Reverse auctions

A number of readers have asked why I described the Paulson plan as "a plan to have a plan" when Bernanke says they want to use a reverse auction to price the securities.

A couple of reasons.  First, because there are a lot of issues with reverse auctions.  Bonds aren't stocks; there are a lot more of them, and they're much more idiosyncratic.  The securities that the government is proposing to buy are even more complicated than ordinary bonds. 

This means there's a gigantic asymmetrical information problem:  the owners of these securities know much more about them than the Fed.  And there isn't (obviously) a large liquid market for the Fed to check against.  So the Fed is likely to overpay, because there won't be a lot of bidders in any one auction.

Second, because there are so many different securities, auctions will take a lot of time.  The whole argument in favor of this bailout is that we don't really have a lot of time.

Third, the reverse auction arguably solves one problem--what are these securities worth?--by adding another:  the aid isn't targeted to the institutions that necessarily need it, or can use it the best.  It's targeted towards those who will bid the lowest.  Maybe that will be institutions in distress.  Or maybe it will be solvent firms looking to clean up their balance sheet.

Relatedly, using a price discovery mechanism means that the bailout may not give distressed firms enough capital.  There's no point in buying distressed securities if the banks go bust anyway.

Thus, while reverse auctions may feature in the final plan, they're unsatisfactory, and probably will not be the primary vehicle for recapitalizing banks.  So I still want to know:  what's the plan?