Sounds like a nice neighborhood to work in. Suspect is a starting point. Adjusting or not adjusting should be born out by data from the market place. At 90% you still have data to prove your case. At 100% REO sales, something which is very unlikely to occur in the real world because some owner occupant is going to eventually need to sell, could not one look backwards to when the make up was less than 100% for clues? Question Mark?
For the grace of exactly ONE non-REO sale a 100% REO market is EXACTLY what I am looking at in one Kenosha neighborhood. In this case the "typically motivated buyer" situation is muddled even more since it is a two-family market and thus the "typical buyer" has always been a mix of owner-occupants and investors. Searching adjacent competing neighborhoods just yielded more (higher priced) REOs. Lots of them. IIRC at last could it was only the 1 non-REO out of over 40 properties examined, which is well over that "90% point", yet since that one was a flip (and very few months between to boot) it is OBVIOUS there is a difference between REO and non-REO sales in that market.
Is that market REO dominated? You betcha!
REO driven? Almost certainly.
Are REOs "THE Market"? Definitely not.
Even if it were 100% I would be doing the same research to determine how close the market was selling to REO prices.
Same market is a completely different story when dealing with 4-family properties. Values there have not dropped significantly and there is a wide range of strata in the market both for REOs and traditional sales based on condition (from "gutted" to "borderline inhabitable" to "inhabitable, but rough" to "move in" and finally "premier properties" (tending yo be in at least "good" condition, superior location, and so forth).
Two markets, both in the 2-4 small income property range, but both total different in what has happened to them since the bust. One prices dropped like a stone (below half of pre-bust), and the other relatively unscathed. Big difference? The buyer ratios changed significantly for one and remained unchanged for the other.
So, when trying to figure out whether or not REOs can represent the market you may want to start looking at who currently tends to be the typical buyer for such properties and who used to be, has it changed, and if so what difference did that make? The differences between fair market value and what REOs are selling at may well be who, traditionally, was the typical buyer for the REO and non-REO properties and whether or not that changed.