The Stress Test Blind Alley

Today, the government will begin discussing the results of its stress tests with heads of the nation's 19 larger banks in meetings in Washington and will release the methodology of the tests to the public. Then, of course, we'll all sit around twiddling our thumbs until May 4th, wondering what comes next.

The number circulating today is $1 trillion -- that's new capital needed -- which is an estimate produced by brokerage firm Keefe, Bruyette and Woods, based on its own stress tests. The number has the advantage of being large and round and largely without competing figures, and so it has made its way into every newspaper preview of the day's meetings. I have no reason to think that it's not in the right ballpark. Certainly, estimates of losses from a range of sources indicate that hundreds of billions of dollars will be necessary.

What's clear is that the test results will be unable to avoid playing favorites. Recommendations will not be one-size-fits-all, but will be tailored based on the quality of a bank's assets, its likely earning power, and its exposure to other, potentially weak, institutions. This creates an obvious problem; the banks in the best condition will need to raise the least capital and will have the easiest time doing it. There's a distressing corollary; those in the worst condition will be required to raise the most capital. They'll have an awful time of it. Private investors will be uninterested, because the possibility of a good return will be slim, and because government could come careening in at any moment, messing up whatever calculation convinced an investor that Citigroup was a promising bet.

Which will leave the government there to fill the hole. But how? If we imagine, not unreasonably, that there will be a need for a few hundred billion between Citigroup and Bank of America, then we find ourselves facing some tricky questions. Where, for instance, does the money come from? How does the government get it to the banks without accidentally winding up with very large ownership stakes? What is the game plan?

And to me, that last is where all the suspense surrounding the stress test lies. The actual results are the opening act. The main event is what the government rolls out as its solution. I have this uncomfortable feeling that if the stress tests reveal anything like the true picture at a Citigroup and the policy response is a combination of an increase in equity stake, fingers crossed for private recapitalization, and the promise of PPIP, markets may be very upset indeed.