Barry Ritholtz says the worst of the housing correction is behind us.

Prices certainly can fall much further; it is possible. However, I am making (what I believe is) a higher probability argument due to fair value. My basis for saying the worst is likely over are prices: We are off 33% from the peak, and as of the end of Q1, were ~5-15% over fair value by traditional metrics.  So a return to fair value -- even a 15% drop in 2011 -- still means the worst is (was) behind us.

If houses were to careen far below fair value -- they were about 40% overvalued, so in theory, they could overshoot 40% to the downside -- then my valuation thesis would be wrong. There are lots of ways house prices could drop much further: If jobs and income plummet from here, home prices will be too high. If interest rates spike, prices will adjust downward. If the mortgage deduction were to be eliminated, prices fall also. IMO, these are smaller possibilities -- say 20-25% chance -- then merely mean reverting towards historic relationships with median income, cost of renting, and home equity as a percentage of GDP.

(via Ezra Klein)