How to regulate CDS

By Felix Salmon
March 10, 2010
Gary Gensler's speech on regulating credit default swaps: it's by far the best thing written on the subject to date. He's absolutely right about pretty much everything, and it would be amazing if the Europeans, who seem much keener to start regulating these animals than we are in the US, were to use it as a blueprint for their own CDS reform.

" data-share-img="" data-share="twitter,facebook,linkedin,reddit,google" data-share-count="true">

Go read CFTC head Gary Gensler’s speech on regulating credit default swaps: it’s by far the best thing written on the subject to date. He’s absolutely right about pretty much everything, and it would be amazing if the Europeans, who seem much keener to start regulating these animals than we are in the US, were to use it as a blueprint for their own CDS reform.

I’m in a little bit of a rush this morning so can’t go through the speech in much detail, but there are a few things worth picking out, starting with this statement:

A comprehensive regulatory framework governing over-the-counter derivatives should apply to all dealers and all derivatives, no matter where traded or marketed. It should include interest rate swaps, currency swaps, foreign exchange swaps, commodity swaps, equity swaps, credit default swaps and any new product that might be developed in the future.

This seems to have been forgotten in a lot of the rhetoric surrounding Greece, but if you crack down on CDS while allowing all manner of other OTC derivatives to remain unregulated, any behavior you don’t like will simply move elsewhere in the OTC market.

Gensler says, rightly, that regulation of derivatives dealers is key to doing this effectively, since you can never wipe out the OTC market entirely, no matter how many derivatives are moved onto exchanges. And standardized OTC derivatives — which is the vast majority of them — should be cleared by clearinghouses. Moving to exchanges and clearinghouses isn’t a panacea, but it is likely to help.

And it makes sense that single-name CDS should be regulated by the SEC, along with corporate stocks and bonds: if investors have long since reached the point at which they write credit protection interchangeably with buying bonds, the SEC should catch up with them. (And for that matter, the SEC should be policing the corporate bond markets much more strongly that it does right now as well.)

Gensler makes a strong case that empty creditors should not be allowed to mess around with bankruptcy proceedings; I suppose the alternative here is to say that writers of credit protection should be given a seat at the table. But Gensler’s solution is more elegant, if arguably harder to implement.

As for bank capital, in the wake of the financial crisis it’s a no-brainer to reduce the degree to which banks can use the CDS market to engage in regulatory arbitrage. If you need capital, you should, as a rule, have capital: you shouldn’t be able to get away with buying CDS protection on some of your assets instead.

The CDS market grew very quickly, without any adult supervision, and it’s inherently a much riskier market than most other derivatives. As Gensler says,

While the value of interest rate or commodity derivatives generally adjusts continuously based on the price of a referenced asset or rate, credit default swaps operate more like binary options. A seller of CDS could one month collect its regular premium with little expectation that the insured company may default and in the next month be on the hook for billions if the insured company goes bankrupt. A credit default swap can quickly turn from a consistent revenue generator into ruinous costs for the seller of protection. This “jump-to-default” payout structure makes it more difficult to manage the risk of credit default swaps.

As a result, CDS do pose systemic risks, and should be carefully regulated. But let’s not unhelpfully oversimplify matters by banning them — or banning “naked” CDS — entirely. For one thing, that won’t ever happen, and it’s a much better idea to try to implement something which can actually work in reality.

Comments
6 comments so far

good piece except this:
“If you need capital, you should, as a rule, have capital: you shouldn’t be able to get away with buying CDS protection on some of your assets instead.”
yes, you should, that’s called hedging.

Posted by alea | Report as abusive

Would a small tax on CDS trades also discourage regulatory arbitrage? The added benefit of actual tax revenues and a second pair of eyes might be useful too.

Posted by polit2k | Report as abusive

The empty creditor problem (one of my favs).

“Gensler’s solution is more elegant, if arguably harder to implement.”

I’ll argue with that, Felix. Gensler’s solution is “to specifically authorize bankruptcy judges to restrict or limit the participation of “empty creditors” in bankruptcy proceedings.” The implementation hurdle here is identifying who among the bondholders are empty creditors. Let’s compare that to your alternative solution “that writers of credit protection should be given a seat at the table.” The implementation hurdle here is identifying who, among all writers of credit protection, is net short protection (net long credit). The latter strikes me as a higher hurdle.

Of course, all debate about how to fix the empty creditor problem presupposes that the problem exists. The Coasean argument suggest that the extent of the problem is driven solely by market frictions (transaction costs, asymmetric info, etc.). If believed, this would suggest clearing/exchange/transparency is a sufficient solution.

Posted by Sandrew | Report as abusive

alea

This is not to say a bank can’t or shouldn’t hedge, in that the point of hedging is (dictionary.com) “to enter transactions that will protect against loss through a compensatory price movement”. It’s to say that under no circumstances should this substantially remove its regulatory capital requirement.

It’s one thing to use interest rate, currency, equity derivs to hedge essentially fungible risks. Not so with some of the deals I’ve seen in my recent career, whereby a large European bank buys a CDS on a pool of private, illiquid and otherwise untransferable loans.

You try and price such risk, and price the trade from both sides.

The size of such deals often ran into tens of billions notional. Ie you replace a granular bunch of credit risk with one big counterparty exposure.

How is this a good hedge?

Surely the idea of capital requirements is for these loans to be supported by adequate capital. To make sense any reduction in the first firm’s requirement should be offset by the increase in the second’s. All of a sudden the trade looks far less attractive.

Posted by vk9141 | Report as abusive

All this dinosaur dung keeps people fat and happy as they push it around, and journalists and bloggers pay rent by talking about it, but isn’t enough enough? Get rid of it. Hedge this, swap that, securitize your mother. Sick little world.

Posted by Uncle_Billy | Report as abusive

“While the value of interest rate or commodity derivatives generally adjusts continuously based on the price of a referenced asset or rate, credit default swaps operate more like binary options. A seller of CDS could one month collect its regular premium with little expectation that the insured company may default and in the next month be on the hook for billions if the insured company goes bankrupt.”–The above describes the asymmetric nature of credit risk in general: loans, bonds, not just default swaps. There is nothing riskier about writing $10MM of protection than in buying $10MM of a bond. If this ‘jump to default’ risk is too tricky for institutions to manage, then they shouldn’t own credit portfolios either, yet noone seems to have a problem with that.

Posted by Chavez | Report as abusive
Post Your Comment

We welcome comments that advance the story through relevant opinion, anecdotes, links and data. If you see a comment that you believe is irrelevant or inappropriate, you can flag it to our editors by using the report abuse links. Views expressed in the comments do not represent those of Reuters. For more information on our comment policy, see http://blogs.reuters.com/fulldisclosure/2010/09/27/toward-a-more-thoughtful-conversation-on-stories/