2009 Real Estate Forecast (BusinessWeek)

Prashant Gopal wrote an interesting article on December 23rd:
2009 Real Estate Forecast: Troubles Spread in BusinessWeek.

Highlights include:

• Wealthier neighborhoods that avoided subprimes may be hurt this upcoming year.
• Already hurt markets (like Las Vegas and California) shouldn’t slide as much as before because they’ve already fallen so much.
• Unemployment will make more of an impact, especially in Texas (energy sector) and Charlotte, NC (financial sector).
• Some places (DC, San Diego, Boston, and Orange County) might start to recover due to people picking up low prices.

Also included were links to the guys that do these predictions:
HousingPredictor.com, Moody’s Economy.com, and Miller Samuel

Whenever I read articles like this (at least after finishing the Black Swan and Practical Speculation) I wonder about the error rates and methodologies of the predictions. For example:

HousingPredictor.com is projecting a 19.4% decline in Manhattan home prices in 2009

But with what error rate (the +/- X%)? If the figure is +/- 50% how useful is it, and with all the wild variables, how does anyone expect to do better? The job analysis is interesting though.

Not that I have any better information nor do I think they’re doing a bad job nor do I lean toward a better or worse view, I just wonder about the inherent limitations of such forecasts.

What do you think?