01 · The numberThe figure every showflat quotes
Where the widely-repeated new-launch premium actually comes from.
Walk into any showflat and some version of this number will reach you: new launches command a premium of forty percent or so over resale. It is quoted as though it were a property of newness itself — a fee you pay for fresh paint and a warranty, recoverable later when your unit becomes someone else’s resale.
In our data the figure is even larger. Across 40,015 new-sale and 61,135 resale transactions, the medians are S$2,440 and S$1,596 per square foot — a gap of 52.9%.
The number is real. What it means is almost entirely misunderstood, and the misunderstanding costs buyers money.
New and resale stock rarely sit in the same building, the same district, or the same decade — which is exactly why a single headline gap misleads.
New launches trade 52.9% above resale on the raw medians. That gap describes two different populations, not two prices for the same thing.
02 · The claimMost of the premium is composition
Stated narrowly enough to be tested against the record.
We are not claiming new launches are overpriced, or that developers inflate values. We are claiming something narrower: the headline gap is a comparison of unlike things, and once you hold the building constant it does not survive.
If newness alone carried a fixed premium, that premium would be roughly constant across the island. It is not. It swings by more than ten percentage points between segments — which is the first clue that we are measuring something other than newness.
03 · MethodHow we measured this
The dataset, the filters, and the comparison we deliberately refused to make.
- The dataset. URA caveats, 2021-07 to 2026-07 — every private non-landed transaction lodged with the state.
- The universe. Strata, non-landed, Executive Condominiums excluded. New Sale and Resale compared; Sub Sale held separately because it is a different act.
- The threshold. Any group we report has at least 40 transactions on each side of its comparison.
- Medians throughout. A handful of penthouses would move an average; they barely move a median.
- What we refused to do. We did not adjust for floor, view, layout or finish. Each adjustment is a judgement call, and the honest way to handle unlike comparisons is to say so — not to bury the difference under a model.
Public caveats, one consistent universe, thresholds stated. The comparison is transparent enough for you to disagree with it.
04 · The evidenceThe gap, measured three ways
By segment, then across time — the same figure from two angles.
Start with where the gap sits. Each row below is one market segment: the grey dot is the resale median, the navy dot the new-launch median.
The pattern is the opposite of what a “premium for newness” would predict. The gap is widest in the suburbs (54.4% in OCR) and narrowest in prime (41.0% in CCR). Newness does not become more valuable the further you get from town. Something else is happening.
Now watch it move through time.
A gap that swings by year is a gap driven by what launched that year — which projects, in which districts, at which price points — rather than by any stable premium attached to being new.
The gap is widest where land is cheapest and moves year to year with the launch pipeline. Both are signatures of composition, not of a premium.
05 · The mechanismWhat the gap is actually measuring
Three composition effects that masquerade as a premium.
Three things separate the two populations, and none of them is newness.
- Age and specification. The resale pool contains stock from the 1980s onward. The new-sale pool is, by definition, current. Comparing them is comparing a 2026 car to the average of every car on the road.
- Location mix. New launches appear where land is released — recently, that has meant suburban plots near new rail. Resale volume is spread across the whole island, including a great deal of older, cheaper stock.
- Unit mix. New launches skew to smaller, more efficient layouts, and price per square foot rises as unit size falls. That alone lifts the new-sale median without any unit being better value.
There is a fourth force worth watching, because it tells you when the market is running hot: the share of transactions that are sub-sales — units resold before completion, the closest thing our data has to a speculation gauge.
Sub-sales currently run at 1.51% of transactions, against a window peak of 12.69%. This is the number to watch if you are worried about buying into froth — it moves before prices do.
Looking at a specific launch? WhatsApp us the project — we’ll send the same comparison against its own resale market →06 · The cohortHold the building constant
The only comparison that isolates newness: a project against itself.
Every objection above disappears if you compare a project to itself — same address, same land, same specification — measuring its new-sale prices against its own later resales.
Across 6 projects with enough of both, the median same-project gap is -8.4%. The comparison does not merely shrink from 52.9% — it reverses.
Be careful with that number in the other direction, though. Within a single project, most new sales happened at launch and most resales happened later, so this comparison carries a timing effect of its own: it partly measures the market moving, not just the launch premium. The honest reading is not “new launches are 8% cheap”. It is that once you hold the building constant, the forty-percent premium cannot be found at all.
Most recent launches have arrived in suburban districts near new rail — which is a large part of why the headline gap looks the way it does.
Same building, both sides: the gap inverts. Whatever the headline measures, it is not the price of being new.
07 · The counter-caseWhere new genuinely wins
The strongest arguments for buying new, stated fairly.
None of this makes new launches a bad purchase, and it would be dishonest to leave that impression. There are real advantages the psf comparison cannot see.
- The payment schedule. Progressive payments spread cash outflow across construction years instead of demanding a full mortgage from month one. For a household mid-upgrade, that is worth real money.
- A full lease and a warranty. A fresh 99 years and a defects period against a building that may already be thirty years into its lease and its plumbing.
- Efficiency. Newer layouts waste less space, so a smaller unit can live larger — part of why psf is higher and total price sometimes isn’t.
- The selection window. At launch you choose stack, floor and facing. In resale you take what is listed.
What the data does say is that you should not expect the gap itself to be recovered. If your case for buying rests on the assumption that today’s new-launch price becomes tomorrow’s resale floor, the same-project evidence does not support you. Buy new for the schedule, the lease and the selection — not for an automatic re-rating.
New launches win on cash flow, lease and choice. They do not win by default on price per square foot recovered later.
08 · Your moveWhat this means for you
Four situations, four numeric rules.
If you are choosing between a launch and a resale unit
Compare the launch against resales in its own district and size band, never against the island-wide resale median. The island-wide comparison will overstate the premium by tens of percentage points, because it is comparing your shortlisted suburb to the whole country’s ageing stock.
Benchmark a launch only against its own district. Using the island median inflates the apparent gap by roughly 52.9 points.
If you are buying for the payment schedule
Price the benefit honestly: progressive payments are a cash-flow advantage, not a discount. Model the years you pay for a home you cannot occupy or rent, alongside your current housing cost, before deciding what the schedule is worth to you.
Count the carry years. If the schedule only works when you ignore paying twice for housing during construction, it does not work.
If you are hoping to exit at TOP
The four-year Seller’s Stamp Duty window and the same-project evidence point the same way. A quick re-rating at completion is a hope, not a pattern our data supports.
Model the exit at a four-year-plus hold. If the numbers need a fast sale, the numbers do not work.
If you are watching for froth
Sub-sale share is the most honest early indicator available in public data, because it counts people leaving before completion.
Track sub-sale share. Today it is 1.51%; the window peak was 12.69%.
Which launch are you weighing up?
We’ll run this same comparison on the project you’re actually considering.
- The launch psf against its own district’s resale market
- Comparable launches nearby and how they’ve since resold
- The break-even hold, with the SSD window applied
One WhatsApp message back — usually same day.
We’ll WhatsApp your launch comparison shortly — usually within the day.
09 · The recordThe gap by segment
The figures behind every chart above.
Median psf for each segment, resale against new sale, with the gap between them. Screenshot it — the figures are dated and we would rather you checked them.
10 · MethodSources, limits and corrections
Data: URA caveat records, 2021-07 to 2026-07; private non-landed only, Executive Condominiums excluded.
Thresholds: minimum 40 transactions per side of any reported comparison.
Known limits: the same-project comparison carries a timing effect (new sales cluster at launch, resales come later), which is why we present it as evidence against the headline rather than as a premium of its own. Caveats lag completion, so the latest month is provisional.
Reproducibility: every figure here is produced by flagships.py against the source dataset. Spot an error and we will correct it in place, dated.
Not investment advice. Analysis of public transaction records, published to be argued with.
A gap is not a price.
What matters is the number on the unit you are actually considering, against the market it will one day be sold into. The Scenario Planner runs that comparison on your figures in about two minutes.
Map my move →