Adverse Selection 101: The Real Reason Why Balance Sheet Detoxification Has Failed
The Treasury's efforts to detoxify banks' balance sheets have been a flop, to put it bluntly. Troubled Asset Relief Program (TARP) funds were initially earmarked to purchase banks' bad assets, but confusion about finding the right price led to the plan's abandonment. And now the much-touted Public Private Investment Partnership (PPIP), a program that would subsidize the private sector's purchases of toxic assets, has also foundered. The FDIC recently shelved its scheme to buy whole loans, and according to the latest news reports, the Treasury intends to scale back its plans to purchase troubled securities.
Two explanations have been given for these programs' failures. Prospective investors say they are put off by the possibility of capricious government interference and caps on compensation. And having passed the government's stress tests and raised additional capital, many banks are no longer feeling pressure to cleanse their balance sheets.
These factors are surely relevant. But there is also a third, more fundamental, reason for the programs' chilly reception: adverse selection. The essence of the problem is that banks do not want to sell their assets at fire-sale prices, while investors do not want to overpay. Buyers and sellers cannot agree on a price, and so no trades take place. This "lemons problem," a generic feature of any market in which the quality of the good is unknown, explains why any market-based detoxification plan—public or private—is likely to fail.
A stylized example will illustrate the problem. Suppose there are two kinds of collateralized debt obligations, or CDOs: nontoxic ones worth 100 cents on the dollar, and toxic ones worth 80 cents on the dollar. And, to make things really simple, suppose that half of the banks hold nontoxic CDOs with a face value of $10 billion, while the other half own toxic CDOs with the same $10 billion face value—but these are only worth $8 billion, of course. Suppose further that the CDOs are the only assets held by both types of banks, and both have deposits and other debt liabilities of $8 billion. Assume that outside investors (including the government) know that half of the banks are stuck with toxic CDOs, but they cannot tell whether any given bank is holding the toxic securities.
- The Treasury's dilemma will be clear once we answer the following three questions:
What is the true value of equity of the two types of banks, and what is the market value of the banks' equity (that is, what an outside investor is willing to pay for)?
- Suppose Treasury Secretary Timothy Geithner offered to buy any CDO from any bank for 91 cents on the dollar, and that banks want to maximize the market value of their equity. Will the owners of the toxic CDOs sell them to Geithner? Will the owners of the nontoxic CDO owners sell?
- Suppose Geithner, worried about paying too much for the assets, instead offered to buy any CDO for only 85 cents on the dollar. As before, assume banks maximize the market value of equity. Will anybanks want to sell their CDOs at this price?
Answer: Good banks are worth $2 billion, bad banks $0. With half of each type, the market value would be $1 billion.
Answer: If a bad bank sells its toxic assets to Geithner, it will see the value of its equity rise to $1.1 billion, regardless of what the good bank does. Understanding that the bad banks will sell at this price, good banks figure that Geithner's offer will entice the bad banks to reveal themselves for what they are. If so, then the good banks will hold on to their nontoxic assets and, once the bad banks fess up, the good banks will see their market value rise to its true $2 billion.
Answer: If bad banks sold their assets for 85 cents on the dollar, they would see their market value fall to $0.5 billion. They would therefore prefer not to sell, and thus maintain a $1 billion market value. Good banks won't participate either, and no assets will be sold.
Geithner's dilemma is this: If he offers a price at or above the market value, only the holders of the toxic assets will sell, in which case the Treasury (i.e., the taxpayer) will overpay for the bad assets and reward failure. But if he pays something closer to the toxic assets' true value, no bank will want to sell and the toxic assets will remain on the institutions' books.
Is it any surprise, then, that neither TARP nor PPIP has succeeded in luring banks into this particular scheme to detoxify banks' balance sheets? Surely not for students in my money and banking class—this example is taken directly from last semester's final exam. Treasury officials are welcome to audit my class when I teach it again this fall.
Kenneth N. Kuttner, professor of economics at Williams College, was a visiting fellow at the Peterson Institute for International Economics in 2005–07.