Bloomberg reported today that, back in July, David Einhorn and some other people decided that (1) betting against European sovereign debt was, and would remain, a good idea, but (2) doing it in CDS form was kind of dumb, so (3) they'd switch to doing it in physical form, by borrowing and shorting the debt. Here's what Einhorn had to say in his July investor letter:
The letter touched on two risks tied to credit swaps on European sovereign debt, including regulators’ attempts to fashion a Greek bailout in a way that prevented the contracts from paying out. The second risk was the possibility that banks that wrote billions of dollars in credit swaps on sovereign debt might not be able to make good on their obligations should a country such as Greece actually default.
Let's talk about that first reason for a minute because I think it's sort of illuminating. The problem is that Europe was in July, and is now, and wow that's depressing, trying to cobble together a "voluntary" debt exchange where holders of Greek debt happily hand it in to Greece and get back a thing with a 50% face value haircut that is also a piece of crap. If you're a European bank who owns Greek bonds and CDS to hedge them, and you feel pressured to accept that deal, then you feel like the "insurance" you bought on your bonds should "pay out," I suppose, though that's all fairly hypothetical. If on the other hand you're David Einhorn and you bought CDS and then Greece haircuts its debt, you feel like your bet against Greek debt has been vindicated so it should pay out. But it doesn't, says ISDA, because the exchange was voluntary and there was no "credit event" under the rules governing your CDS.
Now you can feel some feelings about this - you can think it means the end of the CDS market, or that it's no big deal; you can think that ISDA's determination that this "voluntary" exchange is not a credit event is clearly right, or that it's an evil conspiracy; and of course you can have a range of views about the likelihood of the voluntary exchange happening and about whether, after the voluntary exchange, any un-exchanged bonds will be paid at par or forcibly defaulted.
But you can avoid all that and just short the bonds, which seems to be what Einhorn did. Here Einhorn goes to his prime broker, borrows some Greek bonds, and sells them. In six months, when the bad thing has happened, he buys them back for cheap and returns them to the dealer. That is all straightforward. What could be easier.
Well, you might wonder what happens if, I don't know, the bonds that he borrows are transformed into something else. This might be relevant to you if you are borrowing Greek bonds, which everyone expects to be transformed into something else. (Specifically smaller, crappier Greek bonds.) Securities lending agreements typically provide for this in the expected way - you don't return the securities you borrow, you return "equivalent securities," and
If and to the extent that such Loaned Securities ... consists of Securities that are partly paid or have been converted, subdivided, consolidated, made the subject of a takeover, rights of pre-emption, rights to receive securities or a certificate which may at a future date be exchanged for Securities, the expression shall include such Securities or other assets to which Lender or Borrower (as the case may be) is entitled following the occurrence of the relevant event. ... In the event that such Loaned Securities ... have been redeemed, are partly paid, are the subject of a capitalisation issue or are subject to an event similar to any of the foregoing events described in this paragraph, [long-winded but entirely reasonable things happen].
So, fine, if Greek bonds transform into other Greek bonds, Einhorn returns the other Greek bonds.
But ... they don't. Under the voluntary regime anyway. There, nothing has transformed into anything - some people have just agreed to do a trade. Einhorn's obligation remains what it was - to deliver back the Greek bonds he borrowed, not the ones that those bonds could have been exchanged for.* It is hard to imagine him saying to a securities lender "oh, hey, yeah, I exchanged your bonds for you, hope you like these smaller crappier bonds." He just doesn't get to do that with the lender's bonds. And if he tried - well, I think his prime broker would be a lot more conflicted than the ISDA determinations committee.
You can if you like imagine how this impacts the trade. Einhorn borrows bonds and shorts them. Greece deteriorates and voluntary plan gets uglier, so they drop in value. He has a mark-to-market gain. Voluntary plan is executed. Some bonds are gone. Others remain. When Einhorn wants to close out the position, he has to go buy in those remaining bonds. Their price will depend on whether people now expect them to hard default, or to be paid off, plus I guess on their post-exchange scarcity and how crowded this trade was. If a hard default is likely, or occurs, the short seller makes lots of money. If the stub of bonds not voluntarily exchanged are expected to be paid at par, the short seller loses money.
These are exactly the cash flows you'd get from buying Greek CDS.
None of this happens of course - David Einhorn doesn't enter into a term borrow of bonds, short them, forget all about them, and return a year later to scramble to buy back those bonds. An actual short seller would buy in the bonds and close out his short just before the voluntary exchange, when they're cheap, and even if he didn't want to his borrow would likely be called in by his prime broker who is, after all, a big bank who feels pressured to participate in the voluntary exchange.
But everything is just the present value of the thing it will eventually be, so this should cast a shadow over current trading levels - if post-exchange bonds are gonna do great, then the bonds should trade above their exchange value. (If not, not.) And the equivalency is worth keeping in mind when you see people freaking out about how broken CDS is. (Sometimes. None of this has anything to do with Einhorn's other reason for moving to physical, which is the contagion risk of the banks who write the CDS maybe having written too much of it. That is a whole nother kettle of fish.) CDS levels should more or less track bond yields because they should more or less replicate the experience of short selling bonds - and it's harder to disconnect the two than you might at first think.
* Technically section 6.7 of the standard securities lending agreement provides that on corporate events that require an election, the lender should be able to make the election - so if this falls under that category the prime broker could tell Einhorn whether they want the original bonds or exchange bonds back. It's not at all clear that it does fall under that category, and in any case you'd expect them to elect the original bonds.