Trade exposures with CCPs - CRR Provision
|
Regulatory Capital Anatomy™
The JC’s untutored thoughts on how bank capital works. 306(1). An institution shall apply the following treatment to its trade exposures with CCPs:
|
In which the Capital Requirements Regulation 575/2013 (EUR Lex)) requires that, for a clearing member to achieve a zero weighting on its exchange traded derivatives clearing business, it must first be able to pass losses occasioned by the default of a CCP on to its client.
Where this came from
In the same way that that irritating guy with the Tele in the Redondo Guitar Center lists his influences as Hendrix, Sonic Youth and Mahler, this part of CRR lists as its influences the BIS paper from April 2014, capital requirements for bank exposures to central counterparties. That assumes your brokers are all agents the way US futures commission merchants are.
Twist: European clearers generally act as riskless principals. Cue tons of fun trying to work out what this all agency chat means for a principal.
The BIS Paper
The BIS paper is predicated on the US FCM model where a clearing member or intermediate broker acts at all times as agent. It is in the nature of a disclosed agency that the agent is not personally liable to perform of the contract, and does not therefore have credit risk to persons who are. Thus:
- The starting assumption is that when finally executed, the parties to a futures contract are the client and the CCP. The intermediate entities are all agents and thus not responsible for performance of the contract. Therefore an intermediate broker’s performance and creditworthiness is not really a concern to an executing broker;
- It would be most unusual for the executing broker to take any risk on the CCP, so it would only have to apply a risk-weighting against the CCP where it has actually guaranteed the CCP’s performance. This would be unusual, so the default assumption is that no risk weighting would apply.
This talks about CCPs. What about intermediate brokers?
Well, you would like to think so. It doesn’t make a lot of sense otherwise. But CRR itself is silent on the point. But here's the read across:
- The BIS paper on which CRR provision was modeled is framed in the positive: “Where the clearing member offers clearing services to clients, the 2% risk weight also applies to the clearing member’s trade exposure to the CCP that arises when the clearing member is obligated to reimburse the client for any losses suffered due to changes in the value of its transactions in the event that the CCP defaults.”
- The theory being that unless the clearing member has agreed otherwise, it wouldn’t be liable for those provisions in the first place: ie, it assumes an agency model (Common in the US) for client clearing.
- The European model is a back-to-back principal model, however, so unless you’ve somehow carved it out, you would be liable for it, all other things being equal.
- The BIS does talk about this a little: “The treatment in paragraph 192 to 194 may also apply to exposures of lower level clients to higher level clients in a multi-level client structure, provided that for all client levels in-between the conditions in (a) and (b) below are met.” The conditions (see here for their text) are predicated on all the structures being of an agency kind.
- One is therefore obliged to extrapolate from an agency model to the economic equivalent under a principal model.
What about non-performance by a solvent intermediate broker?
The clearing member’s own negligence, fraud or wilful default
But it will be a generous client indeed who does not insist on a carve-out from that right for defaults caused by the clearing member’s own negligence, wilful default or fraud. Would such a carve-out invalidate an application of 306(1)(c)?
Respectfully, it is submitted, it would not:
It is clear in 306(1)(c) that whether or not the CCP is generally “in default” in the abstract, in the sense of being “insolvent” it must have specifically defaulted under the transaction (that is, it must have failed to pay something that it owed to the clearing member).
Of course, a CCP’s insolvency would be likely to lead to a transaction default.
Any negligence[1], wilful default or fraud on the clearing member’s part under the CCP transaction would, QED, be a default by the clearing member. The CCP then would be entitled to withhold payment under the transaction; ie, it would not be in default in doing so.
What, then, if a clearing member default were so egregious that it caused the total failure of the CCP, meaning the CCP failed to pay even amounts that it was obliged to pay on default by the clearing member?
So a loss to the clearing member which arose out of the CCP’s inability to perform under a transaction which, in turn, came about as a result of the clearing member defaulting to that CCP would not be “loss suffered in the event that CCP defaults”: if the clearing member sued the CCP for that loss, it would fail.
Looking at it another way, if such a carve-out did invalidate 306(1)(c) then the provision would have no application at all, because it would be commercially impossible to remove it.
See also
- The Basel / BIS paper on which this turbulent regulation was based: Capital requirements for bank exposures to central counterparties
- Agency versus [[prini