HomeResearch › Floor premium
Updated 11 Jul 2026 · mature resale (≥5 yrs) vs 2026 new sales · URA caveats

How much is a higher floor worth — and should you pay the developer’s price?

TL;DR
0.29%
market pays / floor (resale ≥5yr)
0.49%
2026 new launches charge / floor
3%
/ floor in a top view stack

The question — and why the price list can mislead

At a new launch the developer hands you a price list where each floor costs a little more. The trap is using that same price list to judge whether height is worth it — because the developer sets the per-floor step. To know what height is really worth, you have to ask what the resale market pays for it years later. So we measured both, separately.

We compared units in the same stack (identical facing, near-identical size), holding market timing constant. For the market premium we used only resale caveats in projects at least five years old — 70,478 of them across 14,973 stacks — so the figure reflects settled market prices, not launch pricing. Separately we measured the 2026 new-sale ladder from 5,294 caveats in this year’s launches.

What you’re charged vs what the market pays

Cumulative $psf uplift by floor, same stack

Uplift versus a low floor. Green = what the resale market pays; amber = what 2026 launches charge; red = a stack where height unlocks a view.
Resale market (~0.29%/flr)2026 new-sale ladder (~0.49%/flr)View stack (~2%/flr)
+0%+23%+46%+69%+93%floor 1floor 10floor 20floor 30floor 40
The height tax: take a 20-floor jump. The developer’s 2026 ladder adds about +10% to $psf; the resale market, on an ordinary stack, later rewards only about +6%. That gap — roughly 4 percentage points of $psf — is what you over-pay for height on a plain (no-view) stack.

What 2026 launches are charging

Across this year’s new sales the per-floor step averages ~0.49%, and it’s steepest in the mass-market OCR — where developers stack the most units and price height hardest:

Segment2026 new-sale $psf/floorCaveats
CCR0.49%807
RCR0.42%739
OCR0.53%3,532

Height isn’t always paid for

Key finding: in established resale the premium is wildly uneven. The middle stack rewards height at 0.34%/floor, but a quarter of stacks sit near 0.04%/floor or below — flat, or even cheaper higher up — while the top tenth reach 1.1%/floor. Height pays only when it delivers light, quiet or a view.

What an ordinary stack looks like

Most stacks aren’t view stacks. These are large, well-traded mass-market projects whose per-floor premium sits right on the market average of ~0.29% — a higher floor is worth a little more, but only a little. On stacks like these, the developer’s ~0.49%/floor launch ladder is hardest to justify.

Project · stackDistrictSizePer floorFloors seen
Sol AcresD23 · OCR1,327 u0.29%1–24
WaterbayD19 · OCR383 u0.28%1–17
Eight RiversuitesD12 · RCR843 u0.30%2–27
Guilin ViewD23 · OCR655 u0.28%1–28
CitylightsD8 · RCR600 u0.31%8–40
Regent HeightsD23 · OCR645 u0.31%1–26

At 0.29%/floor, climbing 15 floors adds only ~4% to $psf — a fraction of what a view stack commands. Same city, very different reward for height.

The view threshold — where paying up is worth it

This is the East-Coast case: below a certain height you face the next block; above it the sea opens, and the price steps up and stays there. It is a threshold, not a slope, and it is where the resale market genuinely rewards height. Tellingly, the steepest-premium stacks in the last five years are almost all on the East Coast waterfront (D15). Real stacks where each floor added 1.5–2.6% to $psf:

Project · stackDistrictPer floorFloors seen
Costa RhuD152.6%1–14
Pebble BayD151.7%1–15
The Ritz-Carlton Residences Singapore CairnhillD91.7%12–33
SilverseaD151.6%1–17
Amber SkyeD151.5%1–20
Eastern LagoonD151.5%2–17

A 3%/floor stack means a unit fifteen floors up trades roughly 47% higher $psf than the same line low down — against ~4% on the ordinary baseline. That is the view premium, concentrated at the floor where the view clears.

So — should you pay up?

On a plain stack, resist the full developer ladder

The market rewards ~0.29%/floor; 2026 launches charge ~0.49%. If the stack has no view, every floor you climb costs roughly 1.7× what you’ll get back — so take the lower floor and pocket the difference, or negotiate the step down.

Pay up only where a view is unlocked — at the threshold floor

Find the floor where the view actually clears the obstruction. Below it you’re buying height with no payoff; at and above it the market pays 4–6%/floor. Don’t pay view money for a sub-threshold floor, and don’t assume it keeps climbing once you’re above it.

Confirm the view is protected

A view premium only holds if it can’t be built out. Check the URA Master Plan and nearby GLS/plot ratios on that facing before paying up — see supply and growth areas.

Sanity-check against real resale

Value the exact unit with our Valuation tool (it floor-adjusts comparables) and read the stack’s own resale spread on the project page, then compare it to the developer’s per-floor step before you commit.

How is this worked out? — same-stack fixed effects, resale-only, 2026 new sales
Market premium
Resale caveats only, in projects completed by 2021 (≥5 years old), so the figure reflects settled resale prices, not the developer’s launch ladder — 70,478 caveats, 14,973 stacks.
New-sale ladder
New Sale caveats dated 2026 (5,294) — what developers charge per floor in this year’s launches.
Model
A within-stack fixed-effects regression of log($psf) on floor + a time control — the floor effect is measured inside each stack (same facing/size) and isn’t contaminated by which units happened to sell when.
Source
URA Data Service caveats via HomeVestor; $psf outliers & en-bloc rows excluded. View stacks are illustrative examples, not unit valuations.
Note: new launches are often taller with more genuine view floors, which explains part of their steeper ladder — but on a like-for-like ordinary stack the gap is a real premium you carry. Verify any specific unit before transacting.
For educational purposes only — not financial advice.
Keep going: all research › · value a specific unit › · work out your offer ›