Teardown Watch for Belmont
Each week, the new Belmont house listings that look like what developers bought before they tore down and rebuilt. It runs inside Peninsula Plays, the free weekly SF Peninsula brief, so one subscription gets you both.
Built from public listing data, county parcel records and our zoning envelopes. Every flag shows the numbers that triggered it, so you can check it yourself.
Get Teardown Watch for Belmont
Free, weekly, on Substack. The new Belmont listings that match the developer-buy signature, plus the week's Peninsula permits and sales.
What gets flagged
- The price rule. The house uses 20% or less of its lot (FAR 0.20 or below), and at the city's usual sale-to-list ratio it is priced at 1.4x or more of the city's median sold price per square foot over the past year. A buyer paying that much for a small house on a big lot would be paying for the land.
- The envelope rule. Zoning allows 2.5 to 4 times the house that stands there now.
- Both skip houses built in 2000 or later and lots that already have a new-house permit on file.
Leads, not predictions
We tested the rules on 14,367 past house sales, 503 of which were followed by a new house within three years. 13% of the sales the price rule flagged went on to a rebuild, 4.5 times the 2.9% rate for all house sales. The envelope rule hit 9 to 11% across all tested cities, and less where rebuilds are rare. So most flags will not become rebuilds, and an owner building their own house counts the same as a developer. One more difference: the backtest used actual sale prices, while on a new listing we estimate the sale price from the list price.
Belmont was not in the backtest. Counting new-house permits per sale puts its rough rebuild rate at 0.9%, an upper bound. The same flag means more in a city where rebuilds are common.
The full method and charts are in What Silicon Valley developers actually pay for.
Every flag is dated
Each flag is logged on the day it goes out. A flagged lot that gets a new-house permit within three years counts as a hit, the same test as the backtest, and one that does not counts as a miss.