We are flattered that a thinker of Nicole Gelinas’s standing and ability has taken note of our suggestion that the banks should be required to disclose their investments in detail, even if she notes it only to disagree.
We hesitate to disagree with her in return, not only because we are gallant fellows but because so much that she says in her piece is right.
First of all Ms. Gelinas is absolutely right that bank size is not the key issue. As she points out, it would hardly matter if all the banks were small if, as a result of conforming their behavior to regulatory standards, they all failed at the same time. Most of the banks that cracked from 1929–1933 were of modest size; that hardly reduced the net impact on the economy of their collective failure.
Score another one for Ms. Gelinas when she points out that regulatory requirements effectively compelling banks to concentrate in AAA paper were a huge factor in the creation of pseudo-AAA mortgage bonds. This is a point free market advocates need to make more often; we promise to do so from now on.
She is also right that our justification for requiring bank transparency — banks are quasi-public institutions — doesn’t really work if we include any firm managing more than $10 million in assets. The problem is where to draw the line without creating a significant advantage for firms just under whatever limit is chosen. We’ll think about that.
Ms. Gelinas is right again that news of high mortgage defaults began to come out years before the banks collapsed. We devote a whole chapter to this issue in our new, highly recommended (by us) book, Panic. The question, as she rightly points out, is why this news did not have a more powerful effect on the people who lend to banks early on.
There are probably several reasons. One was the “structure” of structured finance: Since 100 percent of early losses were absorbed by the lower-rated tranches, default levels could really go quite high before the AAA tranches were affected at all. In a sense, the refusal of markets to mark down AAA tranches in 2005 and 2006 must have seemed like confirmation that the scheme was working. As a result, the AAA indices traded at or near par well into 2007. They suddenly collapsed in the summer as the lower-rated indices headed toward zero, indicating that the cushion for the AAA tranches might be vanishing.
So there was some market-wide complacency about MBS in general. This was compounded, however, by the fact that no one knew which banks owned which MBS. Not all mortgage-backed securities were created equal. With diligent analysis — as both the successful short sellers and later bargain hunters showed — it was possible to sort the sheep from the goats long before the crash. Though that information was useful to direct investors in mortgage-backed securities or index derivatives it was not useful to bank bondholders or, more important, potential short-sellers of bank credit — because no one could tell which banks owned more sheep and which owned more goats.
The real key to the debate, however, is the now nearly universal belief, shared by Ms. Gelinas, that the roots of the crisis are to be found in the “moral hazard” created by the perception that the government would bail out the banks that were “too big to fail.” As a result, say backers of this theory, bank executives were emboldened to behave recklessly and bank creditors were complacent about the risks bank leadership was taking. If moral hazard explains the catastrophe, then the transparency solution we advocate is irrelevant.
It doesn’t. Not even close.
In the case of the bank executives, the moral hazard explanation is obviously implausible. Most top bank executives faced enormous personal losses in the event their own banks were to go under. Dick Fuld lost almost a billion dollars driving Lehman into the ground. He and his colleagues were not reckless, they were clueless; they really believed that structured mortgage securities were safer than old-fashioned mortgage lending. So did essentially every regulatory agency in the developed world.
Then what about bondholders? Why didn’t they cut off the banks? Can “moral hazard” explain their behavior? Not very much of it. It is true that the big banks were able to borrow a bit more cheaply because of their ordinary privileges. Access to the Fed as “lender of last resort” is a kind of ongoing bailout that advantages federally chartered banks in even ordinary times.
But bondholders’ contribution to the crisis was not that they leant to say, Citibank, at AAA rates when they should have been charging Citi at AA rates. The problem was that bondholders treated Citi as an A-class credit when it was really a C-class credit. No investor treats a C credit on A terms just because the government might bail out the C credit and, if it does, might possibly hold bondholders harmless. The mistake bank creditors made is just too big to be explained by moral hazard. It can be explained only on the theory that they did not realize the banks had become C credits.
The ultimate proof that “moral hazard” cannot explain the problem is to be seen in bank share prices. Shareholders don’t expect to be protected in a bailout, so “moral hazard” can’t explain their behavior. If moral hazard dominates creditor behavior but has no impact on shareholders, then the two groups should behave in radically different ways. They didn’t. Citi’s share prices held rock steady until October of 2007, by which time all the damage had been done. As Citi announced a series of write-downs, the share price slid from the 50s to the 20s over the next eleven months. Even then, however, Citi shares remained overpriced for a firm at the edge of bankruptcy until the general crash precipitated by Lehman’s fall.
Like the bank executives, bondholders weren’t reckless. Like shareholders, they were clueless.
Markets failed. Securities markets tend to be dysfunctional under the best of circumstances (though bond markets tend to be saner than stock markets). But to deny them vital information can only make things worse.
The job of reform must be to help markets be all that they can be and that means giving them the information they need.