Its a bit of a dead if you do, dead if you don't, problem. James Kwak explains:

From a regulatory perspective, the goal is to determine which banks will fail a worst-case scenario and force them to take preventive action. But at the same time, the Treasury Department is trying to restore confidence in the financial system. This objective breaks both ways. Confidence is low today in part because markets suspect that the banks are under-capitalized, but are not sure. So in the long term, rigorous stress tests that the public has confidence in should boost confidence by making clear that at least some banks are healthy, even if they reveal that some banks are sick.

But in the short term, labeling specific banks as sick would effectively cause runs on those banks, putting them into immediate danger and possibly triggering ripple effects on other financial institutions. (Although deposit insurance should rule out the possibility of people lining up to withdraw their deposits, runs can still occur in other forms, for example as hedge funds withdraw their business from a troubled bank.) This is an outcome that the Treasury desperately wants to avoid.

And so they have a conundrum.

This is normally why you turn to a policy like temporary nationalization. There's no bank run on a bank that's been taken into receivership. But it's very easy to imagine a bank run on a bank that you've publicly declared to be in poor health and given it 90 days to raise private capital. The outcome for that bank is thick in uncertainty, and thus investors might well try to flee its confines.

It also raises the question of what sort of terms the bank will find its emergency capital. There's a story from Michael Lewis's Liar's Poker that's relevant here. Salomon Brothers is facing down a hostile takeover from corporate raider Ronald Perelman. Perelman is famous for assuming control of high-value, poorly managed companies and then making money by firing bad management.

To save his company, John Gutfreund solicits a massive loan from his friend Warren Buffett on terribly disadvantageous terms. As Lewis argues, "the arrangement had two consequences. It preserved Gutfreund's job, and it cost us, or, rather, our shareholders, a great deal of money." The loan is much worse for shareholders than the takeover. They're essentially paying a large premium -- conservatively estimate at $120 million in 1987 dollars. But it's quickly approved by the board of directors, most of whom were appointed by Gutfreund.

Presumably, much the same thing will happen as these banks dash for private capital. The executives making the decisions will fear the pay caps and firings that come from continued government reliance. So they will cut quick deals on bad terms in order to save their jobs and preserve the company as they currently know it. Over time, that money will be made up by the shareholders, but also by individual investors paying larger ATM fees and overdraft penalties and all the rest. There's nothing illegal about any of that, of course. But it's not an obviously preferable outcome.