To be honest, I'm still trying to understand the Treasury's new "legacy assets" proposal: Krishna Guha's account; Treasury fact sheet. Of course it might fail but my instant reaction was not to think that it is "almost certain to", as Paul Krugman writes with characteristic clarity (and a corresponding refusal to see any merit whatever in the other side's position).
There are (at least) two programs here, not one. Many critics are failing to distinguish between the loans program and the securities program, though these are intended to work in very different ways. And although the great complication of the overall plan seems to have satisfied most commentators' appetite for details, at least for now, crucial aspects of how it will work are in fact still unknown. For instance, as Krisha notes:
Pricing will still be difficult. Many toxic securities are complex and backed by fraudulent mortgage loans. Some are highly idiosyncratic, so discovering a price for one may not convey much about others. Banks may not be willing to part with loans that have not been heavily written down at auction prices, though Ms Bair hinted that regulators may force them to do so.
That banks will simply refuse to play at the prices offered by the PPIFs is probably the main risk for the loans part of the plan--though one that goes away if they are forced to sell. Will they be? Krishna continues:
Moreover, unless credit conditions ease considerably, the price levels established under these schemes will be sustained only by continuing access to lower-cost government finance.
Again, details to follow. The full terms of the FDIC's guarantees under the loans program, and how it proposes to recoup any losses incurred on these guarantees, are not spelt out. The much-expanded TALF will carry the burden of dealing with legacy securities (hitherto regarded by the Treasury as the larger part of the problem) as opposed to legacy loans, but exactly which assets it will cover and the interest rate to be charged by the Fed are also not stated. And so on.