Hearing entitled "The Science of Insolvency"

[Watch the video of all testimonies (including Jeffrey Sachs, Simon Johnson, and David John).]

Thank you, Chairman Miller, for inviting me to testify before the subcommittee and to share my views on the problem of insolvency facing the U.S. banking system. I wish to make three main points in my comments:

  1. There is little logic to the claim that there is no market for “troubled assets,” since in fact such assets are being sold on a regular basis by the FDIC;
  2. There is little reason to believe that the current market prices for these assets are unreasonably low and that they will be selling for substantially higher prices in the foreseeable future; and
  3. If the major banks are fundamentally sound, as suggested by the recent stress tests conducted by the Fed and Treasury, then there can be little justification for the various forms of subsidies, such as the Fed’s special lending facilities, which allow banks to borrow at below-market interest rates.

I will address these issues in turn.

On the first point, it has been widely asserted that the central problem facing the banks is that they have large amounts of assets on their books that are not currently marketable due to the disruptions in national and international financial markets. I would argue that there is no obvious failure of the market. In fact, the “troubled assets” that the banks hold are being sold by the FDIC (among other institutions) on a regular basis, which must auction off mortgages and other assets from the banks that it has taken over in recent months.

There has been considerable confusion about the nature of the troubled assets held by the banks. While banks do hold some amount of mortgage-backed securities, these securities are in fact a relatively small portion of their troubled assets. In its analysis of the bank stress tests, the Fed reported that the 19 bank holding companies it examined collectively held only about $200 billion in non-agency mortgage-backed securities. Furthermore, not all of these securities were of recent vintage or backed by non-prime mortgages, so the amount of these securities that could reasonably be placed in the troubled asset category would be even less than $200 billion.1 The Fed estimated the losses on these assets in the more adverse scenario at $35.2 billion, less than 6 percent of the total projected loss in this scenario. By contrast, the losses on mortgages were projected at $185.5 billion, more than 30 percent of total losses, by far the largest single category.2

In short, the troubled assets on the banks’ books are overwhelmingly mortgages, both first and second or other junior liens, not mortgage-backed securities. The FDIC has acquired large quantities of mortgages from its takeover of several dozen failed banks over the last year. It auctions these assets off on an ongoing basis. The results of these auctions are available on the FDIC website.3 Non-performing mortgages typically sell in these auctions at prices in the vicinity of 30 cents on the dollar.

It is not clear on what basis these auctions can be said not to constitute a market. While the downturn and the constricted credit conditions affect the market, it is simply inaccurate to claim that there is no market for these assets. The major banks are undoubtedly not pleased at the prospect of having to sell off their loans at these prices, but this merely indicates that they are unhappy with the market outcome, just as a homeowner might be unwilling to sell her house at a loss. However, the unhappiness of the seller does not mean that there is no market.

The second issue is whether there is some reason to believe that the prices that these loans currently command is unrealistically depressed and that they will command a substantially higher price in the near future. On its face, there is little evidence to support this view.

Most of the loans that fall in this toxic category were presumably non-prime loans issued to buy homes near the peak of the housing bubble in the years 2004-2007. Most of these loans presumably went to buy lower-end homes in the most inflated bubble markets – places like Los Angeles, San Diego, Miami, and Phoenix.

In these cities, house prices have fallen sharply from their bubble peaks. The table below gives the decline in nominal house prices from their bubble peaks for homes in the bottom third of the housing market, as reported in the Case-Shiller tiered price index. The data show that prices of homes in the bottom third of several of these markets have already declined by more than 50 percent from their bubble peaks. In Phoenix, the most extreme case of the cities included in the Case-Shiller index, the price of houses in the bottom third of the market are already down more than 60 percent from their bubble peaks.


City Decline from Peak Date of Peak
Phoenix 65.9 percent July 2006
San Francisco 58.7 percent May 2006
Las Vegas 54.2 percent July 2006
Miami 52.6 percent March 2007
Los Angeles 51.7 percent February 2007
San Diego 50.2 percent April 2006
Tampa 46.4 percent July 2006
Washington, DC 46.0 percent June 2006

Source: S&P/Case-Shiller Home Price Indices (February 2009).


Furthermore, house prices are continuing to decline rapidly. Prices for homes in the bottom tier are falling at a rate of 3-4 percent per month in the Case-Shiller index. The most recent data in the Case-Shiller indexes was for February. (The data is obtained at closing. Since there is typically more than a month between when a contract is signed and when closing takes place, the February data primarily reflect market conditions in January.) It is therefore likely that the price of houses for homes in the bottom third of these markets are already at least ten percent lower presently (May 2009), than indicated in the February data.

In the peak years of the bubble, 2004-2007, it was common for homebuyers to purchase homes with little or no money down. If a mortgage written against these homes is now non-performing, and the house has lost 60 percent of its value, then it is very plausible that the current market value of this mortgage is 30 cents on the dollar or less, based on the underlying value of the collateral. The costs associated with carrying through the foreclosure are likely to take up a large portion of the proceeds from the resale of the house.

In fact, there have been several press accounts of instances where lenders have stopped carrying through foreclosures in some especially depressed markets.4 They have decided that the money from selling the home would not cover the cost of carrying through the foreclosure. In many former bubble markets or some very depressed non-bubble markets, such as Detroit or Cleveland, prices of 30 cents on the dollar may be high for non-performing loans.

There is little reason to expect prices to bounce back from current levels. The run-up in house prices in the years 2004-2007 was quite obviously an asset bubble. There was no obvious change in the fundamentals of the housing market either nationally or in the most affected cities that could have justified this increase in house prices. Furthermore, the increase in house prices was not associated with any remotely corresponding increase in rents. If the fundamentals of the housing market had been responsible for the run-up in house prices, then there should have been some comparable increase in rents in this period. Instead, real house prices were mostly fairly stable. The plunge in house prices in the last two and a half years is now bringing them back in line with rents. There is no obvious reason that house prices should turn around and go back toward their bubble peaks.

In short, the current market valuation of the banks’ toxic assets seems like an appropriate valuation based on the available evidence. It is understandable that the banks are unwilling to take large write-downs on these loans, especially if it will raise questions about their solvency, but there is no reason to believe that there is any real problem in the market for these assets. In short, in designing plans to relieve the banks of their toxic assets, the Treasury and the Fed are trying to fix a problem that does not exist.

The final point that I want to address is the role of the federal government in the bank bailout in the context of the recently released stress tests of the country’s 19 largest banks. There are serious grounds for questioning the usefulness of the stress tests, most obviously the fact that the economic assumptions in the adverse scenario are now are a relatively optimistic scenario given recent economic data.

However, what is more striking is that policy does not appear to be consistent with Secretary Geithner’s assessment of the stress tests. Mr. Geithner has indicated that the tests suggest that the banks are essentially healthy. While they showed that several banks would need to raise additional amounts of capital, with the exception of GMAC, it seems likely that the capital shortfall could be made up either through capital raised in private markets or by converting the preferred shares already held by the government into common stock.

If it is in fact the case that the banks can weather this crisis without further assistance from the government, then it is reasonable to ask why the government is continuing to provide extraordinary assistance. Specifically, if the banks are able to stand on their own, is there really a need for the special lending facilities that have been created by the Fed and have more than $2 trillion outstanding in loans to the banks and other institutions? The FDIC is guaranteeing several hundred billion dollars of bonds issued by banks in the last eight months and has authorized the banks to issue tens of billions of dollars of additional debt with a government guarantee.

The government continues to fund AIG to pay off counter-parties (mostly banks) who would have incurred large losses without the government’s intervention. And the government stands prepared to subsidize the purchase of as much as $1 trillion in troubled assets from the banks’ balance sheets through the Public Private Investment Partnership (PPIP) program.

These programs all involve substantial subsides from taxpayers to the banks. Arguably, such subsidies are necessary if the survival of systematically important institutions is at risk. However, if these institutions are essentially solvent, as Mr. Geithner suggests based on the stress test results, then it seems appropriate to put an end to these taxpayer subsidies, or in the case of PPIP, to cancel the program before it is put in place. There is an important public interest in maintaining a functioning financial system. There is no public interest in subsidizing banks. If the banks are able to stand on their own without further public assistance, then they should be given that opportunity.

1 The structure of the banks’ assets is discussed in Board of Governors of the Federal Reserve System, 2009. “The Supervisory Capital Assessment Program: Overview of Results,” pages 8-9, available at [http://www.federalreserve.gov/newsevents/bcreg20090507a1.pdf]
2 Board of Governors of the Federal Reserve Board, 2009, Table 2.
3 The FDIC website reporting the results of its auctions can be found at http://www2.fdic.gov/closedsales/LoanSales.asp.
4 For example, see “No Sale: Bank Wrecks New Homes,” Wall Street Journal, May 5,2009;A3.

Dean Baker is the co-director of the Center for Economic and Policy Research (CEPR). He is the author of Plunder and Blunder: The Rise and Fall of the Bubble Economy. He also has a blog on the American Prospect, "Beat the Press," where he discusses the media's coverage of economic issues.