But it's hard to see the compatibility of that with the <$1tn PPIF. Once the insolvent banks are taken under public control (involving capital injection, or following capital injection aka "buffer", I don't quite get), the government can choose what it wants to do with compromised assets, including keep them on the books.

So the banks will be stabilised with a convertible loan. Then the plan is that they'll try to sell off the toxic (um, "legacy" seems to be the new term) assets so the banks can lend again. They'll also try to get the asset-backed-security market going again to kick-start securitisation. What will happen if an when these attempts fail is anyone's guess.

I think it is necessary to stabilize the existing "bad" banks so that their spasms stop causing seizures in the broader economy.

Restarting the flow of credit is a different thing. Apart from the contradiction inherent in saying everyone is indebted beyond their means and immediately laying out a plan to start lending again, there is the fundamental problem that the securitisation mechanism involves selling loans to non-bank entities which are not required to have regulatory capital to cover the risk of default and so should possibly not be allowed to actually buy a loan. Hopefully Geithner doesn't intend for the stabilised banks to start creating SIV's and off-balance-sheet "conduits" again.

So, once the banks are stabilised I think there will still be a credit crisis in the "real economy". I am not quite sure how to deal with that. Also, after a debt binge there must be a monetary contraction unless the debt is inflated away. So it seems to me that a few months from now an expansion of unemployment benefits and of employment through public investment will still be necessary.

Somebody sitting on a large pile of cash could decide to set up a new private bank and start lending from it. There's not going to be much competition from the old banks, and apparently also not from direct government lending.

Most economists teach a theoretical framework that has been shown to be fundamentally useless. -- James K. Galbraith