Thursday, October 16, 2008

The Imperfect Information Excuse

Imperfect information plays a role in our economy. For example, a person diagnosed with a fatal illness may like to purchase life insurance for his family even if he had to pay an actuarially fair (thereby, high) price. But life insurance companies will not sell insurance to such a person because the cost of verifying exactly how fatal is his illness (expensive doctors and testing would have to be hired) exceeds the patient's willingness to be insured.

Imperfect information cannot stop a transaction with billions of dollars of benefits unless the costs of verifying that information is even greater than that.

Professor Kashyap and Diamond are blaming lack of troubled bank equity sales on imperfect information. Specifically, they say that a fairly strong bank would fail to raise equity because investors fear that the bank is actually quite weak (i.e., have some hidden troubled assets). They error both in fact and in logic. In fact, troubled banks have raised some equity from the private sector, which demonstrates that information problems have been overcome in this situation.

Regarding the logic, their story has some sense to it if it were a story about a small bank. But they are applying that story to the large banks (such as those receiving the "equity injections" from the U.S. Treasury) who have hundreds of billions of dollars in assets. Whatever private information those bank managers have could be made sufficiently public (or revealed to a single large scale investor) or sufficiently incentivized for far less than billions of dollars.

I suppose that the Professors would say that "it takes too long to make those discoveries." Where is the evidence for that claim? How many hours did the bank managers currently (and allegedly) holding the private information take to acquire it in the first place? Haven't some of these banks known for months that they would need more equity -- why weren't those months long enough to sufficiently publicize the private information? Why can't some of the managers in-the-know be made part of a new investment group and thereby given an incentive to blow the whistle? A lot of problems can be sufficiently solved in short order when there are billions at stake.

A more coherent story for the lack of equity sales is that the cost of capital is too high (many investors do not want to own bank stock regardless of what a detailed audit would show) and, even if it weren't, subsidized public capital is on the horizon and raising private equity would make them less eligible for the subsidy.


Fitz said...

Casey - You're right that troubled banks have raised some equity from the private sector, but the important question is whether this amount is near the optimum and, if not, the extent to which imperfect information has contributed to a suboptimal outcome. Secondly, another way to inflect the issue of imperfect information in this context is related to the concept of management credibility. A lot of "information" that management possesses could be characterized as subjective probabilistic beliefs about future performance; management typically tries to convince shareholders of such beliefs but tends to succeed only to the extent that management is regarded as credible. And a loss of confidence in management can raise the cost of equity capital to prohibitively high levels.

Sheng said...

I am convinced the markets expected a bail-out in Q1 and Q2. Think about what people are surrounding Paulson and Benanke these days: bankers! I never saw Paulson and Benanke sit down with a bunch of homeowners. Casey, why not do an interest-group analysis? I begin to appreciate Wall Street job is indeed the best job in the world.

Don said...

"and, even if it weren't, subsidized public capital is on the horizon and raising private equity would make them less eligible for the subsidy."

Bingo! Especially if capital is too high.

Don the libertarian Democrat

Tino said...

Last year the financial sector was the biggest dividend payer, with 77 billion of the 250 billion paid by the S&P 500.

Since the crisis their dividend payouts has been reduced by about 20 billion, “But financials are still the biggest dividend payers in the index, shelling out $60 billion annually, according to Howard Silverblatt, senior index analyst at Standard & Poor's.

(since the article came out Bank of America also cut it’d dividend, from 2.8 billion to 1.4 billion per quarter, or to "only" 6 billion per year)

The 9 largest banks who are getting tax-payer money are still paying out almost 30 billion per year in dividends.

The UK banks that were bailed out managed to lobby the government to relax their 5 year ban on dividends to a one year ban.

(Good data on historical dividend payouts of 500 largest major firms, although not by sector.)

Tino said...

I know most people will not like this, but I think part of the mass hysteria in regard to this crisis is politically motivated (consciously or through more subtle mechanisms). The media and the elites have an incentive to exaggerate any problems, in order to get their guy elected. Call it the Bush bad news multiplier.

Compare this to Iraq around 2006. The media declared the war lost. The foreign policy “experts” were collectively convinced Iraq was lost and the US could at best negotiate a retreat with Iran and Syria.

Things were bad in Iraq, but given the subsequent victories (as an example, half way into October the US has had 5 combat casualties in Iraq, the corresponding figure the same period 2006 was 68 combat casualties) it couldn’t have been as hopeless as described. It’s not just striking that the elite was so wrong about an important issue, but how confident they were, and how strong the consensus was.

Another problem is the asymmetric payoff of predictions these days. If someone is excessively pessimistic and continually predicts catastrophe no one cares if they were wrong. But people are afraid of being optimistic and wrong, if things go bad you will be lynched in the public arena. Worse yet, if you are one of all the experts that predict a crisis and you are wrong, the size of the herd alone is enough to shield you.

Look at Bernanke’s personal incentives. First of all he knows the history of the depression too well. He desperately want to be Benjamin Strong, regardless if it costs a trillion dollars, and will come out a hero who saved the US, regardless of whether there is any deep crisis or not. If there really isn’t any crisis we will not even know! They will just say the intervention worked.
Secondly no one will hold Bernanke accountable for the excess loss of the intervention, but everyone will hold him accountable for any cost of a crisis, so even if the probability is 1% he has an incentive to throw everything at it.

I am not knowledgeable enough to know if there really will be a depression or not without intervention. The problem is that the government and expert reaction gives us very little information about the severity of the problem.

With their incentive structure they would have panicked 100% regardless if the objectively probability of crisis was 1% or 50%.

Sheng said...

fitz, I will call it Bernanke minimizing Type I error (probability of not acting when the economy is truly in danger) regardless of how large Type II error will be (acting too much when the economy is not in danger)!