Streetwise Professor

June 8, 2011

When the Levee Breaks

Filed under: Clearing,Derivatives,Financial crisis,Politics,Regulation — The Professor @ 12:52 pm

Reading the BIS paper I just posted about brought to mind an issue that I think is of paramount importance, but which has been almost completely overlooked.  Like many other analyses of the effects of clearing mandates and rules relating to derivatives, the focus has been on providing mechanisms to reduce the likelihood of default on derivatives positions.  The idea is that losses on derivatives positions can be a channel of contagion between important financial institutions.  So to address this problem, legislators and regulators want to raise collateral requirements and other requirements to ensure that non-defaulting derivatives counterparties receive all that they are owed when a firm defaults.

But focusing on derivatives alone is dangerous.  All of the measures intended to increase the safety of derivatives markets–increasing collateralization, multi-lateral netting, etc.–privileges derivatives and derivatives counterparties.  But that privileging comes at the expense of other claimants of a bankrupt firm.  Let’s say that without greater collateralization and netting, derivatives counterparties would have lost X, but with the greater collateralization and netting, they lose 0.  Sounds great: there is no possibility of derivatives losses cascading from firm to firm.  But not so fast: what means, to a first approximation* is that now somebody else loses X more than they did under the previous rules.  In other words, the loss of X has been shifted from the derivatives counterparties to somebody else. There can still be contagion—just via a different channel.

There is something of a debate about the priority of derivatives in bankruptcy.  Mark Roe has questioned this priority privileging.  I’m in London for a conference at the LSE, and one of the conference papers, by Patrick Bolton and Martin Oehmke, also questions privileging derivatives.  I’m not sold on the specific Roe-Boton-Oehmke arguments because I think they don’t see the whole picture, but I do think it is a very important issue.

It is a particularly important issue in the context of the ongoing regulatory debate because all of the rules intended to reduce the likelihood of derivatives defaults effectively increase the priority of derivatives.  Which means that they reduce the priority of other claimants.  That’s not obviously a good thing.  But since the regulators and legislators have failed to look at the problem holistically, instead focusing on derivatives in isolation without seriously questioning how the changes in derivatives rules will primarily shift risk rather than reduce it, whether it is a good thing or not has escaped attention completely.

A couple of analogies come to mind.  One is target fixation.  Many fighter pilots become so focused on the target they are chasing that they fail to see the guy that just jumped their six and is about to shoot them down.  Being fixated on the derivatives target, regulators and legislators are unaware that they are shifting risks elsewhere—risks that may very well jump them of the future while they are still fixated on derivatives.

Another is levees.  Building up a levee around a particular city protects that city, yes, but the rising water has to go somewhere.  So raising the levees in one place makes the flooding problem worse somewhere else.  Legislators and regulators seem obsessed with raising the levees around Derivatives City, without regard as to how this will affect where the water goes and who gets flooded–Commercial Paper Town and Moneymarketville, for instance.  It could well be the case that keeping Derivatives City high and dry results in even more devastating flooding in other, less privileged precincts.

The Corps of Engineers has created many problems in the past with its attempts to control the Mississippi and Missouri Rivers through its efforts to control floods: there is a very real risk that a (metaphorical) Corps of Financial Engineers will also create havoc despite its intention to do good.

There are of course differences.  A river system is more mechanical than a financial system.  But that actually makes the situation more fraught in financial markets.  Market participants will respond to the privileging of one class of claim—such as derivatives—by adjusting their behavior, by adjusting what they trade, their capital structures, etc.

But maybe it’s not that different.  The design of the levee system will affect where people build and how they build.  Designing the levee system properly requires just that—a systemic approach, not an approach focused on just one part of the system, but its entirety; and an approach that takes into account how people will respond to the design of the levee system.  Unfortunately, that’s not the way that it’s being done in the derivatives markets.  The history of the Mississippi River valley suggests that’s not an encouraging thing.

* As I discuss below, this privileging will affect decisions people make, and thus the amount of risk.  That means the problem is even more complex than what I consider here.  But that only makes the derivatives-centric focus of regulators and legislators even more dangerous.

Print Friendly, PDF & Email


  1. (1) I agree it’s the institutionalization of (I think unhedgeable) funding risks which creates more systemic risk than would be saved. Last months’ commodity selloff is a small sample of the possible effects. Any new system would have to have very objective, clearly defined, and time-smoothed funding parameters. What you don’t want are big crowds trying to exit small doors.
    (2) Are papers such as these the very tools Mr Chairman claimed didn’t exist?

    Comment by Derek — June 8, 2011 @ 8:55 pm

  2. I am wondering how come Goldman never “knew” AIG was going bankrupt… Didn’t they naturally try to increase the collateral they held?

    When you say that someone loses “X”, can we ignore things like the total probability of default might be reduced under new rules? If the bilateral mechanism for derivative margining is indeed the weakest link, then one could argue that strengthening it might lead to overall lower default probability because firms might reduce the risks they undertake?

    Comment by Surya — June 8, 2011 @ 9:45 pm

  3. Am I missing something, or is the assumption behind these requirements that large financial institutions are unable to manage their credit exposures without firm instructions from the government? Do we really have any idea of what would have happened if AIG had been allowed to fail? I am reminded of Enron, which was by far the largest provider of liquidity to the natgas markets, and quite possible to several other commodity markets as well. And yet, when Enron failed I don’t believe that a single counterparty failed due to exposure to Enron’s credit. Not only no contagion, not a single firm failed! So is the assumption that the energy companies who were Enron’s counterparties were simply more sophisticated exposure managers than the banks who were on AIG’s hook?

    Comment by JLehman — June 10, 2011 @ 12:00 pm

  4. SWP, thank you for injecting some needed clarity into the debate on derivatives and bankruptcy, but I think your levee analogy misses the point behind “privileging” derivatives (and, for that matter, repo and clearing houses and other financial contracts). The problem close-out seeks to address is the immediate one of rebalancing the non-defaulting firms’ books (or the clearing house’s book) following a default. Failure to do so presents an immediate problem of open positions subject to sudden market moves and the resulting systemic consequences. You might be right that some losses might be transferred to other markets, but this is a longer-term problem that could possibly be dealt with during the insolvency proceedings by, say, ex post claims on the firms that were able to benefit from close-out. This is a good area to ask why the industry developed these procedures and why they have been almost universally accepted by both industry and regulators and questioned only by academics and a few members of the bankruptcy bar. Again, thank you for providing some coherence, which is sadly lacking in the Roe et al. arguments.

    Comment by DrD — June 13, 2011 @ 4:41 pm

  5. […] I return to a metaphor I’ve used before: that of flood controls.  Building up levees to protect one location can increase the risk of a catastrophic flood elsewhere in a river system.  It is good that more regulators and market observers recognize that CCPs are a source of systemic risk, but the levee-building reflex is worrisome in the extreme.  It suggests an inability to think about the financial system as a whole, a system that reacts to changes imposed on one part of it.  These responses are not necessarily salutary, and can be downright frightening. […]

    Pingback by Streetwise Professor » When the Levee Breaks, Redux — April 5, 2012 @ 3:08 pm

RSS feed for comments on this post. TrackBack URI

Leave a comment

Powered by WordPress