In the 12 launches since January 2025 that had at least 400 homes and sold at least 85% of them within two months, 96% of two-bedroom sales were filed in the launch month itself. For four-bedroom homes it was 77%.
- 96% of two-bedroom sales that were filed in the launch month itself 2025-01 to 2026-04 · Truestorey, from URA Data Service — PMI_Resi_Transaction and URA’s development pipeline
- 77% of four-bedroom sales that were filed in the launch month itself 2025-01 to 2026-04 · Truestorey, from URA Data Service — PMI_Resi_Transaction and URA’s development pipeline
- 94% of sales on the middle third of floors filed in the launch month 2025-01 to 2026-04 · Truestorey, from URA Data Service — PMI_Resi_Transaction and URA’s development pipeline
- 88% of sales in the dearest quarter of psf, within each bedroom count, filed in the launch month 2025-01 to 2026-04 · Truestorey, from URA Data Service — PMI_Resi_Transaction and URA’s development pipeline
What was counted, and what was not
The launches are those since January 2025 with at least 400 units that sold at least 85% within two months: The Orie, Parktown Residence, Aurelle of Tampines, Lentor Central Residences, River Green, Springleaf Residence, Penrith, Skye at Holland, River Modern, Pinery Residences, Rivelle Tampines and Tengah Garden Residences. For each kind of home, the figures show what share of its sales in the first two months was filed in the launch month rather than the month after.
URA records the month of a contract, never the day, so booking day cannot be separated from the rest of that month. These are also launches that sold fast. They say nothing about a launch that did not, and nothing about how any future launch will go.
Size sorted the launch month more than floor did
By bedroom count, 96% of two-bedroom sales came in the launch month, against 93% for one-bedroom homes, 91% for three-bedroom homes and 77% for four-bedroom homes. The four-bedroom group is small next to the two-bedroom one, 770 sales against 3,402, but the largest homes were plainly the slowest, with the most of their sales spilling into the second month.
By floor, the differences are smaller. 94% of sales on the middle third of floors were filed in the launch month, against 92% on the lowest third and 91% on the highest.
Price mattered within each size
Within each bedroom count, sales were split into four quarters by price per square foot. In the cheapest quarter, 94% of sales were filed in the launch month. In the dearest quarter it was 88%. Homes priced higher for their size were more likely to sell in the second month.
A wider set of launches says the same about size, and something new about floors
The fastest launches are a select group. A wider set is the 28 launches since June 2023 with at least 12 floors that later sold out. In them, the share of homes sold within the first two months was 90% for one-bedroom homes, 88% for two-bedroom homes, 80% for three-bedroom homes and only 54% for four-bedroom homes and larger.
Floors tell a different story here. 84% of homes on the middle third of floors sold within two months, against 81% on the lowest third and 75% on the highest. In this wider set the top third of floors was the slowest to sell, which the fastest launches do not show.
What a floor costs on the way up
Across the 45 launches since June 2023 that carried enough sales to measure, the median rise in price per square foot was 0.46% for each floor up a tower, with size held constant. A developer prices a launch as one schedule, with a step for each floor and a step for each stack. URA’s records carry a band of floors and a size but no unit number, so the floor step can be measured and the stack step cannot.
What to check before a launch
- The size mix on your list. In these launches two-bedroom homes went first and four-bedroom homes lagged, so a list of one size meets more or less competition depending on which it is.
- How far down your list you can go. The Thomson Reserve launch tool replays booking day against a ballot number and a shortlist, using the same records.
- The month, not the day. Public data cannot say what sold in the first hours, so no figure here should be read as one.
