Inventory of listed single-family homes, condos and TICs in San Francisco fell 24% over the past three weeks driven primarily withdrawn or deactivated listings during the holidays. On average, inventory has fallen 23.3% during the same three weeks over the past four years.
Current listed inventory remains up 34% on a year-over-year basis, up 19% versus the average of the past four years, up 36% as compared to an average of 2006 and 2007 while listed sales this past November (325) were off by 17% year-over-year.
The inventory of listed single-family homes for sale in San Francisco (460) is up 51% on a year-over-year basis while listed condo inventory (609) is up 24%.
Forty-three (43) percent of all active listings in San Francisco have undergone at least one price reduction while the percentage of active listings that are either already bank owned (104) or seeking a short sale (187) has ticked up to 27%, up 2% on an absolute basis over the past three weeks.
Expect listed inventory in San Francisco to steadily climb over the next four months as withdrawn unsold listings return to the MLS along with the new.
The standard SocketSite Listed Inventory footnote: Keep in mind that our listed inventory count does not include listings in any stage of contract (even those which are simply contingent) nor does it include listings for multi-family properties (unless the units are individually listed).
∙ Listed San Francisco Inventory Update: December 13, 2010 [SocketSite]
Comments from Plugged-In Readers
For those who like market-wide stats (I’ve noted my problems with them), the $/sf for listed homes are now at a new 5-year low – by far.
Predictable, given the higher inventories and lower sales we’ve been seeing for many months now.
red is too inflammatory… if SS did that many posters would scream foul.
yellow is too hard to see.
let’s see here… red (no), orange (done), yellow (too hard to see), green (done), blue (done) indigo (done).
I’d say violet but that’s so close to blue..
so I’m going brown.
although it’s nice to have this chart every month, I think it means little until the Superbowl when all the listings get put back on.
as always, the peak months will really tell us what’s happening… May, June, July.
Let’s think outside the box like Bernanke: It’s going to be dotted or dashed, in which case the color is completely irrelevant, until next year 🙂
happy new year!
I think a line intertwined with orange and black so we can continue the celebration.
I just assume that SS keep the current year black and recast 2010 is some other color.
Now that that’s been established, anyone have thoughts on which way the new line will go? My guess is we see spring inventory above historical averages, based solely on the large amount of properties delisted at end of ’10, but just a guess.
Why are we comparing to peak bubble years at the height of the frenzy. Very hard to tell what is normal.
I agree that it would be interesting to see 2010 data plotted against 2005 and earlier data and even better against an average of the last 10 years to better see how much we really are off the norm.
Regarding inventory here is Months’ Supply of Inventory (MSI) for last 24 months through November and some commentary from Paragon.
“MSI is defined as the number of months it would take to sell the current inventory of homes for sale, at the current rate of sale: generally speaking, the lower the MSI, the greater the demand. MSI for all SF homes was 3.8 months in November, which is moderately low. However MSI varies widely by property type: for houses, the MSI was lower at 3 months; for condos, it was 3.9 months; for TICs, 6.3 months; and for 2-4 unit buildings, 5.2 months of inventory. The MSI for luxury homes was 3.8 months.”
I wonder what the data sources here are. Based on SocketSite’s MLS updates and MLS data on rereport.com, I get the following:
Columns are MLS inventory as reported by SS at closest to mid-month, MLS sales as reported on rereport, months of inventory, and trailing three months of inventory:
Nov-10 1834 326 5.6 5.4
Oct-10 1936 372 5.2 5.0
Sep-10 1797 341 5.3 4.8
Aug-10 1594 360 4.4 4.3
Jul-10 1799 385 4.7 3.9
Jun-10 1697 446 3.8 3.9
May-10 1599 480 3.3 3.8
Apr-10 1491 317 4.7 4.2
Mar-10 1361 414 3.3 3.9
Feb-10 1116 247 4.5 3.8
Jan-10 909.5 226 4.0 3.4
Dec-09 1131 412 2.7 3.2
Nov-09 1346 395 3.4 3.5
Oct-09 1460 435 3.4 3.3
Sep-09 1448 392 3.7 3.3
Aug-09 1352 495 2.7 3.5
Jul-09 1569 452 3.5 4.3
Jun-09 1630 389 4.2 5.1
May-09 1685 328 5.1 6.0
Apr-09 1622 277 5.9 6.9
Mar-09 1648 238 6.9 7.7
Feb-09 1500 192 7.8 7.2
Jan-09 1189 141 8.4 7.1
Dec-08 1405 265 5.3 5.8
Nov-08 1788 241 7.4 5.6
Oct-08 1789 381 4.7 4.3
Sep-08 1544 332 4.7 3.8
Aug-08 1388 393 3.5 3.4
Jul-08 1470 478 3.1 3.2
Jun-08 1496 426 3.5 3.4
May-08 1491 478 3.1 3.6
Apr-08 1381 382 3.6 3.9
Mar-08 1329 318 4.2 4.3
Feb-08 1159 292 4.0 4.0
Jan-08 1053 215 4.9 3.8
Dec-07 1077 355 3.0 3.3
Nov-07 1393 408 3.4 3.6
Oct-07 1532 452 3.4 3.3
Sep-07 1408 346 4.1 3.0
Aug-07 1187 464 2.6 2.4
Jul-07 1157 470 2.5 2.3
Trailing 3 months produces the best graph. Months of inventory sometimes looks noisy, and it’s strange that it looks less noisy from Paragon.
Pent up supply inched upwards the past three weeks. Currently, 1604 homes are in some state of foreclosure (NODs, NOTS, bank owned) in Ess Eff. This is compared to 1500 homes three weeks ago. Standard disclosures about noise in the data; information deemed reliable but not guaranteed.
The inventory number is pretty significant if you ask me. It’s all over the place and shows decent demand even in times when you’d think demand would be very weak.
Here is the calculation as I understand it:
MSI = (FS – UC – X) / UC
Where FS is the number of properties For Sale during the month (any property that was Active at least one day during the month), UC is the number of properties that went Under Contract during the month and X is the number of properties that Expired during the month.
“MSI = (FS – UC – X) / UC”
Based on MLS? It’s not strange that the numbers from Paragon diverge from my numbers in differing magnitude on occasion. Since Paragon excludes under contract houses, I would assume the following differences from my compliation: 1) the magnitude of difference from month to month should in affect be amplified going forward because under contract is a forward-looking indicator of sales, and 2) the direction of the difference from month to month should be consistently lower than my number because I don’t exclude under contract houses.
However, what’s strange is that Paragon’s months of inventory numbers are sometimes above my numbers, and sometimes will stay above my numbers for up to 6 months at a time or maybe just for a month or two — nothing consistent. I don’t know how to explain that.
Boy, there quite a few new listings. I see 135 in the past 7 days and over 80 in the last 24 hours. I don’t know how many withdrawals or in-contract places we’ve had, but it looks like we are already coming up from the winter low points. Lots of pent up supply . . . let’s see if demand picks up to match it or if it stays low.
“I see 135 in the past 7 days and over 80 in the last 24 hours. I don’t know how many withdrawals or in-contract places we’ve had, but it looks like we are already coming up from the winter low points.”
This is somewhat normal. Judging from past years’ data, even with the massive spike for winter/spring 2008-2009, it seems like there is a local maximum in trailing 3 months of inventory in November, and then the overall maximum is in March or April. That suggests that either sales are low or inventories are high through January and February. December appears to be a local minimum.
Comments are closed.