Radar Logic puts out a home price composite index (RPX).
It covers the 25 largest metro areas, and fell 1.4% month over month in June — far more than Case Shiller’s 0.5% decline last week. Annual price changes were marginally deeper also: 17.2% year over year for the RPX for versus Case Shiller’s 15.9%.
The two series use very different methodologies; Case Shiller uses repeat-existing single-family home sales price; RPX index uses ‘price per square foot.’ RPX also includes new homes and condos as well as existing homes.
I recently have been hearing people claim that Case Shiller, as negative as it is, actually understates the Housing Price drop, as it lags prices by a year or more, and excludes foreclosures.
Note: I have not taken apart the Case Shiller methodology.
>
RPX Weekly, 2007-08
chart via Radar Logic
RPX Monthly, 2000-08
chart via FusionIQ, Bloomberg
>
Source:
Radar logic home price composite index (RPX)
June 2008
Release Date: September 2, 2008
http://www.radarlogic.com/research/RPXMonthlyHousingMarketReportforJune2008.pdf
Interesting. Price/sq.ft. seems like a good overall measure. Given that prices are still high relative to incomes, I think we have a good ways to go yet. Then, once we hit bottom, we’re in for a very long flat period. On a national scale, 1990 was a blip in comparison, and one can readily see how long it took prices to start rising again. The only way this period could be worse is if it more closely resembles Japan–a long, drawn-out slide to the bottom lasting over a decade.
Speaking solely for myself, until monthly payments drop below rental rates I won’t be buying. Right now I’m renting a 4 BR/2 BA fairly new house in Orlando for $1800/mo. Has a hot tub and thermador grill and the rent includes lawn service. The house would need to sell in the $225k range to be equivalent, but that would be $100k less than what the neighbors are asking. There are plenty of similar-sized houses to choose from in this rental range. Thus I think it’ll be a while before we finally bottom.
Here’s a milestone: There are now more foreclosures on prime mortgages than on subprime ones.
The Hope Now alliance — the lenders’ group put together at the urging of Treasury Secretary Henry M. Paulson Jr — estimates the number of foreclosure proceedings that begin nationally in each month.
The latest figures, for July, put the number at 197,000, the highest for any month since they started keeping track in July 2007.
Of those, 105,000 were prime mortgages, and 92,000 subprime. The June numbers also showed more prime foreclosures initiated.
The Hope Now people report that completed foreclosure sales are still higher for subprime but the gap is narrowing.
There are far more prime mortgages than subprime, of course, and subprime loans are much more likely to get into trouble. But this does show how the foreclosure problem is spreading.
One more indication that things are getting worse. Hope Now tracks the number of homeowners it says its members have helped each month to avoid foreclosures against the number of foreclosure sales completed.
For most of 2007, it figured three homeowners were helped for every foreclosure. Now the ratio is about two-to-one.
It seems to me that, in principle, this housing index should be at least as valuable as Case-Schiller, and in some ways better.
BP – there is a substantial time lag in the Case-Shiller Home Price Indices; we’ve just researched this issue for our subscriber newsletter and feel the data lags by five to eight months (at least) when you logic through their methodology. I’ll have a post up on our blog looking at the issue in detail in the next couple of days.
One of the other criticisms of the Case-Shiller Home Price Indices that holds water in our estimation has to do with what home sales are included and excluded from the database. The indices are designed with very tight inclusion criteria – the tool is really designed to look at detached / semi-detached single family, owner occupied homes. Excluded from the index are: all new construction, condo’s and co-ops, multifamily units – even duplexes, non-arms length transactions (e.g. between family members), transactions immediately preceding or subsequent to substantial physical changes to a property (think rehab for sale or flips), as well as suspected data errors where the order of magnitude in values appears unrealistic.
In particular the exclusion of new homes and condos/ coops – two of the hardest hit single family asset subclasses – further weakens the veracity of the media (and others) when using the CSI to portray the current state of the “national housing market”.
I really like the RadarLogic methodology. They get DAILY price points – and then fold them into a rolling monthly average.
I would suspect their index is much more sensitive to the actual realized prices in the covered areas.