Matthew Yglesias » Home Prices Stop Falling, But Why?

Jul 28th, 2009 at 4:44 pm

Home Prices Stop Falling, But Why?

New Case-Shiller data has the decline in housing prices coming to an end:

Which naturally raises the question: Why? Prices aren’t back down to their baseline level. Felix Salmon says “If housing kept track with CPI inflation, the Case-Shiller index would be at 125 now; in fact, it’s at 140.” What’s more, there’s still lots of unsold inventory. Common sense says prices need to keep falling until all the recently-built, currently-vacant houses can be sold off.

Comments
40

Filed under: Economy, Housing,

40 Responses to “Home Prices Stop Falling, But Why?”

Willie Says:
July 28th, 2009 at 4:52 pm
Common sense says prices need to keep falling until all the recently-built, currently-vacant houses can be sold off.

Not if most of the vacant housing is concentrated in areas with very low demand.
alkali Says:
July 28th, 2009 at 4:56 pm
What Willie said. The housing market collapse in overbuilt Stockton, CA is not going to be resolved for years. Manhattan might be resolved now.
fusion Says:
July 28th, 2009 at 5:02 pm
A single swallow does not the spring make. Let’s see what the data look like over the next few months.
daveNYC Says:
July 28th, 2009 at 5:02 pm
Real vs Nominal. If housing prices stay flat, inflation will take care of bringing the index down a few more notches.
Walker Says:
July 28th, 2009 at 5:04 pm
CR mentioned this earlier, but this graph is not seasonably adjusted. Housing prices always go up in May (trying to get the house before the school year starts) and so the prices must be adjusted seasonally.

There is a seasonably adjusted version of the Case Schiller info, but the NYT went to press before it was finished.

Prices are still going down.
Brent Says:
July 28th, 2009 at 5:05 pm
Right. I mean, we established that regional variations didn’t preclude a nationwide bubble on the way up. But of course not every region experienced the bubble in the same way. Similarly, we’re definitely seeing regional variation on the way down. And the insanity of the Sun Belt just totally distorts all of this.
hankest Says:
July 28th, 2009 at 5:05 pm
They forgot to mention the data weren’t seasonally adjusted…

Barry Ritholtz helpfully explains and links to better explanations.

http://www.ritholtz.com/blog/2009/07/case-shiller-seasonal-adjustments/

“Manhattan might be resolved now.” Really? Based on what? Wishful thinking?
DTM Says:
July 28th, 2009 at 5:07 pm
As pointed out in the linked article, those are the unadjusted numbers. The adjusted numbers show prices still falling, and if ever there was a situation in which you need seasonal adjustments, it is for housing prices.
sacman701 Says:
July 28th, 2009 at 5:07 pm
If inventory is declining every month at current prices, then prices don’t necessarily have to keep falling for the market to eventually soak up all that inventory.

A better measure of whether houses are over- or underpriced is the ratio of home prices to rents or to average incomes, but you also have to adjust for interest rates.
DTM Says:
July 28th, 2009 at 5:08 pm
And I see a couple people were quicker on the draw.
Jonathan Says:
July 28th, 2009 at 5:08 pm
As Alkali and Willie both mentioned, excess inventory is probably resigned to a few areas, and might not have a huge effect on nationwide house pricing.

But I do think there’s a psychological factor involved here. I think people that might be in the market for a home are worried that home prices will shoot up dramatically once there is mild improvement in the economy, whereas home-sellers probably just want to get rid of the home asap to stop having to pay the mortgage on it. Also, compared to the peak of the housing bubble, home prices seem more reasonable than they probably should.

So, the perception that we might be nearing the low point of the recession (perhaps driven by news of finance company profits), people that can buy a home are pulling the trigger early, to ensure that they get the home at what they perceive to be a low price, while they still can.
Adam Kotsko Says:
July 28th, 2009 at 5:17 pm
Thank you for using “raises the question” rather than resorting to the all-too-common abuse of “begs the question.”
Walker Says:
July 28th, 2009 at 5:20 pm
CR has a new post on the housing reports. He thinks that the news is really distorting the data and putting us in danger of a very negative surprise in October.
raylward Says:
July 28th, 2009 at 5:32 pm
Except for places where land is unavailable (e.g., NYC), over time the price of housing equals the cost of housing. And, notwithstanding what some may believe, the sun rises in the east.
WDF Says:
July 28th, 2009 at 5:35 pm
I am a real estate attorney in Columbus, Ohio. My wife is a real estate agent. A lot of the activity we are seeing in the marketplace right now is being driven by the federal first-time home buyer program. It will be interesting to see what happens when (if?) the program expires. My guess is that the administration and/or the federal reserve will try to keep in place policies that have the effect of slowing the rate of price decline. As we’ve seen here this summer, price stability, even if defined as slowly declining prices, can lead to considerably more market activity, which in turn will undoubtedly help stablize the economy.
Rickster Says:
July 28th, 2009 at 5:36 pm
A lot of folks wish this was true, but as others have pointed out Calculated Risk this figure is not seasonally adjusted. In addition, the incentive to use that $8,000 tax credit before expires mean that there will be a pick up in sales to use the credit before it expires. Since to take the credit the house has to close before the end of the year, this will lead to a spike in sales and then a fall. Foreclosures caused by unemployment are picking up again. So although I think the curve of decline is starting to flatten, I expect price declines will go on into at least the Spring of 2011 when much of the inventory overhang will be exhausted and employment will begin to improve.
bob h Says:
July 28th, 2009 at 5:41 pm
Perhaps it is the support provided by 5% 30-year rates.
MikeF Says:
July 28th, 2009 at 6:15 pm
Let me be the first to mention: the data presented is not seasonally adjusted. (CR has more)
pseudonymous in nc Says:
July 28th, 2009 at 6:46 pm
A lot of the activity we are seeing in the marketplace right now is being driven by the federal first-time home buyer program.

Yep. First-time buyers are also less likely to be going after foreclosures, because of the extra work and inside info required to beat the specvestors — they’re going to the MLS listings, perhaps paying slightly more than could be bargained by a savvier buyer — but getting back closing costs and a chunk of their downpayment via the IRS.
RW Says:
July 28th, 2009 at 6:53 pm
as others have pointed out Calculated Risk this figure is not seasonally adjusted.

What’s quite amusing about this is that CR had absolutely no problem whatsoever crowing about Case Shiller’s unadjusted numbers during prior periods when they were indicating declines.

What’s also interesting is that CR didn’t bother to report that the seasonally adjusted numbers likewise showed gains in several markets and improvement from the prior month’s report.

In addition, CR didn’t bother to report that the significance of the CS report is that this is the third month to exhibit some sort of bottoming, which taken as a whole suggests that it may be a change in the overall trend. The comments coming out of S&P are not based upon just a single month’s worth of data.

Blogs such as Calculated Risk would be quite unhappy with any signs of a bottom, as a change in sentiment might provide its readers with less reason to visit the site. They’ve built a brand around a particular perspective, and I would expect them to fight to keep it long after it is no longer accurate.
howard Says:
July 28th, 2009 at 8:06 pm
here i was, happily reading through a set of comments that covered all i had to say, and i come across RW’s cheap cynicism. pretty sad, pal.
DTM Says:
July 28th, 2009 at 9:50 pm
What’s also interesting is that CR didn’t bother to report that the seasonally adjusted numbers likewise showed gains in several markets . . . .

Hmmm. From the post when the seasonally-adjusted numbers came out:

Seasonally adjusted, prices fell in 12 of the 20 Case Shiller cities.

Is the idea that Calculated Risk should be faulted for not subtracting 12 from 20 for us?

Blogs such as Calculated Risk would be quite unhappy with any signs of a bottom, as a change in sentiment might provide its readers with less reason to visit the site. They’ve built a brand around a particular perspective, and I would expect them to fight to keep it long after it is no longer accurate.

Calculated Risk has already suggested several bottoms/peaks may have occurred, including light vehicle auto sales, housing starts, and initial weekly unemployment claims.

Of course I’m not suggesting Calculated Risk doesn’t have a certain general perspective on these matters. But I actually don’t see any hint of intellectual dishonesty.
RW Says:
July 28th, 2009 at 10:11 pm
Is the idea that Calculated Risk should be faulted for not subtracting 12 from 20 for us?

It’s a bit funny that there would be no references to the horrors of seasonality in CR’s posts during the winter months. No cautionary tales of how numbers might be skewed by the fact that hardly anyone moves in January, no reminders of chestnuts roasting on an open fire and how they might deter winter sales.

But now suddenly, after many months of providing the unadjusted figures, there is this compelling need to address this sudden crisis point. Even last summer, I don’t recall any mention of this.

I actually don’t see any hint of intellectual dishonesty.

I recall them predicting the bottom as occurring around 2012. So perhaps it might be considered bad form to overshoot by three years.

It’s also dishonest to not mention the fact that CS numbers have not exhibited these alleged signs of seasonality. Here’s the Composite Index going back to when it began in January 2000; you tell me if you see the peaks and valleys that he implies are typical of it:

January 2000 – 100.00
February 2000 – 100.76
March 2000 – 101.95
April 2000 – 103.50
May 2000 – 105.20
June 2000 – 106.76
July 2000 – 107.77
August 2000 – 108.64
September 2000 – 109.35
October 2000 – 110.04
November 2000 – 110.81
December 2000 – 111.58
January 2001 – 112.39
February 2001 – 113.07
March 2001 – 114.14
April 2001 – 115.29
May 2001 – 116.24
June 2001 – 117.29
July 2001 – 118.20
August 2001 – 119.09
September 2001 – 119.84
October 2001 – 120.31
November 2001 – 120.53
December 2001 – 120.43
January 2002 – 120.64
February 2002 – 121.06
March 2002 – 122.30
April 2002 – 123.92
May 2002 – 125.86
June 2002 – 127.82
July 2002 – 129.66
August 2002 – 131.22
September 2002 – 132.43
October 2002 – 133.55
November 2002 – 134.41
December 2002 – 135.15
January 2003 – 135.64
February 2003 – 136.19
March 2003 – 137.20
April 2003 – 138.56
May 2003 – 140.06
June 2003 – 141.39
July 2003 – 142.99
August 2003 – 144.56
September 2003 – 146.28
October 2003 – 147.82
November 2003 – 149.22
December 2003 – 150.49
January 2004 – 151.69
February 2004 – 153.10
March 2004 – 155.49
April 2004 – 158.47
May 2004 – 161.60
June 2004 – 164.82
July 2004 – 167.43
August 2004 – 169.31
September 2004 – 170.96
October 2004 – 172.41
November 2004 – 173.65
December 2004 – 174.83
January 2005 – 176.44
February 2005 – 178.50
March 2005 – 181.30
April 2005 – 184.24
May 2005 – 187.21
June 2005 – 190.10
July 2005 – 192.67
August 2005 – 194.98
September 2005 – 197.36
October 2005 – 199.40
November 2005 – 200.97
December 2005 – 201.97
January 2006 – 202.44
February 2006 – 203.19
March 2006 – 203.65
April 2006 – 204.82
May 2006 – 205.86
June 2006 – 206.38
July 2006 – 206.52
August 2006 – 206.18
September 2006 – 205.80
October 2006 – 205.41
November 2006 – 204.65
December 2006 – 203.33
January 2007 – 202.31
February 2007 – 201.57
March 2007 – 201.01
April 2007 – 200.54
May 2007 – 200.12
June 2007 – 199.44
July 2007 – 198.72
August 2007 – 197.37
September 2007 – 195.69
October 2007 – 192.98
November 2007 – 188.94
December 2007 – 184.96
January 2008 – 180.68
February 2008 – 175.96
March 2008 – 172.20
April 2008 – 169.98
May 2008 – 168.60
June 2008 – 167.77
July 2008 – 166.36
August 2008 – 164.64
September 2008 – 161.64
October 2008 – 158.09
November 2008 – 154.51
December 2008 – 150.56
January 2009 – 146.34
February 2009 – 143.10
March 2009 – 139.99
April 2009 – 139.21
May 2009 – 139.84
DTM Says:
July 28th, 2009 at 11:15 pm
It’s a bit funny that there would be no references to the horrors of seasonality in CR’s posts during the winter months. No cautionary tales of how numbers might be skewed by the fact that hardly anyone moves in January, no reminders of chestnuts roasting on an open fire and how they might deter winter sales.

Here is an 4/1/2009 post specifically addressing the importance of looking at seasonality in Case Shiller:

Case Shiller House Prices Seasonal Pattern

That is the day after Case Shiller released their January numbers (on 3/31/2009). I’m noticing a pattern here: when you say Calculated Risk hasn’t done something, there appears to be a good chance that is false.

I recall them predicting the bottom as occurring around 2012. So perhaps it might be considered bad form to overshoot by three years.

I think we have ample evidence your recollections are not exactly to be trusted. Anyway, this is what Calculated Risk recently posted:

It is way too early to try to call the bottom in prices. House prices will probably fall for another year or more. My original prediction (a few years ago) was that real house prices would fall for 5 to 7 years (after 2005), and we could start looking for a bottom in the 2010 to 2012 time frame for the bubble areas. That still seems reasonable to me.

We’ll have to see, but I personally don’t think that prediction looks particularly bad right now.

It’s also dishonest to not mention the fact that CS numbers have not exhibited these alleged signs of seasonality.

Here is a graph comparing NSA and SA month-to-month numbers for the Composite 10 going back to 1988:

Case-Shiller House Prices Seasonal Pattern

You tell me if you DON’T see the relevant peaks and valleys. And obviously the seasonal adjustments have helped over time.

I can’t eyeball your (completely unnecessary) column of numbers and do the relevant calculations in my head. But I strongly suspect it will exhibit the same pattern if you do the math.
Aaron Weiner Says:
July 28th, 2009 at 11:25 pm
Here’s the unadjusted line, taken straight from the column of numbers RW posted and filtered through me being slapdash. The black dots are winter months, the blue dots are non-winter months, and the black lines are linear interpolations between winters. As you can see, there are no bumps and troughs, but in all but one case the non-winter prices exceeded what you’d expect from the winters on either side.
DTM Says:
July 28th, 2009 at 11:46 pm
And here are some of my own hand calculations based on the series above. This is percent increases (non-annualized) from December to March and then the subsequent March to June for a few years:

12/01-3/02 1.55%
3/02-6/02 4.51%

12/02-3/03 1.52%
3/03-6/03 3.05%

12/03-3/04 3.32%
3/04-6/04 6.00%

12/04-3/05 3.70%
3/05-6/05 4.85%

Of course this was a time of steadily increasing prices in general, but the price increases obviously always accelerated significantly between these two consecutive seasons.

But as I implied way above, I don’t think there is anyone who knows anything about house prices who actually thinks seasonal adjustments are unnecessary. So we’re really just reinventing the wheel here.
RW Says:
July 28th, 2009 at 11:50 pm
I can’t eyeball your (completely unnecessary) column of numbers

Why would you read someone’s blog but not look at the underlying information? Being that the source data is readily available, you could quite easily look at it yourself, rather than accepting someone else’s interpretation as gospel.

If you were to go directly to the source at CS, download the spreadsheet and plot it out, you will see no peaks or valleys in the month of May in those prior years. Then again, you don’t need to bother, as it is clear in my posting of the index that May is not a particularly disruptive month. During up years, the May figure is above April and below June; during down years, May is below April and above June. No sudden surges prior to a drop off, as implied in the recent revelatory posts. No need to reinvent numbers when you can just look at them yourself.

During these prior posts of the unadjusted data, there was no urge to qualify the horrors of seasonality. When the numbers declined as they did, they were accepted without qualification as being signs of doom. For example, during this discussion of wintry, frigid February, there was no apparent need to make grand qualifications about the impacts of seasonality — after all, the numbers sucked — and the unadjusted figures are cited without any reservation: http://www.calculatedriskblog.com/2009/04/case-shiller-city-data.html

We’ll have to see, but I personally don’t think that prediction looks particularly bad right now.

Perhaps they make lots of predictions so that some of them will be right. Here was a comment from last year:

Also look at the length of the housing bust in the early ’90s. It took over six years from peak to trough in some cities. If this bust takes the same amount of time, prices will not bottom in some cities until 2012 (or there about).

http://www.calculatedriskblog.com/2008/03/real-case-shiller-house-price-index.html
DTM Says:
July 29th, 2009 at 1:13 am
Why would you read someone’s blog but not look at the underlying information?

I meant it was unnecessary to post all those numbers–it is easy enough to link to the Case-Shiller page.

. . . rather than accepting someone else’s interpretation as gospel.

I’m doing nothing of the kind. Saying I have detected no intellectual dishonesty isn’t the same thing as saying I always agree.

During up years, the May figure is above April and below June; during down years, May is below April and above June.

Right, which is why you need to look at the relative magnitude of the month-to-month changes at different times of year. To be brutally honest, it is becoming clear to me you don’t understand the nature of the seasonal adjustments people apply to housing prices.

During these prior posts of the unadjusted data, there was no urge to qualify the horrors of seasonality.

I just provided a link which refutes this claim: Calculated Risk specifically noted the need to adjust for seasonality in Case Shiller less than a month prior to the post you linked.

Perhaps they make lots of predictions so that some of them will be right. Here was a comment from last year . . . .

Note that the following prediction was about “some cities”. Again we’ll have to see, but I wouldn’t be at all surprised if that ended up true as well (that some cities don’t bottom in real terms until around 2012).

Look, it is fine to disagree with various conclusions or predictions at Calculated Risk. But you clearly have no real evidence to support your original charges of intellectual dishonesty, and indeed you have repeatedly misrepresented what Calculated Risk has posted on various subjects, sometimes even after being confronted with direct evidence to the contrary. So when it comes to intellectual dishonesty, I think it is time for you to remove the beam from your own eye.
RW Says:
July 29th, 2009 at 2:05 am
Right, which is why you need to look at the relative magnitude of the month-to-month changes at different times of year.

You keep talking around the point; perhaps you don’t understand it.

In CR’s early post about the May data, there is clearly a mad scramble to confront this little problem of improvements in the data, which was used as a launching pad for this seasonality story that attempts to explain away this unacceptably positive change in the numbers: http://www.calculatedriskblog.com/2009/07/case-shiller-house-prices-for-may.html Much emphasis on the negative, with no references to the fact that there were also improvements exhibited during April and March, meaning that there are three months of data that form a potential base for May’s outcome.

Since you keep harping on seasonality, I’ve asked you to look at the other May results to see how they compare to the April prior to it and the June that follows it — after all, the heart of CR’s seasonality argument is that May behaves in ways that are vastly different from other months.

But as it turns out, during every single year that the Composite 20 has been compiled, there hasn’t been one time when May has defied the prevailing trend. Not once. Not what one would expect, given the alleged seasonality problem.

As noted in the index numbers above, during years when May increases, it improves upon April, while June subsequently increases as well. During down years when May decreases, it declines to levels below that of April and June declines further still. The chart accompanying this article correlates to those figures, and you’ll notice no differences there, either.

So no, contrary to CR’s assertion, there is no history of May behaving like an bizarre aggressively upbeat outlier, and there is no track record of May ticking upward during declining markets.

Now, if you take the trouble to look at this most recent May data in the context of April and March, you can see a pattern of ongoing improvement, as increasing numbers of individual markets change their direction from negative to positive. If that is supposed to be a blip, then it’s a three-month, steadily improving blip, not a one-month ohmigawditsallseasonalandunique! anomaly as implied by . And as noted, May has not once contradicted the prevailing underlying trend since the 20-city index was begun in 2000, so there is no spike in the unadjusted numbers that matches the message of the seasonality story.

If you review the CR data directly instead of relying upon interpretations of same, then it should be evident that the CR posts emphasize certain data points, while sidestepping or deliberately omitting others. For example, they offer this observation — “At the other end of the spectrum, prices in Dallas are only off about 8% from the peak” — while neglecting to mention that Dallas has shown price increases since February, with prices up 3.7% since then. Interesting that three months of recent improvement warrant no mention whatsoever, while the spotlight was placed on the long-run price declines.

CR is trying to tell a story that is consistent with its brand, while trying eagerly to apply this seasonality story as a distraction to what has been transpiring for a full quarter. If June shows improvement as did May, then there will surely be an excuse for that, too.
DTM Says:
July 29th, 2009 at 9:45 am
I’ve asked you to look at the other May results to see how they compare to the April prior to it and the June that follows it — after all, the heart of CR’s seasonality argument is that May behaves in ways that are vastly different from other months.

This is very wrong, and shows that you don’t actually have a basic grasp on how the seasonal adjustments for Case-Shiller work. The seasonal adjustments aren’t somehow just about the month of May: they are about a pattern throughout the annual cycle, a pattern which has in fact repeatedly occurred (see more below).

But as it turns out, during every single year that the Composite 20 has been compiled, there hasn’t been one time when May has defied the prevailing trend. Not once. Not what one would expect, given the alleged seasonality problem.

Again, you are just displaying here your lack of basic comprehension of how seasonal adjustments work with respect to house prices. First, as I noted above, this is not about May in particular. Second, there is no reason to expect that seasonal adjustments will often lead to “defying the prevailing trend”, if by that you mean actually reversing the sign of the nonseasonal trend. That is actually going to be a very rare event: it will only happen on those rare occasions when the nonseasonal price trend is very close to zero–as it happens to be now.

Rather, the normal result of the seasonal effect is just what I briefly demonstrated above using your data series: in certain seasons the nonseasonal price trend will be relatively decelerated, and in other seasons the nonseasonal price trend will be relatively accelerated. That is also the same effect demonstrated by the month-to-month change in the Composite 10 chart I linked above. Incidentally, I find it quite interesting that you have never actually addressed the substance of that chart, which I would guess is either because you don’t understand it, or because it so clearly demonstrates a seasonal effect that you just want to pretend it doesn’t exist.

Anyway, as also indicated in that chart, the seasonal adjustments typically don’t have the effect of changing the sign of the trend, but instead just smooth it out somewhat. Nonetheless, on rare occasions a reversal of sign does happen. In fact, I believe that the last time it happened was in June and July of 2006: the unadjusted price index was still going up, but after seasonal adjustment it had started to go down. Again, this was a period when the nonseasonal tend happened to be around zero.

For example, they offer this observation — “At the other end of the spectrum, prices in Dallas are only off about 8% from the peak” — while neglecting to mention that Dallas has shown price increases since February, with prices up 3.7% since then.

Here is a chart from Calculated Risk clearly indicating that in Dallas, prices have increased in 2009 until May:

Case-Shiller Price Declines from Peak, by Year and City

It really is amusing to me how many times now you have claimed that Calculated Risk is hiding some fact, and then I have easily found that fact clearly represented in one of Calculated Risk’s post.

In any event, all this talk about seasonal adjustment per se is really a waste of time–again, no one remotely serious and well-informed when it comes to housing prices thinks that seasonal adjustments are unwarranted. The much more interesting thing that Calculated Risk recently wrote is that he thinks the historically-based seasonal adjustments may end up being too mild for this particular seasonal cycle. I can see why he thinks that, but it is going to be impossible to tell one way or another until we have the full seasonal cycle to look at.
RW Says:
July 29th, 2009 at 10:08 am
This is very wrong, and shows that you don’t actually have a basic grasp on how the seasonal adjustments for Case-Shiller work.

Er, you apparently don’t realize that the CS numbers that get most of the press attention are not seasonally adjusted!

Maybe you should actually look at the underlying data for a change, instead of allowing CR to digest it for you. http://www2.standardandpoors.com/portal/site/sp/en/us/page.topic/indices_csmahp/0,0,0,0,0,0,0,0,0,1,1,0,0,0,0,0.html

Notice that CS releases two sets of numbers each month, including one seasonally adjusted set. But the seasonally adjusted numbers are given little attention, including by CS themselves. When the press discusses these numbers, they talk about the non-seasonal numbers. In the CR press releases, they emphasis the non-seasonal numbers.

And in past months, just like everyone else, CR was perfectly happy to deliver pronouncements based upon these non-seasonal numbers. When they did discuss seasonality, it was not done with the agenda of discrediting the non-seasonal figures, as has been the case with this latest report.

It’s only this month that the non-seasonal figures are suddenly lobbed into the doubt column. When they were all dutifly downtrending, there was no apparent need to qualify the results.

no one remotely serious and well-informed when it comes to housing prices thinks that seasonal adjustments are unwarranted.

Again, you obviously miss the point that CR had no problem with numbers that weren’t seasonally adjusted — these are the ones that get the press attention — in previous months. But now, they’re a big deal and has thrown the world into a tailspin. You can figure out for yourself why the sudden change of heart.

So sorry, you’re totally missing it. CR is clearly pushing the seasonal angle this month in ways that they have not bothered before because they aren’t pleased by the potential reversal.

Fortunately for them (and presumably for you), I suspect that June numbers will probably worsen, not because of the horrors of the seasons but because rates ticked upward and consumer sentiment have since worsened. Enjoy doing your cartwheels in 30 days, I’m reasonably sure that you’ll feel vindicated.
DTM Says:
July 29th, 2009 at 10:49 am
Er, you apparently don’t realize that the CS numbers that get most of the press attention are not seasonally adjusted!

First, it hasn’t escaped me that you have abandoned any attempt to address seasonal adjustments on the merits. Second, I have no idea if what you are claiming about the press is true, and for obvious reasons I am not going to take your word for it–you have a clearly established track record of being very wrong about what other people have writeen. Third, in any event I really couldn’t care any less about how the press typically reports on financial matters, or really anything involving numbers.

When [Calculated Risk] did discuss seasonality, it was not done with the agenda of discrediting the non-seasonal figures, as has been the case with this latest report.

I’m not even sure what this latest attempt to move the goalposts is supposed to mean. But here is what Calculated Risk wrote back on 4/1/09, when Case Shiller had just released their January numbers:

This graph shows the seasonal pattern for the Case-Shiller Composite 10 house price index. The percentage change is the month-to-month change annualized. It is important to remember this clear seasonal pattern when looking at the Case-Shiller data.

That statement is perfectly neutral: it applied to those bad January numbers, it applies to the better May numbers, it applies to any numbers that come along. Any perceived bias in this statement is a product of your imagination.

Fortunately for them (and presumably for you), I suspect that June numbers will probably worsen, not because of the horrors of the seasons but because rates ticked upward and consumer sentiment have since worsened. Enjoy doing your cartwheels in 30 days, I’m reasonably sure that you’ll feel vindicated.

You are projecting: I don’t have any ego invested in what happens with house prices in the future. In fact, I would be quite pleased if Calculated Risk’s predictions turned out to be too pessimistic, and I certainly think that is at least possible.
RW Says:
July 29th, 2009 at 11:02 am
it hasn’t escaped me that you have abandoned any attempt to address seasonal adjustments on the merits.

It has no merits. Until this month, non-seasonal numbers were routinely cited at face value by everyone, including by Shiller when interviewed, as well as by CR itself. It’s only now that they suggest a reversal that we feel this compelling need to freak out over using them.

You are obviously distracted; CR’s diversion tactic worked on you, so congratulations. You can’t even be bothered to compare May to April and March, and note the changes for yourself.

That statement is perfectly neutral

Not this month, it isn’t. Suddenly, it is being used to modify the veracity of the non-seasonal numbers, which is a new twist. Previously, CR lauded CS’s unadjusted figures for their accuracy; now, we apparently need to lose sleep over them because these numbers really bother CR.

Throughout this change of events, Case Shiller’s methodology has remained unchanged. The process used to produce the non-seasonal numbers that were eminently quotable without reservation last year is still being used now. It’s the output that’s driving this compelling urge to diminish the value of unadjusted numbers.
DTM Says:
July 29th, 2009 at 11:44 am
Until this month, non-seasonal numbers were routinely cited at face value by everyone . . .

Again, you have zero credibility when it comes to what other people have written.

You can’t even be bothered to compare May to April and March, and note the changes for yourself.

Again, this just shows that you don’t even understand what the seasonal adjustments are about. Meanwhile, you can’t be bothered to address the very obvious seasonal pattern in the chart I linked, or the numbers I calculated from your series, or so on. Again, I’m not sure if that is because you are intellectually dishonest, or just can’t understand what you are looking at.

Previously, CR lauded CS’s unadjusted figures for their accuracy . . . .

Again, I can’t tell whether this is a lie, or just a failure to comprehend that 4/1/09 post I linked and then quoted.

Anyway, I’ve debunked all your claims, and you’ve clearly got nothing new to say at this point. But feel free to have the last word anyway, if you so choose.
ostap Says:
July 29th, 2009 at 12:43 pm
Why? Prices aren’t back down to their baseline level. Felix Salmon says “If housing kept track with CPI inflation, the Case-Shiller index would be at 125 now; in fact, it’s at 140.”

What exactly is so magical about the baseline? Maybe the baseline was too low. Or maybe many other possible factors.
DTM Says:
July 29th, 2009 at 1:34 pm
What exactly is so magical about the baseline? Maybe the baseline was too low.

Indeed–in fact I don’t think it is unreasonable to expect a small CPI-adjusted increase in housing prices over the long term. Nothing like what we saw during the bubble, of course, but say something on the order of 1-2% per year.
Lord Says:
July 29th, 2009 at 2:16 pm
Inflation is the wrong index to use for large coastal metropolises. The constraint on urban land, the expense and impracticality of commuting, mean prices in these areas rise with incomes and growth of the areas.
tps12 Says:
July 29th, 2009 at 3:07 pm
One to bookmark for future lulz. Mission Accomplished!
DTM Says:
July 29th, 2009 at 3:47 pm
Inflation is the wrong index to use for large coastal metropolises. The constraint on urban land, the expense and impracticality of commuting, mean prices in these areas rise with incomes and growth of the areas.

Yep, that is the basic explanation for why housing prices can (slightly) outpace general inflation. Of course you can always found new cities, but the startup costs are huge, and we are running out of prime places for cities anyway (hence building big cities in the middle of deserts and such, which is proving to be a somewhat questionable model).
Ivan Says:
July 30th, 2009 at 12:16 am
The following search:

http://www.google.com/search?q=site:calculatedriskblog.com+%22seasonally+adjusted%22

brings up 5 pages of results pointing to past Calculated Risk posts highlighting seasonally-adjusted numbers.

Anyone who claims that CR hasn’t been highlighting seasonality is either so uninformed as to render his perspective worthless or simply lying. Take your pick: ignorant or dishonest?

Matt Yglesias

Home Prices Stop Falling, But Why?

40 Responses to “Home Prices Stop Falling, But Why?”

Blogroll