CDS fight of the day, Seat edition
A couple of weeks ago, Floyd Norris had a column about an obscure European CDS dispute which — just for a change — didn’t involve sovereign debt at all. Instead, it involved Italian yellow-pages publisher Seat Pagine Gialle. Here’s how Norris saw the dispute:
Under the association’s rules, in some cases there is no event if investors “voluntarily” agree to exchanges that in reality cost them money…
Seat has tried to expand its Internet business. But it reported a loss of 33.2 million euros for the first nine months of this year, and on Oct. 28 it said it would delay an interest payment of 52 million euros, or about $69 million, for a month.
Last week, it said it had reached a tentative agreement with a majority of creditors, but that disputes remained with its senior debtholders over how much equity would go to the holders of 1.3 billion euros in bonds. It said that if a final deal were reached and accepted by bondholders, it would make the interest payment by Wednesday…
If the tentative deal falls apart and the interest payment is missed, there would be no doubt that a credit event had taken place. But since Seat Pagine Gialle is trying to get a voluntary agreement for a swap of the bonds for stock, it may be possible that there would be no credit event at all, even though it will be clear that bondholders have suffered a major loss.
In the end, the tentative deal did fall apart, the interest payment was missed, and an event of default was declared. Crisis averted. But Floyd was right that there would have been a big problem if the bond payment ended up being made. He’s just wrong about why there would have been a problem.
Christopher Whittall has an explanation of the real story here — which is actually more worrying than Norris implied. The problem of voluntary exchanges is a bit of a red herring: if Seat bonds were swapped for stock, that would be a credit event, and the CDS would get triggered.
In fact, you see, there weren’t any Seat bonds at all. For obscure reasons, Seat wasn’t actually the bond issuer in this case — the bonds were issued by a Luxembourg-based special-purpose vehicle called Lighthouse. Lighthouse was at heart a passthrough vehicle: it took the proceeds from the bond issuance, and lent them to Seat. And in return Seat made payments on the loan from Lighthouse, which in turn were passed on to bondholders. If Seat were to default, then it would stop paying Lighthouse, and Lighthouse would stop paying bondholders — so it was always perfectly clear that bondholders were taking Seat credit risk when they bought their bonds.
Here’s where things get a bit complicated. Bonds generally come with 30-day grace periods: if you miss a coupon payment, but then make it up within 30 days, then there’s no event of default. And the Lighthouse bonds were no exception — they came with the standard 30-day grace period. If the bonds had been issued by Seat directly, there would be no event of default until 30 days after the coupon payment was missed.
But the bonds weren’t issued by Seat directly, they were issued by Lighthouse. The Seat default was not on the bond payments to the Lighthouse bondholders, but rather on the loan to Lighthouse itself. And it turns out that the loan agreement between Seat and Lighthouse did not have a 30-day grace period: instead, it just had a 3-day grace period.
Somehow, a few holders of Seat CDS managed to get their hands on the loan agreement between Seat and Lighthouse, and used that to say that there was an event of default long before the 30-day grace period was up. In fact, they said, the Seat CDS should have been triggered three days after the company missed its loan payment to Lighthouse. And so even if Seat made that payment in full within the 30-day grace period on the bonds, the CDS should still be triggered and pay out.
This argument did not go down particularly easily on the Determination Committee:
The DC reached a deadlock of eight votes to seven, and the decision was sent for external review…
“Miraculously, some loan documentation that had not been made public appeared and people tried to force ISDA to make a quick decision [that there was a credit event]. If that is not market manipulation then I don’t know what is,” said a senior trader with DC representation, whose firm was initially against the declaration of a credit event.
Eight members of the committee voted that there had indeed been a credit event — Barclays, Credit Suisse, Deutsche Bank, Morgan Stanley, UBS, Citadel, and Pimco. Seven dealers said that there hadn’t: Goldman Sachs, JPMorgan Chase, BNP Paribas, Société Générale, D.E. Shaw, BlackRock, and BlueMountain Capital. Clearly, this was no cut-and-dried case. And so the right decision was reached: to send the case out for a proper external review. A hurried decision to declare a credit event on the basis of a loan default to a single special purpose entity, even when no credit event could yet be declared on outstanding public debt, would have been a decidedly rash thing to do.
The thing to wonder about is what would have happened if, miraculously, Seat had started making its payments again in full, and the CDS question had gone to external review. The external reviewers would then have had to answer: was there an event of default, even though all Seat’s bondholders received all the money they were due within the grace period they had agreed to? Floyd Norris was worried about a state of affairs where bondholders would suffer a loss and the CDS would not pay out; in fact, the state to be worried about was one where bondholders got their coupon payments and yet the CDS was triggered all the same. After all, even if that one coupon payment was made, the bonds would still be worth much less than par, which means that triggering the CDS would cause a lot of money to flow to holders of that protection.
And the bigger worry still, here, is that the ISDA determinations committee was filled with banks who had a dog in the fight, rather than making their decision dispassionately.
The derivatives trade body admitted it was in the process of bolstering standards to ensure dealers cannot “vote their books”…
“The SEAT vote didn’t seem very independent in the way people were voting on legal merit versus their position,” said one senior credit trader, whose firm sits on the DC. “How to enforce independent decision-making may not have been carefully thought out when the DC was set up. It’s a concern that won’t go away and will eventually come up in the context of Greece.”
If banks like BNP Paribas and SocGen are voting their books over a relatively small issuer like Seat, it stands to reason that they would fight quite hard for their books if and when it comes time to decide whether there has been a credit event in Greece.
Sometimes, it’s not obvious whether CDS should be triggered or not. And in such situations, you really don’t want banks with a dog in the fight making those decisions. That’s the real lesson of the Seat fight. And I’m far from reassured that the decisions in Greece are going to be made dispassionately and objectively.