Streetwise Professor

I have deep reservations about the Paulson-Bernanke bailout proposal. In a nutshell, they propose to buy “illiquid” assets from struggling financial institutions. This raises the question: Why are these assets illiquid? When you answer that question, the entire case for the bailout becomes dicey at best.

Illiquidity is first and foremost the result of asymmetric information–a lemons problem. If the owner of an asset has better information about its value than potential buyers, the buyers are very reluctant to make bids for them. They realize they face a winner’s curse: if the owner accepts the bid, it is because the bid is above the owner’s estimate of value. Now, in current circumstances, owners desperate for cash to survive may be willing to sell assets for less than their (relatively well informed) estimate of value. Thus, the lemons problem/winners curse may not cause a complete market failure with zero trading volume (as would be the case if there were no non-informational reasons to trade.) But severe information asymmetry drastically reduces the liquidity of the market for these assets.

Under the bailout proposal, the government will become the buyer of the assets. Regardless of the design of the mechanism for purchasing the assets, it is likely that the sellers will have a substantial information advantage over the government. Thus, the government is likely to suffer from a substantial winner’s curse problem, and end up overpaying for the assets it does buy.

This is not to say that current holders of mortgages and mortgage CDO’s have a very accurate estimate of the true value of these things. It is just likely that they have a better estimate than the government, or most potential buyers. The current owners have had an opportunity to do research on the underlying pools of mortgages. They have also observed the payment experience on what they have in their portfolios. As a result, they have some idea of which of their holdings are (relatively) good; which are bad; and which are truly ugly. The one eyed man is king in the land of the blind. The owners of these assets may have only one dodgy eye, but the government is likely to be nearly blind–so bet on dodgy cyclops.

Substantial competition between participants in an auction can mitigate the adverse impacts of asymmetric information. I am skeptical, however, that there will be a lot of competition here. The assets are so heterogeneous that the competition in the auction for any particular security (and that’s what matters) is likely to be subdued.

Another matter of concern is that although the objective of the auction is to facilitate the recapitalization of financial institutions, there are doubts as to whether the proceeds from auction sales will flow to the institutions that face the most daunting obstacles in recapitalizing in the market through sales of equity or subordinated debt. It is likely true that more desperate institutions are more likely to want to participate, and may accept lower prices for their assets. This would mean that such firms would be the most likely sellers, which would in turn mean that they would be disproportionately represented among the auction winners. Nonetheless, a good portion of the auction proceeds will likely flow to canny asset owners that could probably survive (or recapitalize) without these proceeds. Thus, the government may end up overpaying for assets, and the overpayment may do little to restore the balance sheets of the institutions that really need the help.

So, the bailout aid may not go to where it delivers the most bang for the buck. Paulson and Bernanke appear to be hoping that the bailout will inject liquidity into the market for heretofore illiquid mortgage securities. From the best I can gather, they think that these assets are selling at deep discounts to their actual value due to their illiquidity. Make the market more liquid, asset prices go up, and everybody’s happy.

I really don’t see how this will work. If the illiquidity is due to adverse selection/lemons problems, private buyers will still be very reluctant to purchase those assets the government doesn’t buy. So, in the end, the likely outcome is that (a) the government will overpay for a lot of assets due to the “winner’s curse”, (b) some of the overpayment will go to institutions that really need the capital, but some will go to to those that don’t, or who could raise it on the market, and (c) the market for problem assets will remain illiquid, with the government being the only buyer. That doesn’t seem to be the wisest way to spend $700 plus billion.

The lemons problem arises because of asymmetric information, but also because the seller has the ability to say no to the buyer’s offer. The seller says “yes” when the bidder overpays, and says “no” when the buyer underbids. The only real way to avoid this problem is to deprive the seller of the discretion to refuse offers. That is, Don Corleone (“I’m gonna make him an offer he can’t refuse”) is less subject to the winner’s curse than buyers who employ less coercive means. So, the best way that the government has to avoid the lemons problem is to approach financial institutions and say: we’re paying you X for Y. Your options are “Yes” and “I agree.”

That, of course, presents a whole slew of legal, constitutional and ethical issues. Moreover, there is still the question of to whom the government should extend these tender offers. Presumably, institutions more vulnerable to failure, and whose failure would impose more stress on the system.

But that’s in essence what the government did in the case of AIG. It decided to intervene with AIG because of its central position in the nexus of credit derivatives contracts, and because of its impending failure. It gave AIG a near ultimatum.

That seems a much better approach than playing Uncle Sucker willing to buy anything somebody’s willing to sell them. The seller is likely to have information advantages, and the payments made to winning sellers may not go to where its effect on the stability of the system would be greatest. Moreover, I am skeptical that the government’s purchase of bad assets will suddenly improve the liquidity of the private market for these assets. Instead, focusing on particular companies that are most important to the stability of the system, and which are having the greatest difficulty in financing themselves via the market, and sharply limiting the discretion of those companies to accept the terms offered, is a better way to target government assistance and to avoid the winner’s curse.

How could this restore market confidence today? If it is perceived to be a credible policy. How to make it credible? One thing that would enhance credibility would be for Congress to approve legislation and funding for such a program. This should include, of course, oversight and accountability provisions.

As ugly as they were, the Fannie, Freddie and AIG interventions make more sense than trying to create liquidity in the market for low quality mortgages through the mechanism of a government bailout. If illiquidity reflects the fundamental nature of these assets, and the distribution of private information regarding their value, the purchase program will not restore liquidity in a cost effective way.