How Bad Are Credit Default Swaps?

| Fri Jun. 17, 2011 2:00 PM EDT

Why is the European Central Bank dead set against even a "voluntary" restructuring of Greek debt that would force private investors to take a bit of a haircut on their holdings of Greek bonds? Matt Steinglass thinks their motivations are pretty much what they say they are:

The ECB doesn't believe it's possible to make private holders of Greek debt "share the pain" without precipitating a Greek default. They think if you pressure banks to roll over Greek debt, that debt will become untradeable, which is the same as "worthless"; the ratings agencies will deem the failure to pay at maturity to be a technical default, which may trigger credit default swaps; the Greek banking system will become insolvent, meaning nobody in Greece will have any money anymore; recapitalising those Greek banks will have to be done by governments that actually have money, ie the northern European ones; and ultimately the costs will all fall on the northern European taxpayer anyway. Meanwhile northern Europe's pension funds will be hit by the credit panic, which again will hurt the average citizen. The ECB folks sincerely think there's no way around having taxpayers pick up the bill for saving Greece and the euro.

I would really like to see a more detailed explanation of this. The basic idea here is that credit default swaps are (duh) triggered by a default: when you buy CDS on, say, a Greek bond, you're paying for protection against default. If the issuer of the bond defaults, then you're made whole by whoever you bought the CDS insurance from.

So far, so simple. It's often a little unclear exactly what triggers payment of a CDS, but it's perfectly plausible that even a modest restructuring, whether voluntary or not, would count as an "event" that would trigger lots of CDS contracts. But then what? What are the figures here? Just how much CDS is there on Greek debt? How much would be triggered by default? Who are the main sellers of CDS on Greek bonds? How big would the effect be if they had to pay off bondholders?

I'm curious about this for two reasons. First, I want to know if the ECB really has a good case. Would Greek default trigger a massive wave of renewed insolvency all over Europe as CDS sellers are forced to pay off on their insurance? Second, I want to know if this is yet another reason we should be wary of CDS. One of the big knocks against CDS has been that although it's theoretically a perfectly fine idea, in practice it can act as a huge multiplier, turning a bad default that hurts thousands of people into a catastrophic, systemic payout that hurts millions. In this case, it would turn tens of billions of dollars of Greek default (bad) into hundreds of billions of dollars of rolling CDS payouts (really, really bad).

But is this really the case? If this is what the ECB thinks, I'd like to see the detailed research to back it up. And if they're right, I think it's one more nail in the coffin of credit default swaps in general. If they really do magnify risk this way, it's time to do away with them. If you buy Greek bonds — or anyone else's bonds — maybe it's time to start doing due diligence again instead of just buying CDS and calling it a day.

UPDATE: Felix Salmon has a few numbers here that suggest CDS exposure is fairly small. However, his numbers also suggest that direct exposure to Greek default is fairly small, and in any case, not a problem for the banking industry. So this is still a bit of a mystery.