The lead headline on CNNMoney.com right now is “Home prices headed for historic drop”: we’re told this will be “the first annual decline in nearly 40 years of tracking.”
The headline comes in response to news from the National Association of Realtors: while it thought in February that 2007 prices would rise by 1.2%, apparently it thinks now that they will fall by 0.7%. Apparently all the pain will be concentrated in a few markets:
The group estimates that about three-quarters of the markets nationwide could still see a narrow increase in median sales prices during 2007, but that those gains will be outweighed by the declines in the markets that saw big gains in sales and prices during the record sales years of 2004 and 2005.
This surprises me, since the data I’ve been looking at shows the biggest declines in more depressed, industrial areas such as Michigan and Ohio. But this doesn’t surprise me at all:
The group’s forecast sees an even bigger slowdown in the new home market, as it is forecasting new-home sales will come in at 904,000 this year, down 13 percent from the 1.05 million sold last year.
There’s no doubt that new-home construction is slowing down. But couldn’t that alone explain a large part of the median-price drop? Reader Glen Lineberry emails me:
Everyone assumes this means that houses are selling for less. Isn’t an equally logical explanation, given all the media coverage of the housing slump and mortgage problems, that people are simply buying less expensive houses? Wouldn’t that also shift the median price, if people decided they could live without an additional bedroom, or weren’t willing to pay the freight for someone else’s kitchen renovation?
Now, instead of “less expensive houses,” just try reading “fewer brand-new McMansions”. It’s a well-known fact that the average new-home size has been growing steadily for years, and positively booming of late. So if those new homes bear the brunt of the slowdown, and are sold in much lower numbers, that could do nasty things to the median sales price even if any given existing home doesn’t fall in value at all. The new homes don’t even need to be sold at a lower price, there just needs to be fewer sales.
For even more fun ‘n’ frolics, check out CNN Money’s list of the 100 largest housing markets in the US, complete with forecasts for how those housing markets will do over the next 12 months. The forecasts were obtained by rolling dice from Fiserve Lending Solutions. Apparently McAllen Texas is in for a big boom, San Francisco will see a modest uptick, New York City will fall by almost 4%, and poor old Phoenix, home of Glen Lineberry, will see a 5.5% collapse. Not that Lineberry is particularly worried:
The growth here is just beyond belief. More than 60,000 new housing starts in 2006. It’s down to half that this year, but that’s still the third highest number on record.
A massive light-rail project is underway, to open December 2008, and both ASU and the UA Medical School are building new campuses in downtown Phoenix, so we’re seeing a return to central parts of the city. Lots of infill projects, from single-family homes to condo and apartment complexes, are springing up and median prices are firm or slightly higher.
It’s in the giant tract developments on the outskirts of the city — often an hour’s drive from downtown and the airport — that houses aren’t selling. The builders are holding inventory, investors who’d put down small deposits have walked away, and the only way for a seller to compete is on price.
In the more central neighborhoods, it takes a little longer to sell, but houses offered at last year’s prices are selling. When you remember that prices here went up 25% last year, and doubled over the last three years, that’s no so bad.
The phrase you want is “sales mix”.
Median prices mostly measure sales mix, either regionally (higher sales on the coasts? higher median prices) or as you noted, by segment.
That’s why the only price indexes you should trust are repeat-sales indexes. The OFHEO’s HPI remains the gold standard, despite its being limited to conforming loans. The S&P/Case-Shiller indexes are a worthy alternative, and not thus limited.
Ignore the NAR median-price data. It tells you almost nothing.
Is this sort of like people arguing in the late 90’s that using P/E ratios was an antiquated method for valuing companies?
There are plenty of people (Alt-A’s and subprimes) in a risky position (if not now, certainly 24 months out when adjusted rates and principal payments hit) because they weren’t just expecting a small increase, but likely on the order of 3-5% appreciation for the next couple years.
It’s one thing to fuss over the gap between at 1% increase and half point decrease. 1.5% isn’t fatal, but convenient for CNN headline writers. But the gap between 5% (to say nothing of the fools that were betting on 10-15% increases) and 1.5% or -0.5% is actually large, and looming, since there are no trend lines that would indicate the possibility of a turnabout on the order necessary to offset two years of stagnant growth rates (what would that be — 9-11% for two consecutive years?) many really can’t weather right now.
You can talk all you want about dispersion of risk in the MBS market, but 35% percent of the mortgages written last year just got proven to be failing investment gambles, and that is unprecedented in the history of the housing market.