The research desk: land, supply and prices

What this section covers

Most Singapore property coverage starts at the showflat. Our research starts about two years earlier — at the land tender. The price a developer pays for a Government Land Sales (GLS) site is the single largest input into the eventual launch price, and it is public information the day the tender closes. Read the land market correctly and you can see launch pricing coming long before a price list exists.

The Research section is built around that idea. It holds our GLS pipeline tracker — every confirmed, tendered and awarded site we follow, with land rates and indicative launch price ranges — and our analysis pieces, which connect land tender outcomes to what buyers will actually pay. Project-by-project reviews live under New Launches; historical pricing and resale exit data live under Price Trends. This section is the layer that ties them together.

Perspectives from practitioners

Alongside the desk's own research, we publish clearly-labelled opinion columns from licensed practitioners. These pieces reflect the author's views, carry their byline and disclosures, and sit apart from our independent estimates and models — the separation is spelled out in our editorial policy.

How the GLS programme works — in five steps

If you are new to this, the mechanics matter because each stage produces a data point we track:

1. Confirmed list announcement. Twice a year (typically June and December), the government announces the next half-year's GLS programme. Confirmed List sites will be tendered on a schedule regardless of demand; Reserve List sites are only triggered if a developer commits to a minimum bid. The 1H 2026 Confirmed List carries nine sites and roughly 4,575 homes; the 2H 2026 programme, announced on 3 June 2026, adds around 4,745 more.
2. Tender launch. URA (or HDB, for some sites) releases the tender with the site's specifications: land area, gross plot ratio, maximum gross floor area, and any commercial or integration requirements. From these we can estimate the unit yield before a single bid is lodged.
3. Tender close. Bids are opened publicly. The top bid, expressed in dollars per square foot per plot ratio (psf ppr), becomes the market's clearest read on developer conviction — and the anchor for our launch price estimates. The number of bids matters almost as much as the price: eight bids at Bayshore Road in March 2026 told a very different story from four at Peck Hay Road in June.
4. Award. URA typically awards to the highest bidder within days to weeks. The site moves to "Awarded" in our tracker and receives an indicative launch price range.
5. Preview and launch. Usually 9–15 months after award, the project opens for preview. At that point our estimate meets reality — and we log the outcome, because every launch recalibrates the land-to-launch multiplier we use for the next estimate.

How to read our indicative launch prices

Every estimate on this site comes from one published model — pricing model v2: the awarded land rate plus current construction and development cost (higher for integrated developments, reflecting post-pandemic construction inflation), multiplied by a 10–20% developer margin, then calibrated against comparable launches and incumbent resale pricing. The published range width is set by plot ratio, because taller towers carry a wider floor-to-floor price spread — Vela Bay (plot ratio 4.2) launched with a visibly wider unit range than the mid-rise Pinery Residences (plot ratio 2.6).

Estimates marked ◆ are fully calibrated; unmarked figures in the tracker are raw cost-model bands, so no false precision is implied. The complete formula, cost assumptions, worked example and public back-testing live on our pricing methodology page — one canonical reference that every estimate on the site links back to. When a launch lands outside our range, the outcome is logged and the model assumptions are reviewed; that public feedback loop is the point.

Why we publish this before developers announce

Developers finalise pricing days before launch; buyers commit to viewing plans weeks or months earlier. That timing gap is where most bad decisions happen — people anchor on a neighbour's 2019 resale price, or on an agent's optimism, rather than on what the land actually cost. Publishing transparent estimates early, with the working shown, narrows that gap. When we are wrong, the update log on each page says so, and the multiplier gets recalibrated. That feedback loop — estimate, launch, compare, adjust — is the core of this research desk.

Everything here is built from public primary sources: the URA GLS site listings and tender results, URA REALIS transaction records, developer announcements, and our own compiled dataset of over 32,000 profitable resale transactions and 199 tracked resale projects, which we use for neighbourhood benchmarks. We cite sources in every analysis piece and log every material page change.