CDOM: Flawed Failsafe

Posted by Dave Fratello on Tuesday, October 2nd, 2007 at 5:22am.

Discussion of DOM (days on market) data often opens a chasm between consumers and critics, on the one hand, and real estate professionals, on the other.

The pros tend not to see what all the fuss is about. If we point out that bogus re-listings serve no purpose but to manipulate data the public sees about the RE market, the response is a quiet shrug.

Consider the tin ear displayed by the SoCal MLS consortium when it decided to drop DOM from client reports, and even publicly announced the decision. (MBC chimed in here, and the LA Times ed board here.)

One reason the pros see no problem is that they have their own data field that they trust to cut through the BS of bogus re-listings. It's called CDOM, for "combined days on market." If you work with a realtor, they're confident that CDOM will help them give you the real story.

The rules for our local MLS essentially say you can re-list all you want, reset the DOM for such occasions as a price cut, change in seasons, child's birthday, or any seller whim, but the CDOM field has got to be real. In fact, one of our frequent commenters says it's all but impossible to imagine a way to manipulate the CDOM field.

But it seems the CDOM field is rife with errors nonetheless. Significantly, most errors favor the sellers – meaning the available CDOM figure understates true "CDOM."

Your humble correspondent is not a realtor. MBC doesn't have ready access to the realtors' system.

With a little help, we compiled a snapshot of "CDOM" fields as of Monday, Oct. 1, for active properties identified in our tracking as those that had pulled a bogus re-list since hitting the market. There are 25 such properties (1/3rd of the current inventory), of which 11 had substantial problems with the CDOM field, when compared to MBC's data. Here they are in order of the difference between the True DOM and the CDOM field:
  • 117 Highland – began 5/3/07 – True DOM: 165 – CDOM: 153diff: 12
  • 505 3rd – began 6/25/07 – True DOM: 98 – CDOM: 86diff: 12
  • 1821 Walnut – began 5/22/07 – True DOM: 126 – CDOM: 105diff: 21
  • 869 3rd – began 7/30/07 – True DOM: 62 – CDOM: 37diff: 24
  • 225 39th – began 8/10/07 – True DOM: 51 – CDOM: 21diff: 30
  • 601 Larsson – began 3/20/07 – True DOM: 194 – CDOM: 157diff: 37
  • 742 33rd – began 8/24/07 – True DOM: 38 – CDOM: n/adiff: 38
  • 2709 Oak – began 8/15/06 – True DOM: 410 – CDOM: 346diff: 64
  • 209 42nd – began 10/4/06 – True DOM: 361 – CDOM: 195diff: 166
  • 232 30th Pl – began 4/10/07 – True DOM: 176 – CDOM: n/adiff: 176
  • 844 11th – began 4/26/06 – True DOM: 521 – CDOM: 207diff: 314
We need to note that 232 30th Pl first came up for sale as 3009 Highland, then changed address midstream. (See story here.) Convenient, tho, to have no CDOM entry at all – that avoids the rule saying CDOM is "property-specific." Intriguingly, the same agent repping this listing with no CDOM figure also has 4419 Highland, new construction that has neither a DOM figure nor a CDOM figure.

Can these fields be manipulated? Ask the man.

There was a pleasant surprise in this data. The fortress-like home at 3200 Pacific was on the market last year, dropped off for several months, and came back $450k lower on June 8. (See story here.) MBC shows True DOM as 84 as of Oct. 1, tied to the June 8 start this year. (It has been re-listed since June, on account of some price cuts.) But the CDOM shows 221, reflecting the property's failed attempts last year.

So you see, sometimes the CDOM field can be a well of truth. Just don't count on it.
comments powered by Disqus