• 11 Feb 2013 at 3:01 PM

CDS Contracts Not Ready For The Ways We Go Bankrupt Now

CDS, what is wrong with you? Here is how CDS should work:

  • There are bonds.
  • You buy CDS that is supposed to pay off if something goes wrong with the bonds.
  • Something goes wrong with the bonds, insofar as they default.
  • Like so:

  • So you scoop up a Defaulted Bond, hand it to the CDS seller, and he hands you back the face amount of the bond.
  • He’s stuck with the Defaulted Bond, and effectively loses the difference between the face amount of the bond and the value of the Defaulted Bond.

In the modern world here is what often happens:

  • There are bonds.
  • You buy CDS that is supposed to pay off if something goes wrong with the bonds.
  • Something goes wrong with the bonds, insofar as they poof into some weird garbage-y thing or assortment of garbage-y things.
  • Like so:

  • You can scoop up garbage-y things to your heart’s content, but the contract doesn’t let you deliver them into CDS in a way that achieves the sensible result.
  • Quick defined term:
    • Sensible Result = Face Value of Bond minus Value of Package of Garbage-y Things You Got For Your Bond
  • So you get less than Sensible Result, and are screwed, and the CDS seller has a windfall.1

Greece’s debt restructuring was previously the classic example of this, but Citi has an interesting note out today about Dutch bank SNS Reaal, which was nationalized by the Dutch government in a transaction in which SNS’s subordinated debt was transformed into a particularly odd kind of garbage. You could loosely describe the garbage as “nothing” – as in, the sub debt was “nationalized” or “expropriated” so in exchange for your sub debt you were handed a big ball of nothing – though you could alternatively describe it as “claims against the Dutch government.” As Citi puts it, “for CDS purposes, these claims are not deliverable under the terms of the contract”; the Dutch government adds that they’re worthless, too. “Nothing” is also not a CDS deliverable.

This makes no earthly sense at all; it is, I suppose, an artifact of a time when bankruptcy was a leisurely process in which Healthy Bonds became Defaulted Bonds which, in the fullness of time, became Nothing. Or sometimes became something else – they got Healthy again, perhaps, or they were converted into equity, or were restructured into lesser amounts of new bonds, or were paid out for some less-than-par amount of cash, or whatever. In any case the Defaulted Bonds were both Defaulted enough to trigger CDS, and Bonds enough to be delivered into that CDS. So everyone would settle up, and the CDS protection writer would be stuck with the Defaulted Bonds and the adventure of seeing what they eventually turned into.

An instantaneous step from Healthy2 Bonds into equity, or cash, or restructured bonds, or nothing, doesn’t give you a leisurely period to conduct a CDS auction and get to the Sensible Result. And so you are left with the Silly Result, as here, where the result, per Citi, appears to be that CDS-hedged sub bondholders (1) get nothing on their bonds (they were expropriated) and (2) get nothing on their CDS (since they have nothing to deliver).3

Citi frets that the same sort of problem applies in the broader case of “bail-in” and “coco” bonds, which by design convert into equity – or, on occasion, into nothing – if regulatory capital triggers are breached. Their concern is that equity is not deliverable into CDS contracts, and that the rise of bail-in provisions in European bank bonds will kill off European bank CDS:

These issues over triggers and deliverables could all too easily jeopardise the validity of the CDS market as a hedge. If it were accidentally “killed off”, the consequences for the bond market would be severe. Unlike in sovereigns, where the underlying cash market has normally been more liquid than CDS, in the corporate market liquidity is heavily fragmented. Without the signaling and hedging role of an active CDS market, bond transparency would fall, trading would become more lumpy, and ultimately the cost to bank issuers would increase, as an increased illiquidity premium became factored into spreads.

You might pause to ponder the structural advantages and disadvantages of putting the credit risk of subordinated bail-in bank debt on CDS dealers, rather than on, y’know, the market-disciplining holders of that debt, but leave that aside. Here is Citi’s dire warning:

If lawmakers do not take steps to consciously trigger CDS as part of their bail-ins, the only other alternative we see is redrafting the CDS contracts. New CDS contracts could be drafted explicitly incorporating bail-ins in their credit events’ language, and allowing for the deliverability of a “package” of whatever investors are left with following bail-ins, even if these consist only of equity or claims. Such steps are already underway for sovereign CDS following the Greek experience, and could be adopted for banks too. Even here, though, it may prove difficult to formulate contracts in a watertight fashion (expropriation by definition leaves investors with nothing to deliver).

That’s … that’s not that dire? We’ve talked about this, obviously, in the context of Greek CDS: last year ISDA started the process of pondering fixing CDS so that things like debt-for-equity swaps, or Greek-style debt-for-new-debt swaps, were covered. It may be “difficult to formulate contracts in an entirely watertight fashion,” but getting to a seaworthy concept doesn’t seem all that hard: essentially, define “Reference Obligation” to mean “(i) the Reference Obligation or (ii) whatever package of things the Reference Obligation poofed into in a Credit Event.”4 That sort of fix handles Greece, bail-in bonds that poof into equity, expropriated bonds that poof into “claims,” etc. It covers SNS Reaal, too, if you specify that “for the avoidance of doubt, if a Reference Obligation poofs into nothing, the Reference Obligation will be deemed to be that nothing.” The price of nothing, I think, is uncontroversially zero.5

One thing that is often noted about financial innovation and regulation is that regulators are always a few steps behind private-sector innovators: where there’s profits on the line, bankers are usually faster at coming up with fixes to regulatory obstacles than regulators are at coming up with new obstacles. It’s sort of odd that, in the realm of CDS, the private sector’s responses lag so far behind the public sector’s innovation in finding creative new ways to make bonds worthless.

What bail-in means for CDS [Citi via FTAV Long Room]
Dutch-bottomed bank bondholders [FTAV]

1. There are alternative versions where you can deliver a counterintuitive assortment of garbage-y things into CDS in such a way that you get back more than Sensible Result, though these are curiously rare in practice.

2. In the sense of “un-defaulted,” not necessarily in the sense of “trading at par.” The handwriting can be on the wall before the CDS is triggered.

3. Is that right? It seems exceptionally silly; I feel like the formulas in the ISDA definitions return less “zero” and more “#NUM!” But I suppose the literal reading of like Credit Definitions 7.3 is that you get the greater of (i) Reference Price minus Final Price and (ii) zero, and that Final Price is the price of the Reference Obligation, and if that doesn’t exist then that price is undefined and you’re left with zero? Or is it like, the price of a thing that doesn’t exist is zero, and so you get the Reference Price (which is the Sensible Result)? Or is the price of a thing that doesn’t exist infinite? I suppose that there’s a good argument for that – how much would you charge for a unicorn?

4. This could probably be made less recursive with a small expenditure of effort. But you get the idea.

5. As opposed to the unicorns, which could be #NUM! or infinite.

9 comments (hidden to protect delicate sensibilities)
Show all comments ↓

Comments (9)

  1. Posted by Fightpicka | February 11, 2013 at 3:13 PM

    "Like so" WTF man?! What are you that Ronco guy trying to sell us commintatist a fucking chicken rotisserie?! So wrong dude, so, so wrong. Get back in your little slant six Trans-am and GTFO.

  2. Posted by Son Jazar | February 11, 2013 at 3:20 PM

    Shazar had a Toys R Us mountain bike with pegs growing up.

  3. Posted by Im_a_Dude | February 11, 2013 at 3:23 PM

    Matt,
    good job on the charts. direct and to-the-point!

  4. Posted by guest | February 11, 2013 at 3:40 PM

    And even more silly, GMRA would probably cover this fine (i.e. if you had shorted SNS subs, you would return the resulting claims when the repo rolls off)

  5. Posted by cds quant | February 11, 2013 at 4:48 PM

    actually, the fix would be in amending "Deliverable Obligation" (rather than Reference Obligation) or more specifically, amending "Deliverable Obligation Category" and "Deliverable Obligation Characteristics" but other than that, no concerns…..

  6. Posted by guest | February 11, 2013 at 5:37 PM

    Probably if the language for "Sovereign Restructured Obligation" were carried over to the Bank section, that might work, too.

  7. Posted by Excel | February 12, 2013 at 9:36 AM

    I feel like all we do is Missonary anymore. I need more excitement.

  8. Posted by Dead_cat | February 12, 2013 at 1:10 PM

    I'm afraid it's worse than that. He's cheating on you with powerpoint.

  9. Posted by guest | February 12, 2013 at 2:06 PM

    It seems that the dire consequence is that these defaults are being structured specifically to avoid CDS… what's the purpose of buying CDS, then, if everyone is deliberately (a) avoiding the triggers or (b) eliminating the qualified deliverables.