This is a copy of email I just sent to Peter Coy of BusinessWeek with respect to false claims made last week about deception in the Arizona Regional Multiple Listings Service. Cliff’s Notes: False alarm resulting from “crackpot claims, the end product of a fervid imagination and a Rube Goldberg spreadsheet.”
From: GregSwann@BloodhoundRealty.com
Subject: Problem solved — “Twist” doesn’t know what she doesn’t know
Date: January 8, 2007 2:13:57 AM MST
To: Peter_Coy, Twist
Cc: Jay, Jonathan, John
1. Ms. Averett does not have access to the ARMLS system.
2. Her analysis is based on summary reports issued by the ARMLS staff, presumably for PR purposes.
3. Those reports omit many categories of residential listings, presumably to make the summary fit on a single page.
4. The three columns of Ms Averett’s analysis that buttress her claim — New Listings, Delisted, and Ratio Sales/Delisted are not obtained from these summary reports. There may be some other source, but they’re not in the ARMLS reports — at least not in those I looked at.
5. As has been demonstrated by four Phoenix-area Realtors working independently but directly in the ARMLS system, Ms. Averett’s contentions about the months of November and December of 2005 and January of 2006 are not only false, they bear no resemblance to reality at all. I have also demonstrated that her contentions with respect to the same months one year earlier are also false.
6. Given that the methodology she deploys is dubious at best, and probably completely devoid of meaning, it seems reasonable to surmise that all of the rest of her claims with respect to ARMLS are also false. This is not to imply that the ARMLS system is fault-free, but simply that the fault Ms. Averett claims to have identified does not exist. She does not understand the ARMLS system well enough to make any sort of informed statement about it.
7. The other weblogging Phoenix-area Realtors copied above may have more to add as they peruse Ms. Averett’s work product. John L. Wake surmised from the beginning that Ms. Averett was working from ARMLS-issued summaries rather than directly from the database.
8. My presumption, subject to contradiction, is the Ms. Averett’s claims about the MLS systems in Tucson and Las Vegas are equally erroneous, based in the same erroneous methodology.
9. With respect to the Phoenix-area, I see no reason to conclude that Ms. Averett’s contentions are anything other than crackpot claims, the end product of a fervid imagination and a Rube Goldberg spreadsheet. The actual MLS data incontrovertibly contradict her claims to the last tittle and jot.
10. I am at your complete disposal, now and always, should you like to discuss this or anything relating to the Phoenix real estate market.
Very best,
Greg Swann, ABR, CBR, CRS, E-Pro, GRI
Realtor, Designated Broker
BloodhoundRealty.com
Technorati Tags: arizona, arizona real estate, phoenix, phoenix real estate, real estate marketing
David Saks says:
Great letter, Greg. You sure flushed a monkey out of the woodpile. Glad you threw some water on the fire.
ps…Happy birthday Elvis ! 🙂
January 8, 2007 — 7:18 am
Robert Cot says:
Glad that’s cleared up. The ARMLS internal data, available only to members, is accurate, it is only the data reported by staff to the public that is inaccurate.
What’s the word I’m looking for that describes the reporting of misleading data?
January 8, 2007 — 9:24 am
Greg Swann says:
> The ARMLS internal data, available only to members, is accurate, it is only the data reported by staff to the public that is inaccurate.
This is also untrue. Twist’s false report results entirely from a clumsy error she made in her own spreadsheet. As before: Bubble bloggers are notoriously reckless with numbers.
Yes, you were misled: By Twist.
January 8, 2007 — 10:28 am
Robert Cot says:
I thought you said:
“Her analysis is based on summary reports issued by the ARMLS staff,”
I repeated in my own words:
“…it is only the data reported by staff to the public that is inaccurate.”
In what way do you then not mislead me? Sure Twist made some errors, many more than one. The reason is because the ARMLS is not subject to transparency.
It sure is fun to watch future real estate clerks claim that complaints of witholding information would be proven false if only their critics had access to the data.
I was misled by Twists claims, yes. I resent wasting time tracking the truth. Our difference is that I assign the blame elsewhere. Was not Twist misled by the ARMLS data available to her?
January 8, 2007 — 10:51 am
Greg Swann says:
> Was not Twist misled by the ARMLS data available to her?
No, her error resulted from inventing a forumla that bears no correspondence to reality. She has no idea what she is talking about. You do your own credibility no service by trying to defend her. The actual MLS results for the months cited in the BusinessWeek post — no deceptions, no “fudging,” no conspiracy — are reported here.
Twist was outrageously wrong, and she should admit it. The fault lies with no one but her.
January 8, 2007 — 10:57 am
Doom Fan says:
This issue has been addressed on HousingDoom here:
http://tinyurl.com/wcgxo
January 8, 2007 — 1:19 pm
Robert Cot says:
I say; ” Sure Twist made some errors, many more than one. The reason is because the ARMLS is not subject to transparency.”
You reply with; “She has no idea what she is talking about. You do your own credibility no service by trying to defend her.”
What part of my pointing out multiple errors is defending her? Holding out the chimera of “my credibility” is no bait. You already made up your mind long ago. Just as you decided Twist with an advanced degree in statistics didn’t know math. Just like you feel free to call people with opposing opinions Brownshirts.
The ARMLS is not transparent. The ARMRLS publishes statements that are contradicted by their internal (private) data. It is you and your defending the ARMLS practices that risks loss of credibility. And in case the original issue was lost in all the charges/countercharges, Twist said something was suspicious about the ARMLS data available to the public. You admit as much.
January 8, 2007 — 1:22 pm
John L. Wake says:
Twist made some errors in her analysis, for example, the inventory numbers ARMLS publishes are for one day, usually the 15th of the month the report is published, while the rest of the data in the report is for the previous whole month. It’s very confusing.
However, I also saw that strangeness in the ARMLS published data in 2005 and considered the possibility I had stumbled upon evidence of some vast conspiracy that would explain the rapid rise in prices following the anomalies of August 2004 and January 2005. 🙂
There were a lot of things wrong with that idea, however. In particular the fact that there were spikes in 2 months. That suggests to me a data error.
If there really were some malicious factor going on, it wouldn’t all appear in one month or two months. It would be a trend over several months.
The more likely culprit was a simple data error. ARMLS has made them in the past and it’s reports are so oblique its hard to know what they mean. Prehaps ARMLS was making a correction to their public data and packed them all into one or two months without noting it on the reports.
After I went into ARMLS directly and did an analysis similar to what Greg did, it was obvious that the problem was in the reports and that there was no anomaly in the core data. The only anomaly was in the ARMLS public reports.
The purpose of ARMLS is not to provide data to the public. They have much more important priorities and a very difficult job.
Providing public reports was obviously a very low priority for ARMLS and it showed in their public reports.
ARMLS is, however, doing a much better job on their public reports. Nevertheless, their public reports need a complete makeover.
January 8, 2007 — 5:35 pm
Gary Anderson says:
Greg, if you think the market will recover why worry? All we are doing is educating people not to buy while values are going down and while fundamentals are out of whack. You are becoming the George Bush of real estate, no plan, no concept of reality, etc, etc. Just teasing you Gregg!!!
January 10, 2007 — 10:03 pm
Brian Brady says:
I’ll bite. What specific fundamental indicator allows you to “flip the switch” to a buy recommendation, Gary?
January 11, 2007 — 12:41 am
John L. Wake says:
I’ll give you a “Buy” recommendation.
If you planned to buy in the next several months in the Phoenix area anyway, I recommend that you buy now instead of later.
Inventory is falling and prices tend to go up in Phoenix in the first half of the year.
January 11, 2007 — 11:55 am