When Does a Bay Area Teardown Pencil?
A Bay Area teardown pencils only when land value plus rebuild cost plus carry plus your target margin stays below the achievable finished value of the new home. Rearrange that and it means the existing structure has to be worth far less than the dirt it sits on: the improvement value (the house) must be a small fraction of the parcel value (the land). On the SF Peninsula that condition is routinely met in land-scarce, high-resale markets like Atherton and Palo Alto, where an older, undersized home barely moves the price and buyers pay for the lot. As of July 2026, Atherton’s median sale price runs about $9.98M and Palo Alto’s about $3.30M. It is the opposite in lower-resale submarkets and on lots that already carry a large or recently built house, where finished value cannot clear land plus construction plus margin and the numbers stay red.
The Pencil Test, In One Inequality
Strip away the spreadsheet and every teardown decision reduces to a single inequality. A scrape-and-rebuild only pencils when the total you sink into the deal stays under what the finished home will actually sell for:
If that holds, you have a deal. If it does not, no amount of finish upgrades or optimism fixes it, because the left side is mostly fixed by the market (land) and by physics and labor (rebuild cost and time). The only lever you fully control is the price you pay for the lot, which is exactly why professionals price teardowns with a residual land value model and solve for the land number instead of accepting the ask.
Land Value vs Improvement Value
Every parcel carries two values that the county tracks separately: the land value (the dirt) and the improvement value (the structure sitting on it). A teardown is, financially, a decision to write the improvement value down to zero and rebuild. So the real question is not “is this house old?” It is “how much value am I destroying when I knock it down?”
That reframes the inequality above. Rearranged, a teardown pencils when the existing structure is a small fraction of parcel value. The house you are erasing contributes almost nothing, nearly everything you paid was land, and you keep the land. Put the improvement share on one axis:
When that share is low, the structure is economically irrelevant and the deal turns on land and build cost alone. When it is high, you are demolishing real dollars, and the rebuild premium has to cover not just construction but the value you deleted. As a rough screen, an improvement share under about 20 to 25 percent is where Peninsula teardowns start to look interesting. Treat that band as a rule of thumb, not a bright line: it moves with resale strength and construction cost, both of which are illustrative ranges here, not fixed figures.
A Caveat On The Assessor’s Split
The county assessor publishes a land-and-improvement split on every parcel, and it is the fastest first screen you have. Use it with one large asterisk: under Proposition 13, California assesses property at its acquisition-year basis, not current market. A parcel held for thirty years can show a land value that is a fraction of what the dirt trades for today, which artificially inflates the improvement share and can make a genuine teardown look built-out. Read the assessor ratio as directional, then confirm the land number against recent teardown-comp and lot sales in the same submarket before you underwrite anything.
Where The Threshold Is Routinely Crossed
Two conditions have to line up for the inequality to close: land has to dominate parcel value, and finished-home resale has to clear a real premium over land plus build cost. On the Peninsula, Atherton and Palo Alto are the canonical markets where both hold.
- Atherton. With a median sale price around $9.98M and the priciest land on the Peninsula, an older, undersized house barely discounts the lot. The dirt is the asset, so the improvement share sits low almost by default, and even a livable home is frequently worth more scraped. See the full worked example in the Atherton teardown ROI guide.
- Palo Alto. A median near $3.30M plus strong new-construction resale means small mid-century homes on standard lots regularly read as land with a structure attached rather than the reverse. High land value, low improvement share, and a resale ceiling that supports the rebuild premium.
The common thread is not that the houses are worse. It is that the land is scarce and expensive enough that the structure stops mattering to price.
Where It Does Not Pencil
The same math turns negative in two distinct situations, and both are easy to misread as opportunities:
- Lower-resale submarkets. In Peninsula cities where finished new-construction resale sits closer to land plus construction cost, the achievable finished value cannot clear the full stack. The house might be cheap and dated, but the exit price is capped, so Land + Rebuild + Carry + Margin lands above finished value. A cheap house is necessary but never sufficient.
- Lots that already carry a large or recent house. A well-built, recently constructed, or simply large home has a high improvement share. Knocking it down destroys value you paid full freight for, so the rebuild has to overcome both construction cost and the deleted improvement. Even in an expensive zip code, a newer 4,000 sq ft house is usually a bad teardown.
Reading The Threshold Across Submarkets
The pattern below is illustrative, meant to show how land share and resale interact rather than to quote figures for any one parcel. Confirm every line against current comps before underwriting.
| Profile | Land share of value | Resale headroom | Teardown pencils? |
|---|---|---|---|
| Small old home, Atherton / Palo Alto | Very high | Large | Routinely |
| Small old home, mid-tier Peninsula | High | Moderate | Sometimes |
| Small old home, lower-resale submarket | Moderate | Thin | Rarely |
| Large or recent home, any market | Low (high improvement share) | Irrelevant | No |
Illustrative. “Resale headroom” is the gap between achievable finished value and total project cost. Land share and headroom both vary by parcel, finish level, and month; verify against recent comps.
Check Teardown Probability
Enter any Peninsula address to instantly calculate its hidden developer value.
From Framework To Deal Flow
The framework tells you what to look for: small, older homes with a low improvement share, sitting in submarkets where finished resale clears the rebuild premium. Screening for that by hand across two counties is the hard part. Our underbuilt-lot screener ranks active San Mateo and Santa Clara County listings by exactly this signal: how far below its allowable floor area a home sits, which is the built-environment mirror of a low improvement share. Once you have candidates, run the numbers with the residual land value ROI model to solve for the land price that keeps the inequality on the right side.
Frequently Asked Questions
When does a teardown make financial sense in the Bay Area?
A teardown makes sense only when the finished (after-build) home value exceeds land acquisition plus demolition, hard costs, soft costs, carrying costs, and your target profit. If Land + Rebuild + Carry + Margin lands above achievable finished value, the deal is upside down no matter how tired the existing house looks. In practice that inequality only closes when the lot is worth far more than the structure on it.
What is the land-value vs improvement-value framework?
Every parcel splits into land value (the dirt) and improvement value (the structure). A teardown converts improvement value to zero, so you are only paying for it if you demolish something worth keeping. The framework says a scrape-and-rebuild pencils when the improvement is a small share of total parcel value: the house you are erasing contributes little, and almost all of what you paid was land you keep. When the structure is a large share of value, you are destroying real dollars and the rebuild premium rarely covers it.
How do I tell if a house is worth more as land than as a home?
Compare the improvement share of value: improvement value divided by total value. A low ratio (as a rule of thumb, roughly under 20 to 25 percent) flags a home contributing little beyond its lot. Small, older houses on sizable lots in expensive zip codes are the classic profile. The county assessor publishes a land and improvement split you can use as a first screen, but Proposition 13 assesses property at its acquisition year, so a long-held parcel understates current market land value. Treat the assessor ratio as directional, then confirm against recent land and teardown-comp sales.
Why do Atherton and Palo Alto teardowns pencil when other cities do not?
Both markets pair extreme land scarcity with very high finished-home resale, so land is the dominant share of parcel value and new construction clears a large premium over land plus build cost. An 80-year-old 1,800 sq ft house barely discounts the lot, and a finished large home resells for a multiple of total project cost. In lower-resale Peninsula submarkets the finished-value ceiling sits too close to land plus construction plus margin, so the same rebuild math stays negative.
Does a low-value house always make a good teardown?
No. A cheap or run-down house is necessary but not sufficient. The lot still has to be worth enough, and the neighborhood’s finished-home resale still has to clear the full stack of land, demolition, hard and soft costs, carry, and margin. A tired house in a market where new construction resells near cost is a bad teardown; a solid house in Atherton can still be one, because the land is the asset either way.
Find lots where the math already works
Stop eyeballing tired houses. Filter Peninsula properties by their teardown and underbuilt-lot signal, then underwrite the winners.
Launch Alpha Screener →