From the WSJ:
The Treasury's bank strategy is twofold. One, get enough capital into the 19 biggest banks so everyone believes each can withstand a really bad recession. Two, get toxic assets off their books so banks will pick up the pace of new lending, and savvy big-money investors will put money into the banks and help achieve the first objective.
The unfortunately named "stress test" -- which conjured up images of Citigroup collapsing on a treadmill -- was meant to be a confidence builder, though announcing it seems instead to have magnified doubts and uncertainty about the banks. The notion is to figure out by the end of April how much capital cushion each of the 19 big banks needs to survive a bad recession (that's the "stress test") and then give those that need more capital six months, until Oct. 31, to raise it privately or take a bigger taxpayer investment on different terms than former Treasury Secretary Henry Paulson offered. Until then, the Federal Deposit Insurance Corp. will guarantee bank debt so no one need worry about lending to them, or so the Treasury hopes. None of the 19 banks will flunk the test; the only question is which will need taxpayer capital in the fall.
Let's think this through.
The Treasury plan wants the individual banks to act privately before more taxpayer money comes into the equation. On the pro side, this will prevent the use of more taxpayer money if its successful. On the bad side, its takes time to raise capital during which a number of bad things could happen. In addition, who would invest money in these banks? However, I think the third quietly implied option of this part of the plan is there will be mergers between weak and strong banks of one sort or another in reaction to the "stress test".
The last piece of the Geithner plan comes soon: Buying toxic loans and securities, mostly linked to real estate, from the banks and others. One challenge is putting a fair price on them. The Paulson Treasury spent months trying to fashion auctions in which the government would buy these assets. It never bought any. The Geithner Treasury decided that approach wouldn't work. What's more, it hasn't nearly enough taxpayer money to buy enough of the assets to make a difference.
So the plan is to form joint ventures between the Fed and money managers like Pimco or BlackRock. The Treasury kicks in, say, $1 for every $1 the private guys put in. The private investors, not the government, decide what securities to buy from the $1 trillion or so in securities linked to real estate or consumer loans. The private guys decide what price to pay. That's their business. Taxpayers and the investors would share the profits, if any. If the Fed lends to these ventures, they'll be able to buy more securities and pay more for them.
A separate set of joint ventures will shop among the $1 trillion or so in toxic loans on bank books. This effort will be leveraged by FDIC lending. The hope is that some banks will make themselves more attractive to private investors by selling toxic assets to the new joint ventures, and thus ease both parts of the banking problem simultaneously.
The central problem with this plan is the central problem with all plans dealing with the bad assets: will the banks be willing to sell and at what price? There is no guarantee anyone will be willing to sell or buy at a rate the market would like. I think this plan has an OK chance of working, but that is hardly a ringing endorsement.
I will think the best idea we've had so far is to make one big bank out of the remaining taxpayer funds, use that bank to buy the good assets from the troubled institutions and then let the banks keep the bad assets to manage. That still makes the most sense to me, largely because it only involves the creation of one structure which would be easier to monitor.
While this is not the best plan, we're left with a question of "what else can be done?" While there are still calls for nationalization of the banks, the current problems with the TARP plan -- that is, the increasing number of politically motivated bailouts -- indicates that my fears the process will be politicized are well founded. If we can get rid of that problem I think it would be appropriate. But until that question is answered, I still don't think it makes sense.