The U.S. government, faced with the country’s worst financial crisis since the 1930s, has announced a $700 billion plan to purchase bad mortgage assets from banks. Is this a good bailout or a bad one? The difference is important, because history shows us that bad bailouts can actually make financial crises worse.

The Japanese government’s handling of its massive bank crisis in the 1990s is an example of bad bailout policy. At first, the Japanese government ignored the problem it was facing, then it decided not to enforce its own rules requiring sufficient bank capital (“forbearance”), then it began investing government funds into insolvent banks through preferred stock and subordinated debt. This caused the crisis to only grow larger, which contributed greatly to Japan’s poor economic performance during the 1990s and for several years thereafter.

A good bailout focuses on liquidity. The government acts appropriately when it helps institutions with liquidity at a time of crisis by lending against collateral or by purchasing bank assets at fair market prices. A bad bailout is one that tries to help with solvency. The government acts badly when it invests money directly into failing banks without closing them. If a bank — even if there is more than one — is actually insolvent (i.e. its liabilities exceed the value of their assets), then best practice dictates that it should be closed.

Closing a bank is less draconian than it sounds. A closed bank is not simply blown up or thrown away; on the contrary, every effort is made to preserve the bank’s franchise value and maintain continuity with customers and employees. Normally, the government cleanses the bank of its bad assets after closure and transfers the cleaned-up business to new owners as rapidly as possible.

Good bailouts wipe out the shareholders of insolvent banks and dismiss their senior management. Why? Because these are the people who created the problem, and they must be seen to pay a high price. Remember that most banks are conservative, well-run and solvent; only a minority get over-extended.

The problem with the Japanese banks in the 1990s was not just the bubble economy of the 1980s but a continuing unwillingness by banks and the government to acknowledge bad lending practices and change them. A few banks were eventually closed; however, one of these names, Long-Term Credit Bank, eventually became Japan’s greatest banking success after it was closed and sold to an American buyout fund, which resurrected it under the name Shinsei Bank.

From 1986-1992, the FDIC closed over 2,300 banks and thrifts. The Resolution Trust Corporation (RTC) was established to move the bad real estate assets back into the economy as promptly as possible. The RTC, much admired for its speed and efficiency, did not try to buoy up failing banks; it handled their assets after the failing banks were closed.

So is the current administration plan a good bailout or a bad one? At this point the proposal is written in a very general way, and the devil is always in the details. The test will be whether the government plans to buy assets at something resembling their fair market value.

Suppose a bank has made a $10 million loan that is actually worth about $6 million. If the government buys the loan at $6 million, then it is providing liquidity assistance, which is fine. But if the government buys the bad loan for $10 million, it is in essence giving the bank a $4 million gift. Handing out gifts to misbehaving banks is characteristic of a very expensive, bad bailout.

Secretary Paulson has spoken of buying the bad assets at a deep discount, but this is not written into his proposal. It should be a requirement. That would minimize the ultimate cost to the U.S. taxpayers and increase the chances that this bailout is a good one.

Photo credit: Adam Fagen