An Analysis of the Financial Crisis

University of Chicago economist Robert Shimer analyzes the financial crisis:

Let me explain what I think is happening in credit markets. This is my assessment, formed through numerous discussions with colleagues.... As everyone now knows, financial institutions hold significant assets that are backed by mortgage payments. Two years ago, many of those mortgage-backed securities (MBS) were rated AAA, very likely to yield a steady stream of payments with minimal risk of default. This made the assets liquid. If a financial institution needed cash, it could quickly sell these securities at a fair market price, the present value of the stream of payments. A buyer did not have to worry about the exact composition of the assets it purchased, because the stream of payments was safe.

When house prices started to decline, this had a bigger impact on some MBS than others, depending on the exact composition of mortgages that backed the security. Although MBS are complex financial instruments, their owners had a strong incentive to estimate how much those securities are worth. This is the crux of the problem. Now anyone who considers purchasing a MBS fears Akerlof's classic lemons problem. A buyer hopes that the seller is selling the security because it needs cash, but the buyer worries that the seller may simply be trying to unload its worst-performing assets. This asymmetric information ... makes the market illiquid. To buy a MBS in the current environment, you first need an independent assessment of the value of the security, which is time-consuming and costly. Put differently, the market price of MBS reflects buyers' belief that most securities that are offered for sale are low quality. This low price has been called the fire-sale price. The true value of the average MBS may in fact be much higher. This is the hold-to-maturity price.

The adverse selection problem then aggregates from individual securities to financial service institutions. Because of losses on their real estate investments, these firms are undercapitalized, some more so than others. Investors rightly fear that any firm that would like to issue new equity or debt is currently overvalued. Thus firms that attempt to recapitalize push down their market price. Likewise banks fear that any bank that wants to borrow from them is on the verge of bankruptcy and they refuse to lend. This is the same lemons problem, just at a larger scale. No firm that is tainted by mortgage holdings, even those that are fundamentally sound, can raise new capital.

With a theory of the problem, we can now ask whether the Paulson plan would solve it. My understanding is that the $700 billion would be used in a series of reverse auctions. In such an auction, the government would announce its intent to use some amount of money to purchase a particular class of security. Financial institutions would then compete by offering the most securities at the lowest price. I think we can agree that it is implausible that the government would be better than other buyers at determining the current value of the stream of payments from those securities. This gives financial institutions a strong incentive to sell the government their lowest quality securities at the highest possible price. Indeed, the government seems to want sellers to unload their worst assets so as to improve their balance sheet, so there really is no conflict of interest here.

This program does not solve the lemons problem. The government purchases a lot of lemons at an inflated price. This improves the balance sheet of the firms that can sell their worst securities. It also improves the balance sheet of firms that own better securities because the market price of those securities will increase. ... But this is fundamentally no different than giving taxpayers' money to owners, managers, and debt-holders of firms that made the worst decisions.