Contingent convertible bonds, a.k.a. contingent capital, are the latest fad to hit the optimistic technocracy in Washington and London. A contingent convertible bond is a bond that a bank sells during ordinary times, but that converts into equity when things turn bad, with "bad" defined by some trigger conditions, such as capital falling below a predetermined level. In theory, this means that banks can have the best of both worlds. They can go out and borrow more money today, increasing leverage and profits (which is what they want). But when the crisis hits, the debt will convert into equity; that will dilute existing shareholders, but more importantly it means the debt does not have to be paid back, providing an instant boost to the bank's capital cushion. In other words, banks can have the additional safety margin as if they had raised more equity today, but without having to raise the equity.

Tett is skeptical for all sorts of reasons — defining the trigger point (remember, Bear and Lehman were well-capitalized on paper when they collapsed), finding people willing to buy these things, the impact on the market of triggering a conversion, etc.

I'm skeptical for a more basic reason. Contingent capital, like any other type of capital requirement, assumes that we can predict in advance how bad the crisis will be and therefore how much capital will be necessary to avert a bank-killing panic. That means we have to be able to predict (a) just how fat the fat tail is, based on virtually no data points, and (b) how panicked people can get and for how long. That seems to me technocratic hubris of the first order.

So why is contingent capital so popular? (It's even mandated by section 107(b)(1)(D) of the Dodd bill.) Well, the people don't matter don't listen to me or to Gillian Tett. Here is Tett's explanation:

"Even amid all those hurdles, the CoCo idea currently has many fans, not just among investment bankers touting for business, but some western regulators too. The reason stems from a big, dirty secret stalking the financial world: namely that while global policymakers have spent a year wailing about the ‘Too Big to Fail' problem, they have hitherto done almost nothing tangible to remove that headache in a credible manner."

Tett says what we need is a cross-border resolution system for bank failures. I'm a little skeptical of that too, for reasons I think I've outlined elsewhere. In case I haven't, this is the problem: When push comes to shove, would the U.S. government use whatever "resolution" powers it has to take over JPMorgan Chase or Goldman Sachs against its will (or let an international body do so)? Leaving aside the issue of political connections for the moment, the political hit it would take from the right (SOCIALISM!!!) would make the health care debate look like a friendly game of flag football. If we can't even get meaningful derivatives regulation in 2009, what makes us think that any government would have the political capital to take over one of America's biggest banks when it needed to? More technocratic hubris.

But I agree strongly with Tett on why contingent capital is suddenly so popular. Policymakers in Washington are looking for something, anything that will allow them to declare victory over the TBTF problem — without having to break up the banks. The idea that any clever regulatory scheme we come up with today, which by definition will be untested, can be counted on to come through in the next crisis seems hopelessly naive to me. I think it would be more honest to admit that there are really only two choices:

1. Break up any institution that is too big to fail.
2. Leave them in place (because "big companies need big banks," or whatever other nonsense justification you want to use) and admit that we've done nothing to solve the TBTF problem.

That's the real choice.