Wednesday, February 27, 2008

Case Shiller impacts


I consider my real estate emotions article my signature article. Its the previous article to this one. However... you have to look at this Graph. It is the rate at which home prices are declining in a select number of large markets.

Notice something? The west coast markets are all burning down at 3.0% to 3.5% lost equity per month. For some reason Miami is burning down slower; I wonder if its due to fewer resales due to all the fresh construction being sold cheap. Washington DC is joining the crowd too.

What's that mean? Simple. The national credit crunch is causing all 'bubble markets' to converge on the same loss rate. Since all areas covered by Case-Shiller are declining, we can state with certainty that the entire US residential real estate market is now in decline.

The loss rate is fractionally higher than what I predicted. I was predicting that the bubble markets would converge on a loss rate of 2.5% to 3.0%. So the question is, why is it worse?

Got Popcorn?
Neil

19 comments:

sandman said...

Wow. That's my second reaction, after utter speechlessness.

Phoenix takes over the lead, and is the only city with an increasing rate of price declines. To put things in perspective, that's roughly a 40% yearly rate. I just don't see this being sustainable - maybe it's the fact that these numbers came from the holiday season?

Give it until at least Feb's numbers are in your graph, then we'll have a good idea of the sustainable trend. But let's assume that 3-3.5% holds. How long can that continue? If it stays through 2008 then the current foreclosure mess will look like a walk in the park. That, plus a recession, will result in an absolute bloodbath.

Here's to hoping that this is a kneejerk spike, for everybody's sake.

Chazu said...

It has been stated many times; in the case of DC: Case Shiller includes counties in West Virginia and other outlying areas that have nothing to do with the DC metro market.

Go ahead and look at an actual map of the United States. Note the distance between West Virginia and the nearest DC suburb that is arguably in the DC market; Vienna, VA

GT said...

yawn, dc includes WV. big deal. as if that is a substantial portion of the dc market? get over it. fact is, people who live in WV and even PA work in dc and commute everyday. but them being .3% of the case shiller numbers makes it irrelevant

Chazu said...

fact is, people who live in WV and even PA work in dc and commute everyday. but them being .3% of the case shiller numbers makes it irrelevant.

If you knew anything about Case Shiller calculations, you'd know that they don't include commute distances or time. They include homes located in very specific geographic locations. Such as including West Virginia in the DC Metro area. (which it is not) Get it?

Are you with Raytheon also?

Chazu said...

PS: someone who lives in Morgantown, West Virginia and works in town at the local Burger King sells his home, and his stats get included in Washington DC's Case Shiller index. That is a fact. See the problem now?

sandman said...

I fail to see why it matters either way. On a national level, a higher sample size only improves the calculations.

But maybe you're concentrating on just DC. Fair enough. But to claim that the numbers are invalid you'd have to show 2 things. First, that the WV numbers are statistically significant (i.e. if it's 0.3% of what's used in DC, it doesn't matter), and that there's a significant difference in the C-S index for the two areas.

Also, remember that Case-Shiller uses repeat sales of the same home. So absolute price differences between the two areas are irrelevant. The only number that matters is the difference in price declines between the areas, and my gut says that WV would only help prop up DC.

wannabuy said...

Case-Shiller is driver by sales. So the further you get away from a major urban center, the fewer sales... so the lower the weighting.

As Sandman has already noted, put it into perspective. How much are the ex-urb numbers impacting the total trend? Not enough. They might be helping hold up DC. ;)

Oh, Virginia has 15 counties listed as distressed markets by Wells Fargo. 13 more in Maryland. Hmmm...
http://biz.yahoo.com/bizj/
080228/1597695.html?.v=1

It comes down to one constant story, prices are dropping fast. DC is falling apart. The inner areas always fall later... and then faster in any historically similar downturns. However... this downturn has broken the trend.

It is scary. If that spike persists... we'll have 30 million homes under water.

Got Popcorn?
Neil

Chazu said...

Gerrymandering.lookitup.

CRT said...

"Case-Shiller is driver by sales. So the further you get away from a major urban center, the fewer sales... so the lower the weighting."

Not exactly (at least not in DC). Right now, DC consists of 3 markets, if you will: (a) areas inside the beltway for which median prices are flat or up slightly since the peak (b) suburban markets which are down a good bit, and (c) exurban markets which are a bloodbath.

Sales in the suburbs far exceed that of either the exurbs or the urban core. If you look at MRIS sales numbers you typically see 5X as many sales in the suburbs than in the urban core. My guess is its about the same for suburban to exurban (5 to 1 weighting).

So in essence, in DC Case Shiller is the "story of the suburbs". The exurbs drag DC down a bit, but the core brings it back up a bit too. In each case however, the sales are so heavily titled toward the burbs that this is really all CS can tell us.

CRT said...

Since I posted this, Neil made a slight deviation from this argument over on the NOVA bubble blog - frustrated, I actually took the time to see how many sales were urban, suburban and exurban (instead of just guessing as I had above). The results -

Urban (Arl/Alex & DC) 248
Suburban (Ffx, Mont. & PG) 1,066
Exurban (all the rest) 963

So instead of the 1-5-1 "story of the suburbs" argument made above, it looks like Case Shiller is actually closer to a 1-4-4 story of the suburbs and exurbs.

Thus, using Case Shiller to tell you what is going on "IN DC" gives you an even more distorted picture than I previously thought.

Neil - Case Shiller gives us the best picture out there of whats happening in the entire MSA. But if you think it tells you what is going on in the urban core, you couldnt be further from the truth. Cant we just finally put this thing to bed?

wannabuy said...

CRT,

It is a diverse region. But it is accurately showing how the whole region is dropping. Otherwise, why did Wells Fargo just raise the down-payment limits on DC as if it were as bad as Florida, San Diego, or Phoenix?

There is some smoothing of Case-Shiller by where the existing housing stock is located. Their weighting system is pretty good.

Too many large metropolitan areas are dropping fast. With 1-4-4, assuming its weighted that way (which I don't believe it is), there would have to be a crash in the suburbs to cause that price drop. Step back. Its next to impossible for that level of free fall to be occurring. That implies the whole region is dropping.

Do not its also re-sales, not sales. So the ex-urbs will be weighted even less than their sales.

So its not distorted. A precise gauge of a neighborhood? Of course not. But sales help tell the story.

Got popcorn?
Neil

wannabuy said...

Take a look here:
The Michigan survey of consumer confidence is down at 70.8.

The last time it was that bad was during a recession in Feb of 1992.

My... 16 years just went bye-bye.

http://paper-money.blogspot.com/

Got Popcorn?
Neil

CRT said...

Neil – you are killing me. OK lets go over this based on your comments.

"It is a diverse region. But it is accurately showing how the whole region is dropping." Correct - CS takes all the counties in the DC area together and spits out a single regional number. There is no question that on the AVERAGE the region is down and down a lot.

"Otherwise, why did Wells Fargo just raise the down-payment limits on DC as if it were as bad as Florida, San Diego, or Phoenix?" I don’t know for sure. My guess is since they aren’t from here and don’t know the area, they do as you do, and rely on broad indicators such as CS – even their Wells Fargo indicator included counties 2 states away in West Virginia. Moreover, as you pointed out in your “redlining laws” entry, when one area is bad, they have to report the whole MSA as being in the same boat.

"There is some smoothing of Case-Shiller by where the existing housing stock is located." Correct – again, the housing stock is in the suburbs and beyond. The population of DC proper is less than 600,000 people. The population of the DC MSA (which Case Shiller uses) is 5,300,000 (coincidentally close to my weighting don’t you think). Thus, there is no question that the housing stock is in the suburbs and beyond.

"Too many large metropolitan areas are dropping fast. With 1-4-4, assuming its weighted that way (which I don't believe it is)," Neil – I know you have been provided this link before, but my guess is you aren’t actually looking at it – here it is again, http://www.mris.com/reports/stats/ This is where Harriet gets her County by County “decade of sales” that you have no problem referencing – go look it up yourself.

"there would have to be a crash in the suburbs to cause that price drop." BINGO – you finally are getting it – price drops in the tiny urban area are modest or non-existent, the Big suburbs have big drops, and almost as big exurbs ARE CRASHING!

"Step back. Its next to impossible for that level of free fall to be occurring. That implies the whole region is dropping." No – its EASY for the small drop reported by the tiny inner areas to be DWARFED by the large drop in the HUGE suburbs, and the freefall in the NEARLY AS HUGE EXURBS). AGAIN, here is my math based on resales reported by MRIS

Urban (Arl/Alex & DC) 248
Suburban (Ffx, Mont. & PG) 1,066
Exurban (all the rest) 963

You say you don’t believe this, Again, show me where this is wrong
http://www.mris.com/reports/stats/

"Do not its also re-sales, not sales. So the ex-urbs will be weighted even less than their sales." I didn’t understand this sentence – please restate if it is necessary.

Look – in any event, all I am doing is using Harriet’s data (which you have quoted yourself) is that the housing stock is basically evenly split between the suburbs and the exurbs, with the inner core being a tiny afterthought. I hope you finally see this and finally put this incorrect argument about it accurately showing what is happening “in the city” to bed.

wannabuy said...

CRT,

Have you ever stress tested large mathmatical models? Complex ones for CFD, finite element, chemistry, or other similar concepts? I have. You're trying to argue like a lawyer; I'm pointing out that it doesn't matter. The core is going down too.

Harriet's data has sales. Good. But the inner core isn't an afterthought. It part of the whole mess.

There isn't some magically protected area. Have you ever heard of the substitution effect? Housing is the #1 example market for it! If housing 5 miles out drops 20%, assuming no bridge has collapsed, then it will put down the value of housing 5 miles further in.

Nothing wrong with the data. But look at it as a stress tested model. If the core of DC was holding steady, to have the losses we're seeing, would imply worse economic devistation that what is being reported.

Got Popcorn?
Neil

wannabuy said...

I should have pointed out that housing prices follow a diffusion model format. While known gradients in prices can vary, a change in one areas prices will "diffuse over" into another area. Thus, with Case-Shiller showing the declines that it has been for all of DC, its safe to assume the 'inner core' has also lost value.

Median prices don't mean anything, we housing bears have known that for a long time. How much will that home really sell for that someone might buy?

Mortgages keep getting tighter and tighter, so the January and February results will be interesting.

Did everyone see the car sales?
Ouch!

Got Popcorn?
Neil

CRT said...

Neil – let me paraphrase the entire argument:

Neil – CS is driven by sales. So the further you get away from a major urban center, the fewer sales… so lower the weighting.

Neil (on the nova bubble blog) - Case-Shiller is being driven down by the actual re-sales; that should be mostly in the inner area.

CRT – Not exactly, the re-sales (and therefore weighting) is in the suburbs and exurbs – generally on a basis of 1-4-4 (urban, suburban, exurban). Therefore it doesn’t give you a good picture of whats going on in urban areas.

Neil – it cant be weighted 1-4-4.

CRT – it is weighted 1-4-4 and heres the sales data that shows it.

Neil – Im an engineer – you don’t know how to scrutinize this stuff – im sure the whole region is going down.

WOW – quite a divergence in that last response there. While you earlier had no problem engaging me on mutually understandable terms, you now are arguing with terms like “stress testing", "finite element", "CFD" and now "dissusion model". It looks like you are trying to add new elements to the discussion because you assume (correctly) I don’t understand how these things work.

If you want, I can respond to you by saying “you are wrong because I’m a real estate attorney – the driving factor here is the doctrine of worthier title, subordination, attornment & estoppel, res ipsa loquitor, etc. -Maybe this is correct and proves my point, or maybe it is inapplicable here (how would you as a non lawyer know)? Look, until now, you and I were speaking in mutually understandable terms – lets keep it that way.

Now back to the principal issue - the claim - Case Shiller is accurate picture of the inner core because re-sales (and weighting) are in the inner areas. Here are some of the things you said and here are my responses:

“The core isnt an afterthought. It is part of the whole mess”. I don’t disagree with you. In fact, if it helps out let me say for the record. THE CORE ISNT AN AFTERTHOUGHT. All I was trying to do was refute your claim that re-sales were “mostly in the inner areas”. Based on my reading of Harriet’s data, it looks to a simple guy like me that your claim that re-sales were “mostly in the inner areas” is false.

“There isnt some magically protected area. Have you ever heard of the substitution effect”. Hyperbole aside, I challenge you to find any statement, any where, throughout all time, where I said prices in the urban core wont go down. If it helps let me say for the record, THE CORE IS NOT IMMUNE, I BELIEVE PRICES WILL GO DOWN. As to the substitution effect, fortunately, I did have some economics “lernin” in law school so yes I do understand that principle and I do believe it will affect the core prices. That’s not the point though. The question is whether CS paints an accurate picture of the inner core because (your claim) re-sales (and therefore weighting) are mostly the inner area. Again, you haven’t provided me with anything here to support your claim of re-sales being “mostly in the inner areas”, so lets move on.

Now, you did say – “Nothing wrong with the data.” Perhaps this is as close to an admission that re-sales are not “mostly in the inner areas” as I am going to get. At the same time however, it looks like you are hedging a bit because you then claim the data is not “stress tested”. This lack of stress testing doesn’t seem to be a problem for you when you cite Harriets other data. However, that aside, perhaps this should be a challenge for you. If this is so critical, why don’t you go ahead and stress test it and get back to the rest of us with the results. A word of warning though, you will need to explain your results and all these sophisticated concepts in language us regular folk who follow the blogs can understand.

CRT said...

To follow up, let me restate my argument about Case Shiller in relatively simple terms. First - because it looks to me like most (i.e. at least 75% and perhaps as much of 90%) of the re-sales are in the suburbs and exurbs, the claim that re-sales are “mostly in the inner areas” looks false. Second - given that most of the re-sale data used by CS is based upon suburban and exurban sales, it looks to me that the claim that CS is an accurate view of what is happening in the urban core of DC is also false. Is this correct? If not, can you explain to me in relatively simple terms (like I have to you) why it is not?

Now, for a related but separate argument that I have made elsewhere. For those of us that live around here, there seems to be a general consensus that to date (and I stress “to date”) prices have fallen the most in exurban areas, less so in the suburban areas, and even less so in the urban core of DC. This argument is based on

(a) info from local blogs like Harriet’s, data from MRIS and elsewhere that show (in general) a larger median price drop in the exurban counties, a lesser one in the suburban counties, and the smallest drop in the urban core

(b) multiple reports from the Washington Post and other local MSM which have said in sum “the outer counties have fallen the most, the inner ones not so much” and

(c) anecdotes from those of us who live around here – the people farthest out reporting the worst drops – the people closest in, the smallest.

If you disagree and want to debate the merits of this inner/outer argument then go right ahead – just don’t base it on Case Shiller data until you can show us why what appears to us simple bloggers as the weighting towards the suburbs & exurbs is false or doesn’t matter.

On a related note, what I really find insidious is affirmative statements on the accuracy of Case Shiller like this one on Harriet’s blog:

“Neil Said - Prices in DC have fallen 13.3% so far. Don't forget Case-Shiller is based on resales. New suburbs with few (if any) resales are not weighed in. Ex-urbs with few sales don't count for much either. Case-Shiller is being driven down by the actual re-sales; that should be mostly in the inner area.”

I’m sorry but since I haven’t stress tested or otherwise scrutinized CS, this just looks to me (and I suspect most other simple bloggers) like a statement trying to prove a point based on something that is at best grossly inaccurate or even worse patently false. That said, if you can explain to me (via stress testing, CFD, finite element, etc) in laymens terms why the weighting thing doesn’t matter, I will be the first one to say "I was wrong and you were right".

Finally, as you can probably sense, I really dislike it when people resort to their areas of expertise in an effort to (what looks to me as) bully the other - stating in essence, “you don’t understand, but rest assured I am right”. I'll be the first to admit, I am guilty of this from time to time as well. That said, just understand if you go to guns, I will too – but its not productive. All that happens is you wont understand me, I wont understand you, and no one will understand either of us except for the various cadre of nerd lawyers and nerd engineers who sit on the sidelines and say, “you got him Neil” or “you tore him up CRT”. Thus in the future, please try keep it as simple as you can, and I will too. Agreed?

wannabuy said...

CRT,

You keep arguing strongly against Case-Shiller, but it is the best index out there. I've looked into the details as best I can on the index and believe it is the most accurate representation of home value trends.

I'm certainly not trying to do anything but get buyers to do their own research and think before buying. I hear people in the 'nicest areas' say how prices won't go down... and then suddenly the price drops are unmasked. My wife and I saw this last weekend.

One issue with 'the core' is that high end buyers always buy. Its just what are they getting for their money?

If you have better information blog it! But I go out to open houses and I see in the best areas homes coming back onto the market that were sold within the last 12 to 18 months. Too many homes for said areas to be immune.

Arlington Inventory is already at last year's end of March level. (I don't have Alexandria's value.) Sales are down. That means the velocity of money is down and thus pricing pressure is negative.

Do you want me to say areas fall at different rates? Sure! But the argument that 'its different here' is old and untrue. All real estate is driven by economics and the no money down liar loans are going the way of the dodo. No housing metropolis should have home prices falling at 2.5% to 3.5%/month. Then again, it was a bad thing to have it rise that quick too!

The last time I saw that was in California; Beverly Hills and Palos Verdes were 'immune' in 1993 and down 40% by 1996. If you look back, they weren't really immune in 1993... people were getting more home for the same money. In effect, the local Case-Shiller was falling.

If the core of DC is not falling, why did Wells Fargo put it on the critical market list? I know VP's at Wells Fargo. They use quantitative judgement for their policies. Over half the population in the US wasn't put on the declinign market list. Why was the core of DC put on that list if it isn't declining? Wells Fargo is giving up business by putting DC on that list; so why did they cut loans to the market?

You're trying to get me to say that since I don't have data for a very small area I don't have data... that's not true. We have data. The argument is over the precision. Its good enough. Does it matter if one area really dropped 1.5% versus 2.5%? No. It doesn't. Not to a buyer who is going to be underwater and unable to retrieve their down payment (if they had one).

Got Popcorn?
Neil

CRT said...

Neil - this answer is so disappointing, it is as if you dont even read what I say.
Lets unpack your comments - I will put them into 2 separate parts (I) items Neil asked (again) which I already answered and (II) items Neil refuses to answer.

PART I - ITEMS NEIL ASKED THAT I ALREADY ANSWERED:

NEIL SAID - "You keep arguing strongly against Case Shiller, but it is the best index out there".
CRT PREVIOUSLY SAID - "Case Shiller gives us the best picture out there of whats happening in the entire MSA."
CRT's ANALYSIS - So, it looks like we agree fully on the accuracy of CS, we just disagree on its weighting (see "items Neil refuses to answer" below). In either event there is no disagreement on the overall accuracy of CS. As such, PLEASE DONT EVER REHASH THIS ISSUE AGAIN.

NEIL SAID - "If the core of DC is not falling..."
CRT PREVIOUSLY SAID - "I challenge you to find any statement, any where, throughout all time, where I said prices in the urban core wont go down. let me say for the record, THE CORE IS NOT IMMUNE, I BELIEVE PRICES WILL GO DOWN."
CRT's ANALYSIS - Once again it looks like we agree fully on this point - and for God's sake, how much clearer can I get? NEIL, PLEASE DONT EVER REHASH THIS ISSUE AGAIN.

NEIL SAID - "why did Wells Fargo put it (DC) on the critical market list"
CRT PREVIOUSLY SAID - I dont know....redlining?
CRT's ANALYSIS - Lets ignore for a second I never once stated that Wells Fargo was wrong in doing what they did, nor did I really care since it has little to do with my principal issue about the weighting of Case Shiller (see "items Neil refuses to answer" below). It it makes you feel better, point to you - this is a devastating attack on this central issue to our entire debate. You won I lost. AS SUCH, PLEASE DONT REHASH THIS ISSUE AGAIN.

PART II - ITEMS NEIL REFUSES TO ANSWER:

What started this whole debate is when you said:

"CS is driven by sales. So the further you get away from a major urban center, the fewer sales… so lower the weighting."

AND

"Case-Shiller is based on resales. New suburbs with few (if any) resales are not weighed in. Ex-urbs with few sales don't count for much either. Case-Shiller is being driven down by the actual re-sales; that should be mostly in the inner area.”

NEIL THIS IS THE PRINCIPAL ISSUE I WANT YOU TO FOCUS ON - NO MORE NO LESS - CAN YOU DO THIS?

I have pointed out, time and time again, that the re-sales in the urban core are few. In the D.C area, resales are mostly in the suburbs and exurbs. Let me paraphrase our debate on this point thus far:

NEIL - Resales are mostly in the inner areas of DC (or some derivation of this).

CRT - Around DC, CS is weighted more towards the Suburbs/Exurbs than Urban Areas.

NEIL - I dont believe thats right.

CRT - Around DC, CS is weighted more towards the Suburbs/Exurbs than Urban Areas.

NEIL - You dont know how to review data - I do. It doest matter - the core is going down too.

CRT - Around DC, CS is weighted more towards the Suburbs/Exurbs than Urban Areas.

NEIL - CS is the best thing out there. If you have something better, Blog it! The argument "its different here" is old and untrue...liar loans...Palos Verde...Wells Fargo...We have data...Its good enough...

Its a rather sad exchange isnt it? You continue to flail about, but don’t admit your mistake. It reminds me of when Lance could not admit that he mistook a negative sign in front of a number for a hyphen and he was ridiculed because he could not admit it. Let me show you how it could have gone, based on a debate between Doug (a notorious bull) and myself on Harriet's blog: DOUG - build a modular home.
CRT - dont build a modular they all suck.
DOUG - CRT, not all modulars suck take a look at these
CRT - I looked at those, and they arent half bad - I stand corrected.

See how easy that was? I didnt quailfy it, I didnt admit - but then restate my position, I just came out and said, I was wrong. We then moved on to debate other matters.

You will notice my tone has changed throughout this exchange. At first I was very civil with you, but as time went by, I got more terse. And if you now think I am mocking you with alot of this - you are right. The thing is, I cant stand it when people are backed into a corner and still they cannot admit that they are wrong. It really destroys ones credibility in my book.

I think I've badgered you enough on this so I will not respond further to this thread. You can continue to maintain that despite its limitations, CS still is generally accurate as it relates to the inner core, and we can debate this further in future days. My issue with you now is much simpler than that - leaving general accuracy of CS aside for a second, you earlier stated (in sum) that in the D.C. area, most re-sales re in the urban core. Keeping that in mind, Please tell me which is a correct statement:

(a) In the D.C. area most re-sales are in the in the urban core

or

(b) In the D.C. area most re-sales are in the suburbs and exurbs.

Again, the above does not say anything about accuracy of CS on inner areas - all it asks is where are the resales. This is a fair question it asks you to state which is true (a) or (b) - no more no less.

If you pick respone (b) and dont try to quailfy it, say it doesnt matter, or that CS is still accurate, etc, etc, etc. I will accept that answer. If so let me say thank you for being willing to admit your mistakes, and I look forward to a continued civil discussion (on your blog and elsewhere) where you and I have a spirited debate on how this all shakes out.

If you choose not to respond at all, I will accept that you cannot admit you have made a mistake, and I will deem you to be a generally un-credible source for any future debate.

If you respond and say (a) or if you say "(b) BUT it doesn’t matter CS is still accurate" or some other qualified response, please note I have saved this whole exchange, and you run the risk that I may or may not continue to point out here and elsewhere that you cannot admit to a mistake again, and again, and again.

Also, no matter what your response, from here on out, I dont ever want you to say here or elsewhere CS is accurate as it relates to the core of DC "because most sales are in the core". That statement is just plain wrong and you know it.

I'm dying to see how you respond as I dont think you have the guts to make an honest unqualified response. My prediction is you will say (b) but then qualify it some way or the other mentioning something about Case Shiller. What do you think - fellow readers of Neil’s blog? Can he just flat out admit, it or will it be an admission with qualifications? I guess we shall see!