Stats & Numbers: 2-4 Unit April ’07 versus April ’08

From Garrett, as always, with love:

District Map (PDF)

2-4 Units
District 1 Apr-07 Apr-08
Number of Sales 7 8
Median Selling Price 1,450,000 1,380,000
Average DOM 95 90
District 2 Apr-07 Apr-08
Number of Sales 8 7
Median Selling Price 1,037,500 1,200,000
Average DOM 47 61
District 3 Apr-07 Apr-08
Number of Sales 1 0
Median Selling Price 800,000
Average DOM 26
District 4 Apr-07 Apr-08
Number of Sales 0 1
Median Selling Price 1,072,500
Average DOM 57
District 5 Apr-07 Apr-08
Number of Sales 11 9
Median Selling Price 1,675,000 1,500,000
Average DOM 55 35
District 6 Apr-07 Apr-08
Number of Sales 6 2
Median Selling Price 1,612,500 1,250,000
Average DOM 55 49
District 7 Apr-07 Apr-08
Number of Sales 2 2
Median Selling Price 1,685,000 1,875,000
Average DOM 185 15
District 8 Apr-07 Apr-08
Number of Sales 1 1
Median Selling Price 600,000 1,520,000
Average DOM 32 10
District 9 Apr-07 Apr-08
Number of Sales 7 3
Median Selling Price 1,160,000 978,000
Average DOM 126 70
District 10 Apr-07 Apr-08
Number of Sales 1 2
Median Selling Price 770,000 742,500
Average DOM 83 54

Data Provided by SFAR

9 thoughts on “Stats & Numbers: 2-4 Unit April ’07 versus April ’08

  1. Because of the wacko local ordinances on 2-4 units, I am not sure how comparable the data is for such buildings. A 2 unit both vacant, a 2 unit with both units occupied with no protected tenants, a 2 unit both occupied with one protected tenant, a 2 unit with one unit vacant and one occupied by a non protected tenant, a 2 unit with one vacant and one occupied by a protected tenant, a 2 unit with a designated OMI unit, blah blah blah. Each with its own potentials and pitfalls and ‘value profiles’ that can vary sometimes by 20% or more depending on its profile. It is a maze worse than the Minotaur’s labyrinth.

    To do apples-to-apples comps, one will have to break down the data into each of the ‘profiles’ — like for 2 units both vacant in 07 April, in district 2 a total of 7 units sold for average $1.1m DOM 47, and in 08 a total of 4 units sold for average 1.2m DOM 61, etc… As an owner of a 2 unit I’d love to see such breakdowns.

    SFRs and condos are considerably easier to compare therefore they are closer to apples-to-apples comparisons (at worst it is only granny smith vs fuji)

  2. Fuji….hands down. Granny Smith no. And certainly the worst of all, Red Delicious…chances are less than 25% that you’ll actually get a red delicious, juicy apple. They’re always mushy!

  3. if you are going that way… I dont care. What matters is apples from Washington State vs apples from NewZealand. (and spring is a bad season)

    It’s not about taste, it’s about carbon footprint ;-)

    -> locavore/locovore (locovore seems to be the oldest word, but it’s locavore which was included in the Oxford dictionary)

    (Ps: fuji is an easy pick here too)

    (PS2: anon8mizer: GREAT POINT. and if the market slows down, those differences will matter more and show more in direct $$$)

  4. To do apples-to-apples comps, one will have to break down the data into each of the ‘profiles’

    Utter nonsense. Apart from being an exercise in complete wankerdom, the data sample sizes would be so small as to make any additional analysis impossible. Remember the three rules of RE: “Location, location, location.”

    From a statistical standpoint, it doesn’t matter if the downstairs apartment is occupied by old man with arthritis or a young couple with a baby. Or if both units have “up to date” kitchens and baths or just one or if it located on a south-facing street, etc, etc. All of these things work themselves into the final price but you cannot aggregate uniqueness. Location is the only thing that matters. Anything else looks like anxious justification.

    And none of this would have even been contemplated when SF, or D7, or unoccupied two-unit flats on a “nice” street in D7 were “on fire.”

  5. I don’t disagree that the data size would be so small that it would become statistically insignificant. However, I disagree that this is anxious justification. When you are a buyer, you always contemplate it. When I bought my unit, one unit was empty and one unit was occupied; and because the occupied unit had a long term tenant nobody wanted to touch it and it sat on the market for 3 months — in a market that was on fire, where 2 duplexes (similar to mine) across the street (both vacant & not remodeled) sold after 2 weekend showings that closed about 1 month prior.

    If one didn’t know how to value the unit in such a profile and got caught up in the frenzy, one would have offered to purchase the unit at a price close to the vacants across the street, according to your “location” theory. But I got the units at 22% less than the ones across the street because of the ‘complications’ of the unit. I would say 5% of that is due to the comparable conditions of the units (which puts it into the granny smith vs fuji category), but 17% is due to the fact there is a long term tenant that appeared to have no interest in moving. This 17% makes it less of an apple-to-apples comparison to the 2 units across the street.

  6. Statistics is not about apple-to-apple comparison.

    If you want apple-to-apple, the best data is CS, at least they try to be. Statistics is about an overview of the market. Nobody claimed it is apple-to-apple. However, it is useful data, and if the standard deviation doesn’t change too much, it is almost as good as apple-to-apple.

  7. FYI – to illustrate the high % of variations: a VACANT duplex at 1249 9th Ave, with 1820 sqft, asking 999,000, sold for $1.2mm – 20% over, according to SFnewsletter. It’s $655 a foot.

    It’s a dump! And it’s right at the mouth of the GG park on 9th next to the stinky and dirty gym.

    I am not sure the new owners want to keep it as a duplex. They might just poke a hole in the wall and connect the two units to make it into a single family home.

Leave a Reply