Add Kevin Drum to those who think that a bit of CDS demonization is not such a bad thing at all. Unfortunately, he’s a bit shaky on the facts:
It’s a little hard to get a handle on an exact figure, but within the U.S. banking system total losses on subprime mortgages themselves probably total around half a trillion dollars. That’s a helluva lot of money, but it’s nowhere near enough to crash the system. That can only happen if the losses are magnified several times over via derivative losses.
This is wrong on a number of levels. For one thing, mortgage losses of half a trillion dollars are enough to crash the system. Remember that banks, by their nature, are leveraged institutions: they only need $8 of capital for every $100 of loans they extend, and the rest of the money can be borrowed. That’s the regulatory minimum: they can’t go below that. But when losses arise, they come straight out of that precious capital. And if you have half a trillion dollars of capital wiped out in the space of a few months, that can take you dangerously close to the regulatory minimum very quickly.
Now layer on top of all those losses the fact that we’re entering the deepest recession in living memory, and that defaults and loan losses on banks’ non-mortgage assets are sure to rise sharply from an uncommonly low level. That alone would be enough to rock many financial institutions back onto their heels; combined with the fact that they’re already very weak after being hit by mortgage-related charges, and you have more than enough to knock them out.
More to the point, however, derivatives didn’t magnify any losses at all. Name one bank — just one — which has taken any significant charges from its credit derivatives operations. It hasn’t happened. (Soc Gen lost a lot of money on stock index futures, which are derivatives, but they’re exchange-traded — which means they’re supposedly safe — and the money was lost due to a rogue trader breaking the law.)
It’s not surprising that CDS desks haven’t lost a lot of money, because CDS, like all derivatives, are a zero-sum game. That’s why Nathaniel Baker concentrated on monolines defaulting on their CDS obligations: that kind of counterparty event is the only way for absolute CDS losses to materialize. But as I pointed out yesterday, unsecured monoline CDS exposure is a tiny fraction of any bank’s balance sheet — in the case of Bear Stearns, it was just $330 million, and was frankly the least of the bank’s worries.
And what about those notorious super-senior CDO tranches? Aren’t they, fundamentally, credit default swaps, wrapped up and tranched out in seemingly clever ways? Yes. But remember what the banks did here. They took an enormous amount of mortgage debt onto their balance sheet, and then bought protection against that debt going sour: the banks used the cash flows from the homeowners to insure themselves against default.
Now there are two ways this deal can go wrong when the mortgage market implodes. The first is if the people you bought protection from are unable to pony up the cash. That’s counterparty risk, it’s something the banks were pretty much on top of all along, and so far it hasn’t been a big deal, thanks to margining requirements. But the second way that the deal can go wrong is if you’re not fully hedged: your counterparties will make up any initial losses that your portfolio suffers, but their maximum payout is capped at a level less than the total amount you paid for the mortgages in the first place.
That’s what happened with the notorious super-senior CDO tranches: they represented mortgage payments which were thought to be so safe that no one ever conceived they might default. So while the banks were hedged on their initial losses, they found themselves taking hits to their own balance sheets that none of their models had ever anticipated.
To put it another way, if it wasn’t for all of the credit protection that the banks had bought, the magnitude of their losses would be vastly greater, not less, than what they eventually suffered. Thanks to the wonders of default protection, the banks are much better off now than they woud have been without it.
Of course, it’s not quite as simple as that: if it weren’t for CDS technology and the ability to hedge their downside risk, the banks would never have taken all those mortgages onto their books in the first place. That’s what I meant when I said that CDS made it easier to lever up. But in no way does it make sense to say that derivatives magnified the banks’ mortgage exposures. The way that Kevin puts it, he makes it sound like there were half a trillion dollars of mortgage losses, and then trillions more in derivatives losses on top. And that just isn’t true.
Maybe bankers would have found some other way to lever up. But in the event… they used CDS. So why then is it unfair to say that CDS exposure was a huge driver of the financial meltdown?
It’s unfair because net CDS exposure, as we saw with Lehman, is generally de minimis. CDS exposure isn’t the problem. The problem is exposure to real-world credit, such as mortgages, or the debt of financial institutions like Lehman or Kaupthing. When Lehman defaulted, the real-world repercussions were so great that one of the biggest money-market funds in the world broke the buck. The CDS-market repercussions were almost invisible to the naked eye, for all that lots of pundits spilled much more ink on them than they did on the actual default. Bondholders holding some $100 billion of Lehman debt have been left with pennies on the dollar; the payout from writers of default protection on Lehman totaled about 5% of that figure.
In other words, the driver of the financial meltdown was good old-fashioned credit markets: not the CDS market at all. A few buy-side investors, to be sure, lost a lot of money selling default protection, among them the two Bear Stearns hedge funds which went bust in the summer of 2007 and really got this credit crisis going. But financial institutions, with the exception of the monolines and AIG, were not net sellers of default protection, and therefore did not lose money on CDS when spreads started gapping out and people started defaulting on mortgages. Insofar as banks have lost money, they’ve lost money on real-money loans to real-world individuals and companies. They have not lost money by speculating in the CDS market.
Salmon’s larger crusade is to defend dervatives in general, and CDS in particular, as useful devices when they aren’t abused, but it’s not clear to me that this is especially controversial. The question is, how should they be regulated in the future to ensure that they aren’t abused? I agree that leverage itself should be the primary target of regulatory reform, but surely, under the circumstances, some reasonably strict trading rules on derivatives of all kinds will end up being part of that. Leverage is a hard thing to get at directly, after all, and we’re going to need to attack it from a variety of directions. A bit of CDS demonization might not be a bad place to start from.
The first question is to determine, in a sober and responsible manner, whether CDS were abused. So far, I’ve seen little if any indication that they were. The second question is how to regulate the amount of leverage that banks take on. And far from being "a hard thing to get at directly", that’s actually very easy to measure: you just look at how big the bank’s balance sheet is, and how much capital it has to support that balance sheet.
To be sure, hedge funds and other shadow financial institutions can use CDS to replicate a bank’s balance sheet without regulation, and that’s a problem. So maybe hedge funds should be regulated — although now that their cost of funds is much higher than that of the banks, that might no longer be necessary.
As far as the banks are concerned, most of the losses they’ve taken have come straight out of their balance-sheet assets, or maybe out of SIVs which once were off-balance-sheet but which now are much more transparent. Occasionally you’ll find some weird and wonderful instrument like the liquidity put, which seemingly comes out of nowhere to saddle a bank with enormous losses. But the liquidity put had nothing to do with CDS, and regulating derivatives would have done nothing to prevent it from happening.
This is why the CDS demonization meme is dangerous: it’s basically the financial equivalent of all the security theater at airports. Regulating CDS might trick the public into feeling safer, but it won’t do any real good at all. And if there’s one thing we’ve learned over the course of this crisis, it’s that the more we think we’re safe, the riskier things actually are.