"The Case Against Case-Shiller"

Bad Numbers??

The S&P/Case-Shiller report for October (there's a one-month lag) is due to be released tomorrow, and the numbers are likely to be ugly: the Sept. numbers were down 17.4% year-over-year, and down 1.8% month-to-month, respectively, and there's plenty of anecdotal evidence that things got worse in October. So, to repeat, it's not going to be pretty.

However, whether the national housing market is actually as bad as Case-Shiller is likely to indicate is debatable.

Twin Cities realtor Pat Paulson has an excellent piece, "The Case Against Case-Shiller," which, although a year old, does a very nice job of dissecting weaknesses in Case-Shiller's methodology:

http://activerain.com/blogsview/370858/The-Case-Against-Case

The gist of his (and others') argument is that, by exclusively focusing on "sales pairs" -- back-to-back sales of the same home -- Case-Shiller overweights homes that change hands frequently.

Which homes tend to be frequently resold these days? Foreclosures. Nationally, "lender-mediated sales" (short sales and foreclosures) account for about 50%(!) of recent sales volume -- more in particularly distressed markets like Miami, parts of Southern California, and Las Vegas.

To no one's surprise, in a down market, abandoned, neglected houses lose a lot more value, faster, than ones that are owner-occupied. Even though Case-Shiller uses a variety of techniques to adjust for foreclosures, it's not clear that they're effective.

Which leaves the question, how accurate is Case-Shiller?

Generalizing about the national housing market from a sample pool dominated by foreclosures is like estimating Americans' wealth looking only at bankruptcy filings.