Posted this on another thread, but thought it deserved its own. Good piece by Jonathan Miller that really explains why contract data is not that great. Closed sales prices are lagging, but those data are more reliable to show true movement at this point. That's the beauty of the market movement threads. http://matrix.millersamuel.com/?p=3043 "One of the most sought after trending tools for housing... [more]
Posted this on another thread, but thought it deserved its own. Good piece by Jonathan Miller that really explains why contract data is not that great. Closed sales prices are lagging, but those data are more reliable to show true movement at this point. That's the beauty of the market movement threads.
http://matrix.millersamuel.com/?p=3043
"One of the most sought after trending tools for housing markets is contract data. Not listing data, not closed data. Contract data.
Compile a lot of data across all regions, property types and price strata and you are golden. You are observing the market as close to the “meeting of the minds” as is humanly possible - you have its proverbial pulse.
I thought to write about the concept of reporting contract data after I got a call from The Real Deal about a new contract-based real estate market report. Their founder is a very creative, very smart and very successful marketer of real estate, first as an agent and then as a marketing expert for new developments. Visually, the report is beautifully done, consistent with the quality of their firm’s marketing materials and online presence. However, they might consider dropping the name of “real-time” from the report. It’s monthly. I understand the intention, but the use of the phrase “real-time” infers a live feed, which this report is not. Isn’t “monthly real-time” an oxymoron?
A quote from The Real Deal article:
It tracks contracts info. To me, that’s what reflects the marketplace and where we are currently, not closed information, which is actually a look back in history.
Another company attempted “real-time” a few years ago by treating real estate listings like the stock market and began publishing a “ticker” type interface. I have to give them credit for the innovation, but it never really got people’s attention.
But I digress
What is contract data exactly?
It’s a property sale with an executed (both parties signed) contract - It is usually 45-60 days ahead of a closing date if new development data is excluded. Actually this 45-60 day time frame is currently expanding as lenders become more difficult to deal with. New development data in the mix could lag the market by 1 to 2 years.
I sort of dealt with contract activity in the most recent market report numbers in my 4Q 2008 Manhattan Market Overview but not in the traditional sense of aggregating contract data and trending it.
Our appraisal firm began to see a pattern in late September 2008 where current contracts of properties we were appraising, were clearly lower than contracts signed in the summer of 2008. The range was roughly 15% to 20%. My 20% number has been widely referenced by the Fed, Goldman Sachs and others, and in fact, page one of AM New York published the number “20%” in red on the entire cover. But our conclusions were based on more of a case by case analysis, similar to a repeat sales analysis.
I don’t currently issue contract reports but I certainly aspire to, but only when I have credible results. Periodically I’ll see one of my appraisal competitors distribute a press release with their own contracts tabulated. I’ll see real estate brokers and marketing agents issue contract reports.
Readers oooh and ahhhh over the relevancy of contracts because the data is perceived to be fresh and current. In principle it is current, but in practice it is much more subject to skew than other data.
I also wonder why methodologies are never fully provided, especially those prepared by marketing groups or departments.
Here are the issues that make much market analysis of contract reports suspect, despite perhaps the best intentions of the authors.
Quantity of data — the key issue that makes much analysis unreliable - absent from the public domain.
Location of the data — contract data tends to be sourced from a few institutions or entities so its availability and the potential for skew is very serious.
Unit mix of the data — This is subject to skew depending on the source of the data - what type of business they have - who their customers are (low end, high end, studios, 3-bedrooms, etc.)
Source of the data — The four largest real estate brokerage firms probably account for 80% of all sales in Manhattan. I know each of the senior management teams so I am fairly confident they will not release contract data in bulk to anyone outside their company, especially to a competitor.
I have never met a broker that will share contract data in bulk because it can jeopardize their company’s sales and commissions. We are able to get contract data periodically, but not in bulk. If producers of contract reports can win me over on these key issues, I am ready to jump in with two feet. NAR publishes a pending contract index and frankly, not many people I know believe the results.
In other words, contract data is the Holy Grail, but I am not convinced it’s yet achievable as a reporting tool.
Now give me a sales contract specific to the appraisal we are working on and I am happy ’cause that’s a whole ‘nother story."
[less]
Response by bjw2103
about 17 years ago
Posts: 6236
Member since: Jul 2007
bump for nyc10022's pleasure
Ignored comment.
Unhide
Response by nyc10022
about 17 years ago
Posts: 9868
Member since: Aug 2008
Interesting... so he's complaining about the BROKER data.
Get that, bjw, the guy from the apprasial firm complaining about the BROKER data.
Now, why does that sound familiar again? I can't remember... hmm.. uh... maybe this is deja vu?
Ignored comment.
Unhide
Response by bjw2103
about 17 years ago
Posts: 6236
Member since: Jul 2007
Except that this is the data you LOVE to cite. It's suspect, and even the very same guy you quoted says so.
Ignored comment.
Unhide
Response by nyc10022
about 17 years ago
Posts: 9868
Member since: Aug 2008
Considerably less suspect than using data from an irrelevant period... I think the dow went up in 2000... should I buy now??
BTW, there were two separate appraisal firms releasing the 20% down contracts stats.... you forgot about the other one...
Ignored comment.
Unhide
Response by bjw2103
about 17 years ago
Posts: 6236
Member since: Jul 2007
2000? Your exaggerations are ridiculous. Closed sales data lags - we all know this - but so does contract data, and it's getting even worse. As for your other appraisal firm, Miller addresses that very clearly in this article. It's all suspect!
Ignored comment.
Unhide
Response by nyc10022
about 17 years ago
Posts: 9868
Member since: Aug 2008
Again, no more suspect than using data from a period that doesn't matter.
Yes, 2000 is an exagerration, but its just as inaccurate as using ANY pre-crash data... that being pretty much 100% inaccurate.
Ignored comment.
Unhide
Response by tina24hour
about 17 years ago
Posts: 720
Member since: Jun 2008
Hey folks, can we agree that there is no single datum that tells us what is happening in the market at each nanosecond? I think that's what we were all getting at in the other thread. Jonathan Miller just lays it out in a dispassionate, comprehensive manner - as usual. Data is his game.
Ignored comment.
Unhide
Response by bjw2103
about 17 years ago
Posts: 6236
Member since: Jul 2007
Well, no, otherwise Miller would use the contract data (and he says as much here). And once again, your claim really only applies to the Q4 reports - the point is, contract data is always pretty poor. At least until they fix the issues he lists.
Ignored comment.
Unhide
Response by bjw2103
about 17 years ago
Posts: 6236
Member since: Jul 2007
Exactly, tina24hour. Just trying to dispel the myth about contract data that was being propagated.
Ignored comment.
Unhide
Response by nyc10022
about 17 years ago
Posts: 9868
Member since: Aug 2008
> contract data is always pretty poor.
And still much better than datsa from an irrelevant period. Sorry.
Ignored comment.
Unhide
Response by bds
about 17 years ago
Posts: 187
Member since: Jan 2009
In this enviornment, I would love to know the percentage of condos that close, compared to the number in contracts....now that is valuable. Because when a bldg says that it has sold 50 or 60%, my concern that because of the climate, people will be unable to close..
Ignored comment.
Unhide
Response by bjw2103
about 17 years ago
Posts: 6236
Member since: Jul 2007
"Because when a bldg says that it has sold 50 or 60%, my concern that because of the climate, people will be unable to close.."
I agree, and it's yet another knock against contract data. nyc10022 just doesn't get that.
Ignored comment.
Unhide
Response by nyc10022
about 17 years ago
Posts: 9868
Member since: Aug 2008
Meaning the market past that point will be even worse off than the curren contracts suggest...
making looking at outdated data an even bigger mistake...
Not to mention, the contracts still show where prices were agreed upon. That things are getting worse just shows how bad the outdated reports really are...
Ignored comment.
Unhide
Response by bjw2103
about 17 years ago
Posts: 6236
Member since: Jul 2007
You're making a whole bunch of unfounded assumptions there, but I get it. You won't back down even when the guy you quote tells you not to read too much into what he said. Cool.
Ignored comment.
Unhide
Response by nyc10022
about 17 years ago
Posts: 9868
Member since: Aug 2008
Hey, we're all aware of the perils of data, particularly in moving markets.
But I'm not sure in what world valid data "you can't read too much into" is trumped by completely invalid/irrelevant data.
But, if that's your thing...
Ignored comment.
Unhide
Response by nyc10022
about 17 years ago
Posts: 9868
Member since: Aug 2008
Actually, it just hit me... he actually disagrees with your point.
You made the claim a couple weeks back that contract data is problematic because it is contract data, that it itself is fundamentally flawed.
I actually agree with him on all of these challenges:
Quantity of data
Location of the data
Unit mix of the data
Source of the data
absolutely. But those are problems with the data SET. Same way median data for closings gets skewed.
Your whole claim was "they're just contracts, not sales".... and this actually does a pretty good job of showing that to be a lacking argument.
I agree that you need a GOOD set of contract data.... but your argument that contract data itself isn't meaningful, well, I think you're alone in that camp.
Ignored comment.
Unhide
Response by bjw2103
about 17 years ago
Posts: 6236
Member since: Jul 2007
"You made the claim a couple weeks back that contract data is problematic because it is contract data, that it itself is fundamentally flawed."
Uh, no. I may never have been as eloquent about it as Miller, but I have always believed contract data to be bad because it is inherently not as good as sales data, supplemented by the fact that there's not as a good a historical record on contract pricing. Sales data represent actual prices paid for actual apartments traded. Not maybes, renegotiated, or canceled deals. Done deals. You can't beat that. And your chief argument against those data is that they lag - for the 100th time, we know this, and it doesn't make it "invalid" or "irrelevant." Miller clearly says that contract data lags as well, and it's getting worse! Listen, if contract data were really better than sales data, Miller would use it in his reports! The fact that he's not speaks volumes about the quality of it and the conclusions you can take from it. And I'd much rather trust someone who does this for a living than some anonymous guy on a message board who has a hell of a time admitting he's wrong about anything.
Ignored comment.
Unhide
Response by nyc10022
about 17 years ago
Posts: 9868
Member since: Aug 2008
"I may never have been as eloquent about it as Miller, but I have always believed contract data to be bad because it is inherently not as good as sales data, supplemented by the fact that there's not as a good a historical record on contract pricing. "
You're not as eloquent because, well, he didn't say that.
Thats the point... he actually doesn't note that as an issue. Your "evidence" on why you were right doesn't actually support you... well, being right.
Nice try, though!
Ignored comment.
Unhide
Response by bjw2103
about 17 years ago
Posts: 6236
Member since: Jul 2007
"You're not as eloquent because, well, he didn't say that."
He said the first part, but not the second. But that one is also true. Spin and backpedal all you want.
Ignored comment.
Unhide
Response by nyc10022
about 17 years ago
Posts: 9868
Member since: Aug 2008
> He said the first part, but not the second. But that one is also true.
ok, now you're just being too funny....
you want to call it backpedalling, the guy DOES NOT support your claim, and the part he does say - which was not part of your argument, I said I agreed with.
Thats not backpedalling, thats you trying a bait and switch. He didn't agree with your assertion, and the thing he says, I never said otherwise.
Nice try, though...
Ignored comment.
Unhide
Response by bjw2103
about 17 years ago
Posts: 6236
Member since: Jul 2007
You are hilarious man. Let's dumb this down for you:
You champion contract data.
Miller titles his post "Contract data can mean nothing" and says "I am not convinced it’s yet achievable as a reporting tool."
You got it wrong.
Q.E.D.
Ignored comment.
Unhide
Response by nyc10022
about 17 years ago
Posts: 9868
Member since: Aug 2008
and to think, all this started with your childish chiding...
"bump for nyc10022's pleasure'
Ignored comment.
Unhide
Response by nyc10022
about 17 years ago
Posts: 9868
Member since: Aug 2008
> You are hilarious man. Let's dumb this down for you:
bjw, your lack of comprehension is amazing here. I would expect better of you.
> You champion contract data.
Yes, I like it.
> Miller titles his post "Contract data can mean nothing"
Yes, and gives specific reasons why contract data can be pad. I agree with them. Hell, I quoted them above. They're the same reaons ANY data can be bad, including closed sales data. Those are data set problems
The reason you've been citing before, and are still citing now.... NOT supported by this article at all And still lousy logic in my book.
Add this to the fact that you started this one off being COMPLETELY childish... WOW, i would expect better of you.
You begain this with being childish. Then moved the name calling. All that, and the guy DOESN'T AGREE WITH YOUR POINT.
Wow, seriously, bjw, get a grip.
And save the insults for someone dumber than you.
Ignored comment.
Unhide
Response by nyc10022
about 17 years ago
Posts: 9868
Member since: Aug 2008
really, wow, for someone who spends SO much time being hall monitor around here, you've been quite childish and petty here.
Ignored comment.
Unhide
Response by bjw2103
about 17 years ago
Posts: 6236
Member since: Jul 2007
nyc10022, I have not once "insulted" you here. Don't be so touchy. If you willfully want to ignore what this article is saying - that the data you like is bad - fine, but don't make false accusations. Yes, those are reasons any data can be bad, but he's specifically talking about the data YOU use. The truth is, he doesn't agree with YOUR point.
bump for nyc10022's pleasure
Interesting... so he's complaining about the BROKER data.
Get that, bjw, the guy from the apprasial firm complaining about the BROKER data.
Now, why does that sound familiar again? I can't remember... hmm.. uh... maybe this is deja vu?
Except that this is the data you LOVE to cite. It's suspect, and even the very same guy you quoted says so.
Considerably less suspect than using data from an irrelevant period... I think the dow went up in 2000... should I buy now??
BTW, there were two separate appraisal firms releasing the 20% down contracts stats.... you forgot about the other one...
2000? Your exaggerations are ridiculous. Closed sales data lags - we all know this - but so does contract data, and it's getting even worse. As for your other appraisal firm, Miller addresses that very clearly in this article. It's all suspect!
Again, no more suspect than using data from a period that doesn't matter.
Yes, 2000 is an exagerration, but its just as inaccurate as using ANY pre-crash data... that being pretty much 100% inaccurate.
Hey folks, can we agree that there is no single datum that tells us what is happening in the market at each nanosecond? I think that's what we were all getting at in the other thread. Jonathan Miller just lays it out in a dispassionate, comprehensive manner - as usual. Data is his game.
Well, no, otherwise Miller would use the contract data (and he says as much here). And once again, your claim really only applies to the Q4 reports - the point is, contract data is always pretty poor. At least until they fix the issues he lists.
Exactly, tina24hour. Just trying to dispel the myth about contract data that was being propagated.
> contract data is always pretty poor.
And still much better than datsa from an irrelevant period. Sorry.
In this enviornment, I would love to know the percentage of condos that close, compared to the number in contracts....now that is valuable. Because when a bldg says that it has sold 50 or 60%, my concern that because of the climate, people will be unable to close..
"Because when a bldg says that it has sold 50 or 60%, my concern that because of the climate, people will be unable to close.."
I agree, and it's yet another knock against contract data. nyc10022 just doesn't get that.
Meaning the market past that point will be even worse off than the curren contracts suggest...
making looking at outdated data an even bigger mistake...
Not to mention, the contracts still show where prices were agreed upon. That things are getting worse just shows how bad the outdated reports really are...
You're making a whole bunch of unfounded assumptions there, but I get it. You won't back down even when the guy you quote tells you not to read too much into what he said. Cool.
Hey, we're all aware of the perils of data, particularly in moving markets.
But I'm not sure in what world valid data "you can't read too much into" is trumped by completely invalid/irrelevant data.
But, if that's your thing...
Actually, it just hit me... he actually disagrees with your point.
You made the claim a couple weeks back that contract data is problematic because it is contract data, that it itself is fundamentally flawed.
I actually agree with him on all of these challenges:
Quantity of data
Location of the data
Unit mix of the data
Source of the data
absolutely. But those are problems with the data SET. Same way median data for closings gets skewed.
Your whole claim was "they're just contracts, not sales".... and this actually does a pretty good job of showing that to be a lacking argument.
I agree that you need a GOOD set of contract data.... but your argument that contract data itself isn't meaningful, well, I think you're alone in that camp.
"You made the claim a couple weeks back that contract data is problematic because it is contract data, that it itself is fundamentally flawed."
Uh, no. I may never have been as eloquent about it as Miller, but I have always believed contract data to be bad because it is inherently not as good as sales data, supplemented by the fact that there's not as a good a historical record on contract pricing. Sales data represent actual prices paid for actual apartments traded. Not maybes, renegotiated, or canceled deals. Done deals. You can't beat that. And your chief argument against those data is that they lag - for the 100th time, we know this, and it doesn't make it "invalid" or "irrelevant." Miller clearly says that contract data lags as well, and it's getting worse! Listen, if contract data were really better than sales data, Miller would use it in his reports! The fact that he's not speaks volumes about the quality of it and the conclusions you can take from it. And I'd much rather trust someone who does this for a living than some anonymous guy on a message board who has a hell of a time admitting he's wrong about anything.
"I may never have been as eloquent about it as Miller, but I have always believed contract data to be bad because it is inherently not as good as sales data, supplemented by the fact that there's not as a good a historical record on contract pricing. "
You're not as eloquent because, well, he didn't say that.
Thats the point... he actually doesn't note that as an issue. Your "evidence" on why you were right doesn't actually support you... well, being right.
Nice try, though!
"You're not as eloquent because, well, he didn't say that."
He said the first part, but not the second. But that one is also true. Spin and backpedal all you want.
> He said the first part, but not the second. But that one is also true.
ok, now you're just being too funny....
you want to call it backpedalling, the guy DOES NOT support your claim, and the part he does say - which was not part of your argument, I said I agreed with.
Thats not backpedalling, thats you trying a bait and switch. He didn't agree with your assertion, and the thing he says, I never said otherwise.
Nice try, though...
You are hilarious man. Let's dumb this down for you:
You champion contract data.
Miller titles his post "Contract data can mean nothing" and says "I am not convinced it’s yet achievable as a reporting tool."
You got it wrong.
Q.E.D.
and to think, all this started with your childish chiding...
"bump for nyc10022's pleasure'
> You are hilarious man. Let's dumb this down for you:
bjw, your lack of comprehension is amazing here. I would expect better of you.
> You champion contract data.
Yes, I like it.
> Miller titles his post "Contract data can mean nothing"
Yes, and gives specific reasons why contract data can be pad. I agree with them. Hell, I quoted them above. They're the same reaons ANY data can be bad, including closed sales data. Those are data set problems
The reason you've been citing before, and are still citing now.... NOT supported by this article at all And still lousy logic in my book.
Add this to the fact that you started this one off being COMPLETELY childish... WOW, i would expect better of you.
You begain this with being childish. Then moved the name calling. All that, and the guy DOESN'T AGREE WITH YOUR POINT.
Wow, seriously, bjw, get a grip.
And save the insults for someone dumber than you.
really, wow, for someone who spends SO much time being hall monitor around here, you've been quite childish and petty here.
nyc10022, I have not once "insulted" you here. Don't be so touchy. If you willfully want to ignore what this article is saying - that the data you like is bad - fine, but don't make false accusations. Yes, those are reasons any data can be bad, but he's specifically talking about the data YOU use. The truth is, he doesn't agree with YOUR point.
> nyc10022, I have not once "insulted" you here.
Really? .. "Let's dumb this down for you"
> Don't be so touchy.
Hey Pot....
http://www.merriam-webster.com/dictionary/dumbdown
Wow.
Wow, another link that didn't prove your point.
Really, bjw, grow up. Take responsibility for yourself for once.
nyc10022,Happy Valentine's Day @---- @----
@---- @----
bjw2103, quit it pls.