The two functions of a CCP – and why they don’t have to be united March 3, 2012 at 5:35 pm
In a previous post, I introduced the notion of a central margin custodian: a body which administers and settles both initial and variation margin for OTC derivatives. You might ask what CCPs do above and beyond this – the answer is that they act as a guarantor of derivatives receivables. As a technical matter the CCP is actually a party (the clue is in the ‘P’) to cleared trades, but it need not be. You would get the same functionality if the CCP acted as an insurer, where the thing insured is the cleared derivatives receivable.
Thus we have CCP = CMC + derivatives receivable insurer
Of course, insurers need financial resources, and that is what the CCP’s default fund (and the wafer-thin slice of equity they have) are for. They are, to be precise, amounts of money that are available to meet credit losses above IM. Given that the CMC covers losses up to IM, the decomposition is precise.
Now thinking about things this way has the key advantage that we can see the relationship between IM and DF more clearly. From the CMC’s perspective, it doesn’t care how high (or low) IM is: it is just an amount of money to be collected. (Well it might have a preference for it to be higher so that it has more funds under management, but that is hopefully a secondary effect.) For the insurer, though, it matters a lot. The tranche of loss it has insured is attached at IM, so the lower IM is, the more risk it has. It will need more financial resources – more DF – to support that risk. In fact one thing that would make me feel more comfortable about CCPs would be knowing that the insurance premium charged was commercial: would Berkshire Hathaway, say, write credit insurance on the cleared derivatives receivable if the premium available was the default fund contributed charged by LCH, and IM was determined by LCH’s model? How about ICE or CME? And if Berkshire wouldn’t trade at that level, what does that tell you about the safety of the leading CCPs?
Or, to put it another way, suppose that the credit insurance part of a CCP was a standalone business, with the CCP’s capital and DF behind it. How would it be rated?
[...] CCPs perform mainly two functions, or possibly [...]
True dat. CCPs default funds are buttons in comparison to the risk they take on. Obviously as you can’t insure everything. At the end of the day someone has to take risk no matter how you obfuscate it. Much the same with deposit guarantee schemes. The funds are nonsense in comparison to the risks, yet their existence gives a false sense of security and allows participants to go “Look! We have this wonderful insurance thingy. No need for more regulation. We’ve good boys now and have learned our lesson”. Which is all fine until a CCP blows up (or monoline or Icelandic banking system) and in steps the tax payer in the name of firewalling systemic risk. Plus ce change.
“to put it another way, suppose that the credit insurance part of a CCP was a standalone business, with the CCP’s capital and DF behind it. How would it be rated?”
The whole point of a clearing house is that it is mutual organization that is backed by much more than the default fund and equity. E.g. at the LCH between the exhaustion of the default fund and the exhaustion of equity lie clearing member pro-rata contributions, any insurance that may be available, and other clearing member indemnities. It’s my understanding that new clearinghouses also have mutual guarantees behind them — but please correct me if I’m wrong.
On LCH, see here:
http://ftalphaville.ft.com/blog/2010/06/30/274766/will-derivatives-reform-take-us-from-ccps-to-gses/
That said, it’s clear that the conservative pricing of this insurance is extremely important. In fact, I believe that’s the point of a clearinghouse — and the reason the capital of the big banks is at risk before the equity of the clearinghouse.
Carolyn
You are somewhat right, but the limits are interesting. You say
“The whole point of a clearing house is that it is mutual organization that is backed by much more than the default fund and equity. E.g. at the LCH between the exhaustion of the default fund and the exhaustion of equity lie clearing member pro-rata contributions, any insurance that may be available, and other clearing member indemnities. It’s my understanding that new clearinghouses also have mutual guarantees behind them — but please correct me if I’m wrong.”
First, extra default fund assessments are capped at 100% of the original. So yes, a CCP can basically double its capital if it makes a capital call and everyone pays up (something one might at least be a little sceptical about given the kinds of conditions under which it is likely to be calling). But it can’t go further than that: CM liability is limited (ignoring the problematic case of Japan at least).
Second, as I understand it CME is the only large CCP with a big chunk of insurance behind it ($6B from memory but I might well be wrong). I don’t think LCH or ICE have material credit insurance policies.
Third, this capital has mostly come from the very banks that were big OTC participants pre-clearing. In other words it isn’t as if CCPs have attracted substantial NEW capital to support counterparty credit risk: they have just reallocated it (in arguably a more efficient arrangement, but still).
Thank you for the reply, David. I spent some time trying to determine the source of my information on clearing.
“extra default fund assessments are capped at 100% of the original”
I’m not sure there’s a cap at least in the US CCP structures. Economics of Contempt claimed there usually a clause in the CM agreement allowing the clearinghouse to call on the CM for a contribution if needed. http://economicsofcontempt.blogspot.com/2009/12/lynch-amendment-bizarre-and-confused.html
And whenever I’ve checked the documentation, this seems to be correct. E.g. for ICE, bottom paragraph of page 5 here: http://www.federalreserve.gov/newsevents/press/orders/orders20090304a1.pdf
“this capital has mostly come from the very banks that were big OTC participants pre-clearing”
I agree that this can be a problem — especially if the banks believe they can cry “systemic crisis” and get a government bailout whenever such capital calls are actually made. But then if the financial system is relying on the government to bail it out, it’s far from clear that any structure can solve the problem.
Thanks for the comment.
The key question is not ‘how much can a CCP assess’ but ‘how much can they get out of a CM before the CM can resign its membership and leave’. This is because a CM is not locked into a CCP that is heading for failure; they can always port their trades to another one. (Well operationally that is going to be challenging, but in principle it can happen, and in practice I am assured by several CMs that this is what they would look to do were a CCP to be heading for disaster). That is where the 2x figure comes from.
Thanks for the FED document on ICE. I agree that they (and LCH and CME) can call for more, but I still believe that there is a cap on how much more.
Interestingly I think I was wrong about insurance at CME. See here: http://www.cmegroup.com/clearing/files/financialsafeguards.pdf – there is no mention of insurance in that document.
A passage of interest with respect to assessments in the same reference says: “assessment powers cannot exceed 2.75 times the aggregate Base guaranty fund requirement across all clearing members for a single default. For multiple defaults in a five-day period, assessment powers are capped at 5.5 times the aggregate guaranty fund.”
That’s higher than 2x, but of course it all depends on how the DF is sized in the first place (i.e. the balance between funded and unfunded capital).
BTW, there is a very real need to cap DF assessment powers, and that is bank regulatory capital. Capital requirements are calculated based on the notional at risk. Now the capital requirement for unfunded DF (aka DF assessments) isn’t yet clear, but it is not going to be zero, and anything more than zero percent of infinity is infinity…
One thing that I thing would at least help, would be for CCPs to be forced to raise substantial [multi billion dollar] equity on the public market. That would (a) introduce new capital from outside the system which supported counterparty credit risk, (b) validate that CCPs’ business model is at least sound enough to allow them to raise money and (c) provide (via standard disclosures for listed companies) rather more information to the market and to supervisors about CCP health. And, frankly if I were a clearing member, I would then consider hedging my DF position by buying equity puts…
Thank you! I find this a very interesting discussion.
You are welcome; I am glad you found it interesting.