What is it Really Going to Cost?

Love it or hate it, the true cost of Treasury Secretary Hank Paulson's proposed rescue of the financial system is not the sticker price of $700 billion. Conceivably, the government could make money; with glum assumptions, the losses would probably be less than $250 billion. No one knows the correct answer -- not Paulson, not Federal Reserve Chairman Ben Bernanke nor anyone else -- but here's how to think about the problem.

Under Paulson's proposal, the Treasury could buy distressed mortgage-backed securities. Consider a batch of hypothetical securities originally worth $100 million and paying an interest rate of 6 percent. They're no longer worth $100 million because half of the homeowners have stopped making their monthly payments. Suppose, then, that the government buys the mortgages for $50 million. It earns 6 percent on its $50 million, and if it borrowed money at 4 percent to buy the securities, it would make a tidy profit. If the government holds the securities until maturity and all the remaining homeowners repay their mortgages, the government would come out ahead.

Would something like this happen? It could, and Pimco's Bill Gross argued in the Washington Post that it might, but there are several reasons it might not.

First, we don't know what price the government would pay for the mortgage-backed securities. There are conflicting goals. On the one hand, the government wants to minimize the bailout's costs to taxpayers; that would favor paying the lowest possible price. In my example, the profit would be greater if the government paid only $40 million. On the other hand, the whole idea of the bailout is to help banks and other financial institutions get rid of risky assets and replace them with cash that would encourage a resumption of normal lending and investing. That favors a higher price. If the government paid $80 million instead of $40 million, say, it would lose money.

Second, we don't know how a weakening economy will affect future mortgage repayments. The worse the economy gets, the more homeowners will default. At the end of June, about 2.75 percent of home mortgages were in foreclosure, and an additional 6.4 percent were at least 30 days behind in their payments. The unemployment rate was 6.1 percent in August. If it rose to 7 percent or higher, defaults and delinquencies would climb. In my example, if only 25 percent of borrowers repaid their mortgages, the government would lose money.

No wonder members of Congress -- and the public -- are confused. My simple example captures the main unknowns, but in practice there are many more. What bonds and securities would Treasury buy? Would the government hold them to maturity or later try to resell them to private investors? To all questions, Paulson has said in effect: Trust us.