Sure, the banks are servicing many more loans than they actually hold in portfolio. But they’d still have a motivation to use their servicing rights to manipulate the value of their own holdings.

Think about it this way. You’re a bank, and you’re managing 1000 REO properties, of which 100 are “yours” and 900 are serviced on behalf of investors. If you drag your feet on marketing those 900 REOs and on pushing other defaulted loans through the foreclosure pipeline, you expose yourself to some tiny risk of being sued by MBS investors … but you also prop up the value of the 100 houses you do own.

The “cleanest” way to do this would be to keep your loss mitigation department chronically understaffed. This gives the appearance of earnest effort, as long as no one asks why you can’t just hire more people to handle the load.

In the last year I’ve noticed this pattern:

Fall 2008 – winter 2009: Banks dump the worst of the REOs on the market, presumably to save the best properties for the expected rebound. This backfires by driving down comps.

Spring 2009: Banks put some of the better properties on the market, sometimes even trashing them out or even repainting. This stimulates demand and creates an apparent market bottom, but keeps inventory low. Throw in the effects of government cheese and market frenzy results.

Early summer 2009: Banks respond to the frenzy by putting a trickle of the best REOs on the market, often throwing in new paint and carpet. This stimulates demand even more. Inventory reaches new lows. Buyers go wild, to the point where appraisals become a problem.

Late summer 2009: Banks release a bunch of so-so inventory. New higher comps have been set, so perceived value and asking prices are higher. Demand still high enough to support the higher asking prices.

Here’s what I’m hoping for fall 2009: Banks continue to release decent-quality REOs. Supply and demand more in balance, therefore non-cash buyers are able to get decent homes at fair prices.