Rich Toscano is much better at deciphering the Case-Shiller Index, click here to see how he sized up the latest June numbers – plus here are his charts to illustrate the trends:
He sums it up in his last paragraph:
As of June, despite having risen 17 percent off their April 2009 trough, low-tier home prices were still down 45 percent from their 2005 peak. The high tier was down “only” 28 percent, with the middle tier and the aggregate index both down in the 34 percent range.
Last night a reader suggested Encanto, 92114 as an area that has been red hot, and a good place to see data that could influence the Case-Shiller index in a positive direction. The first 52 of Encanto’s 70 sales in June, 2009 were REOs, short sales, flippers, or probates, with sales prices ranging from $85,000 to $253,000:
38 REO listings
7 short sales
4 flipper end-sales
There were 11 of the 38 REO listings, and one of the flipper sales that have resold again. But because the Case-Shiller Index discards any sales pair that is less than six months old, all but three of 12 would be disallowed.
Here are the three that would qualify for the C-S Index:
|Sales price in June, 2009||Second sales price,date||%chg|
These are the gains used in the C-S Index that would offset declines elsewhere, and help cause the county-wide, non-seasonally-adjusted Index to be positive for 14 months in a row.
In the Case-Shiller methodology, it did appear that they include all single-family sales (not condos), no matter how much time elaspsed in between the sales.
On their page 19, they say that typically 85% to 90% of the data used receives no down-weighting, but they also include an interval weighting procedure that accounts for the increased variation in the price changes measured by sale pairs with longer time intervals between transactions. For larger metro area markets, the interval weights for sale pairs with ten-year intervals will be 20% to 45% smaller than for sale pairs with a six-month interval.
In the end (and this will be the end), I guess it is possible for them to examine the 628,531 houses in San Diego County (they use the 2000 census count), weight all the sales pairs they can find that are older than six months, and track an index over time. But it’s a far reach from what’s actually happening to values in your local neighborhood.
I think your point is especially apt for NCC property. The top tier starts at $475K, which means 1/3 of houses sold in San Diego County. This tier is the one showing the most weakness, but the smallest drop from peak. Looking at just this tier is still an imperfect tool, but a much better indicator for NCC than the entire CS-index.
Even still taking a tier that starts at $475K to represent what’s happening in areas where houses start at 700K is dubious at best.
“These are the gains used in the C-S Index that would offset declines elsewhere, and help cause the county-wide, non-seasonally-adjusted Index to be positive for 14 months in a row.”
C-S also kicks out the statistical outliers, so that 100%+ gain in 12 months would probably be eliminated from the data set used for the published data, as would the other two.
Your last paragraph sums it up perfectly. The effects of the subprime boom, and then of the various government interventions, have created what I would bet is an unprecedented disparity between different neighborhoods in SD.
So while CS is the best way to understand what’s happening with all sales in an aggregated manner (which is of analytical interest, at least to some), it is completely useless for deciding how to price a house in a given neighborhood.