Tuesday, May 20, 2008

CDS Counterparty Risk and the Bear Stearns Bailout

(Felix Salmon) Jesse Eisinger points me to David Evans's 4,400-word Bloomberg article on credit default swaps; he likes it a lot. Me, not so much. If it were shorter, it wouldn't bother me so much. But if you're writing at that sort of length, you should really take advantage of the opportunity to get things absolutely right, and not fudge complex issues.

The piece starts out with an anecdote about Tim Backshall, a Walnut Creek CDS trader worried about counterparty risk in the days leading up to the Bear Stearns bailout. And it loses no time in painting that bailout as the result of a Fed fearful of precisely that counterparty risk:

"There's always the danger the bank selling you the protection on Bear will fail,'' Backshall says. If that were to happen, his clients could spend millions of dollars for worthless insurance...
The Fed bailout of Bear Stearns on March 17 was motivated, in part, by a desire to keep that sword from falling, says Joseph Mason, a former U.S. Treasury Department economist who's now chair of the banking department at Louisiana State University's E.J. Ourso College of Business.
The Fed was concerned that banks might not have the money to pay CDS counterparties if there were large debt defaults, Mason says...
The Fed was worried about the biggest players in the CDS market, Mason says. ``It was a JPMorgan bailout, not a bailout of Bear,'' he says.

Now I'm sure that the Fed was worried about counterparty risk in the CDS market when it orchestrated the Bear bailout. But to read Evan's story, the big worry was something like this: Bear goes bust; other banks have to pay out a lot of money on CDS protection they've written on Bear's debt; they don't have the money to do that; chaos results.

Really? Does Evans really think that JP Morgan had written so much unhedged credit protection on Bear Stearns that it couldn't pay out on its obligations were Bear to go bust? That's the message he's sending, but I don't buy it for a minute.

The thing that the Fed was worried about was not the CDS written on Bear, it was the CDS written by Bear. If Bear were to go bust, then no one would have a clue how to value all the credit protection that Bear had written. Overnight, all the banks who thought they were hedged (they'd bought protection from Bear, sold it to someone else) would find themselves with some large and (worse) impossible-to-calculate net exposure. Eventually, inevitably, the Fed and the Treasury would pay or cajole a major Wall Street institution into taking on Bear's CDS book - after all, the CDS desk at Bear was not one of the areas losing money, and there would probably be quite a few financial institutions interested in buying it. But in the interim, coherent risk management would be all but impossible.

Much worse, of course, would be if there was a major default during that interim period. If Bear wasn't able to pay out on its obligations, then there could be an extremely nasty domino effect, leaving other institutions also unable to pay their obligations. But note that in order for that to happen, someone else - other than Bear itself - would have to go bust: Bear didn't write protection on itself.

Evans also never gets into the crucial question of recovery value. He writes that the CDS market works

as if many investors could buy insurance on the same multimillion-dollar home they didn't own and then collect on its full value if the house burned down.

But of course people who buy protection don't collect in full at all. They receive the difference between full value and the cheapest-to-deliver security - and as Alea points out, if the recovery value is high, then CDS buyers can lose money even in the event of a default.

Suppose that you bought protection on Marconi at 250 bp sometimes in the late 90s, in 2001 the spread had widened to 4000 bp. You are rich or so you think. The unwind price depends entirely on recovery assumptions, at 30% recovery you would get 51% of par, at 99% you would get 1% of par,that is you would "lose" even though you got the scenario right.

This is where I part ways with Jesse, too. He quotes with approval this passage of Evans's piece:

For traders who bought protection swaps just a few days earlier -- when prices were in the 600s to 800s -- the Fed bailout is crushing. Their investments have turned to dust.

It's a bit weird to consider buying credit protection an "investment" - it's more insurance than investment. But in any case the key question isn't where the spreads were, but the value of the CDS contracts themselves. And while Evans can call those spreads "prices", they're not prices at all: the relationship between spreads and prices is very complex indeed, and not nearly as simple as it is in the bond market.

So when Jesse says that the Fed's bailout of Bear "harmed prudent buyers of insurance," I wonder what he's talking about. I might be a prudent buyer of earthquake insurance, but that doesn't mean I'm harmed if someone manages to avert an earthquake. (Hey, it happens, I'm sure I saw it in a James Bond movie once.)

More to the point, a genuinely prudent buyer of insurance would have been insuring something: Bear Stearns liabilities which he owned and which would have plunged in value in the event of a default. Yes, if Bear had defaulted, the value of his credit default swaps would probably have risen. But the value of his Bear Stearns bonds would have fallen. So it's a bit weird to say that such an investor was harmed by the bailout - if anything, it was those investors who the Fed bailed out more than anybody else.

Evans continues in this vein. The CDS market "is larger in dollar value than the New York Stock Exchange," he writes, for all the world as though comparing notional principal amount in a derivatives market against actual tradable value in a stock market is an apples-to-apples comparison. It's not, and I get annoyed when space-constrained columnists make that mistake. When someone has over 4,000 words at their disposal, there's no excuse.

As a result, the good points that Evans makes - and there are some - get devalued. Yes, it's worrying that hedge funds, which aren't generally required to post collateral against the CDS protection that they've sold, have written 31 percent of all CDS protection. Yes, it's also worrying that no one really knows where most of the risk lies in this market: everybody simply hopes that everybody else is, more or less, hedged.

But Evans seems less interested in the nitty-gritty of such matters than he is in random color:

The night of Thursday, March 13, Backshall can't sleep. He lies awake worrying about Bear and counterparty risk. The next morning, he arrives at work at 5 a.m., two and a half hours before sunrise.
Through the window of his ninth-floor corner office, he takes a moment to watch the distant flickers of light in the rolling foothills of Mount Diablo. Across the street, he sees the still-dark Walnut Creek train station, about 30 miles (48 kilometers) east of San Francisco.
Backshall, wearing jeans and a blue, button-down shirt, sits at his desk, staring at a pair of the 27-inch (68.6- centimeter) monitors that display swap costs.

It's worth noting that 5am in Walnut Creek is 8am in New York, a perfectly normal time to start. But I do like the bit about "the distant flickers of light in the rolling foothills of Mount Diablo," it makes Walnut Creek seem almost beautiful. (It isn't.)

Evans ends with an ominous warning about counterparty risk:

The sword of Damocles will remain poised to fall, as banks, hedge funds and insurance companies can only guess whether their trillions of dollars in swaps are covered by anything other than darkness.

It's poetic, and one can forgive him for not being empirical, since it's impossible to quantify counterparty risk. But I would at least have liked Evans to draw a distinction between the risk of a broker-dealer going bust, on the one hand, and the risk of a hedge-fund protection seller going bust, on the other. Both constitute counterparty risk, but the way that each of the two cases might play out are decidedly different.

Update: Alea, in the comments, points out that Evans's piece spends much time on Backshall, and then bemoans the fact that the CME's nascent CDS exchange hasn't worked out so well, without mentioning the very close relationship between Blackshall and the CME. That's a disclosure Evans really should have made.

No comments: