Hoisted from Mark Thoma's comments

Mark Thoma was kind enough to refer to my post on forecasting housing prices. ONe of the comments came from Fishy Math:

Fishy math says...

Wait a second! This formula is based on a perpetuity! Of course you are going to get huge swings resulting from relatively modest changes in assumptions. If you believe that housing prices will decrease by 5% per year INDEFINITELY, then you're correct. More likely, we will experience sharp depreciation for a couple of years, followed by a resumption of whatever the normal level of appreciation in housing in perpetuity.

Assuming 2 years of 10% declines followed by a resumption of 5% annual increases, I get a multiple of about 16x capitalized rents.

I want to make it clear that I am NOT forecasting declines of 5 percent in perpetuity. My point was that expectations, which overshot on the upside before, could now overshoot on the downside, and that we have rather poor estimates of what expectations are--although the CSW housing futures index suggests that those who are putting money on the line think that prices are going to fall for quite awhile. People also seem to be myopic when they form expectations.

The most important point, as Mark notes, is that small changes in expectations can lead to big changes in prices. To use a less extreme example, if expected appreciation drops from five percent to two percent per year, values would fall by about 40 percent. I think that is quite enough.