Streetwise Professor

The deputy governor, financial services of the Bank of England, Paul Taylor, gave a speech warning of the dangers inherent in clearinghouses. All of his major points were raised in my ISDA discussion paper (and which I’ve been flogging in my research and here on SWP for going on 3 years now).  Most notably, he highlighted the dangers of procyclical margins, and the very real possibility of a CCP failure and the devastating consequences such a failure would have. Been there, said that.

Such realism is good to hear, but as I said when Bernanke echoed similar concerns, it would have been a damn sight better had people like Mr. Tucker piped up before the clearing mandate train left the station with Congressional and Parliamentary Casey Joneses at the throttle.

But here’s the truly ironic part.  As I did in my ISDA piece, Tucker grapples with how to resolve a failed CCP:

Two strategies come to mind, which I am airing for debate. The first would be ‘recapitalising’ the CCP so that it can carry on.  The second would be to aim to bring off a more or less smooth unwinding of the CCP’s book of transactions.

On the former route, in the banking sphere the basic strategy is to develop regimes for bondholders to share in the losses after the equity is wiped out; a distressed bank gets ‘recapitalised’ in some way through a reconstruction of its liabilities.  But CCPs do not issue debt, so there is not obviously an economic equivalent of recapitalisation by way of haircutting the debt claims of bondholders.  Recapitalisation would seem to entail an injection of new outside capital.   That should not come from the public sector.  So the only options are the surviving clearing members themselves or a new ‘owner’.  On the face of it, the clearing members would have to be involved, as surely the stricken CCP’s default fund would also need replenishing in order for it to continue.

If the default fund were to be replenished by its surviving clearing members, that could in principle be sorted outat the time or it could be pre programmed.  In practice, making it up at the time must be dangerous.  A pre-programmed allocation of losses would be akin to the old “Down To The Last Drop” rule that used to be employed by a Chicago clearing house [the CME], under which surviving clearing members ultimately had collectively to absorb all losses: mutualisation.  (It was dropped because, as banks became futures-exchange clearing members, bank supervisors didn’t like the unlimited exposure of CMs who were banks.  An example of unjoined-up thinking.) [That’s not the way I heard it explained by those involved in the decisions.] [Emphasis added.]

Breathe deep the irony.  Remember that the primary justification advanced for mandating CCPs was to create a firewall that would limit the exposure of major financial institutions to counterparty losses arising from derivative defaults.  That is, the concern was that the failure of one large financial institution would cause it to default on derivatives trades, imposing losses on other large financial institutions, potentially resulting in their failure.  CCPs were supposed to contain default risks and prevent the from spreading, contagion-like, through the financial system.  But Mr. Tucker argues that for clearing mandates to work, it is necessary to breach the firewall and make big financial institutions the residual claimants of counterparty risk.

Mr. Tucker has just revealed the dirty little secret.  Well not a secret exactly; I’ve been pointing this out for years, notably in my posts on AIG.  Specifically, clearing does not make risks go away.  (No, Gary, there is no magic box.)  It just shifts risk around.  It has to go somewhere.*  Collateralization, by moving derivatives counterparties to the front of the creditor line, shifts risks to other creditors (who may well be large financial institutions, in part).  If collateralization is insufficient, the risks go to the other members of the clearinghouse–which means, it goes to other financial institutions.  Put differently, it vectors the risk right back to the same financial institutions that clearing was intended to protect.

Following the inexorable logic of the situation, namely, that the risk has to go somewhere, Tucker looks for where the risk can go.  Can’t go to bondholders, because a CCP doesn’t have any.  So it must go to clearing members–large financial institutions.**  Which means we’re right back where we started.

Which is another echo.  Several times I’ve predicted that clearing mandates would likely spur a process similar to that in the apocryphal story of the Indian village infested by mice, that brought in cats to eliminate the mice, that brought in dogs to deal with the proliferation of cats, that brought in elephants to drive out the hordes of dogs . . . and then brought back the mice to rid themselves of the rampaging elephants.  Clearing mandates were predicated on the chimerical belief that it was possible to make risk go away, when in fact, its primary effect is to shift risk around.  Which raises the question of where it has to go.  The BOE’s Mr. Tucker has apparently concluded, quite logically, that ultimately it has to rest with large financial institutions where it was in the first place.

In other words, bring back the mice.

* As I’ve noted repeatedly, clearing also affects incentives and therefore can affect the level of risk in the system.  There are a variety of reasons why CCPs may lead to an increase in total risk exposure, mainly by creating moral hazard, and increasing capital efficiency through netting.

** An important caveat here.  The nature of clearing members is not exogenous, and independent of the sharing rule.  Imposing good to the last drop is likely to make big banks reluctant to be clearing members with effectively unlimited exposure to other members.  Instead, GTTLD would likely lead to the creation of affiliates with limited capitalization.  By not recognizing this likelihood, Mr. Tucker is guilty of some “unjoined-up thinking” of his own.

Update.  Whoops.  In my hurry, wrote “Taylor” instead of Tucker.  Thanks MDC.