There's an evolving consensus that we're no longer seeing a liquidity crisis, but rather a solvency crisis. Banks don't just lack cash; they're fundamentally insolvent, with their assets (things like mortgage-backed bonds) worth less than their liabilities (including their deposits). Improving liquidity, through rate cuts or swap lines or discount windows or buying unsecured commercial paper or any other means, can no longer get us out of our present hole. What we need is a recapitalization of the banking system.

The thinking behind the TARP is that the solvency crisis is a direct result of the liquidity crisis. (Remember that it was conceived before the most recent downward lurches in global markets.) Why are banks insolvent? Because they're sitting on worthless assets. Why are the banks' assets worthless? Because there's no bid for them. So create a $700 billion bid for those assets, get a two-way market in them moving again—and presto—the solvency disappears.

There are more than a few problems with this line of thinking. Firstly, the logic is a bit screwy: Even if insolvency was a direct result of illiquidity, there's no guarantee that you can simply run the movie backward and hope that restored liquidity will result in solvency.

What's more, just because the government is willing to spend 40 cents on the dollar for a mortgage-backed bond at the end of a descending clock auction doesn't mean anybody else is: The market won't necessarily really believe those bonds are worth 40 cents. Sure, the banks will use that mark when officially valuing their balance sheet. But investors in the banks' stocks and bonds might well still be skeptical and mark those instruments down accordingly.

In order for people to really start believing in the price, you need a two-way market, with multiple sellers and buyers at or around the 40-cent level. If Treasury remains the only buyer in town, that's not going to happen.

But more important, what if the banks really are insolvent at that level of 40 cents on the dollar? The price revealed by the TARP would make crystal clear what many investors already suspect—that the banks are worthless. Far from rescuing the market from crisis, if the TARP auctions cleared at a very low level, they could actually precipitate it.

In that event, Treasury would have to try a very different tack. They could use the TARP, still, and try deliberately overpaying for bank assets: Instead of paying 40 cents, they'd bid 80 cents. That would help to recapitalize the banking system, since the banks would be getting the fair-value 40 cents for their toxic loans, plus another 40 cents on top, which they could use to stay afloat.

Under that kind of system, Treasury would presumably be much more aggressive in terms of the amount of equity that banks would have to give them in order to participate in the auction. Or maybe the auction would be a decreasing-equity auction, rather than a decreasing-price auction: Treasury would set the bid at 80 cents, and then see if there were any bidders willing to hit that bid and give up 10 basis points of equity per $10 million of bonds sold. If there weren't, they would bring the level down to 9 basis points, and so on.

The big problem with any kind of recapitalization auction, however, where the Treasury deliberately pays above the market rate for toxic securities, is that it defeats one of the main purposes of the TARP: to set a transparent mark for illiquid bonds. If everybody knows that Treasury is overpaying, then no one will believe for a minute that the securities still left on banks' books are worth nearly that much—and the market in such securities will remain frozen.

It's probably much simpler and easier to take the route chosen by Gordon Brown, the British prime minister: Ignore the asset side of banks' balance sheets altogether, and just start writing multibillion-dollar checks to big entities in the financial system in return for shares and/or preferred stock.

That's essentially what the U.S. government did with American International Group; it's a bit weird that it chose a very different route for the banking system as a whole.