Fun with numbers
Redfin is all kinds of awesome. What I like the most about their service is that they not only show everything they're allowed to show by the idiotic, consumer-hostile MLS(r) rules they have to abide by, they actually let you download big chunks of their listing information as CSV spreadsheets straight from their search results page, including MLS(r) numbers, days on market, price, square footage, etc. This kind of transparency is unheard of in the secretive world of real estate, where an informed buyer is an annoyance, and it should be applauded.
Because it's Friday night and there's no better time to do basic statistics, I downloaded the 94501 search results into Excel and found the moderately interesting facts below.
Source data: 86 single-family home listings in the 94501 ZIP code, as listed for sale on Redfin on 8/14/2009. I excluded 2 houseboats from the result set, and had 84 listings to play with.
Method: I divided the data into quartiles according along price, price per square foot, and days on market, and computed averages and medians.
Because it's Friday night and there's no better time to do basic statistics, I downloaded the 94501 search results into Excel and found the moderately interesting facts below.Source data: 86 single-family home listings in the 94501 ZIP code, as listed for sale on Redfin on 8/14/2009. I excluded 2 houseboats from the result set, and had 84 listings to play with.
Method: I divided the data into quartiles according along price, price per square foot, and days on market, and computed averages and medians.
Price quartile data
Price / sqft quartile data
DOM quartile data
If you look at the bottom quartile (i.e. most recently listed properties), the median list price is the lowest of all quartiles, but the average is the highest of all. The high average is due to a few very expensive houses, and the low median is due to a mix of list prices that's skewed towards cheaper properties. This sharply contrasts with the more homogeneous distribution you can see in the other quartiles, and it is largely due to two attributes of the bottom quartile:
All in all, nothing terribly enlightening, but it'll be interesting to see what it's like after the selling season ends, and at the same time next year.
- There's a slight inverse relationship between price and price per square foot; higher-priced houses are cheaper per square foot than lower-priced houses. That's no surprise (although Manhattan goes the other way; the more expensive (and larger) the home, the more you pay per square foot)--there's a base line cost to any building and lot that adds more to the price per sqft when distributed over a smaller building v. a larger house.
- The difference between median and average DOM is considerably larger in the top quartile, pointing to a handful of very expensive outliers.
- The top half (by price) has much longer DOM than the bottom half. Again that's not terribly surprising; higher-priced houses take longer to sell because the buyer pool is smaller.
| Price quartile | Avg $ / sqft | Avg price | Median price | Avg DOM | Median DOM |
| top | $367 | $1,061,645 | $975,000 | 73 | 42 |
| high-mid | $379 | $734,314 | $736,944 | 80 | 61 |
| low-mid | $400 | $582,208 | $588,500 | 49 | 34 |
| bottom | $392 | $397,480 | $399,000 | 63 | 51 |
Price / sqft quartile data
- The median DOM tends to increase as price per sqft decreases. This is expected, since a lower price per sqft is often the mark of a more expensive property, and more expensive properties take longer to sell.
- All four quartiles show large discrepancies between average and median DOM. This reflects a few outliers in each quartile pushing the average up.
- Average DOM is roughly the same in the bottom three quartiles, and quite a bit higher in the top slice. This is due to five outliers pushing the average extra high with 138, 158, 162, 163, and 336 DOM in the top quartile.
| $ / sqft quartile | Avg $ / sqft | Avg $ | Median $ | Avg DOM | Median DOM |
| top | $477 | $757,664 | $663,750 | 74 | 37 |
| high-mid | $390 | $756,614 | $694,500 | 63 | 38 |
| low-mid | $317 | $791,345 | $789,500 | 66 | 46 |
| bottom | $244 | $466,570 | $399,000 | 62 | 51 |
DOM quartile data
If you look at the bottom quartile (i.e. most recently listed properties), the median list price is the lowest of all quartiles, but the average is the highest of all. The high average is due to a few very expensive houses, and the low median is due to a mix of list prices that's skewed towards cheaper properties. This sharply contrasts with the more homogeneous distribution you can see in the other quartiles, and it is largely due to two attributes of the bottom quartile:
- Properties listed at very low prices tend to sell very fast, so they wouldn't be represented in the higher-DOM quartiles, which makes the difference between the average and the median price in the bottom quartile artificially larger.
- At the high end of the price distribution, properties that have been on the market longer have had a chance (or two or three) to drop their prices, and so the higher average price in the bottom quartile is mostly an artifact of the listings being new (i.e. not reduced yet).
| DOM quartile | Avg $ / sqft | Avg $ | Median $ | Avg DOM | Median DOM |
| top | $370 | $697,220 | $688,500 | 165 | 136 |
| high-mid | $385 | $702,270 | $697,000 | 63 | 63 |
| low-mid | $381 | $656,180 | $634,250 | 32 | 33 |
| bottom | $398 | $704,620 | $589,000 | 9 | 5 |
All in all, nothing terribly enlightening, but it'll be interesting to see what it's like after the selling season ends, and at the same time next year.
Miscellaneous