San Francisco Sales Volume And Median Price: May 2011 (www.SocketSite.com)
Recorded home sales volume in San Francisco fell 20.1% on a year-over-year basis last month (492 recorded sales in May 2011 versus 616 sales in May 2010), up 16.6% as compared to the month prior which was down 1.4% year-over-year.
Keep in mind California’s Homebuyer Tax Credit program created an incentive to push April closings into May last year.
That being said, May sales figures for San Francisco from 2004 to 2009 were 938 (2004), 773 (2005), 691 (2006), 616 (2007), 593 (2008), and 498 (2009), averaging 685 sales per May over those six years. And on average over the past seven years, sales volume has increased 13.8% from April to May.
San Francisco’s median sales price in May was $660,000, up 3.7% compared to May 2010 ($636,500), up a nominal 0.8% compared to the month prior.
For the greater Bay Area, recorded sales volume in May was down 15.4% on a year-over-year basis, up 2.9% from the month prior (6,988 recorded sales in May ’11 versus 8,264 in May ’10 and 6,789 in April ’11) as the recorded median sales price fell 9.3% year-over-year, up 3.3% month-over-month.

Last month’s sales were the lowest for the month of May since 2008, when 6,216 homes sold, and the third-lowest on record, behind May 1995 and 2008. May sales have ranged from a low of 6,216 in 2008 to a high of 13,567 in 2004, while the average is 9,693. Last month’s sales fell 27.9 percent below the May average.

Distressed home sales made up about 45 percent of the Bay Area’s resale market last month.

Foreclosure resales – homes that had been foreclosed on in the prior 12 months – accounted for 26.9 percent of resales in May. Last month’s figure was down slightly from 27.9 percent in April and up from 26.7 percent a year ago. Foreclosure resales peaked at 52.0 percent in February 2009. The monthly average for foreclosure resales over the past 15 years is about 9 percent.

Short sales – transactions where the sale price fell short of what was owed on the property – made up an estimated 18.3 percent of Bay Area resales last month. That was up from an estimated 17.5 percent in April, 18.2 percent a year earlier, and 12.5 percent two years ago.

At the extremes, Santa Clara County recorded a 23.6% drop in sales volume (a loss of 510 transactions) on a 5.1% decline in median sales price, no counties recorded an increase in sales volume. Solano recorded a 13.7% decline in median price on a 7.7% decline in sales (50 transactions) while San Francisco was the only county to record an uptick in median.
As always, keep in mind that DataQuick reports recorded sales which not only includes activity in new developments, but contracts that were signed (“sold”) many months or even years prior and are just now closing escrow (or being recorded).
Bay Area May Sales: Median Price Up From April; below 2010 [DQNews]
San Francisco Recorded Sales Activity Down 1.4% In April [SocketSite]
San Francisco Recorded Sales Activity Down 20.1% In May [SocketSite]
San Francisco Listed Sales Volume Down 14% In May (YOY) [SocketSite]
He’s It’s Back: California’s $10,000 Homebuyer Tax Credit Returns [SocketSite]

25 thoughts on “San Francisco Recorded Sales Activity Down 20.1% In May”
  1. The chart makes it clear why sales are down: there was a peak last May. Since it then dropped steeply in 2010 expect sales to be “up” YOY for July.

  2. Medians are not prices, as is oft quoted.
    yet its interesting how steady the 12mth moving average has been for around 2 years now.

  3. Sorry to be off-topic, but did I miss something? Or has no one commented on the fact that Zillow just changed its algorithm for computing “Zestimates.” Big change in SF property “values”, at least for the properties I follow. If you haven’t checked your home’s “Zestimate” lately (and you care), you may want to check it. By way of example, a friend’s Haight condo was re-Zestimated from $1.1m to $865k overnight.

  4. sanfrantim, I haven’t looked at Zillow in a long while but it is linked in to my Yodlee account, so I have historical data.
    I can see that they have dropped the valuation of our place about 10% over the last couple of weeks and have done so for their graphs as well — so they have lowered what they are listing as your estimated value a year ago, even though a year ago they presented higher numbers on their web site. Does that make sense?
    The new numbers are much closer to what I think are accurate, btw and even line up pretty well with the appraisal we got done about a year ago when we got a refi.

  5. “Or has no one commented on the fact that Zillow just changed its algorithm for computing “Zestimates.””
    I heard something about this, but I don’t really have any comments, since I don’t look at Zestimates too carefully. Do they seem like they are more accurate? I’ve heard about swings in the opposite direction too, albeit probably smaller than the $235K you pointed out, sanfrantim.

  6. “So they have lowered what they are listing as your estimated value a year ago, even though a year ago they presented higher numbers on their web site. Does that make sense?”
    I don’t think it’s that uncommon to revise historical estimates based on new data or algorithmic changes. A great deal of economic data gets revised that way.
    I read that the new algorithm better incorporates remodel data which when previously ignored would show up as increased market appreciation. Previously your unremodeled house would get a boost if comp homes were remodeled and sold since all the comp value changes were attributed to market changes.

  7. I agree with NVJ that the new estimates are more accurate and, generally, lower in SF. There are some exceptions. A friend’s home in Sonoma jumped up overnight by 10% as a result of the new algorithm.
    As much as it is fun to mock these Z- values, they play a very influential role in this market. My friend, who is out of work and whose condo just “lost” $200k in value, has been using the prior inflated Z estimate as false assurance and as a basis for making important decisions in his life — whether to keep paying the mortgage or not (I advised him, not).

  8. I should add, as NVJ points out, that Zillow has made the adjustment post hoc – lowering the historical data points of the past year or so, so that the value graphs do not show a recent major dip. But if you’ve looked before last week or so, and look again today, you will see a profound difference.

  9. Interesting sanfrantim, I wonder if this recent algorithm adjustment is going to result in a mini-wave of zillow induced walkaways.

  10. Look everyone the markets crashing. Medians are down everywhere. Look out below. Except in San Francisco, where medians shouldn’t factor into the equation anyway. It’s all just new kitchens, marble baths and chopped pillows.
    On a more serious note, Zestimates are a total joke, in general, but especially in San Francisco. I have no idea really how they fare in more stable neighborhoods that do not vary in quality block to block but here in SF there is hardly any use in these things. The new algo seems to give more credit to recent proximity home sales and applies that to neighboring properties. Seems like a good idea but it’s all local here. I also don’t buy the conspiracy theory that Zillow made these changes to skew the market metrics. They could care less.

  11. According to Alexa, zillow.com has a US traffic rank of 118 among all websites – phenomenally high. [For comparison, socketsite.com’s is about 37,000.] Meaning a lot of people read and presumable rely to some degree on Zillow. It is a force to be reckoned with. I would venture say that Zillow’s new algorithm will do as much to move the market than most anything else discussed here. And, yes, to speculation about zillow-induced walkaways. (But, no, to the theory proposed above about a zillow conspiracy.)

  12. “They could care less.”
    They obviously care: They are trying to be more accurate. And Kudos to them for improving their estimates. Nothing wrong with more and more accuracy as they revise their algorithms over time.
    As real estate prices sink, I would think people would want a more accurate estimate of how much money they are losing.
    I, of course, get to see that every day when I do my stock quotes for the 10,000 shares of LinkedIn I bought at $122.70 (about a month later, it’s at $71, for a stunning $500,000.00 one month loss) and the 10,000 shares of Pandora I bought yesterday at $26 (One day later it’s at $16, for a devastating $100,000 one day loss), so it’s only fair that people who were counting on the IPOs of these stocks to INCREASE the value of their homes to be able to see their homes plummet in value as a result of the decline of these same stocks.
    Kudos to Zillow. As long as they aren’t making any money, and have no real prospects for doing so, I’ll be eagerly awaiting their IPO so I can buy more shares with my dwindling funds I had saved for a downpayment.

  13. “Meaning a lot of people read and presumable rely to some degree on Zillow. It is a force to be reckoned with.”
    I’d agree with that. Even my mom, who is barely able to use the internet, has mentioned using Zillow.

  14. I, of course, get to see that every day when I do my stock quotes for the 10,000 shares of LinkedIn I bought at $122.70 (about a month later, it’s at $71, for a stunning $500,000.00 one month loss) and the 10,000 shares of Pandora I bought yesterday at $26 (One day later it’s at $16, for a devastating $100,000 one day loss), so it’s only fair that people who were counting on the IPOs of these stocks to INCREASE the value of their homes to be able to see their homes plummet in value as a result of the decline of these same stocks
    Why?

  15. zillow.com has a US traffic rank of 118 among all websites – phenomenally high. [For comparison, socketsite.com’s is about 37,000.]
    That’s like comparing apples with blueberries. Zillow is national, covering a 300M+ population. SS covers only 820K or ~350 less. Pretty much in proportion.
    Another thing about SF peculiarities in regard to Zillow. For most of the US, prices are so much depressed that not many sellers will perform expensive redos. Therefore Zestimates are decent proxis. The 2 main reasons I think: 1 – You might not recoup your costs because buyers are in charge + prices might have dropped when you’re done with the redo. 2 – Many resales are distressed which means very limited cash for redos. Maybe enough for some lipstick, but that’s it.
    In SF it’s a different story. Some wealthy buyers do not care that much about paying a premium to get a place in mint condition. They’re looking for perfect and wow, not something you’re waiting 6 month or a year to show your friends. Plus the cost of some redos by pros fall into the sweet spot where contractors/renovators can make a decent profit while buyers still get a very decent bang for their money.
    A small data point. Paris is now at $1200/sf on average (not median) whereas SF is around 2.5 times less. A tech worker up there will make 20 to 50% less than in SF, not including the money from stock option if he is in the right spot.
    SF is a very affordable city when viewed from this prospective.

  16. Tipster obviously is shorting these stocks and is manipulating these equity prices just like he shorts the housing markets and manipulates those as well. Pure genius.

  17. Paris? Seriously? The one in France?
    I haven’t done the arithmetic, but I’d guess that SF is a very affordable city when compared with the Upper East Side of Manhattan. Doesn’t mean that’s a meaningful comparison.

  18. Is there a way to short the SF housing market? How? If there is, I’ll do it in a heartbeat. Could have made a LOT of money doing so in the last 4 years and still more to be made (if there were a way to do it).

  19. Brahma,
    it does apply in a few important ways:
    – A city 5 times smaller than its suburb where the core is very different from the rest
    – A place that attracts regional, national and international talent
    SF is competing for knowledge workers, is part of the global wage arbitrage war and has founds its niche. All these could also apply to Paris or London or New York.

  20. A.T. – You can short the Case-Shiller at various points I think.
    I don’t kniw whym, but alwaysa find the link for this hard to find to will leave this for someone more dedicated than me 😉
    I seem to recall the SF futures weren;t showing much of an increase or decrease last time I checked but that was some time ago now (I remember a nify graph) so, yes, if you think the index will contine to fall, and it does, there’s gold in ’em hills.
    [Editor’s Note: They’re Betting Against Us (San Francisco) On The CME.]

  21. It looks like one main change was the remodel data. User submitted though, which might have issues.
    The measure their accuracy by comparing sales price to the Zestimate just before the sale. They claim the recent change brings median error down from 12.7% to 8.5%.
    While it wouldn’t seem unreasonable to think that SF Zestimates fair worse, it actually turns out that SF has a median error of 7.2%, better then average. Nationally 33.2% of homes Zestimate to within 5% of sales price, for SF 37.7% are within 5%
    http://www.zillow.com/howto/DataCoverageZestimateAccuracy.htm
    Just conjecture, but possibly accuracy is higher in places where land value is a higher fraction of purchase price. I’d think that issues like remodels and home condition would add more error to the structure value component.
    What kinds of improvements did you make to the Zestimate?
    We improved the algorithm models that better incorporate user-submitted data and market data. This has improved the Zestimate median margin of error to 8.5 percent nationwide. Previously, it was 12.7 percent.
    What do you mean by user-submitted data?
    Zillow users have made updates on more than 28 million homes, adding information such as remodel and home fact information that’s not reflected in public records. These updates are used to calculate more accurate Zestimates on more homes.

    http://www.zillow.com/wikipages/Zestimate-Improvements/

  22. “A.T. – You can short the Case-Shiller at various points I think.”
    My understanding is that both the Case-Shiller options trading and the market ETF’s (UMM, DMM) were commercial failures. I think the ETF’s closed out and last I looked open interest in SF options were in the single digits.
    Conjecture, but I’d guess that since housing recently has had such a strong national component institutions can go long/short housing via the already existent mortgage backed securities and derivatives. Which in many cases are already conveniently tranched up for different risk levels. You can think up some uses that individuals and small time builders could put these to, but it doesn’t seem surprising that this wouldn’t appeal to large numbers of these groups.

  23. SF Zestimates appear to have benefitted from the change. They have a blog post showing before and after numbers for selected cities.
    Under the old algorithm SF had a median error of 10.1% (Better then the old national average) which dropped to 7.2% with the new one. Homes selling within 10% of Zestimate for SF increased from 50% under the old method to 62% under the new.
    Additionally, they note that they have only recomputed historical Zestimate data using the new algorithm back to 2006 as of now.
    They plan to recompute back to 1996 in the future, but for now there’s probably a discontinuity if you look past 2006.
    http://www.zillow.com/blog/research/2011/06/14/upgrading-the-zestimate/

  24. I think the comment about Paris is really relevant and it’s driving my personal decision about not purchasing in San Francisco. Except, my comparison is with Paris, Texas with a population of about 25,000 and an average home price of $73,000. I think that SF is overvalued and I refuse to buy in Pacific Heights until I can get a SFH for $73,000. As a value investor, it only makes sense . . .

Leave a Reply

Your email address will not be published. Required fields are marked *