URA Pipeline Data: How to Read Incoming Condo Supply

The headline number is almost never the right one

URA's pipeline dataset, published quarterly under the PMI (Property Market Information) feed, lists every private residential project URA tracks: pre-launch, under construction, just TOPped, and long since completed. Today's snapshot carries 1,364 rows totalling 377,486 units. If you read that as “incoming supply,” you have already made the most common error in Singapore property analysis.

Of those 1,364 rows, only 96 projects with 40,222 units have an expected TOP (Temporary Occupation Permit) of 2026 or later. The remaining 1,254 projects at 332,600 units have already completed and are part of the existing housing stock; they were “pipeline” in 2018 or 2022, but they are not pipeline now. A further 14 rows have no usable TOP date.

That single filter cuts the number by an order of magnitude. Anyone quoting “hundreds of thousands of incoming units” from URA pipeline is reading the file wrong.

A worked example: Treasure at Tampines

Treasure at Tampines is the largest single private development in Singapore at 2,203 units. It still appears in the URA pipeline file with district 18 and a 2023 expected TOP. It also already exists. The carpark is full, the residents have moved in, the resale market for it is active.

If you keep Treasure in your “District 18 incoming supply” tally, you double-count it: you count it once as existing stock (because it is) and once as pipeline (because URA still lists it). The fix is mechanical: drop any row whose expected TOP is in the past. A few other completed mega-projects make the same trap. Piccadilly Grand (2,026 units, D8, TOP 2025), Normanton Park (1,862 units, D5, TOP 2023), D'Leedon (1,715 units, D10, TOP 2014). None of these are incoming supply. Filter them out before you compute anything else.

What “future TOP” actually looks like

With the past-TOP rows stripped, the genuine national pipeline by year of expected TOP is roughly:

• 2026 TOP: 20 projects, ~4,531 units
• 2027 TOP: 23 projects, ~6,520 units
• 2028 TOP: 20 projects, ~9,820 units
• 2029 TOP: 17 projects, ~9,455 units
• 2030 TOP: 15 projects, ~9,300 units
• 2031 TOP: 1 project, ~596 units

The arc is gently rising into 2028 to 2030, then thins because URA's visibility into projects that have not yet broken ground is limited. The 2031 number will keep growing as new GLS sites get awarded and booked into the pipeline file.

The denominator buyers forget: district transaction volume

Even after filtering for future TOP, an absolute unit count tells you nothing without a denominator. 1,000 incoming units in a district that absorbs 2,000 sales a year is a ripple. The same 1,000 units in a district that absorbs 200 sales a year is a tsunami.

The cleanest denominator is annualised three-year transaction volume in the district, restricted to non-landed private (apartment, condominium, executive condo). It is a measure of how much demand the district has historically converted into completed sales. The ratio of incoming units to annualised volume is the number worth tracking. Call it the supply pressure ratio. A ratio of 1.0 means the pipeline equals one full year of district demand. A ratio of 3.0 means three years of demand sitting in the pipeline today.

Two reasons to prefer transaction volume over the alternatives. URA also publishes private residential stock by district (a count of standing units), which is a tempting denominator, but it conflates owner-occupied stock that rarely turns with active resale liquidity and overweights mature districts with large old condo estates. Rental contract counts are another option, but the URA rentals feed is reported by project, not always by district, and short-tenancy churn inflates raw counts. Three-year sale-side caveat volume is the cleanest read on how much the district digests per year.

One refinement worth making: if your target district is sensitive to a single launch (D6 with Canninghill Piers, D2 with One Marina Gardens), strip the dominant launch out and recompute. The supply pressure ratio is most useful when no single project contributes more than half the numerator, otherwise you are really measuring one launch's take-up, not the district's.

The most-pressured districts

Computed against URA caveat volume in the three years to April 2026, and pipeline filtered to expected TOP of 2026 or later, the supply pressure leaderboard looks like this:

• D24 (Lim Chu Kang, Tengah): 2,582 incoming units against 519 sales a year. Ratio 4.97x.
• D1 (Raffles Place, Marina): 1,986 units against 416 a year. Ratio 4.77x.
• D20 (Ang Mo Kio, Bishan): 1,674 units against 480 a year. Ratio 3.49x.
• D26 (Mandai, Upper Thomson): 3,290 units against 1,099 a year. Ratio 2.99x.
• D22 (Boon Lay, Jurong): 1,684 units, 643 a year. Ratio 2.62x.

Read these as years of demand sitting in the pipeline, assuming future demand looks like the recent past. D24 and D1 each have roughly five years of district absorption already lined up; D20, D26 and D22 have between two and three. None of these are catastrophic; they are signals that buyers should expect more competition for tenants and slower resale velocity than buyers in lower-pressure districts.

The District 6 trap (and why ratios need a sense check)

A naive supply pressure run will return District 6 (City Hall, Clarke Quay) with a ratio of 39.40x: 696 incoming units against 18 annual transactions. That is not a useful number. It exists because D6 is dominantly commercial, holds almost no resale condo stock, and the entire incoming figure is a single project (Canninghill Piers) that sits on a mixed-use podium. The denominator is too small to support the math.

The lesson is that the supply pressure ratio is informative when both numerator and denominator are large enough to be stable. As a working rule, ignore districts with fewer than 200 transactions over three years; the ratio there is dominated by single-project noise, not market dynamics.

Where the ratio understates pressure

Three caveats make the supply pressure ratio a floor, not a ceiling.

First, executive condominiums are excluded from the denominator on most clean reads of caveat data because they sit in their own 5-year MOP and 10-year privatisation track. If your target district (D27, parts of D18, D23) carries an EC-heavy launch slate, the denominator understates relevant demand. Add EC stock to both sides for an apples-to-apples count.

Second, the URA pipeline file is conservative on TOP year. Projects often slip by 6 to 12 months versus their initial target; some pull forward. The annual bands published above should be read as plus or minus one year, not exact.

Third, “incoming supply” in the URA file does not equal “hits the resale market on TOP day.” Most launch units are pre-sold under the Sales of Building Under Construction (SBUC) licence years before TOP. They show up as URA caveats at sale, not at TOP. The pipeline matters most for rental supply (because owner-occupiers move in, but investor units hit the lease market on TOP) and for resale velocity in years 5 to 7 (when speculative buyers exit).

A fourth caveat is GLS visibility. URA's pipeline file only captures projects with a granted planning permission and an assigned project name. Sites that have been awarded under the Government Land Sales programme but not yet through planning are absent. A buyer looking five to seven years out should pair this guide's pipeline read with the GLS slate on the launches page. The two together approximate true forward supply; either alone understates it.

A concrete reading: D15 in 2026 to 2029

District 15 is the country's most-transacted RCR district. The URA pipeline lists twelve future-TOP projects totalling 3,862 units against 1,959 sales a year. The ratio is 1.97x, just below two years of district demand. Most of the volume sits in three already-marketed mega-projects: Grand Dunman (1,008 units, TOP 2028), Emerald of Katong (846 units, TOP 2029) and The Continuum (816 units, TOP 2027).

None of these is an unsold inventory shock. They are pre-sold under SBUC and will trickle into the resale market individually as owners decide to exit. The boutique residue (Ardor, Claydence, Parq Bella, Koon Seng House, Straits at Joo Chiat) adds prestige PSF prints without meaningful unit count. A D15 buyer in 2026 should plan around the fact that 2027 to 2029 will see steady tenant supply entering the market as new units TOP, which favours owner-occupiers over yield investors.

A checklist for reading the next URA pipeline release

URA refreshes PMI Pipeline once a quarter, usually within four weeks of quarter end. When the next release lands, run the same five steps:

1. Drop any row with expected TOP in the past.
2. Drop any row with no parseable TOP year (text like “na”, “tba”, blank).
3. Group remaining rows by district and sum total units.
4. Divide by annualised three-year non-landed caveat volume in that district.
5. Discard ratios for any district under 200 transactions in the three-year window; treat the rest as years of district demand sitting in the pipeline.

If your target district's ratio drifts above 3.0x quarter on quarter, expect rental yields to compress and resale liquidity to slow during the TOP year band. If it sits at or below 1.5x, the district is absorbing supply faster than it accumulates, and resale velocity should hold up.

One discipline buyers rarely apply: track the quarter-on-quarter change in the ratio, separately from its absolute level. A district moving from 1.2x to 2.0x over two quarters is a louder signal than a district holding steady at 2.5x. Quarter-on-quarter deltas reveal whether the slate is filling up faster than absorption can grow, which is the early sign of pricing pressure 12 to 18 months out. URA caveats are filed within two to three weeks of OTP, so the data itself is near real-time; the lag is in price discovery, where a district median needs enough fresh prints to move. Read the ratio in deltas and it leads price discovery by a quarter or two; read in levels and it sits coincident with the current price level.

Tools and related pages

• Launches dashboard with the live pipeline filtered to future TOP, plus per-project detail and the curated GLS pipeline.
• District analytics for transaction volume, median PSF and yield by district, refreshed daily.
• Mortgage and TDSR calculator for sizing the cash stack on a target launch unit, including BSD, ABSD and joint-buyer IWAA.
• Watchlist to track price moves on the launches you care about as their TOP year approaches.