An analysis of some psf trends from July 2008 to June 2009 using NYC data
Started by joedavis
about 17 years ago
Posts: 703
Member since: Aug 2007
Discussion about
I went over and analyzed the linear trend in real estate prices from July 2008 to July 1 2009 using the NYC real estate data (from the NYC site -- deleted all transactions less than $100 psf and over $5000 psf so that no strange things remain in the data I did this by category (as a regression of psf of the property vs the date and am reporting the slope of this regression) and by per sq foot so... [more]
I went over and analyzed the linear trend in real estate prices from July 2008 to July 1 2009 using the NYC real estate data (from the NYC site -- deleted all transactions less than $100 psf and over $5000 psf so that no strange things remain in the data I did this by category (as a regression of psf of the property vs the date and am reporting the slope of this regression) and by per sq foot so that it is easy to figure out the change in price for any property you are interested in: ALL MANHATTAN irrespective of the type of property. There are a total of 5943 transactions over the year and they translate into an annual psf decrease of $60 CHELSEA All types 462 transactions and a decline of psf by $263 Financial district 509 transactions and a decline of psf by $164 Flatiron 182 transactions and a decline of psf by $153 Greenwich Village 195 transactions and a decline of psf by $310 Harlem Central 395 transactions and a decline of $72 Harlem East 136 transactions and a decline of $234 Harlem Upper 35 transactions and a decline of $184 Kips Bay 81 transactions down $122 Lower East side 46 transactions UP $18 Manhattan Valley 71 transactions DOWN $423 -- this really stands out since the base psf is not that high MIDTOWN 856 transactions down $233 MUrray Hill 285 transactions UP $29 Soho 66 transactions down $343 Tribeca 457 transactions UP $98 -- most remarkable a few recent high priced closing and a reduction in low priced closings marks the trend Upper East Side 735 transactions UP $5 Upper West Side 850 transactions down $146 Washington Heights 84 transactions down $105 For some of the areas I have been following more closely, I broke this down by category Summary for Harlem - Central (I have not done other areas in Harlem but am happy to share the spreadsheet in case someone wants to do it) 1) 1, 2 or 3 family buildings, i.e. basically brownstones 19 total transactions with a drop of $126 psf for the year. So if you are interested in a 3000 sf brownstone this is $378k drop in value 2) Rental -walkup apts (1 person owns the building and rents) 37 total transactions with a drop of $18 psf for the year 3) Condos - walkup apts (building is divided up into separate owners) 32 transactions with a drop of $500 psf per year This is screwy -- beginning of the year is marked by a few transactions at nearly $1000 psf and then end of the year has very few transactions and they are around $300-400 psf. so essentially the sales associated with the high end brownstone condos evaporated. Probably a good market to avoid in a downturn 4) condos elevator and condos 2-10 unit residential prices went UP $16 psf!! Again , here the high end sales >$1000 psf are gone. The new high end is $700 psf. However, the low end that was probably the lottery units has also vanished so the low end climbed from 200 in 2008 to 300 psf in 2009 So, very interesting how the declines are distributed across the different categories If you look at the detailed data (can't seem to attach spreadsheets or figures here but can email if requested), we see that there is a uniform shrinkage in volume and a sheer drop off in sales in the higher priced properties and the middle range is holding up -- looks like the high rollers and the low enders dropped off, leaving the middle class schlups to cough up the sales. Empirical evidence is that many of the speculators who bought to flip and sitting on their properties and planning to ride this out. They may have quite a wait the way things are going. By te way I ran the same analysis on the upper west side. For those who are interested 1) Rental Walk up -- actually appreciated $124 psf for the year 106 transactions 2) Condos elevator dropped $166 psf for the year based on 746 transactions Of course the average psf in UWS is higher than the top psf in Harlem [less]
Add Your Comment
Recommended for You
-
From our blog
NYC Open Houses for November 19 and 20 - More from our blog
Most popular
-
3 Comments
-
38 Comments
-
28 Comments
-
33 Comments
Recommended for You
-
From our blog
NYC Open Houses for November 19 and 20 - More from our blog
Thank you Joe. I found your analysis very insightful. It's posts like these that make this board worth coming to. I wish there were more of them.
Wow, great work!
Great info. How about the Turtle Bay area?
Thanks very much for sharing your research. It was very informative.
Excellent work. I am surprised by the UP findings in some areas. Murray Hill has been undervalued for awhile, in my most humble opinion, but why it should catch up a bit right now, I can't say (bargain hunters? Well nothing in Manhattan is really a bargain, so relative-bargain-hunteres) Why Soho and the Village should fall so much, while Tribeca gains, I don't know either. Of course, square foot isn't the whole story, we all know that, but it is important. Fascinating stuff! Thanks for your hard work.
{Manhattan real estate agent}
joe, great stuff.
fluter, it could always just be a case of type of sales. more condos than coops, or more families looking to buy to obtain a spot in a decent public school (ps116, 234) rather than singles looking for studios, greater or less percentage of new development units closing.
the numbers are quite interesting, but at this level can only tell you so much.
also, in the neighborhoods that have longer ownership trends, you have far more estate sales, and they really mix up things.
joedavis, where'd you get square feet for the co-ops?
Thx very much for the comments
Key one is that I should have explained 1 aspect better
There were a total of 17613 transactions. After I applied the criteria of keeping only those where the psf was less than $5000 and greater than $100, only 5943 transactions were left.
I had to do this to eliminate all the silly $0 and $10 transactions that are just transfers and all the divide by 0 errors for the cases for which no sq ft information is provided.
Unfortunately all the Co-ops disappear in the process of doing this.
My apologies for not making this clear to begin with.
So if you are focused on Co-ops this post has no useful information for you.
I suppose I could run an analysis on Co-ops only but then it would be hard to know what size of unit we are dealing with.
I am not a Streeteasy insider so if someone has the ability to download that sort of data, I'll be happy to run the analysis for you.
You'd still be left with no useful co-op info, as SE's square-feet are taken from the listing. As we know, those are useless except in very broad terms.
As it is, only 32 of the 79 recorded co-op sales for the last two weeks have associated square feet. (The other brokers wisely decline to include that number in their listings. That way they don't have to listen to whining from buyers who think they're purchasing carpet or fabric.)
Even among the 32, the number is meaningless. E.g., 4A and 7A at 330E79 both sold the other day. The floorplans and maintenance tell us they're the same size, yet one is listed at 1300 sf and the other at 1400 sf.
For coops, some people normalize by the number of rooms. So instead of price per square foot, it would be price per room. It is not a good a metric as sq ft, but it is better than nothing. The idea is if your data set is large enough, the variations in different room sizes will average out.
Room-count would get you closer, but even then you're dependent on whether/what the listing reports. Of the afore-mentioned 79 co-op sales, only 51 include an associated room-count.
No clue how MillerSamuel does its categorization. Maybe they interpolate from the big firms, who presumably supply consistent data.
I think Miller uses room count for the coop data.
Fantastic Joedavis!!! Thanks!!!!!!
Nice work my brother.