Every new launch price starts from one number: the developer's breakeven. That is everything the project must recover, divided by the floor space it is allowed to build. Add a profit margin and you have an estimate. We add 15% to 30%. That comes from 42 launched projects here, where the price developers got was a median 23% above breakeven. This page shows the costs behind the breakeven number, where the margin range comes from, and where the model has been wrong.
- Land is the winning tender price for a state site, or the purchase price for a collective sale. It includes betterment charges and lease upgrading premiums where those apply. This figure is public, and we never estimate it.
- Construction is the main building contract. It rose sharply from 2020 and has not come back down.
- Financing, fees and marketing covers several things. The interest on the land while it sits. The architects and engineers. Legal costs and taxes. Then the showflat, the agents and the advertising. Together they are roughly a fifth of the bill.
- Margin is what the developer adds on top. We publish 15% to 30%, measured from what launches have actually achieved.
1. What is a breakeven rate, and why start there?
Breakeven is the developer's total cost divided by the floor space it is allowed to build. We quote it in dollars per square foot per plot ratio, the same unit as the land rate, so the two sit side by side. Here is a worked example. Say a site costs $1,027.8 million all in and allows 35,369 square metres of floor space. That is 380,706 square feet. Divide one by the other and the breakeven is $2,700 psf ppr. Sell below that and the project loses money. We reproduced that sum ourselves; it checks out.
We hold a cost breakdown for 73 projects in our land-sale register. Almost all come from our own cost model, set out below. Every breakdown adds up to its own stated breakeven. Where a project was put together from more than one tender, we show the most recent breakdown.
How our cost model builds each breakdown
- LandOur register’s price, plus betterment and lease premiums
- Construction$300 (2018), $330 (2019–21), $390 (2022–24), $546 (2025 on) per sq ft; ECs about 75%
- Land financing8.75% of land before June 2022; 15.76% after
- Professional, legal and taxesIRAS stamp duty on the land, plus 3.54% of land and 6.82% of construction
- Marketing and other6% on top of everything above
- BreakevenSum of the above ÷ GFA in sq ft, in $ psf ppr
We tested these rates against 130 published cost estimates from 2018 to 2026. For 95% of sites, our breakeven came within 1.2%. Sites marked † keep EdgeProp Singapore’s estimate.
2. Where does the 15% to 30% margin come from?
From the record, not from a rule of thumb. We multiply the breakeven rate by 1.15 to 1.30. That range is what 42 launched projects actually achieved. For each one, we compared its average selling price against its estimated breakeven. The median outcome was +22.8%, and 22 of the 42 landed inside the band.
The multiple is doing two jobs at once, and it is worth being clear about that. Part of it is the developer's profit. The rest is a difference in measurement. Breakeven is stated per square foot of gross floor area, which includes corridors, lift lobbies and plant rooms. Homes are sold per square foot of strata area, which is only the space inside the units. Strata area is the smaller number. So a project can sell 20% above its breakeven rate without the developer earning anything close to a 20% profit. Read these numbers as pure profit and you will overstate how much room a developer has to discount.
3. When does this model get it wrong?
Half our launches land outside the band we publish. That is the honest reading of the spread. Two patterns explain most of the misses, and both are predictable enough to flag on the project page rather than pretend away.
| Segment | Projects | Median over breakeven | What it means |
|---|---|---|---|
| OCR | 19 | +19.3% | OCR buyers are the most price-sensitive and the pipeline is deepest, so developers have least room. The lower half of our band fits here. |
| RCR | 15 | +20.6% | RCR sits close to the middle of the band, which is where the model is most reliable. |
| CCR | 8 | +33.1% | CCR land is often long-held and the buyer is less price-led, so achieved prices run well above the band. Treat a CCR estimate as a floor. |
The second pattern is the age of the land. A developer prices against today's market, not the costs it locked in years ago. So a site bought cheaply and launched late clears far above its breakeven. The extreme cases in our dataset are all of this type: long-held freehold and prime sites where the cost breakdown is simply out of date. Where we know a project falls into that group, we say so on its page and treat the published range as a floor rather than a middle estimate.
There is a mirror image at the other end. A few projects launched at or below breakeven. That is what a developer does when the deadline for clearing unsold stock is close and selling matters more than the margin. So a low estimate is not a bargain signal on its own.
A worked example: Thomson Reserve
UOL and CapitaLand Development bought the old Thomson View site together, in the 2024 cycle. Here is the estimate from start to finish.
Thomson Reserve — step by step
- Land, including premiums$1,247.6m
- Construction$412.9m
- Land financing$196.6m
- Professional, legal and taxes$209.5m
- Marketing and other$124.0m
- Total development cost$2,190.6m
- Gross floor area98,391 sqm = 1,059,073 sq ft
- Breakeven$2,068 psf ppr
- Margin applied (1.15–1.30)$2,378 – $2,688
- Published, rounded to the nearest $10$2,380 – $2,690 psf (est.)
Note what the land rate alone would have told you. At $1,178 psf ppr, the old method of multiplying land by 1.6 to 1.8 gives $1,880 to $2,120. That is below the developer's breakeven, so it is a price at which the project could not be built at all. That failure is why we retired the multiplier method.
What do we currently estimate?
Below is every site in the pipeline that has a cost breakdown, ranked by breakeven. These are the same numbers that appear on each project page and in the land cost tracker. One model, one set of outputs, no adjusting by hand.
| Project / site | Land (psf ppr) | Breakeven (psf ppr) | Land as % of cost | Published estimate |
|---|---|---|---|---|
| Peck Hay Road GLSCCR · D11 | $1,865 | $3,194 | 58% | $3,670–$4,150 (est.) |
| Bukit Timah Road GLSCCR · D11 | $1,820 | $3,132 | 58% | $3,600–$4,070 (est.) |
| River Valley Green (Parcel C) GLSCCR · D09 | $1,730 | $3,008 | 58% | $3,460–$3,910 (est.) |
| Dunearn Road (Plot 2) GLSCCR · D11 | $1,625 | $2,862 | 57% | $3,290–$3,720 (est.) |
| Dover Drive GLSRCR · D05 | $1,556 | $2,767 | 56% | $3,180–$3,600 (est.) |
| New Upper Changi Road GLSOCR · D16 | $1,537 | $2,741 | 56% | $3,150–$3,560 (est.) |
| Berlayar Drive GLS†RCR · D04 | — | $2,700 | 56% | $3,100–$3,510 (est.) |
| Holland Plain (Parcel B) GLSCCR · D10 | $1,491 | $2,677 | 56% | $3,080–$3,480 (est.) |
| Tanjong Rhu Road GLSRCR · D15 | $1,455 | $2,628 | 55% | $3,020–$3,420 (est.) |
| Amberwood at HollandCCR · D10 | $1,432 | $2,596 | 55% | $2,990–$3,370 (est.) |
| Kallang Close GLSRCR · D14 | $1,415 | $2,572 | 55% | $2,960–$3,340 (est.) |
| Dorset GardensRCR · D08 | $1,338 | $2,466 | 54% | $2,840–$3,210 (est.) |
| One Chuan GroveOCR · D19 | $1,355 | $2,456 | 54% | $2,820–$3,190 (est.) |
| Bedok Rise GLSOCR · D16 | $1,330 | $2,455 | 54% | $2,820–$3,190 (est.) |
| Telok Blangah Road GLSRCR · D04 | $1,326 | $2,450 | 54% | $2,820–$3,180 (est.) |
| Bayshore Drive GLSOCR · D16 | $1,323 | $2,445 | 54% | $2,810–$3,180 (est.) |
| 8 Thomson LaneCCR · D11 | $1,297 | $2,410 | 54% | $2,770–$3,130 (est.) |
| Lentor Central (Plot 4) GLSOCR · D26 | $1,278 | $2,383 | 54% | $2,740–$3,100 (est.) |
| Tan Boon Liat BuildingRCR · D03 | $1,245 | $2,338 | 53% | $2,690–$3,040 (est.) |
| Hougang Central GLSOCR · D19 | $1,179 | $2,247 | 52% | $2,580–$2,920 (est.) |
| Lucerne GrandOCR · D22 | $1,132 | $2,182 | 52% | $2,510–$2,840 (est.) |
| Upper Thomson (Parcel A) GLSOCR · D26 | $1,062 | $2,085 | 51% | $2,400–$2,710 (est.) |
| Thomson ReserveRCR · D20 | $1,178 | $2,068 | 57% | $2,380–$2,690 (est.) |
| Chencharu Central ResidencesOCR · D27 | $980 | $1,972 | 50% | $2,270–$2,560 (est.) |
| Dairy Farm Walk (Plot 3) GLSOCR · D23 | $962 | $1,947 | 49% | $2,240–$2,530 (est.) |
| Loyang ValleyOCR · D17 | $936 | $1,911 | 49% | $2,200–$2,480 (est.) |
| Clovelle of WoodlandsOCR · D25 | $794 | $1,612 | 49% | $1,850–$2,100 (est.) |
| Wynwood GrandOCR · D25 | $782 | $1,595 | 49% | $1,830–$2,070 (est.) |
| Solano GrandOCR · D23 | $771 | $1,580 | 49% | $1,820–$2,050 (est.) |
| Miltonia Close (EC)OCR · D27 | $732 | $1,526 | 48% | $1,750–$1,980 (est.) |
| Sembawang Road (EC)OCR · D27 | $692 | $1,471 | 47% | $1,690–$1,910 (est.) |
Sites still open for tender have no cost breakdown, so they get no estimate. They appear in the pipeline tracker with land data only. We would rather show less than imply a precision we do not have.
How do we check ourselves against real launches?
The back-test below is the evidence for the whole model. For every launched project where we hold both a cost breakdown and enough sales to work out an average price, it compares what the developer needed against what it got. It is sorted from the thinnest outcome to the fattest.
| Project | Segment | Breakeven | Achieved avg psf | Over breakeven | Sold |
|---|---|---|---|---|---|
| Canberra Crescent Residences | OCR | $1,537 | $1,526 | -0.7% | 91% |
| The Lakegarden Residences† | OCR | $2,073 | $2,168 | +4.6% | 99% |
| Narra Residences | OCR | $2,027 | $2,158 | +6.5% | 38% |
| Grand Dunman | RCR | $2,277 | $2,527 | +11.0% | 91% |
| Pinetree Hill | OCR | $2,234 | $2,519 | +12.8% | 100% |
| Hillock Green | OCR | $1,948 | $2,197 | +12.8% | 98% |
| Lentoria | OCR | $1,978 | $2,245 | +13.5% | 94% |
| Vela Bay | OCR | $2,535 | $2,929 | +15.5% | 74% |
| Chuan Park† | OCR | $2,229 | $2,581 | +15.8% | 97% |
| The Orie | RCR | $2,320 | $2,695 | +16.2% | 95% |
| Tembusu Grand | RCR | $2,115 | $2,459 | +16.3% | 99% |
| The Arcady At Boon Keng | RCR | $2,227 | $2,598 | +16.7% | 62% |
| Nava Grove | OCR | $2,131 | $2,497 | +17.2% | 100% |
| Ardor Residence | RCR | $2,094 | $2,494 | +19.1% | 91% |
| Tengah Garden Residences | OCR | $1,752 | $2,090 | +19.3% | 100% |
| Meyer Blue | RCR | $2,710 | $3,235 | +19.4% | 79% |
| 8@Bt | OCR | $2,268 | $2,714 | +19.7% | 72% |
| The Continuum† | RCR | $2,296 | $2,764 | +20.4% | 98% |
| Arina East Residences | RCR | $2,354 | $2,840 | +20.6% | 79% |
| Dunearn House | CCR | $2,566 | $3,125 | +21.8% | 55% |
| Hudson Place Residences | RCR | $2,050 | $2,512 | +22.5% | 66% |
| Sora | OCR | $1,833 | $2,255 | +23.0% | 51% |
| Norwood Grand | OCR | $1,690 | $2,080 | +23.1% | 92% |
| Lentor Gardens Residences | OCR | $1,889 | $2,357 | +24.8% | 54% |
| One Marina Gardens | CCR | $2,378 | $2,973 | +25.0% | 71% |
| Upperhouse At Orchard Boulevard | CCR | $2,675 | $3,398 | +27.0% | 82% |
| River Modern | CCR | $2,579 | $3,277 | +27.1% | 94% |
| Bagnall Haus | OCR | $1,945 | $2,505 | +28.8% | 94% |
| Faber Residence | RCR | $1,684 | $2,179 | +29.4% | 97% |
| Springleaf Residence | OCR | $1,691 | $2,207 | +30.5% | 98% |
| The Hillshore | RCR | $1,853 | $2,434 | +31.4% | 15% |
| Coastal Cabana | OCR | $1,331 | $1,790 | +34.5% | 82% |
| Penrith | RCR | $2,035 | $2,801 | +37.6% | 98% |
| Pinery Residences | OCR | $1,828 | $2,534 | +38.6% | 94% |
| River Green | CCR | $2,271 | $3,158 | +39.1% | 94% |
| Promenade Peak | RCR | $2,242 | $3,122 | +39.2% | 72% |
| The Sen | OCR | $1,603 | $2,357 | +47.0% | 41% |
| W Residences Marina View - Singapore | CCR | $2,147 | $3,205 | +49.3% | 2% |
| Zyon Grand | RCR | $2,102 | $3,198 | +52.1% | 90% |
| The Collective At One Sophia† | CCR | $1,768 | $2,786 | +57.6% | 27% |
| Artisan 8† | RCR | $1,359 | $2,361 | +73.7% | 62% |
| The Giverny Residences† | CCR | $1,909 | $3,444 | +80.4% | 50% |
The median is +22.8% and the mean is +27.1%. The mean sits above the median because a few long-held prime sites pull the top end up. That is exactly why we publish the median band and flag those cases separately, rather than widening the band to swallow them. When a launch lands outside what we published, the record stays up on that project's page.
A land multiple is still a useful cross-check, but never a substitute. The reason is that the multiple itself changes with the land rate. It falls from roughly 2.50 times on sites below $900 psf ppr to about 1.91 times above $1,600. Our Chuan Grove analysis works that correction through from start to finish. It also shows how to adjust for the floor-area rule change of June 2023, which matters whenever you compare a newer project with an older one.
What are these numbers, and what are they not?
- They are ranges for a whole project: typical low-floor to typical high-floor pricing at launch. Single units can price outside any published range. That includes penthouses, ground-floor units with private gardens, and premium stacks.
- They rest on our own cost model. Nobody publishes the building contract, the loan terms or the marketing budget. So the breakdown is a well-informed rebuild of the sums, not the developer's own books.
- They are estimates of a price nobody has announced yet. Developers set final prices days before launch, against conditions no one can see in advance.
- They change. When a cost breakdown or the margin band is revised, the estimate moves, and we log the change below.
The short version — read this first
Six things to know before you use any estimate on this site.
How our estimates work
- We start at breakeven, not landTotal development cost divided by the floor space allowed. Land is only a median 54% of that cost
- We add 15% to 30%Measured, not assumed. Across 42 launched projects the median achieved price was 23% above breakeven
- The margin is not all profitBreakeven is quoted per square foot of gross floor area. Homes sell per square foot of the smaller strata area
- Half our launches land outside the band22 of 42 landed inside it. We publish that rather than widen the band to hide it
- Where we run lowCCR (Core Central Region) sites and land held a long time. Those cleared a median 33% above breakeven, against under 20% in the OCR
- No stack, no estimateSites still open for tender show land data only. We would rather show less than imply precision we do not have
So what should you do with this?
- Read our estimate as a range, not a price. The developer sets the real price days before launch.
- If a project page flags a long-held or prime site, treat our range as a floor rather than the middle.
- Before you commit to anything, check the developer's official price list. Our numbers are research, not an offer.
Four terms used on this page
- psf pprDollars per square foot per plot ratio. A cost divided by the floor space a developer is allowed to build there
- Gross floor area (GFA)All the floor space in a project, including corridors, lift lobbies and plant rooms
- Strata areaThe floor space inside the homes themselves, which is what you buy. Always smaller than gross floor area
- BreakevenThe price per square foot a project must sell at just to cover its costs
Frequently asked questions
How does PropertyInsider.sg estimate new launch prices?
We start from the estimated breakeven rate, which is the developer's total cost divided by the floor space it may build, and we add a margin of 15% to 30%. Across 42 launched projects, the average price achieved was a median 23% above breakeven. That is where the band comes from.
What is a breakeven rate, and why start there instead of the land price?
Breakeven is everything the developer must get back. That means the land, the building work, the interest on the land, fees, legal costs, taxes and marketing. Divide the lot by the floor space it may build and you have the breakeven rate. Land is usually the biggest single item, but rarely more than half the total, and its share varies a lot from site to site. Starting at breakeven counts the whole cost base, instead of assuming everything scales with the land price.
What goes into the cost stack?
Five parts. Land, including betterment charges and lease upgrading premiums where they apply. The building work. The interest on the land. Fees, legal costs and taxes. And marketing. We publish the dollar figure for each one on every project page where we have it.
Why 15% to 30%?
Because that is what the record shows. Across 42 launched projects with a known breakeven, the median achieved price was 23% above it, and 22 of the 42 landed inside the band. The multiple also absorbs a second effect. Breakeven is quoted per square foot of gross floor area, while homes sell per square foot of strata area, and strata area is smaller.
When does the model get it wrong?
It runs low on CCR sites and on land held a long time. CCR projects cleared a median 33% above breakeven, against under 20% in the OCR. Long-held freehold sites can run higher still, because the developer prices against today's market rather than costs locked in years ago. We flag both cases on the project page.
Do you still use the old land-multiplier method?
No. Multiplying the land rate by 1.6 to 1.8 was retired in July 2026, and the model that replaced it was itself superseded on 12 August 2026. A land multiple is now only a sanity check, because the multiple changes with the land rate itself.
How often do you review this?
Every six months. We also review it straight after any launch that lands well outside our estimate. Every change is logged at the bottom of this page.
Data sources
Land prices come from URA and HDB land tender results and reported collective-sale deals. Cost breakdowns and breakeven rates come from our own cost model, using IRAS stamp duty rules and BCA’s Tender Price Index. Sites marked † use EdgeProp Singapore’s estimate. Achieved launch prices and take-up come from our own new launch dataset, URA caveat records and developer announcements. Where a figure is an estimate rather than a published fact, we mark it (est.).
Disclaimer. Estimated launch prices are research projections, for general information and education only. They are not financial, investment or property advice. Developers alone set actual launch prices, and those may differ a lot from any estimate here. Cost breakdowns are our own estimates and may be revised. Do not make a purchase, sale or financing decision based on an estimate. Check the official price list and seek professional advice. PropertyInsider.sg is an independent research publication. See our editorial policy.
What has changed on this page
- Wording pass. Region names now use URA's CCR, RCR and OCR labels in place of prime centre, city fringe and suburbs. No figures changed.
- Cost model v1. Breakevens now come from our own cost model; † marks EdgeProp estimates. Estimates moved $20 psf or less.
- Model v3.0. Estimates are now anchored on the estimated breakeven rate — the full development cost stack divided by GFA — with a margin band of 15–30% applied on top. This replaces the v2 land-plus-build-cost-plus-margin construction published on 7 Jul 2026. The band is calibrated against 42 launched projects rather than assumed. Cost stacks for 71 sites were merged into the land-sale register, and every pipeline estimate was recomputed; the largest movements were Bedok Rise, Lentor Central (Plot 4) and Chencharu Close. Tan Boon Liat Building and Loyang Valley were added to the register.
- Bayshore Drive re-estimated on the confirmed $1,323 psf ppr top bid (tender closed 15 July).
- v2 published: land plus build cost plus margin, checked against nearby launches, with range width set by building height. It replaced the original method of multiplying the land rate by 1.6–1.8×, which construction inflation had made unreliable.
