A belated post on yesterday’s Case Shiller release:
Through June 2008 the S&P/Case-Shiller Home Price Indices showed "continued broad based declines in the prices of existing single family homes across the United States, a trend that prevailed throughout 2007 and has continued through the first half of 2008."
The month-to-month drop was slightly better than expectations, dropping 0.5% versus the 0.8% consensus forecast. As the chart below shows, the index experienced its greatest year-over-year price drop at -15.9%:
>
Source: Standard & Poor’s
>
Negative contributions came from all of the usual suspects: Phoenix, San Francisco, Miami, Las Vegas and San Diego, all of which are down anywhere from 23%-30% YOY.
Note that the lack of accelerating price drops does *not* intimate accelerating price gains. Inventories remain elevated and foreclosures are only increasing.
What is the opposite of Cassandra? She was the cursed prophetress who had the ability to foresee the future, but an inability to convince anyone of the coming dangers. With housing, the usual cadre were once again calling the bottom. They are the anti-Cassandras — have a distinct inability to see the future (we’ve documented this all too many times) yet for some inexplicable reason, there seems to be a swath of people eager to lose money (again) believing them. It truly is astonishing to watch the lambs get led to the slaughter month after month by the same siren call. (Note that the perennial bottom calls is the basis for our guest post by Mark Thoma later this morning).
Dan Greenhaus of the Equity Strategy Group at Miller Tabak notes that "Like a bad hangover, time is the best medicine to help work through the glut of housing." We agree.
>
Source:
National Trend of Home Price Declines Continued through the First Half of 2008
S&P, August 26, 2008
http://tinyurl.com/june2008caseshiller
New home sales bounce in July was less than impressive
Many commentators seemed to turn a bit euphoric over the new home sales number that was released yesterday, but here are the facts, as we see them, beneath the headlines: The pundits are happy because new home sales appear to be bottoming out. Well, it is probably true that they are no longer sliding at an accelerating rate, but because of the downward revisions to May and June, the bounce in July (+2.3% to 515k) was less than impressive, in our view. The average of the past three months in this notoriously volatile indicator otherwise known as new home sales is 511,000, which is the weakest since September 1991. Adjust for size of the population, and home sales are at their lowest level since the summer of 1982 when the economy was trying to pull out of the worst recession of the post-war era.
Improvement in sales seems to be of the ‘spec’ variety Digging beneath the surface, sales have been concentrated in “units not yet started”, which rose 9% (caution: these data are not seasonally adjusted). So, all this improvement in sales seems to be of the ‘spec’ variety. Sales of “units under construction” were flat on the month. But sales of completed product plunged 18% and, yes, while the data are not adjusted for seasonal vagaries, this still went down as the second weakest July in the past twenty years.
Con man.
Barry, I think you’re missing an emerging turning point here (although I just was told Cramer said same thing last night, so I could be wrong). Markets move based on second derivatives generally, and the second derivative of housing prices in June improved. 9 of the 20 markets saw prices improve in June and 7 saw prices improve in the 3 months ended June. Moreover, in those markets where prices are still falling month over month, they are doing so at slower rates. My model shows declines decelerating in 9 markets and accelerating in none, with 2 markets mixed. This is a vast improvement over October to Feb when all 20 were accelerating to the downside.
Up Slower Faster Mixed
0607 – – 20 –
0707 6 3 6 5
0807 1 3 10 6
0907 – 3 14 3
1007 – – 20 –
1107 – – 20 –
1207 – – 20 –
0108 – – 20 –
0208 – – 20 –
0308 – – 16 4
0408 5 – 10 5
0508 6 3 2 9
0608 9 9 – 2
That table didn’t post well, hopefully this will be better.
MMYY_Up__Slower_Faster_Mixed
0607_0_____0______20_____0
0707_6_____3______6______5
0807_1_____3______10_____6
0907_0_____3______14_____3
1007_0_____0______20_____0
1107_0_____0______20_____0
1207_0_____0______20_____0
0108_0_____0______20_____0
0208_0_____0______20_____0
0308_0_____0______16_____4
0408_5_____0______10_____5
0508_6_____3______2______9
0608_9_____9______0______2
Regarding the glut in housing:
http://trendingrealestate.typepad.com/trendingrealestate/2008/08/inventory-conti.html
It’s all about inventory. Until inventory declines meaningfully, the bottom callers will be revealed to be what they are – tools.
The recent slowing in the pace of fall of home prices (Case Shiller composite) is an illusion.
The data is not seasonally adjusted. On average, monthly price gains in the Q2 are 0.64% higher than Q1. This number is calculated pre bust. Since the bust started, the Q2 monthly changes are still 0.54% higher than Q1. So prices falls should be less severe in Q2 than in Q1 and that is what is showing up in the numbers. IMPROVEMENT IN THE HOUSING MARKET IT IS NOT.
Guys, if you would like the monthly seasonal factors that I have calculated, it makes for a pretty chart:
Month 1987-2005 Last 2 yrs
Jan-Dec monthly seasonals
-0.34%,-0.26%,0.01%,0.28%,0.47%,0.58%,0.33%,0.14% ,-0.13%,-0.25%,-0.41%,-0.42%
Jan-Dec monthly changes since bust started:
-0.84%,-0.93%,-0.85%,-0.42%,-0.29%,-0.31%,0.26% ,0.04%,0.05%,-0.16%,-0.57%,-0.80%
@GDM – uh, guy… you’re honestly trying to compare April – June to October – February? Lol, lol, lol!
If you compare retail sales from October – December versus Feb – April, I bet you’ll see some stunning differences, too!
In many areas prices are starting to stabilize. In Florida and California, it will take years for things to stabilize.
Things definitely could change, but the fact that this crisis is becoming regional is a positive.
At ReturnFreeRisk’s suggestion, I ran the data for the 20 markets through BLS’s X12-Arima seasonal adjustment program. I then created a diffusion index to measure strength. A reading of +100% means all 20 markets have prices rising at an accelerating rate, -100% means all 20 have prices falling at an accelerating rate.
It bottomed at -92% in Nov-07, bounced to -77% in Feb-08 retested at -93% in Mar-08 and has recovered all the way to -2%. Of course we did a similar thing in Nov-06, rebounding from -70% to +7% by Mar-07. But this feels and looks like the BEGINNING of a bottoming process that, barring further dislocations in the mortgage market, could bode well for homebuilders in a year and homebuilding stocks over the next 6 months.
If anyone wants the model, give me your email and I’ll send.
Kerry and GDM,
I have analyzed these data and came to a similar conclusion as you. Even with controlling for seasonal variance (by anchoring the data to the same month from a previous year), the rate of decline is decelerating. I think it is even possible to see a positive value for the composite 20 in the next couple of months – somethiing we haven’t seen since July 2006. The seasonal uptick would be a part of that, but it still would never happen if we were experiencing the rapidly declining prices we were 7-8 months ago.
I am a statistical analyst by trade and do not work in the real estate profession
While it’s easy to adjust for seasonality, that’s the wrong approach. The C-S Index reflects seasonality – and for the next 6 months prices decline relative to the annualized trend. We have just finished the good months – and now we are facing the time of year that historically is slow.
Over the past 20 plus years of the index, every one of the next 6 months faces a lower rate of housing appreciation. Add the fact that we are coming from a period when prices movement is negative in almost every market, and you can expect to see continued deterioration until February data is released.