Random Housing and Economy Stuff

PRINT     E-MAIL STORY    SHARE

FACEBOOK

REDDIT

NEWSVINE

DIGG

DEL.ICIO.US

BLINKLIST

Just a few quick bits today.

First, the New York Times informed us last week that "It's not just subprime anymore." It's nice that the mainstream media is coming around, but as I've been ranting about for the better part of this year, it was never just subprime.

Second, Señor SLOP pointed me to a U-T article about a recent study contending that the housing bust will "erase" $1.5 billion from San Diego's economy next year, resulting in a local economic growth rate of 2.1 percent instead of the 3.0 percent it should have been. The article is an interesting read but I'm skeptical of the ability to forecast the housing downturn's effect on economic growth with such precision. There are a lot of factors here -- not just home equity extraction and housing employment, but second-order effects like a potentially more widespread credit tightening, the housing bust's impact on stock prices and consumer confidence, the inflationary impact of the Fed's attempts to ease the downturn, and who knows what else. It seems like a long shot to quantify the impact of all these moving parts to within a tenth of a percent, but maybe that's just me.

Finally, just to make this an apparently continuing series on the Case-Shiller Home Price Indexes, I wanted to address some questions I've gotten about how the indexes are created. To start with, each month's index number is created by collecting all home sales (except for invalid ones like sales between family members, etc.) in the prior three months. So August's HPI number is actually measuring price changes for all sales that took place in June, July, and August. The point of using three months' of data is to smooth out movements in the index, though obviously it also induces a bit of a lag -- and that's on top of the two-month delay before a given month's index figure is released. I had forgotten that they use three months' worth of data until I brushed up again on their methodology. I will make sure to be more clear about that in the future.

Moving on, each sale that took place during the last three months is paired with the prior sale of that same home to ascertain the change in its price. These "sale pairs" are the basis of the index and are used to model how home prices have moved over time.

The approach to dividing sale pairs into three price tiers is very simple. The sale pairs are divided into three equally-sized groups based on the earlier sale price of each home. The resulting three sets of sale pairs comprise the low-, mid-, and high-tier indexes. This method insures that a given home will never jump categories between the time of the first sale and the time of the second. Dividing the data into three equal chunks also explains the seemingly random cutoff points between categories.

Hopefully that extremely brief overview answers some of the questions out there. Folks who'd like more detail might be interested in the methodology paper on S&P's site.

While we're on the topic, the September Case-Shiller numbers came out today. Kelly Bennett has a first look at the figures, and I will put a chart up shortly.

Well, that didn't turn out to be very quick at all. I promise I will be back to my laconic, visual aid-heavy ways by Thursday at the latest.

-- RICH TOSCANO