March 17, 2009

Idle Question of the Day

Exactly what bad consequences would follow if laws were passed by the relevant countries rendering credit default swap contracts void henceforth? (That is, canceling all the outstanding wagers because the bookies went bust.)

Update, 22 March: Well, one bad consequence would evidently be agreeing with Ben Stein. A bit from that link (by Felix Salmon, not Stein) is worth quoting:

There's a good chance, just for starters, that every major bank in America would go bust overnight: after all, they've been packaging up and selling off the credit risk on their multi-trillion-dollar loan portfolios [for years]. If Stein got his way, all that credit risk would suddenly reappear on the banks' balance sheets, and there's nothing they could do about it. Genius. Remember that those super-senior CDOs were the safest bits of the credit that they sold off. Just imagine what their balance sheets would look like if all the risky bits reappeared.

The issue he's raising is that if the banks can't say that they're covered for the risk of their loans defaulting (via the credit default swaps), they need to hold more capital as a protection against default. So as a legal or regulatory issue, ending the swaps would make the banks worse off. Substantively, however, this only makes sense if the swaps would, in fact, protect banks in the event of defaults — if they actually shifted the risk to the swap-sellers. Since we have just had pretty dramatic demonstrations that this is not something to be counted on, it's not at all clear to me that the banks ought to be able to keep that risk off their balance sheets. (In other words, the real value of the swaps to the banks is zero, or next to zero.) In any case, this objection could be countered by combining ending credit default swaps with public guarantees of the banks' existing positions — which is effectively what's happening anyway, only without making it harder to repeat the mistake in the future.

More broadly, ending credit default swaps would mean that those who sold such swaps would lose their stream of payments (a flow) but gain back their collateral and reserves (a stock); conversely buyers of default protection would gain a cash flow but take a hit to their capital stocks. Right now one imagines that even those selling the "end of the world trade" might prefer to get out of the game; I'd be interested to see an estimate of the effects of this on the stability of the financial world right now.

There is also the possibility that eliminating the swaps would deprive us of information about how risky different debts are. The value we should place on this, however, depends on how well these markets actually succeed in aggregating information about risk. I'd say there is abundant cause for skepticism about this — especially when things are, in fact, dangerous. Economic theory does not, in fact, provide any reason to think that such markets will be dominated by those with the most accurate beliefs (or even that the market as a whole will be more accurate than the best-informed trader), unless you assume a complete set of markets, which is a reductio ad absurdum if ever there were one. (When markets are incomplete, more markets are not necessarily better.)

To be clear, I am not asserting that credit default swaps should be ended. I honestly don't think I know enough to have an opinion about that, and while I'm obviously skeptical about their value, some serious and credible people (i.e., ones who do not have a vested interest in the matter) who've studied them in more depth see merit to them. If this is what a world with efficiently-allocated risk looks like, though, I'd hate to see a messed-up one.

(Thanks to readers D.H. and son1 for comments and pointers.)

The Dismal Science; The Continuing Crises; Modest Proposals

Posted at March 17, 2009 22:22 | permanent link

Three-Toed Sloth