If no large bank is found insolvent (or to use the same euphemism as the Treasury, in need of additional capital), the markets will not believe the tests are credible, and the reaction may be very difficult to manage.

If on the contrary the results are really bad – say, more than half of the 19 tested banks are found insolvent – then it’s the Treasury’s responsibility to prepare the marketplace. The recent signals being on the positive side (“most US banks have sufficient capital”, etc.), this would require a difficult U-turn, but the longer Geithner waits, the more difficult it becomes, and the more chaotic the reaction may be.

If, as seems most likely at the time of writing, only a few banks are insolvent, then the least bad option – as was implemented in Sweden in 1992, admittedly with a much smaller and simpler banking system – is to let the banks themselves announce the news, good or bad. They have the investor relations skills and know how to manage expectations. There is precious little time left to do so. In this scenario, Geithner should exert pressure on the banks so that when he announces stress test results next week, the marketplace already knows most of what’s in there. Those banks which are found insolvent will be massacred by the stock markets, of course, but there is no way they can avoid this outcome. This has already been illustrated on Tuesday 28 April, when City and BofA expressed disagreement with the Treasury only to see their stock prices take a plunge.

Former Senator George Aiken of Vermont had a suggestion concerning the slightly less expensive project known as the Vietnam War: declare victory and quit.

It would have been wiser and less expensive to let failing banks fail and solvent banks get on with life $700 billion ago.

But better late than never.

All this flailing around by the Obama administration is making things worse, delaying a recovery.

Updated at 10:22 a.m. on April 27.