Noe Valley Is Not Immune (Noe Valley Median Slides Along With SF As a Whole)

Woah! Where the hell has theFrontSteps been? A few things…I did, in fact, have a birthday (29 for those wondering) and I’ve been slammed with real estate. Thank goodness for the people behind theFrontSteps. This graph again from Misha Weidman:

noe-valley-vs-sf-all-districts-percent-change

This is median prices, and SINGLE FAMILY HOMES ONLY.

If this isn’t ammo for the haters, we don’t know what is. Thanks Misha for keeping the data coming!

26 thoughts on “Noe Valley Is Not Immune (Noe Valley Median Slides Along With SF As a Whole)”

  1. …and so must Cole Valley, too, eh? (are you just pointing out Noe due to that SFGate article awhile back?) Just curious.

    1. You’ll have to ask Misha. I just posted the graph. We’re hoping he does the same thing for all the districts, but that would take years to complete, so don’t hold your breath. Maybe Cole Valley should be next.

      Never saw the SFGate article. Been neck deep in writing offers, and negotiating contracts.

  2. DanRH: Yes, my guess is that Cole Valley has probably taken the same course. I will take a look there too, but I was thinking of looking at the high end like Pac Heights next.

  3. I agree that prices are down in Noe somewhat. But look at say February 2006. Is that data point really indicative of what the market looked like then? Or were there only a few sales? The reality is that certain months only contained a few sales, and that there were news events within all of this. In the fall of 2006, a lot of people were convinced that Noe had peaked. In late August 2007, financial markets were roiled. Numerous buyers backed out of contracts, and a two month mini-scare set in.

    As of right now, there are six properties in contract of pending which exceed the median, most by a wide margin. If they were to all close in close succession, guess what? The data likely shoots north. Or if any of the higher priced properties should sell, including 465 Hoffman at ~4M. And that is one to watch. Because despite all of this, Noe never got to a point during which it was routinely producing 3M+ sales.

  4. Bottom line is that April – when anecdotes here were that things were moving – saw 8 sales in the whole of D5 fot SFHs.
    Thats a 79% Decrease in sales, and an 84% drop from 04.
    That can only lead to substantial discounts as that is the only way houses wil sell.
    So yes, I agree fluj median prices where data is so scarce is pretty volatile.
    The real story in in sales volumes I think.

  5. First, what a great looking chart. What program was used to create it. Very nice.

    Second, great analysis too. Thanks.

    Third, this chart excludes what I consider to be a highly relevant time period from 97-02 where there was a whole other rapid appreciation bubble (as noted in Misha’s posting). Include this data and you would see how rapidly and massively prices have increased; and how far potentially they could go. I understand Fluj’s comment above but despite his points it is still very difficult to dispute the raw data. I ran some current Noe numbers just now(data via cleanoffer):

    53 SFH’s on MLS for $71.5M w/ avg = $1.35
    34 Avail SFH for $48.8 w/ avg = $1.43
    19 In Escrow SFH for $22.7 w/ avg = $1.20

    The $1.2M for the in escrow properties assumes these sell at asking which is optimistic in this market and right in line with Misha’s Musings. It would also suggest that on average the available SFH’s are overpriced slightly.

    Lastly, I think it’s safe to assume that Noe is a “Prime” area of SF and not likely to retread too far back down the line. Wish I lived here in ’97!

  6. Yeah, but sales are taking longer now. Things did move in April compared with the preceding few months. There are 19 SFRs in contract or pending for 5-C alone right now.

  7. I think another good anecdote to maybe keep an eye on is 587 Jersey. It’s a fixer with an attic and an oversize garage that appears to have pretty good bones, on a great block, decent lot (25 X 89) with views. Two years ago it sells for the low 9s, or maybe even ~1M. Now I think it goes somewhere in the 8′s. If that happens, to my mind the fixer market in Noe has reset to 2005 type levels, when every single one seemed to be priced at 699 and sell for 850K.

  8. “LOL. Yeah, and the markets are up because “Pending Home Sales” are at an all time high.”

    Markets are up, eh? You said it. Not me. Well, you and the NAR.

    To be clear, I said April saw more movement than preceding.

    Seriously man. Take that shit to you know where.

  9. Well, I obviously didn’t appreciate yours either. Leading with LOL? If you’re the same Eddy I think you are, it wasn’t all that in keeping. Regardless, I apologize. You were talking about the odd way the markets are bending over backwards for any bit of halfway decent information.

  10. Same Eddy, and my LOL was not directed at you; it was directed at the NAR and the markets in general. Apologies here too. All good; thanks.

  11. Whew, glad I stayed out of the cross-hairs on that one. Fluj, you are absolutely right to point out that the medians can be misleading from one month to another because of a small number of sales. I mention that, along with some other methodology issues, in the full post on my blog. However, I do think that the trend since last March is pretty incontrovertible and ultimately shouldn’t really be a surprise. Why should Noe Valley be immune?

    Eddy, thanks for the compliments on the charts. They’re all Excel-generated, but with a special add-in that I purchased to generate medians in a “pivot table” type format. That’s what allows me to dice the data fairly fine.

    I have a couple of comments on recent sales/listings available in Noe but I’m going to just post those on my own blog since I started tracking them there originally.

  12. Noe Valley is not immune. Nothing is going to be “immune.” That’s why the very first thing I said was that I agree values are down somewhat. Since last March, you think? I bet if you look at last March there will be a jumble of sales in the high 2′s. I’ll post last March later. Out the door …

  13. Yeah, just like I thought. Of the 11 SFR sales in 3/2008, five of them were above 2M, four of those above 2.5M, and one above 3M. Of course it’s gonna look like a spike.

  14. Fluj, I’m not sure what your point is. Given the danger that using less data points (sales) will always pose to statistical analysis — a danger I’ve already acknowledged — why discount the significance of a median based on 11 sales, a substantial portion of which were at high values? I haven’t gone back to verify, but my recollection is that 11 sales is not a particularly low number for a month’s worth of sales for this subdistrict. In fact, I believe it’s on the high side based on current rates. Additionally, my numbers are 3 month moving averages, which should further reduce anomalies in any particular month.
    Perhaps most importantly, I’ve gone to a fair bit of trouble to use medians rather than averages to generate these charts precisely because medians are less distorted by values that are unusually low or high.

    If I were looking to dispute the 30% fall in Noe Valley values, I think I’d probably be looking at different price segments of the market and trying to differentiate between, say, what’s happened to homes that were selling for over $2 million, vs. those (were there any?) that were selling in the $1.5 million range.

    Seb, I’m going to stick my neck out and say that –barring a global economic melt-down (I’d put the chances of that at less than 10%) — prices will be higher. Why? A. People have short memories. B. There is or will be pent-up demand. C. San Francisco’s denizens — and certainly those that can afford to live in Noe Valley — are relatively well-placed to ride out the rough economy and be on the front end of its recovery.

  15. Misha,

    Noe has not fallen by 30 %. I specifically looked at March 2008, because that’s the departure in your chart. Eleven sales was probably a decent enough figure for the time, but the cluster of high end sales produced a distorted statistic. And you are on this very thread saying that the trend was since last March. It’s not. Look at October 2007 versus now, for example.

  16. Fluz: I guess we’ll have to agree to differ then. Noe Valley has indeed fallen by 30% from its all-time high, as my chart indicates. If you wish to argue that March is an aberration, that of course is a different story.

    I’ve looked (again) at sales for March 2008 and I don’t see anything unusual or aberrational about them. I’ve also looked at sales for October 2007, as you suggested, and I note that there were only 6 sales, so if anything drawing a median from that month is more questionable because there are less data points.

    I’ve sent Alex another chart that simply plots median prices (rather than % change). Hopefully he’ll post it. IMO it shows a pretty clear trendline up in prices from Feb 2006 through March 2008, notwithstanding a significant dip during the Fall of 2007.

    Anyway, I’m done with Noe Valley for now. Moving on to another subdistrict. It’s always good to have a tough interlocutor — keeps you honest!

  17. OK then. My argument is three fold. One, that median with small datasets is a poor indicator. Secondly, March 2008, with 5 out of 11 over 2M, was without question an aberration. (That month even contained the bloginfamous 3.038M sale of 2212 Castro street! Again, Noe has seen precisely three ever 3M+, and the other two wer double lots and much larger.) Last, due to small datasets and relative stasis with Noe’s sub ~1.2M SFR homes, your chart is mostly only tracking a difference in the highest priced Noe houses.

  18. I have to agree with fluj.
    the problem is not the size of the data – but the fact that the data doesn’t consistently covers the three noe market.
    top market: staying strong in breathtaking new supplies year after year (fire station, google house etc). middle market: standard noe moving condition 3br 1-2ba – 1500-2200sqft and low market: all the fixers which are often sold as tear downs.

    I don’t have time to specifically back up my assertion, but I’ll state it anyway: “the middle market has been about absent from the radar since the spring 2006″ (I’m not sure about the exact date – but its when there is no move-in property in the sold lists in the 1.2-1.7 range -> this was creating even more problem, the few houses in that range being snapped way over a reasonable value)

    the whole foundation of a good data representation is when the Gauss is looking like a
    [img]http://static.daysofwonder.com/lang/english/images/gf_ranking5.gif[/img]
    not like
    [img]http://ninaherenorthere.com/wp-content/uploads/2008/08/twin_peaks-san_francisco.jpg[/img] —> everything breaks, from average to median to dispersion of the data etc… Stats require to have more than 2/3 of the sample data in the close range of the median.

    I’m not saying that the post is not GREAT – I just find it difficult to use it as the only chart into making any analysis or prediction for NoeValley.
    as usual, if you need some hard core data regarding one specific market (to sell or buy), only an experienced agent in that market will be able to sort data out to get you a reasonable but tailored data sample to help you nail down a property value or value change

    PS: for the fun of it, I visited a property asking 40% more than 2.5years ago when I put it at – 5% – a 45% difference! (and no, it should be nowhere to -30%)
    PS2: re a 30% drop. if this is indeed a FACT – then give me specific properties that were at X and are at Y today (same property) – and vice versa, properties that are at Z and would have had an asking of W then.
    example: if 123 TheFront street closes today at 1.7M. Would that property have been at 2.4 in that month/year?
    if 456 Steps street closed 2 years ago at 2.4, would that property be asking/closing today 1.7?

    this is very different from 2008 only sells 2+M houses while 2004 only sold 695 asking 850 closing houses. Those houses are not the same houses – nor the same market nor the same buyers nor anything in common.

  19. I am not proclaiming any sort of rebound, or even offering this one up as an overall indicator of resiliency. But 261 28th street has gotten into contract after about four days on the market. What I can say is that if it closes, it too will fall into an act.cont/pending Noe group of SFRs which should make Misha’s chart spike north once again.

  20. Misha is using medians, and 3 month MA of median at that. So high dollar amount transactions won’t move the needle much, unless there are a lot of them, which I guess is what you’re implying. As always, we shall see.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s