financial system, capitalization, etc.
Some blokes* presented a paper at Jackson Hole last month with the idea of having banks buy insurance against a financial systemic risk; this wouldn't give them extra capital if they got themselves in idiosyncratic trouble while the rest of the financial system was holding up, but if it was determined that there was a systemic crisis these policies would infuse banks with a lot of extra equity capital. They went through various contortions to avoid defaults on the insurance policies and to deal with the sheer scope and correlation of these things -- you have hundreds of billions of dollars in insurance claims that, by their design, will either never pay out or will all pay out at exactly the same time -- that seemed to me to be a complicated way of issuing catastrophe bonds, but with extra transaction costs. I don't know why the banks don't issue catastrophe bonds that knock out on the relevant trigger, suddenly moving from the liabilities section of the balance sheet to the equity section. I suppose those would have idiosyncratic credit risk if the bank went under before they triggered, but I can't imagine that's a big impediment to their use for this purpose.

Anyway, I didn't read past about page 10 of the paper, and maybe I should have, but the tricky thing in my mind to this sort of thing -- through insurance or a knock-out liability -- is defining the trigger. You could concoct something around a sustained rise in a swap spread, I suppose; what the regulator might prefer is that the regulator simply be able to declare an event, and that that would be it, but if you're going to actually place these risks with investors, the regulator is going to have to have reasonable credibility with them that he isn't going to declare an event because GDP comes in below 2%; there needs to be a sense that the regulator won't abuse this discretion, but also that what the regulator sees, in good faith, as systemic crises isn't going to include a broader set of possible outcomes than what the investor sees that way, or than is necessary or appropriate. (Regulators are, on some level, paid to be anxious.) Presumably such credibility could be developed; I don't know whether Bernanke has it now. If such a system weren't simply still-born, though, due to this problem, presumably some modus vivendi would develop in which what was priced in by investors was about what regulators actually used as their criteria, but it could well take a long time for things to get to that point.

I wonder, though, whether something could be structured this way to protect an individual financial firm, whose operations depend on its ability to find counterparties to deals -- counterparties who have to trust its ability to make financial commitments. A loss in creditor confidence is a problem for a manufacturer which, even if solvent, might have to be creative in handling its finances while it tries to pay down debts for a while out of its operating profits. For a financial firm, however, a loss in creditor confidence shuts down the operating profits as well. I was suggesting (before the credit crisis began, in fact) that banks develop a subordinated debt cushion, possibly with PIK coupons that would allow the firm to give investors stock warrants instead of coupon payments; the idea, ultimately, is that financial commitments made in the course of operations, which are at a senior unsecured level, would have enough of an extra buffer above them that the firm could continue to make money from operations while trying to restructure its finances. What would be really nice, though, is if there were a way to abruptly repudiate some debt when things are going badly without threatening the standing of the operations or senior unsecured debt. This creates moral hazard issues, and I suppose is ultimately the purpose of equity per se; the large subordinated debt cushion itself would hopefully give way to a relatively nondisruptive pre-packaged bankruptcy if it became clear that the notional debt, even in its subordinated form, was preventing the firm from doing business. (Giving existing shareholders warrants on the post-bankruptcy equity might even be possible in such a situation, if the bankruptcy were likely to improve operating cash-flow to the point that it heavily benefited the subordinated debt holders.)

There may be a certain length of paragraph that should indicate to me that I'm probably babbling.

Update: *The paper to which I refer is called "Rethinking Capital Regulation", by Anil Kashyap and Raghuram Rajan at the University of Chicago and Jeremy Stein at Harvard. I have a copy on my computer, and it's probably on the web somewhere.

Labels: financial catastrophe bonds

- posted by dWj @ 12:22 PM