Two big points on Financial Reform:
As I've mentioned before, the idea that you can avoid too-big-to-fail by splitting up banks is stupid, because there's not a substantial difference between bailing 1 of 5 big banks or 3 of 15 medium size banks, so if there's a systemic shock and all the banks are in trouble, you're going to have to engineer a systemwide bailout anyway. The way to avoid those types of bailouts is to reduce the interconnectedness of banks, not the size of them.
Taking that idea, this article leads to a dangerous conclusion. Splitting up too-big-to-fail may actually INCREASE moral hazard in some circumstances. If one bank acts irresponsible but the others are fine (a variant of the Long Term Capital Management collapse), then splitting big banks makes it worse, but in avoiding system-wide mania, a larger number of banks means they'll ALL go along with it. 1 large bank may be able to ignore systemic pressure and do what it thinks will work (large portions of Goldman, for example), because if it's wrong and everyone else is right, they'll know they have a bailout... so you end up in a situation where everyone does what they think is best and the "too-big-to-fail" moral hazard actually increases system stability by making the banks independent. A small bank that can't count on a bailout if it screws up may be better off just following everyone else cuz they win if they're right (as they always do), but if they're wrong, there's way more safety in numbers. Thus, the systemwide effects of too-big-to-fail can potentially be the reverse of the individual level effects. Increasing the risk of one independently stupid bank to reduce the risk of a stupid system seems like a smart tradeoff every time, and should require less public capital.
2) Speaking of moral hazard, I question significantly whether it's stability-enhancing to enforce moral hazard in bad times if it wasn't clear if you would do so in prior good times. Certainly, Lehman is an empirical example of that, but the few people out there who defend the Lehman bailout argue that in the future, it shows banks that we'll let them fail (though I'd argue that the reaction to the Lehman death has been so bad that it has just cemented for the future the idea that we CAN'T let big banks fail, so I actually think the LEH bailout actually made moral hazard risks worse). The interesting answer, as usual, is the mechanism.
Bill Miller (of Legg Mason) has argued that government's harsh treatment of shareholders in some troubled institutions (in the name of preventing moral hazard) led to an unwillingness by shareholders to provide equity capital to other troubled institutions, functionally guaranteeing that the government had to intervene and provide capital to other banks. In other words, enforcing moral hazard in bad times had massive negative externalities - it cratered the equity base of every other financial institution amidst tremendous uncertainty about which financial institutions were troubled and which weren't. This is something that, to my knowledge, no academic economist has ever examined.
This problem is certainly most acute for banks, where the strength of their equity capital base is literally a requisite for survival. However, more generally, it should apply to any industries that need to raise capital. The government's constitutionally-questionable wipeout of a number of GM's creditors is going to make shareholders a lot more jittery when providing capital to GM, Ford or Chrysler (the latter, of course, in a private equity context, or after a re-IPO) in the future. These three companies are still significantly troubled, with Ford being the only one to have any sense of direction (but still massive problems with image, quality, cost control, organizational structure, etc). They're all interconnected as much as the banks are, by relying on a common group of suppliers, many of whom would go bankrupt if any of the three went under. Functionally, that's 2 million jobs tied up in three car companies whose survival in the medium run relies on each other. Enforcing moral hazard on GM's creditors will make the automakers significantly less stable in the future and more likely to require government transitional help, especially under a Democratic or populist administration (moral hazard was less the aim with GM and more a side effect. The original aim was coming to an agreement with the unions, so administrations more likely to make concessions to unions over capitalholders will be viewed with more skepticism by capitalholders).
It's much easier to enforce moral hazard when nobody needs to presently raise any capital, because then companies can adjust. In periods of capital raising, enforcing moral hazard may cause more trouble than it solves. Of course, you also run into the problem of self-fulfilling prophecies - investors and creditors think that the government will bail out a bank if it gets into trouble, so they never divest and the bank requires a much smaller (or no) government bailout. The possibility of moral hazard is also useful there.
I wonder if the way to exploit that would be a focus on the principal-agent problem - an understanding that creditors and shareholders will be fine, but management who were around during the misbehavior will get REAMED. Again, that's not something you can enforce after-the-fact, for a number of reasons - respect in contracts, retaining management talent, higher costs of uncertainty, etc - but before the fact, management can act accordingly. The problem is to prevent a management merry-go-round in bad times (Firm A and B both have problems, so the CEOs swap companies and are "newcomers" at their new companies and so don't get reamed), but one would think that mechanisms exist for dealing with that.
No comments:
Post a Comment