do say “No”, they cannot be legally bound to accept a restructuring. It doesn’t talk about “voluntary” or “coerced” it is all about being bound by the terms of the bond.
A Repudiation/Moratorium Event Is More Likely To Be Triggered
Well, if you have made it this far without shaking your head and mumbling something about what a ‘bizarre’ product CDS is, this should put you over the top.
Section 4.6. Repudiation/Moratorium.
(a) “Repudiation/Moratorium” means the occurrence of both of the following events: (i) an authorized officer of a Reference Entity or a Governmental Authority (x) disaffirms, disclaims, repudiates or rejects, in whole or in part, or challenges the validity of, one or more Obligations in an aggregate amount of not less than the Default Requirement or (y) declares or imposes a moratorium, standstill, rollover or deferral, whether de facto or de jure, with respect to one or more Obligations in an aggregate amount of not less than the Default Requirement and (ii) a Failure to Pay, determined without regard to the Payment Requirement, or a Restructuring, determined without regard to the Default Requirement, with respect to any such Obligation occurs on or prior to the
Repudiation/Moratorium Evaluation Date.
….
(c) Potential Repudiation/Moratorium. “Potential Repudiation/Moratorium” means the occurrence of an event described in clause (i) of the definition of Repudiation/Moratorium.
(d) Repudiation/Moratorium Extension Condition. The “Repudiation/Moratorium Extension Condition” is satisfied (i) if ISDA publicly …
So Greece needs to be very careful about saying anything about not being willing to pay all of their debt. Saying that they will not pay their debt is not enough to trigger an actual Credit Event, because it has to be followed by a Failure to Pay or Restructuring (though the conditions are slightly easier to satisfy if the Potential Repudiation/Moratorium has occurred). The messy part is that the Potential Repudiation/Moratorium event affects the scheduled Termination Date of the CDS contracts. The details are quite complicated but the main impact, as far as we can tell, is that the CDS contracts are extended for at least 60 days. Our interpretation is that if Greece creates a Potential Repudiation/Moratorium Event, then CDS contracts would remain outstanding past their originally Scheduled Termination Date. So if Greece repudiated their debt, they would enter the Potential Repudiation/Moratorium Event phase. No CDS contract would expire during that time period. It is the later of 60 days, or when payments are due under one of the repudiated bonds. If by that time there is a Failure to Pay, there would be a Credit Event.
In our opinion, the CDS curve should be even flatter (on a price basis) than it already is, since the risk of some official threatening bondholders and triggering a Repudiation/Moratorium Extension Provision has increased dramatically. So far Greece is saying all the right things and not triggering, but as negotiations heat up, it will get harder and harder for some official with the authority not to say something that is strong enough to trigger the Repudiation/Moratorium section.
We are less confident on our analysis in this section, as it is extremely confusing and could have been documented much better, but it is worth reviewing since direct Greek involvement does make it more likely that the provisions of this section get triggered.
December 19, 2011 Bonds and Grace Periods
The December 19th bond is particularly interesting. It is not yet clear that Greece will have the IMF money, and we do not believe that they will have a deal in place with bondholders by this date. That means there is a chance that Greece will not pay that bond on the 19th. If they have the money from the IMF, they may still choose to withhold payment as a negotiating ploy against the banks. Assuming for a moment that Greece does not make that payment, is that a Credit Event? The answer is “not immediately”.
Greece could skip paying principal and interest on the 19th of December on the bond that is maturing and it wouldn’t be a Credit Event? Correct, but it is more complex than that.
We could not find the Offering Circular for that particular bond, but we did find one for another Hellenic Republic bond issued under Greek Law. It has a “cure” period of 7 days for principal payment and 30 days for interest. Assuming the December 19th bond has similar provisions, non-payment on this bond would commence a “grace” period or “cure” period. If during that period, the bondholders received the payments due to them, there would be NO Credit Event (nor would it be an Event of Default under the bond). So that at least make sense. If the bond itself has a cure period, then so long as payment is made within the prescribed contractual timeframe, it is NOT a Credit Event because the issuer had that right.
This would not be a big deal if it wasn’t so close to the CDS Scheduled Termination Date. A lot of CDS will be expiring on December 20th as it is a standard maturity date. So what happens to the CDS that expires on the 20th if the bond is not paid on the 19th but then is paid on the 23rd for example? It depends on whether Grace Period Extension is applicable or not. Our understanding is that the standard for sovereign CDS is Grace Period Extension is applicable (we are trying to get additional confirmation of this since it is potentially critical).
Section 1.11. Grace Period Extension Date. “Grace Period Extension Date” means, if
(a) Grace Period Extension is specified as applicable in the related Confirmation and (b) a Potential Failure to Pay occurs on or prior to the Scheduled Termination Date (determined by reference to Greenwich Mean Time (…), the date that is the number of days in the Grace Period after the date of such Potential Failure to Pay. If Grace Period Extension is not specified as applicable in the related Confirmation, Grace Period Extension shall not apply to the relevant Credit Derivative Transaction. If (i) Grace Period Extension is specified as applicable in the related Confirmation, (ii) a Potential Failure to Pay occurs on or prior to the Scheduled Termination Date (determined by reference to Greenwich Mean Time (…)) and (iii) an Event Determination Date in respect of that Failure to Pay does not occur on or prior to the last day of the Notice Delivery Period (including prior to the Trade Date), the later of the Scheduled Termination Date and the Grace Period Extension Date will be the Termination Date (even if a Failure to Pay occurs after the Scheduled Termination Date).
Under this provision, we believe that the CDS would “extend” past the originally Scheduled Termination Date under this assumed case and follow through until the bond is either “cured” or has a Failure to Pay. So if the bond didn’t pay on the 19th, the CDS would remain outstanding (even past its originally Scheduled Termination Date) to determine whether the payment was cured within the Grace Period specified (which would follow the bond documentation but there are provisions for a minimum of 3 days even if the bond itself doesn’t have a cure period). If on the 23rd the bond paid as in our prior example, there would be NO Credit Event and the CDS would terminate.
If, after the cure period of the bond, or the minimum allowed by ISDA, the bond is still not paid, then a Credit Event would be triggered. It would be a Failure to Pay Credit Event and would occur after the Schedule Termination Date. Again, the documentation is not light reading, but that is how we interpret the document and it is consistent with trying to maintain the “sanctity” of the basis package. Personally we have always preferred owning CDS longer than the bonds with a margin for error, but it seems that if Grace Period is applicable (as does seem to be the standard) the CDS would extend past its originally Scheduled Termination Date to determine if a Potential Failure to Pay turned into a Failure to Pay Credit Event.
Could A 100% Agreement Rate On Any One Bond Be Considered “Bound” And Trigger A Credit Event?
The document is a bit sloppy and there may be a loophole. What happens if 100% of the bondholders agree to extend the maturity of a specific bond? We believe the intention of the document was that this would NOT be a Credit Event, but it might well be in someone’s interest to challenge that interpretation. If every holder voted to make a change, would that then bind all holders to the change? If you agreed to something, you are bound to it, so someone might argue that 100% of people agreeing to something meets the requirement that (after the agreement) sufficient number agreed such that all bondholders are subject to the terms. It seems a little bit of a stretch, against the intention of the document, and creates a strange scenario where 99.9% of the bondholders could accept amended terms without causing a Credit Event, but 100% would cause trigger a Credit Event. Those arguing against it being a Credit Event when 100% of people agree would fight about what ‘bind’ means and what ‘sufficient number … to bind’ means. They would argue that it was meant to pick up only cases where some bondholders did not agree to a change but were subjected to it regardless.
Maybe the document should have expressly stated that a 100% vote was not a trigger of a Credit Event. You might wonder why this language is as confusing as it is. Part of the reason is that a restructuring is a bit like pornography, it is hard to define but you know it when you see it. That might be good enough for Supreme Court opinion, but for a financial contract with a stated maturity it really should have been defined better. The other reason is that the U.S. market in particular was already moving to a modified restructuring world (where the consequences of a Restructuring Credit Event were muted by maturity limitations placed on Deliverable Obligations) and has ultimately moved to a No Restructuring world with the implementation of the SNAC protocol. Full Restructuring actually applies to a very limited universe of outstanding CDS trades, but sovereign debt, and Greece in particular, is one such area. With so few people having a vested interest in Restructuring, it is easy to see how it has evolved less over time than other areas of credit derivative trading.
DISCLAIMER
In no way is this article meant to provide any legal advice. This is our interpretation of the available information, our reading of the rules and standard documentation, and our understanding of current market standards. It should not be construed as legal advice to be applied to any specific factual situation. If you require (or your are unsure whether your particular circumstances require) legal advice, you should consult your own lawyer for legal advice.