What’s the Deal with Days On Market (DOMs)?

Anyone who’s watched the Lower Mainland real estate market through the lens of Twitter knows a bunch of the usual suspects, and I was lucky enough to get a suggestion from @FIVRE604 regarding breaking out the stats a little differently. That led to a tweet from me notifying @FIVERE604 of the change, which led to a days on market issue I never really considered:

There are some interesting claims there, not all of which I accept, but which deserve some discussion.

DOM Doesn’t Count Re-Listings.

That is correct. A re-listing is a new listing, in contrast to an extension of an existing listing. Listings are contracts, and so have certain conditions on them as a function of them being legal documents, but like all contracts they can be changed in a couple ways. They can be amended, extended, or terminated and re-done.

A “re-listing” is similar to a “counter-offer”; arguably neither one is a amended version of a previously existing thing, but rather a completely new thing altogether.

  • 123 Main Street gets listed with Listing #1;
  • Listing #1 ends one of three ways: it expires, it gets cancelled or it sells;
  • Assuming it expires or gets cancelled it can get listed again;
  • This is actually a new listing, Listing #2;
  • It could be listed with the same agent at the same or a different price and terms;
  • It could be listed with a new agent at the same or a different price and terms;
  • It is not Listing #1.

Is that too much inside baseball? Maybe, but it matters when we move on to implied motivation of different parties in the conspiracy rife tale that is #VanRE.

A listing is a contract. A contract that is ended, for whatever reasons, ends and no longer exists. In some cases a brand new contract takes the place of the first contract.

What’s the alternative? Maintain the original contract but amend the parties and terms? That’s not always workable.

However, even if I convince Rayndall that a re-list is a brand new contract the observation still stands: in those situations where an active listing is ended and replaced immediately or very soon after with a brand new listing contract, the DOMs on the second listing contract are or could be argued to be, deceiving. Fair point.

(It’s very easy for realtors and buyers with agents to see past that, but to someone not seeing the full realtor version of a listing it’s hard to see).

“BCREA Prefers the DOM to Look Shorter Than It Actually Is”

Without a smoking gun I’m going to say “That’s not the case”.

Strictly speaking I don’t know what BCREA prefers in relation to DOMs, and I’m pretty sure there isn’t a policy statement on it (and if there is it better be along the lines of “We like DOMs to accurately reflect what our members and their clients are actually doing”) but even if BCREA did want DOMs to look low for some reason, their preference is irrelevant.

A listing is a contract between a seller and a brokerage. BCREA is not a party to the contract and cannot make changes to it. Only the brokerage and the seller can make those changes. An agent may advise the seller to cancel a listing and enter into a new one in order to make it look like a fresh new listing to the untrained eye, and the brokerage that the agent works for may go along with the tactic, but the action is driven by the agent and seller, not BCREA. It’s just that simple (unless I’m really out of touch, which could be possible).

“Look at every single Vancouver West Side sale in past 2 weeks. 18 homes in all. I drilled down to individual histories… ACTUAL Days on Market (DOM) is on average 7 times longer than BCREA claims”

Now this is an interesting take, and one I like, because of the numbers involved. The stars may lie, but the numbers never do.

Last things first: BCREA isn’t claiming anything about REBGV or FVREB sales on the MLS daily stats downloads. The respective boards have data entry staff that enter data from forms sent to them by individual realtor members who enter into contracts with individual sellers. The boards are just reporting what the data entry staff entered. The graphic (unless the numbers are manipulated rather than calculated) just crunches the numbers that the staff enters.

DOMs for sales comes from two documents: the original listing, dated on Day 1, and the sales report, dated DAY 1 plus whatever number of days to took to sell. It’s possible that the data entry staff are in on a conspiracy to tweak DOMs, but that’s a big stretch, and frankly, it’d make agents mad, especially those who aren’t getting their property sold and have to explain why it takes much longer than the …manipulated average DOMS.

Anyway, I just downloaded yesterday and today’s sales, stuck them in a spreadsheet and took a look. I don’t know where 18 sales in all for the Westside over two weeks comes from, but today and yesterday saw 26 sales.

Here’s what the Hotsheet report on Paragon looks like:

What does Excel say?

High days on market is indeed 82. Low days on market is 0 (there’s only one of those). Total DOMS is 403, which, divided by 26 is…15.5. Ok, they rounded down. Median is, indeed, 9.

What does that tell us? It tells us that the boards are not manually manipulating the data entered by the data entry staff to give us lower DOMs through tweaking of the computer generated average calculations.

But what about “ACTUAL days on market”? I looked at the history of all 26 sales. 11 were listed, terminated, re-listed anywhere from one day to four months later at a lower price. About half were listed with completely new agents, indicating that the seller was making the decisions on cancelling, re-listing and price change (if any). Some of the prices changes were substantial (well over $100k) while others were minimal (under $20k).

Is 11 out of 26 a substantial number? Only if you think 42% is substantial.

So, What Are We Seeing?

@RayndallBell is identifying a real thing, but what is it? It’s not, I submit, BCREA or any of the boards manipulating DOMS to get a desired result that will mislead the public. It’s a price change by another name.

Does it render DOMs faulty? That’s an opinion question, so the answer will vary depending on who’
s doing the answering. Personally I’d say it means look at DOMs in tandem with price changes. Most price changes are reductions, even in hot, hot markets. Look at how many price changes there were for the search in question:

That’s 20 official price changes and 11 more price changes masquerading as new listings. 26 sales, 31 price changes. That’s worth burning a few brain cells on, if not a ton of them.

Could I argue that DOMs that include grossly over-listed list prices from start to finish, not matter how many traditional price changes occur screw up DOMs? Sure I could, but what’s the point? We’ll never get away from agents taking over-priced listings, whether its because the agent priced it wrong, the agent and owner priced it incorrectly, or the owner was too aggressive and the agent wanted to control the listing knowing that the market would dictate price in the end like it always does.

Whaddaya Gonna Do?

  1. Don’t diagnose conspiracy where none exists;
  2. Keep an eye on market practices like “freshening” a listing;
  3. Get occasional looks at the raw data;
  4. Don’t concentrate on single metrics – look at average, median, DOMS, Price Changes…everything;
  5. Ask the questions.

What does a price change posing as a cancellation and re-list look like? It looks like this:

A traditional price change would have a different original and list price:

8 comments

  1. I personally wasn’t aware of that. A “Days since trying to sell” would be helpful but I can see how the “parameters” of the data point would be hard to establish. I appreciate your work and the data your providing.

  2. “BCREA isn’t claiming anything about REBGV or FVREB sales on the MLS daily stats downloads. The respective boards have data entry staff that enter data from forms sent to them by individual realtor members who enter into contracts with individual sellers. The boards are just reporting what the data entry staff entered. The graphic (unless the numbers are manipulated rather than calculated) just crunches the numbers that the staff enters.”

    This, exactly. BCREA does not see individual listings or sales and we certainly can’t alter them. Moreover, I can’t even remember a time when we used Days on Market for anything in our analysis. Thanks for pointing this out.

  3. “I’d have to drill down to the individual listing level on each listing”

    This is where you are wrong. It would be entirely trivial with the access you have to this data to write a small script that tracks relistings, and generate a much more useful market analyzing statistic, as another put it “Days since trying to sell”, or perhaps “Days since initially listed”, DSIL for short.

    I have been annoyed by the lack of such a statistic for far too long. But i’m not going to go out and become a realtor just so I can create it.

  4. Adrian – I’m not wrong. I’m right. You just think I’m wrong because you don’t know me.

    I can’t write much of a script in PHP or anything else. I’ve had them written for me, but I can’t write anything but the simplest script and I’d have to keep flipping back and forth to YouTube.

    Now, can you write a script that does this or anything else useful? Let’s do it.

  5. How does said ‘DSIL’ differ from the already available ‘Cumulative DOM’ (CDOM)?

  6. How are you defining DSIL and CDOM?

  7. In response to Alex and Adrian’s requests for a ‘Days since initially listed’ script…

    ‘Cumulative Days on Market’ already exists as a search parameter in MLS.

    Is this not the same thing?

    Is CDOM a better market indicator than DOM?

Leave a Reply

Your email address will not be published.