Once again, data from the S&P Case-Shiller Home Price Index is providing a confusing picture of what’s happening to U.S. home prices and, once again, the Standard & Poor’s website is incapable of handling the 9AM EST traffic on the last Tuesday of the month.
From accounts like this report at MarketWatch, it looks like seasonally adjusted prices rose 0.3 percent in January while unadjusted home prices fell by 0.4 percent as shown below.
Note: The image above is an animated .gif – it should changing.
When you think about it, this shouldn’t be surprising. First, the Case-Shiller data comes out a month later than most economic reports, so, the data reported today is for January, not February. More importantly, this index is not a reading for just a single month – it is a three-month moving average – and nearly every home price index shows prices rising last fall as the first round of the homebuyer tax credit was about to expire.
Perhaps more important than either of the above, the unadjusted data is likely providing a truer gauge for home prices than the one that is adjusted for normal seasonal variations because the variations that were seen last fall were anything but normal.
![[Most Recent USD from www.kitco.com]](http://www.weblinks247.com/indexes/idx24_usd_en_2.gif)
I’ve been a long time reader of your blog, but I’m disappointed in your superficial analysis of the differences between seasonally and non seasonally adjusted numbers. The non-seasonally adjusted numbers are not any more right or more wrong then the seasonally adjusted numbers. They are just come out of the black box of statistics in a slightly different way. I’m not sure of the practical usage of the NSA numbers. While they reflect real time prices, they are simply an average for a very large area and shed no light on purchasing or selling property in a local neighborhood. If you are looking for a real estate trends over a broad area, then you do indeed want to look at the SA numbers. The SA numbers say that this January’s market is doing better then last January’s number. It’s not important to know that prices dropped between December and January because that we all know that prices face a head wind in the winter and a tail wind in the summer.
if you want to claim that the unusual market patterns are causing a disruption in the SA numbers, then it would be more useful to analyze the numbers to see what exactly the effect of a seasonal abberation does to the SA numbers. Simply claiming that we should now look at the NSA numbers as “more reliable” is like picking the data series to fit the world view and is exactly what the Realtors did when they switched their marketing from YOY numbers to MOM just as the market was turning.
“The non-seasonally adjusted numbers are not any more right or more wrong then the seasonally adjusted numbers.” ?????
The NSA numbers are real, the SA numbers are not. They are simply a way of making sense of a volatile data series with regular seasonal patterns, which, in this case, I think are quite misleading.
I guess I should have mentioned that virtually every other home price indicator that is not based on the even more severely flawed “median” price also shows home prices declining in recent months.
“if you want to claim that the unusual market patterns are causing a disruption in the SA numbers, then it would be more useful to analyze the numbers to see what exactly the effect of a seasonal abberation does to the SA numbers.”
I’ll have something more on this in an hour or so – the fact is that the biggest positive seasonal adjustments of the year are now coming at a time when prices had the homebuyer tax credit surge last fall.
I think what Tim is referring to is that the NSA series appears to indicate the beginning of another downward trend, while the SA trend remains ambiguous. The SA series is low pass filtered (averaging of some kind) to get rid of ‘noise’. However, at any turning point, a filtered or averaged series is always slower to respond than its unfiltered companion. Another downleg is exactly what we should expect as the next set of defaults (prime/jumbo/commercial/states/etc.) begins and as the misguided ’stimulus’ peters out.
Anecdotally, in my neck of the woods, prices haven’t moved at all. Also, nothing’s sold either. There are still, two, brand new condo buildings sitting empty as they have for the last two years. I see “For Sale” signs go up on houses, and then come down, then go back up, then come down, then go back up… parlor tricks by RE agents trying to keep the listings “fresh”. Despite what the indices read, the market is still highly dysfunctional, from my particular, geographical point of view.
See here for more on seasonal adjustments.
[...] case you were wondering about the thoughts offered earlier today on the impact of seasonal adjustments on the latest Case-Shiller Home Price [...]
I have to agree with Mike you can paint the numbers anyway you want to. Give us facts not opinion.
“Give us facts not opinion” ?? At a blog?? Did I miss something??
From a report by Diana Olick at CNBC:
“The Federal Housing Finance Agency’s (FHFA) adjusted figures show a housing price decline of 2 percent in December and 0.6 percent decline in January—reversing some regional price increases in 2009.”