What Is the Teardown ROI for a Home in Atherton?
One fully underwritten Atherton teardown rebuild, built from recorded Atherton sales, pencils at about 19% of total project cost at mid-range build costs and the resale our comps support. Move either input across its measured range and the same lot spans −16% to +47%, losing money in two of nine cases and clearing 15% in five, so 19% is a base case rather than a midpoint.
That is one deal, not a market rate, and we do not publish an Atherton-wide ROI because we cannot measure one: as of 19 August 2026 AddressIntel can price just 9 active Atherton listings, of which 2 clear a 15% return, and a distribution from single digits would be noise. What we can measure is the wider SF Peninsula: of 498 active listings the model can price, 68% are underwater at their asking price and the median is -10%, while the 79 that clear 15% run a median of 33% (middle half 24% to 51%). Those Peninsula figures use the screener’s flat cost constants, so the like-for-like version of the example is not its 19% but the 39% the same lot returns on that basis, slightly above the 33% clearing median.
Atherton land dominates the proforma by size rather than by unit price: measured across teardown-candidate sales, Atherton ranks 17 of 22 Silicon Valley cities on land price per square foot at about $273, well under Palo Alto at $526. What it has is big parcels: the median teardown lot is about a third of an acre, more than twice a Palo Alto parcel, and the acre-plus class a full-size rebuild needs runs about seven times one. Cheap dirt, bought in bulk, is what sets the size of the cheque.
The figures below are built from recorded sales and a worked proforma rather than an average. As of August 2026, Atherton’s median asking price across active listings is about $13.75M, homes sell in an average of ~36 days at about 103.7% of list, and developers pulled 4 new-construction permits there. In AddressIntel’s recorded Atherton sales, teardown-candidate homes (small, older houses under ~2,500 sq ft) have a median sale price of $4.50M, with the middle half between $3.68M and $7.80M across 33 recorded sales from July 2024 to July 2026, while finished large homes resell at a median near $2,100/sq ft, and at $2,116/sq ft across the comps our model selects for a full-size rebuild. AddressIntel computes ROI with a Residual Land Value model, ROI = projected profit ÷ total project cost, where profit = net sale proceeds − (land + demolition + hard costs + soft costs + holding costs) and net sale proceeds are the finished home value after a 5% cost of sale. The worked example below is that one coherent deal: a 1-acre teardown lot, built out to what Atherton zoning allows on it, priced against comps matched to that build.
What the Recorded Data Shows
The land and resale figures come from recorded Atherton sales. Teardown-candidate homes, small, older houses under ~2,500 sq ft, the ones developers buy for the lot, have a median sale price of $4.50M, with the middle half between $3.68M and $7.80M (33 recorded sales, July 2024 to July 2026). The full spread runs from $2.5M to $37M, so the median is the only figure here worth underwriting to. Finished large homes resell at a median near $2,100 per square foot, and the top decile reaches about $2,510/sq ft. Individual sales print higher, but above $3,000/sq ft our model treats the comp set as broken rather than hot and declines to price it, so figures beyond that are not something the screener will underwrite to.
Atherton is a small market by permit count, which is worth stating plainly before any ranking: developers pulled 4 new-construction permits there in the tracked window, which runs cumulatively from 18 May 2026 rather than monthly. That is well behind the volume leaders (San Mateo is at 15), so read Atherton’s standing as intensity per parcel, not as regional permit volume. See the full ranking in our Bay Area teardown & new-construction activity guide.
How the ROI Is Calculated: Residual Land Value
In a high-barrier market like Atherton, developers don’t price a teardown off the existing structure, the house is worth nothing to them. They use the Residual Land Value (RLV) model, working backward from the finished home value to solve for the most they can pay for the lot and still hit a target return.
For the full step-by-step framework, ARV comps, the five cost buckets, and how holding cost erodes returns, see how to estimate developer ROI on a luxury teardown rebuild. The Atherton-specific math is below.
A Worked Example: 1-Acre Atherton Teardown, Built to Zoning
The one thing a worked example has to get right is that the lot, the build and the comps all describe the same deal. Atherton’s median teardown-candidate lot is about a third of an acre, and R-1A zoning allows roughly 3,100 sq ft on a lot that size, so pairing that lot’s median price with a large rebuild prices two different properties at once. This example therefore picks one lot class and stays in it: a 1-acre R-1A teardown lot, at the measured median price for that class, built out to the floor area Atherton actually permits on an acre, and resold against comps matched to that size. Every line traces to a measurement or to a sourced range.
| Line item | Basis | Amount |
|---|---|---|
| Finished home value (ARV) | 7,800 sq ft × $2,116/sq ft (median of the 12 comps our own selection returns) | $16,504,800 |
| Cost of sale | 5% of ARV (commission + closing) | −$825,240 |
| Net sale proceeds | What you actually bank | $15,679,560 |
| Land acquisition (teardown lot) | Median of 11 recorded teardown sales on lots of 0.9 acre or more | −$8,000,000 |
| Hard costs | 7,800 sq ft × $550/sq ft (middle of the $400–$700 Peninsula builder band) | −$4,290,000 |
| Soft costs | ~12% of hard costs | −$514,800 |
| Demolition | Clear existing structure | −$50,000 |
| Holding costs | 18 months × $15,000/mo (base case; see the fast and slow cases below) | −$270,000 |
| Total project cost | Land + build + carry | $13,124,800 |
| Projected profit | Net sale proceeds − total project cost | $2,554,760 |
| Developer ROI | Profit ÷ total project cost | ~19% |
Where each line comes from
Land. The median of the 11 recorded Atherton teardown-candidate sales that sat on lots of 0.9 acre or more, the class that can actually carry this build. The smallest lot in that set is 0.93 acre and the median is a flat 1.00. The all-lots median of $4.50M buys a much smaller entitlement.
Build size, and the one input carrying real provenance risk. Atherton R-1A floor area on the class’s median lot of 43,573 sq ft, a shade over an acre (Muni Code 17.32.040), the same build-to-zoning rule the per-property proforma uses. We model R-1A, which covers about 85% of Atherton’s residential parcels; R-1B allows slightly more below an acre. The Town’s own code site blocked a direct read, so the coefficients come from two secondary mirrors that agree verbatim. We hold the rule at medium confidence and would re-verify it before relying on it for an actual entitlement.
One limit matters at this size: floor area above the first storey is separately capped at 7.5% of lot size, which our rule does not model. On a 1-acre lot that is 3,267 sq ft upstairs, so a 7,800 sq ft house has to put at least 4,533 sq ft on the ground floor and then answer to lot coverage. Treat 7,800 sq ft as the zoning envelope, not as a design that has been shown to fit.
Resale. The median of the 12 comps the product’s own selection returns for this build: Atherton sales of the same property type, inside a ±25% size band, capped at the 12 most recent. We run the product’s routine rather than an approximation of it, down to rounding the build to the same 100 sq ft step the calculator uses (the raw R-1A envelope is 7,843 sq ft) and applying the $3,000/sq ft sanity check to the median rather than to each comp. The two tail columns in the grid below read off a wider population, the 25th and 75th percentiles of all 53 in-band sales, because a percentile over 12 rows is too jumpy to publish.
One difference from a live property page. The product only builds to zoning when the allowable area is no more than four times the existing house, and otherwise falls back to 1.5× the existing house. On a lot this size that gate trips for any existing house under about 1,950 sq ft, which is a large share of the teardown candidates this page describes. So a property page for a small old house on an acre will model a smaller rebuild than this table does.
Carry, which is an assumption and cannot be anything else. The permits we ingest carry an issued date but no application date, so no one can derive a true application-to-approval duration from the public record. Run it as a range: at 12 months the deal returns 20.3% and at 24 months 18.7%, so a full extra year of review costs 1.6 points, about 0.14 points a month. That makes the timeline the smallest lever on this page, well behind a $500K land overpay at about 4 points and an order of magnitude behind build cost and resale.
The cost lines are ranges, not Atherton measurements. Hard costs, soft costs, demolition and holding come from our teardown ROI guide, and the table runs mid-range values from each.
The same deal across the ranges
A single number implies a precision this model does not have. Two inputs carry almost all of the uncertainty, so here is the same lot across both, in both directions: build cost across the sourced $400–$700/sq ft builder band, and resale at the 25th percentile, the ARV our comp selection returns, and the 75th percentile of the wider in-band pool.
| Hard cost | Resale $1,630/sq ft (25th pct of the 53-sale in-band pool) | Resale $2,116/sq ft (ARV, median of the 12 selected comps) | Resale $2,343/sq ft (75th pct of the 53-sale in-band pool) |
|---|---|---|---|
| $400/sq ft (band floor) | 2% | 33% | 47% |
| $550/sq ft (band midpoint, table above) | −8% | 19% | 32% |
| $700/sq ft (band ceiling) | −16% | 9% | 20% |
Read the grid before the headline. Five of the nine cases clear the 15% bar, two land between zero and 15%, and two lose money, on a deal where the land price never moved. The spread from corner to corner is 63 points. Until this grid carried a downside column it looked far safer than the deal is: at the 25th percentile of resale, the lot only just breaks even at the cheapest build in the band and loses money at every other. Resale is the wider of the two levers, worth 40 points at mid-range build cost against build cost’s 24, and it is the one you do not control. Build cost is the one you can negotiate. If you cannot buy the build near the middle of the band, the deal wants a lower land price, which is what underwriting the lot as a residual is for.
Reconciling this against the number on a property page: the per-property proforma runs the same lot at 39%, because it is deliberately uniform across every listing rather than tuned per deal. It holds hard costs at $400/sq ft, soft costs flat at $150,000 instead of a percentage, demolition at $25,000, and no carry at all, since it assumes a cash buyer. Those four choices are worth about $1.83M of cost, which is the entire gap. Neither figure is wrong; the screener is a comparable screen across thousands of lots, and this table is one deal underwritten with mid-range assumptions.
The free weekly Peninsula Plays brief
The week's best teardown and spec-build plays across the Peninsula, with the developer ROI worked out from permits, parcels, and sold comps. One email a week, no spam.
What Moves Atherton ROI the Most
- Resale: The widest input, and the one you do not control. Across the 25th to 75th percentile of the size-matched comp pool, and nothing else, the example above swings 40 points and crosses from underwater to comfortably clear.
- Build cost: The widest assumption you can negotiate. Moving across the sourced $400–$700/sq ft builder band, and nothing else, swings the example by 24 points.
- Land price: The lever you control at the point of offer. On the example above, every $500K of overpay costs about 4 points, and unlike build cost it is settled with a single signature.
- Lot class: Match the lot to the build before anything else. Atherton’s median teardown lot is about a third of an acre and zones to roughly 3,100 sq ft, so a large-rebuild proforma run against the all-lots median land price is pricing two different properties.
- Buildable square footage: Capped by zoning. Confirm what you can actually build under local Floor Area Ratio (FAR) and lot-coverage limits before trusting any ARV.
- Entitlement timeline: Atherton’s discretionary design review is slow, and every extra month is ~$15,000 in carry. On the example above that is only about 0.14 points of ROI per month, so a year of delay costs 1.6 points against the build-cost band’s 24. Slow review is a cash-flow and risk problem here rather than a return problem. See the Peninsula entitlement process and the demolition permit process.
- Resale comps: Atherton resale runs at the top end of the Peninsula per square foot, which is what makes even land-heavy deals pencil. We have not measured it against every Peninsula city, so treat that as a description of Atherton, not a ranking.
Automating the Math with AddressIntel
Running the Residual Land Value model by hand for every Atherton property that hits the market is impossible, by the time you pull comps and estimate carry, the lot has sold off-market. That is why we built the Teardown Predictor: it scores parcels across 15 Peninsula and South Bay cities, Atherton among them, giving each a 0–100 teardown score, and carries a build-to-zoning proforma covering purchase, demolition, construction at a flat $400/sq ft and resale against size-matched comps, so the RLV arithmetic above is computed for you rather than by hand. Two limits are worth stating plainly: those 15 cities are not the whole of San Mateo and Santa Clara counties, and the proforma needs an asking price, so an off-market parcel carries a teardown score but no ROI until it lists. You can filter Atherton inventory directly in the screener, or browse the live Atherton permit & teardown dashboard. And for winning the sourcing race itself, our guide to finding off-market teardown opportunities on the SF Peninsula covers the owner-outreach, probate, and listing-lifecycle channels that surface Atherton lots before they trade.
Frequently Asked Questions
What is the teardown ROI for a home in Atherton?
AddressIntel does not publish an Atherton-wide teardown ROI, because Atherton inventory is too thin to measure one: as of 19 August 2026 the model can price 9 active Atherton listings, of which 2 clear a 15% return, against a median asking price of about $13.75M across active listings. What we publish instead is one fully worked deal and the range around it. Developers underwrite to a target margin rather than to a measured average, and in Atherton the land line dominates that arithmetic because the parcels are big, not because the dirt is dear by the foot. Using real Atherton comps: the all-lots median teardown-candidate sale is $4.50M, but the median teardown lot is only about a third of an acre and Atherton zoning caps what can be built on it, so the representative deal below prices a 1-acre lot at its own measured median of $8.0M and builds the 7,800 sq ft that zoning allows. At mid-range build costs and $2,116/sq ft size-matched resale, net of a 5% cost of sale, that pencils at roughly 19%, and it spans −16% to +47% across the sourced build-cost band and the 25th-to-75th percentile resale range, losing money in two of nine cases. Set that against the wider Peninsula, where the same model puts 68% of 498 active listings underwater at their asking price, with a median of -10%. The 79 that do clear 15% run a median of 33%. Those Peninsula figures come off the screener’s flat cost constants, not this table’s, so the honest comparison re-prices the same lot the same way: on the screener’s basis it returns 39%, a little above the 33% median of the listings that clear. Read the two numbers for what they are. The 19% is the stricter figure, carrying real build cost and eighteen months of carry, and it is what you would underwrite to. What makes Atherton hard is not the margin, it is the size of the cheque: $8.0M of land and $13.1M of total cost before anything is sold.
Why does land dominate the proforma on an Atherton teardown?
Lot size, not land price. Measured across teardown-candidate sales, Atherton land runs about $273 per square foot, 17th of 22 Silicon Valley cities we can price and well under Palo Alto at $526. The dirt is not expensive by the foot. The parcels are big instead: the median teardown lot is about a third of an acre, more than twice a Palo Alto or Menlo Park parcel, and the acre-plus lots a full-size rebuild needs run about seven times one, so the absolute acquisition price is a far larger share of total project cost. Atherton’s resale is strong, with the top decile of large-home sales reaching about $2,510/sq ft against a median near $2,100/sq ft, and that is what carries a deal this land-heavy. What the acre does is raise the absolute stake, not sink the return, which is why disciplined buyers still solve the land price as a residual instead of paying the median and hoping.
How much does it cost to build a spec home in Atherton?
Hard costs (materials + labor) at Peninsula builder pricing run $400–$700 per square foot, so the 7,800 sq ft rebuild in the example below is roughly $3.1M–$5.5M in hard costs alone. The $800–$1,200+ figure often quoted is a retail general-contractor price for a one-off custom build, not what a spec builder pays. The worked example below runs the middle of the builder band, $550/sq ft, and shows the $400 and $700 ends alongside it, because we have no Atherton-specific hard-cost measurement and will not assert one. Add soft costs (architecture, engineering, permits, financing) of 10–15% of hard costs, demolition of $30,000–$60,000, and holding costs near $15,000/month, the same rate our teardown ROI guide uses. Atherton’s slow discretionary review shows up in the number of months you carry, not in a higher monthly rate. In recorded Atherton sales, large finished homes resell at a median near $2,100/sq ft; individual sales print higher, but above $3,000/sq ft our own model treats a comp set as broken rather than hot and declines to price it.
How does AddressIntel calculate developer ROI for Atherton properties?
AddressIntel ingests Atherton’s issued building and demolition permits from the town’s own eTRAKiT portal, refreshed on a twice-daily pipeline, with the back catalogue filled in from a bulk public-records delivery, so the record runs from early 2023 to the present. Those permits and the recorded sales are the inputs. It then runs the Residual Land Value model per lot: it pulls local comps for the finished home value, matched on city, property type and size to within 25% of the modelled rebuild. Those comps carry no age filter, so a recent sale of an older house of the right size counts alongside a new build, which pulls the resale estimate below what a true new-construction-only comp set would give. It then applies the same flat cost constants it uses everywhere: $400/sq ft hard costs, $150,000 soft costs, $25,000 demolition, a 5% cost of sale and no carry, because it assumes a cash buyer. Those are deliberately uniform so lots stay comparable, which also means the per-property number is a screen rather than an Atherton-tuned underwriting. Each lot gets a 0–100 teardown score and a build-to-zoning proforma, so you can filter Atherton properties by return instead of modeling each deal by hand.
Find your next Atherton deal
Stop guessing on Atherton lot values. Use our screener to instantly screen Peninsula properties by teardown score, then read the per-lot proforma on the ones worth underwriting.
Launch the Screener →