As Greece prepares to default on its new bonds, now seems as good a time as ever to fix the problems that occurred when it defaulted on its old bonds. Remember that? Basically there was this thing where if you had a Greek bond with a face amount of €100 and CDS on that Greek bond, and that Greek bond got poofed into a new Greek bond with a face value of €20 that traded at par, then your CDS would pay out not the expected €80 that you lost on your first bond but rather €0 because the second bond was deliverable into CDS and it traded at par. Which makes no sense if you view CDS as hedging your losses on the first bond, which to a reasonable approximation you do.
Fortunately, though, in the particular case of Greece, the new bonds were split into lots of little tranches and one of them basically looked like the old bonds, value-wise (though not otherwise), and so everything worked out and actually made CDS buyers a little bit of extra money. So that was nice for them, but otherwise it was all just terrible.
So this gets a yay:
Credit-default swaps market leaders will meet this week in New York and London to discuss changes to the contracts in what may be the biggest revisions since 2009.
The International Swaps and Derivatives Association’s credit steering committee will meet May 11 to discuss changes, said Steven Kennedy, an ISDA spokesman. Possible amendments to standard contracts, which are governed by ISDA, include how debt-for-equity exchanges would be treated after a bankruptcy, specifying that credit swaps only cover losses from defaults that occur after their purchase, and clarifying how the date of a so-called credit event is determined, according to people familiar with the situation.
The solution to the Greek-exchange problem for Portugal or whoever is actually pretty simple: just revise the ISDA definitions to make clear that when it refers to Obligations it means (1) the Obligations or (2) after the Obligations are involuntarily exchanged, whatever they are involuntarily exchanged for. That is the idea proposed by these Stanford folks but don’t fall over yourself in racing to congratulate them for their ingenuity because it’s kind of an obvious idea. I mean, even I had it.
Don’t congratulate me either, because I didn’t even know it was an idea – it’s just what is done in every other instrument ever so I assumed CDS had to work that way too. In derivative instruments like convertible bonds, equity options, securities lending (including of Greek bonds) – kind of everything but CDS – if the underlying security changes into something else, the derivative instrument changes into the same sort of derivative on the something else. The fact that CDS does not work that way was sort of puzzling to me and, it turns out, to ISDA and CDS market participants. But it’s all better now, or will be soon.*
Back when Greek CDS was maybe not going to work, there was much discussion of the possibility that this could kill the sovereign CDS market, and specifically of the possibility that that was exactly what European leaders wanted, what with the CDS market being the province of nasty hedge funds who were speculating against their debt, driving up their borrowing costs, and generally being unsportsmanlike.**
That speculation always suffered from one small flaw, which is that in the final analysis European leaders don’t get to decide how CDS pays out – ISDA does. And ISDA is a creature of banks and institutional investors and lawyers. Lots of lawyers. And so it’s not surprising to see ISDA rushing to fix this payout problem, after waiting a decent amount of time so as not to make it too obvious that it was a huge embarrassing problem in the first place. First of all, in my experience derivatives bankers and traders and investors and lawyers do place some stock in just plain being right: even aside from who has what economic incentives to have CDS work or not work, if you’re in the business of designing CDS you want it to work out of sheer pride of craftsmanship.
But second, ISDA – in the person, particularly, of its dealer members – wants CDS to work because it makes money trading CDS, and if the product doesn’t work no one will buy it. Plus: banks buy CDS! When they have bought CDS, they want it to pay out. Plus plus: banks sell CDS! When they have sold CDS, they might prefer to have it pay out the wrong amount of that amount is too low, but they would prefer to have it pay out the right amount rather than the wrong amount if the wrong amount is too high. And the way CDS currently works, instead of the potentially too low Greek payout, CDS buyers could maneuver to get too high a payout in the next Eurozone defaulter.***
I suppose it’s possible that governments might try to reduce the prevalence of various sorts of financial transactions just by introducing uncertainty and doubt into those transactions. That is one model you could have for the Volcker Rule, which “allows market-making” but requires a look into your soul to determine if you are in fact market-making. But if that’s what they want, it’ll have to be through their own rules, of which there are plenty. The financial industry, left to its own devices, prefers its rules predictable.
ISDA to Begin Biggest Revisions to Credit Swaps Since 2009 [Bloomberg]
Redesigning Credit Derivatives to Better Cover Sovereign Default Risk [SSRN]
* Duffie and Thukral at Stanford are specifically addressing sovereign CDS and the Greek exchange problem, but an “exchange property” concept would fix not only the Greek thing, but also a similar problem with corporate debt-for-equity exchanges like at GM, which are also on the agenda. Here, if the company does a prepack bankruptcy and poofs its bonds into equity, then there are no bonds to deliver into CDS and so the CDS never pays out. This can be fixed by just doing the auction before the bonds poof, but you may not want to rely on that always happening.
** There was also a residual sense that European banks were the net writers of Greek CDS and so preventing them from paying out would be a mitzvah from a systemic stability perspective, on the theory that European banks needed money more than hedge funds did. That theory was probably overblown but it existed, anyway.
*** Basically: €100 old Bond A exchanges into €20 face amount new Bond B trading at par plus €1 face amount new Bond C with 0.1% coupon and 100-year maturity; new Bond C trades at one cent on the dollar and is delivered into CDS, resulting in €99 payout on CDS plus the €20 new Bond B, for a package worth €119 to the holder while costing Portugal or Spain or whatever only 20-21 cents on the dollar.