A couple of quick, semi-random thoughts.
First. Many of the difficult to value assets in bank portfolios are tranched claims on payments from mortgages. The securitizers took pools of mortgages, and created securities that had different priority claims on the cash flows. The more senior tranches have first priority on cash flows, and lowest exposure to a default by the borrowers; the less senior tranches have lower priority, and the highest exposure to borrower default.
These complicated structures have option-like characteristics, and complicated relations between default rates and timing on the one hand and security cash flows on the other. This makes each piece more difficult to value than the underlying pool of the loans. Put differently, even if you have a good idea on the default risk of the underlying pool of loans, it is more difficult to determine how this default behavior will affect the cash flows on the different tranches. Therefore, it is harder to value the individual tranches than the pool of loans underlying them.
One way to facilitate the valuation of these securities, and hence the liquidity of the market, would be to reverse the tranching process and create vanilla MBSs. Equivalently, if the bailout plan proceeds, and the government starts bidding for assets, it could solicit offers only for bundles of the tranches collateralized by a given pool of loans, where each bundle would be equivalent to a vanilla MBS on the underlying pool.
Operationally this might be a challenge because, by design different tranches were typically marketed to different clienteles. But it is worth considering as part of the strategy, as pools of whole loans should be easier to value than the sliced and diced CDO tranches created from them.
Second, the WaMu bailout seems to be preferable to the buy assets approach of the Paulson plan. It is targeted at an institution that needs help, and utilizes the existing powers of the government. Indeed, other than the monoline insurers and the investment banks (notably Goldman and Morgan Stanley), most of the financial institutions that warrant concern are banks subject to FDIC regulation, and therefore standard FDIC approaches can be relied upon rather than taking a wholly new approach that doesn’t necessarily address the problems of the institutions posing systemic risk.
Third (I guess this is a bonus, as I set out promising only “a couple” of comments). The essential intellectual underpinning of the Paulson plan to buy assets is that the auction process–or whatever process that the government will use to buy assets–will improve market transparency. That is, the prices “discovered” in the auctions/market purchases will reflect private information about the value of the assets, and other market participants can use this information to mitigate asymmetric information problems that seriously impair market liquidity for problem assets. Even though sellers will receive an information rent (i.e., the government will overpay due to its information disadvantage), the creation of information is a public good that improves efficiency that is an offsetting benefit.
Perhaps. I am skeptical that the auction prices will be all that informative, given the heterogeneity of the underlying assets and the (related) lack of competition that is likely to characterize the auctions. Discovery of a noisy price in a not-very-competitive auction for a particular CDO is unlikely to improve the precision in the valuation of another CDO that is not bought in the auction.