DataQuick is a popular source for tracking the California housing market, and their numbers are widely used and quoted. As we’ve been considering the effects of faulty MLS data on the market, I received an interesting email from SD. SD points out that DataQuick has its issues as well–one of which is their definition of a sold property. SD states:
You are not going to believe this! I live in Shasta County, California. The local newspaper, the Redding Record Searchlight publishes addresses and dollar amounts of sales each month. Their source is DataQuick.
On a lark, I decided to track foreclosures for their impact on sales. The Record Searchlight publishes "Notices of Trustees Sales" (potential foreclosures) in the legal section of the paper. When the lender gets the property back (foreclosure) a "Trustees Deed" is issued . I found FORECLOSURES being counted as SALES.
DataQuick does not distinguish "Deeds" (regular property transfer) from "Trustee Deeds" (foreclosure). These are counted together as sales. Yes, that is correct, FORECLOSURES, are counted as sales. For Shasta County 141 sales were reported for November, of those 12 were foreclosures [8.5%]. That means that in reality there were 129 sales.
Data Quick did not respond to my email. The Record Searchlight contacted DataQuick. Data Quick indicated their program,"Prospect Finder Farm Services", has NO function to distinguish Deeds from Trustees Deeds (foreclosures) and it would be expensive to fix this gross error.
Bottom Line: The inclusion of foreclosures means that ALL CALIFORNIA SALES ARE OVERSTATED BY DATA QUICK.
Here are a couple of foreclosure examples which you can check on zillow.com. Zillow is doing the same thing as DataQuick.
- 1025 Chardonnay Walk, Redding, CA 96001 shows sold $635,000 on 11/17/06. (Foreclosure currently on the market with Coldwell Banker for $520,000, listed as a lender owned property.)
- 6740 Southgate Drive, Redding, CA 96001 shows sold $249,474 on 11/13/06. (Foreclosure currently on the market with Bears Den Real Estate, listed as a lender owned property for $213,000).
This must also be happening in Nevada and Arizona.
Thanks SD for pointing this one out to us. As foreclosure rates rise in CA, this issue will increasingly affect sales and pricing numbers.
As always I turned to the ever-patient L for his opinion. He said:
Boy you sure make me work! [He's right--L has been very generous with his time!]
This has a yes and no answer. In actuality it is a sale–ownership transferred, so the winning bid bought the property. In these cases the winning bidder was the beneficiary. Now how do you use this transfer? I’m sure most appraisers would throw it out, but when the house sells for the lower price it is now listed at, would we want them to use that number….? The sellers in that area would say no, the buyers in that area would say yes. As you can see, its going to get interesting, isn’t it?
Now lets say Mr.& Mrs. Smith bid at the trustee sale and they had the high bid and paid that much for that house to live in. It would be prudent to show that as a sale because it was. Just because it sold in a different manner than FSBO or listing it on the MLS doesn’t mean someone didn’t buy it for the same purpose, does it…………..?
Most all data sellers like Net Value, DataQuick. Night Owl and many more, report sales from deeds. Most all deeds are sales, and how you would separate them I don’t know. By law that was and is a sale.
Shasta CA. is a small area and using a service like DataQuick is the simplest way for the local paper to gather data. I’m not sure how the Arizona Republic picks up their sales but it’s fairly accurate most of the time. I don’t think we will be seeing anything like this here in Phoenix that would cause any great imbalance.
I’ve spent some time myself mining data from the Maricopa County Recorder’s Office–most of it involved highly repetetive work wading through lots of PDF documents. Our sys admin said that he can write a program that would count the deeds, but not one that could easily differentiate between them.
While there are more online tools than ever for consumers to use to evaluate the market, the complexity of it all means these tools have their limits. If you know and understand the problems, you can make adjustments. But as SD so nicely demonstrated, you can’t just take your data at face value–you need to dig a little deeper.
Thank you SD and L–your input is much appreciated.
I’m much more sanguine. Dataquick reports recorded transactions. They call recorded transactions “sales” and even then aren’t incorrect. Isn’t a sale the transfer of controlling interest? Any attempt to call into question their definition of including nonmarket not arms length transactions leaves open the prospect that things like median and averages also not include the same. When something changes hands and no guns are involved that to my mind is a sale. This isn’t an error or misreporting as long as it is acknowledged.
This differs greatly from the ARMLS exposé. There the methodology is neither transparent nor ultimately supportable. Here the DQ numbers are what they are. Personally, since the absolute value is of little use to me I’d be more upset if they changed and thus destroyed relative continuity of the dataset.
That said, DQ is; damn sloppy, late, random, dogmatic. I’ll take a consistent bedevilment over a random one anyday.
Techscan-
As I indicated in my post, I’ve found trying to figure out transactions at the county recorders office a royal pain. Even if you had the time to have a human being review every transaction, how do you know which sales have huge kickbacks, whose selling to a family member at a huge discount, the list goes on…
As long as you know the faults and foibles in any system you can work around it. I just like to see the discussion out there, so people have a better grasp of what it means.
I think techscan is confusing a sale by the bank to a new owner(REO sales) with the foreclosure itself; foreclosures are most certainly done at gunpoint. Just go try to stop the auction at the courthouse steps and see what happens. And there is no ’sale’ in the market sense. It’s just the bank seizing their collateral.
I believe Techscan is on the money. Deeds passing from one party to another is probably the best, particularly since all sorts of
“transactions” could be segregated if we start. For example, related party transfers (divorced spouse extinuishing interest of other;family trust to child) create all sorts of different wrinkles in the “arms length” test. And for historical comparisons, the devil you know is better than the new improved devil.
Upgrayed, however, is, I think, confusing the “feelings” of coersion by the debtor in default in a foreclosure from the fact that generally the deed passes(and thus becomes a “sale” )only after an opportunity to redeem and an auction, the later being absolutely the standard for a fair market price. The debtor cannot stop the auction on the court house steps of course, but they can bid, and it is no longer the length of their arm that matters. Rather than a foreclosure being at gun point, it is a far more structured and civil (and expensive)method of legitmiate debt collection than, say, repossing an auto in very similar circumstance.