In the sociology of "The Life of Brian," a world in despair is a world casting about for messiahs. Despair over the problem of big banks has some now reaching for the messiah of higher capital standards.
If banks had more capital, goes the argument, they wouldn't have needed bailouts. This is true—if you make an unwarranted assumption that the riskiness of bank assets would have been unaffected, or even improved, under higher capital requirements.As we discussed in previous posts, under higher capital requirements, both Merrill and Citi could have substituted super senior tranches of CDOs for Treasury securities and improved their return on equity without significantly changing their risk adjusted capital ratios.
... So the equation more capital = safer is not as straightforward as it seems.
Happily, the complexities are not lost on the gnomes of Basel ... Their latest solution, Basel III, would seriously hike the amount of capital banks must hold, but otherwise persists with the basic strategy of setting different levels of capital against different assets precisely to preserve the incentive of banks to invest in assets perceived as safe.
The fatal conceit, of course, is "perceived." Triple-A mortgage securities once were seen as safe. Greek bonds were safe. Under TBTF, when banks receive no discipline from their own creditors who expect to be bailed out, it falls on regulators not only to guess which assets are safe but to lean against the incentive of banks to categorize risky assets as safe in order to hold less capital against them.
How well regulators have performed this function can be guessed from a succession of global financial crises...
Nor is it pound-foolish to avoid banking panics at the cost of increasing moral hazard. But the fact remains: Everything we do ends up increasing moral hazard.Actually, everything that the regulators have tried so far increases moral hazard.
... Lacking, meanwhile, has been any willingness to grasp the nettle of moral hazard directly. We want a messiah, but apparently not one who asks anything radical of us or challenges us to relinquish any sacred cows.
All but dismissed, for instance, has been the idea of contingent capital—requiring banks to sell a bond that would convert to equity if the bank gets in trouble.
Forget the trifling argument over whether such convertibles should count as "capital." What matters is the creation of a class of tradable debt exempt from bailout, giving speculators an incentive to scour for early signs of trouble.The issue with contingent capital is that there is no readily apparent market for it. Without asset and liability-level disclosure, investors have no way of assessing the riskiness of what they are buying.
It is one thing to make an investment when you know what the bank's exposures are, but quite another to make an investment on what may or may not be there.
Or how about requiring government-insured deposits to be 100% backed by Treasury bills?We already have this. Since the beginning of the credit crisis, it is called a money market mutual fund.
Don't underestimate the importance of FDIC deposit insurance to the edifice of TBTF. If government were seen acting in advance to protect its own claims in a bank failure, and devil take the hindmost, it would have a powerful effect on the thinking of the hindmost.
Bank creditors might begin to doubt their own rescue and demand more transparency and less complexity from bank CEOs.Following the messiah of transparency comes market discipline and the government acting in advance to protect is own claims in a bank failure.
As this blog has discussed at length, market participants need to have access to all the useful, relevant information in an appropriate, timely manner. For banks, this information is current asset and liability-level data.
With this data, market participants can determine if a bank is solvent or not. With this data, market participants can determine how much risk a bank is taking and adjust the pricing and amount of their exposure to the bank accordingly.
No comments:
Post a Comment