Sunday, October 19, 2008

Why the CDS Market Didn't Fail

(Felix Salmon) Jane Baird has the latest on what Alea calls "the non-event of the year": the Lehman Brothers CDS settlement on Tuesday. The upshot is that there's very little to worry about: the worst-case scenario is limited to the failure of a small hedge fund or two, and even that seems improbable. The reason's simple:

The standard practice in the CDS market is that hedge funds and other counterparties must adjust collateral on a daily basis as the value of a contract changes.
As Lehman CDS fell in value, before and after it filed for bankruptcy, protection sellers would have had to provide increasing amounts of Treasury bonds or other cash-like investments as collateral for those contracts.
"The mark-to-market on the CDS is margined daily as a credit event draws near, and that mitigates a large, lumpy payment at the end," said Peter Goves, another Citigroup strategist.

This bears repeating: if you take credit risk by writing credit protection, your position is marked to market daily, and is margined daily. Compare that to the behavior of banks, say, which took billions of dollars of credit risk by holding super-senior CDO tranches and didn't -- couldn't -- ever mark them to market. It's hardly a surprise that the banks have been stunned by the magnitude of their losses, while writers of credit protection were forced to face their deteriorating positions on a daily basis.

But hang on, I hear you say, what about AIG? What about the monoline insurers? Weren't they undone by CDS? Yes -- and they're the exception which proves my point. AIG and the monolines had something no other writer of credit protection had: a triple-A credit rating. As such, they were the only sellers of credit default swaps who didn't need to put up collateral as the market moved against them. The minute they were downgraded, they suddenly needed to come up with billions of dollars of collateral, and they failed.

This is basically the issue I have with Jesse Eisinger, who in his latest column claims that "the roots of this year's financial crisis" lie in the creation of the CDS market:

Credit derivatives aren't, of course, solely to blame for the pandemic that has helped bring down Wall Street. They didn't single-handedly force Bear Stearns and Lehman Brothers to bulk up on toxic debt, dooming them to collapse. But they made the financial world more complex and more opaque.

More complex? Yes. More opaque? No. In fact, credit default swaps, being much more liquid than most debt instruments, are thefore also more transparent than most debt instruments. Try to sell a CDO tranche, these days: you can't. There's no two-way market in such things. But if you want a price on a credit default swap, that's very easy to obtain. Eisinger, remember, is the chap who presciently worried about the magnitude of banks' level-three assets more than a year ago. Credit default swaps aren't impossible-to-value level-three assets; they're not even hard-to-value level-two assets. They're transparent level-one assets: it's harder to think of a credit instrument which is easier to value.

I'm reminded of the great Dutch urban planner Jan Gehl, who made pedestrians safer by seemingly making streets riskier. The NYT explains how his invention, the woonerf, works:

Pioneered in the Netherlands -- the word roughly translates as "living street" -- the woonerf erases the boundary between sidewalk and street to give pedestrians the same clout as cars. Elements like traffic lights, stop signs, lane markings and crossing signals are removed, while the level of the street is raised to the same height as the sidewalk.
A woonerf, which is surfaced with paving blocks to signal a pedestrian-priority zone, is, in effect, an outdoor living room, with furniture to encourage the social use of the street. Surprisingly, it results in drastically slower traffic, since the woonerf is a people-first zone and cars enter it more warily. "The idea is that people shall look each other in the eye and maneuver in respect of each other," Mr. Gehl said.

The CDS market, it turns out, is a bit like a financial woonerf. If you're buying credit protection from a vendor who himself has credit risk, you go more slowly, tread more cautiously. Rather than trusting the system designers to keep you safe and whizzing through green lights just because the rules say it's OK to do so, you actually spend much more time interacting with the other participants in the system and making sure that they have they will do in practice what they're supposed to do in theory.

And if the CDS market is a woonerf, the interbank market is an autobahn. Historically, did all the banks in the interbank market perform due diligence on each others' creditworthiness on a daily basis? Of course not -- just as when you're driving down the freeway at 80 miles an hour, you can't check to see whether the car in front of you has dodgy brakes. And the more you travel on the freeway without incident, the more you learn from that lack of incident, and the more confident you are that there won't be any problems. Which means that the distances between cars remain quite small, even at speeds significantly in excess of the speed limit.

And then, of course, one fateful day, there's a massive pile-up, and the entire freeway comes to a screeching halt, and the authorities have to mount an elaborate and expensive attempt to clear up the wreckage. After which people still are nervous about driving on that freeway.

A lot of people don't get this at all, Karen Shaw Petrou among them. She wants to cordon off the woonerf and stop anybody driving through it, despite the fact that it's the one part of the credit markets which is still actually functioning.

Regulators around the world should put a clicker into the CDS slots and hold speculative trades in it as is. Then, an orderly work-out of the $60 trillion market can be gradually implemented--without the forced selling and resulting liquidity cataclysm currently caused by desperate CDS traders covering "naked" bets made without the necessary backstop of actually holding the bond against which the CDS are pledged.

This wouldn't work: the whole reason why the CDS market works so well is the fact that if the markets move against you tomorrow, you can scale back your positions tomorrow. If you set today's CDS positions in aspic and stopped people being able to change them as the markets moved, the total losses would skyrocket.

But never mind the fact that Petrou's plan is unworkable: the bigger point is that it's based on a false premise. There was no "forced selling and resulting liquidity cataclysm" caused by "desperate CDS traders". People like Petrou think that there was: I've heard many times that Bear Stearns and Lehman Brothers were brought down by their CDS desks. But they weren't. Their CDS desks were actually functioning perfectly well, and as far as anybody can tell were making money all the while.

Yes, there is systemic risk embedded in the CDS market. And yes, it's a good idea to move CDS trading onto an exchange so as to minimize that systemic risk. But the cataclysmic chain of counterparty failures that everybody's so worried about hasn't happened yet: the proximate cause of today's financial meltdown was not the CDS market. And in fact there's a strong case to be made that the very visibility of the systemic risks in the CDS market was responsible for the fact that it so far hasn't failed. Banks were well prepared for the big obvious risks, including counterparty risk in the CDS market. They weren't prepared at all for the big non-obvious risks, like their super-senior CDO tranches being marked down by 40 or 50 or 60 cents on the dollar.

So by all means fix the CDS market, and make it safer. But bear in mind, too, that people drive faster when they're wearing seatbelts.

No comments: