By Gary Lim — ERA Senior Division Director, CEA Licence R009877B · 17+ years, 500+ deals · BuyCondo Team · Updated 13 November 2026

At a glance: the new launch take-up rate in one table

Short version, before the detail. Every new launch take-up rate you read this year came from one of two different sums, and nobody tells you which. Here is what I found when I lined up every 2026 Singapore launch that has a published launch-weekend number.

The weekend that triggered this article: Lucerne Grand, 61.4%

Saturday 3 October 2026, sales commenced at Lucerne Grand. By the end of the weekend, 350 units were sold. EdgeProp put it precisely: 350, or 61.4%, of the 570 units. Average price, $2,480 psf.

The project itself: 570 residences across five 17-storey towers, directly linked to Lakeside MRT on the East-West Line. Pricing started from S$1.498m for a 2-bedroom, S$1.988m for a 3-bedroom and S$2.788m for a premium 4-bedroom. CDL won the site in June 2025 at $1,132 psf per plot ratio. It is the first new private launch in the Lakeside precinct in a decade, which matters more than people think — there was no recent comparable project in that pocket for buyers to anchor against.

Buyer profile was unusually domestic: 93% Singaporean, 7% Permanent Residents, with the PR buyers coming from Malaysia, China, India, Bangladesh, Myanmar and South Korea. That is a project sold to people who intend to live in it.

The industry read it the way you would expect. Sherman Kwek, CDL’s group chief executive, said the developer was “encouraged by the healthy response at Lucerne Grand”. Mark Yip, chief executive of Huttons Asia, described the buyers as “mainly owner-occupiers living in western Singapore”. Justin Quek, deputy group chief executive at Realion, added that “HDB upgraders also formed part of the buyer pool”. Kelvin Fong, chief executive of PropNex, called the “buoyant take-up” something that “bodes well for the new home sales market”.

I do not disagree with any of that. 61.4% on a 570-unit project in the west, with no recent neighbour to benchmark, is a solid weekend. But “solid weekend” is where most coverage stops, and that is exactly where the useful part begins. If you want my unit-by-unit read on the project itself, I wrote it up separately in my Lucerne Grand review. This article is about the number, not the project.

Because here is the thing. 61.4% sounds like a verdict. It is actually an average of several very different outcomes, divided by a denominator somebody chose. Both of those facts are hidden in plain sight, and both are fixable in about thirty seconds once you know what to ask.

The 2026 new launch take-up rate scoreboard

This is the full-year table. Eleven projects, every one with a published launch-weekend figure. I have put the source and the date in the row, because with this metric the date is half the information.

Credit where it is due, and I am not going to be shy about it: six of those eleven rows — Rivelle Tampines, Pinery Residences, River Modern, Coastal Cabana, Newport Residences and Narra Residences — come from Stacked Homes, specifically Ryan J. Ong‘s Q1 2026 table published on 2 April 2026. Two more, Tengah Garden Residences and Vela Bay, come from Timothy Tay at Stacked Homes, dated 26 April 2026 — and I then confirmed both independently at EdgeProp the same day, which published “About 853 units (98.8%) of the 863-unit Tengah Garden Residences were sold on launch weekend” at “an average price of $2,120 psf”, developed by a joint venture of Hong Leong Holdings, GuocoLand and CSC Land Group. EdgeProp did not break out Vela Bay separately, but it did report “combined sales from Tengah Garden Residences and Vela Bay — totalling about 1,224 units”, and 1,224 minus 853 is 371, which is exactly the Vela Bay figure in my table. That subtraction is mine, not EdgeProp’s, but it reconciles the two reports to the unit and is the reason I am comfortable printing that row. Without their reporting there is no 2026 scoreboard to build. My contribution is extending it to the full year to date and then stress-testing what the percentages mean, which is a different job from collecting them.

One honest flag on the Lentor Gardens Residences row. That 282 of 499 is from my own earlier record of the launch, not from a published report I re-verified today, and I do not have a reliable average psf for it. Treat it as indicative. If you are looking at that whole corridor, my views on supply there are in my piece on Lentor condo oversupply after seven launches, and the short version is that the launch-weekend number was never the interesting part of that story.

Which totals are reported, and which ones I divided myself

This is the kind of thing most sites quietly skip. For the six Q1 2026 projects, what was published was the units-sold count and the percentage. The project total was not. So if you want the denominator, you have to divide — and that is arithmetic, not reporting.

I am going to show you my arithmetic rather than pretend it is a source. Below, the “implied total” column is my own division, rounded to the nearest unit. It is approximately right and it is not a reported figure. Do not quote it as one.

Why does it matter whether a total is reported or implied? Because the moment the total is implied, you cannot tell whether the published percentage used the project total or the phase released. If 529 units at 92.5% was computed on units released rather than the whole project, then my “roughly 572” is not the project size at all — it is the phase size, and the project could be far larger. I flag it, I do not resolve it, and I would rather tell you that than hand you a clean-looking number I cannot defend.

The denominator problem: what a new launch take-up rate is actually counting

If you read one section of this article, read this one. It is the single most useful paragraph I can give you about new launch reporting, and almost nobody states it consistently.

A take-up rate is a fraction. Units sold on top. Something on the bottom. And in Singapore new launch coverage, that something is one of two completely different numbers:

• Units sold ÷ total units in the project. The honest, hard denominator. 350 of 570.
• Units sold ÷ units released in that phase. A denominator the developer chooses, and can choose again next month.

Nobody is lying when they use either one. They are just not telling you which one they used, and the gap between them is not a rounding difference. It can be threefold.

The proof: one project, one weekend, two answers

Amberwood at Holland launched over the weekend of 25–26 September 2026. Sim Lian Group sold 23 of the 212 units, or 11%. That is the project denominator.

In the same article, EdgeProp also reported this: “based on the 70 units released in the first phase, the take-up rate was 33%”. That is the phase denominator. You can read the EdgeProp report on Amberwood at Holland’s launch weekend yourself — and I want to credit EdgeProp specifically for publishing that phase sentence, because it is rare and it is what makes this entire argument provable rather than just something I assert.

Three times. Same project, same two days, same 23 buyers. One number says the launch struggled. The other says a third of what was on the shelf walked out the door, which reads like a respectable result.

Why this is not an accident

Now think about the incentives, because they run in one direction and they are entirely predictable.

A developer who expects a strong weekend releases a big block of units — sometimes everything — and quotes the project number. Of course they do. 853 of 863 is a headline you want in print, and it only exists if you put all 863 on the table.

A developer who expects a soft weekend releases a small first phase. Seventy units out of two hundred and twelve. Then the phase number becomes the quotable one. And nothing about that is improper — staged releases are normal practice, they help with price discovery, they protect pricing on later stacks, and any sales director would do the same. But it does mean the percentage you read is partly a product of a marketing decision made before a single buyer walked in.

So the asymmetry is this: strong launches tend to be reported on the harder denominator, weak launches on the softer one. Which systematically compresses the apparent spread between good and bad weekends. The real spread is wider than the headlines suggest.

Which denominator should you actually care about?

Depends what you are using it for, and this is where I will give you a straight answer rather than both sides.

If you are deciding whether to buy, you care about the project denominator. The question underneath your question is: how many unsold units in this development will be on the market, in some form, over the next several years? Unsold inventory is future competition for your unit, and it also tells you how much pricing power the developer retains. 23 of 212 means 189 units are still coming. The phase percentage does not tell you that.

If you are trying to judge pricing accuracy on the specific stacks launched, the phase denominator has real information in it. A high phase take-up on a tight release says the developer priced that release well for the buyers who came. That is a genuine signal — it is just a narrower one than the headline implies.

What you must not do is compare a phase percentage from one project against a project percentage from another and conclude that one development is doing better. That comparison is meaningless, and I see it made constantly, including by people who should know better.

One project, two answers: Lucerne Grand’s 87.5% and its 51%

The denominator problem has a second floor, and this is my own finding from the Lucerne Grand numbers rather than something I read anywhere.

Lucerne Grand’s headline is 61.4%. Here is what that 61.4% is made of.

Look at the range inside one building. The standard 2-bedroom cleared 87.5%. The 4-bedroom as a category cleared 51%. And then, strangely, the premium 4-bedroom cleared 87.2% — almost identical to the standard 2-bedroom, and from a much smaller pool of 47 units.

So the 61.4% is an average over outcomes that span 51% to 87.5%. It is a true number and it is also a blur. If you were the buyer who wanted a standard 2-bedroom, the market you were actually competing in cleared 87.5% in two days — nearly everything gone — and “61.4%” badly understated your urgency. If you wanted a non-premium 4-bedroom, you had roughly half the stock still available on Monday morning, and “61.4%” overstated how much pressure you were under.

Same building. Same weekend. Same price list. Two completely different experiences, and one headline number that describes neither of them well.

What to do with this as a buyer

Ask for the breakdown by unit type for the configuration you actually want. Agents have this. Developers produce it. It is not confidential. And it is the only version of the take-up rate that bears on your decision, because you are not buying 570 units — you are buying one, in one configuration, in one stack.

A practical note from the Lucerne Grand pattern that I think generalises: the premium 4-bedroom clearing 87.2% out of only 47 units is what scarcity does. Small pools of distinctive units move fast at launch because the buyers who want exactly that thing know they will not get a second chance. The ordinary 4-bedroom, with 95 units and no scarcity story, took its time. If you want a unit type where there are many of them, you usually have more room to think than the headline suggests. If you want the one type there are few of, you have less.

The spread is the story — and it tracks unit mix more than psf

Now step back and look at the scoreboard as one object instead of eleven rows.

2026 launch-weekend results run from 11% to 98.8%. That is nearly a ninefold difference, inside a single calendar year, under the same cooling measures, the same Total Debt Servicing Ratio, the same Additional Buyer’s Stamp Duty schedule, the same interest-rate environment, the same economy.

Which kills a claim you will hear at every coffee shop and in half the headlines: “the market is hot” or “the market is cooling” cannot explain this. A market-wide condition, by definition, applies to every project in the market. If the condition were doing the work, the results would cluster. They do not cluster — they sprawl from 11 to 98.8. So the explanation has to be project-specific. Something about the individual development is doing nearly all of the work.

The tempting answer, and why I am not going to give it to you cleanly

The obvious story is price band: cheap projects sell out, expensive ones lag. Let me group the rows and we can see how well that holds.

If price band were the engine, those bands would stack neatly downward. They do not. The $2,000–$2,599 band contains both the best result of the year (98.8%) and the second-worst (23%). The $3,000-and-above band contains a 90%. So I am not going to tell you “cheaper sells better” and dress it up as a finding, because the rows I just put in front of you contradict it.

Two rows break the simple price story outright, and I would rather name them than bury them:

• River Modern, $3,266 psf, 90%. The second-most expensive project on the board, and it cleared 90%. A pure price-band theory has no explanation for this.
• Narra Residences, $2,148 psf, 23%. One of the cheaper projects of the year, and it managed 23%. From the published figures alone I cannot tell you why, and I am not going to invent a reason. Something project-specific — location, unit mix, launch timing, competing launches that weekend — is doing it, and none of that is in the data I have.

The variable that fits better: is there a small, low-quantum unit to buy?

What does hold up across the rows, in my reading, is not psf but entry quantum and small-unit availability. Look at where the high numbers are. The two executive condominiums. Tengah Garden Residences. Pinery Residences. And inside Lucerne Grand, the standard 2-bedroom at 87.5% against the ordinary 4-bedroom at 51%.

Then look at the bottom of the table. Amberwood at Holland, 11% — and, as I will get to in a moment, a project with no 2-bedroom units at all.

Price per square foot is the number everyone quotes because it is comparable. But almost nobody buys psf. They buy a total price they can finance, which is psf multiplied by size. A $3,266 psf unit at 500 square feet is about $1.63m. A $2,990 psf unit at 1,100 square feet is about $3.29m. The second one is “cheaper” per square foot and more than twice the cheque. The launch-weekend crowd is overwhelmingly the group that can write the smaller cheque, so projects with small units convert faster. That is the mechanism I think is operating.

Amberwood at Holland: 11% that might not be a failure

Amberwood at Holland is the most instructive row on the board, so let me give it proper space.

Sim Lian Group sold 23 of the 212 units, or 11%, over the launch weekend of 25–26 September 2026, at an average price of about $3,019 psf. Of those 23 units, 11 were four-bedroom, nine were three-bedroom, and the remaining three were five-bedroom.

Read that last row again, because it is the cleanest piece of evidence in the entire dataset.

Amberwood has no two-bedroom units at all. The smallest configuration you can buy is a three-bedroom starting at $2.516m. And at $3,019 psf average, Amberwood is actually the cheapest of the 2026 prime launches on a per-square-foot basis — cheaper than River Modern at $3,266 and Newport Residences at $3,370, both of which cleared far higher percentages.

So here is a project that was not the most expensive per square foot, and still recorded the lowest take-up of the year. The thing it lacked was not competitive psf. It was a low-quantum entry point. Kelvin Fong of PropNex said it directly — “the project has limited entry-level options” — and that one sentence does more explanatory work than any amount of talk about market sentiment.

If you want the sharpest version of the argument: psf tells you what a project costs per unit of space. Entry quantum tells you how many people in the room can actually transact. Launch weekend is a test of the second thing.

The counter-argument, which is fair and which nobody bothers to make

Now let me argue against my own framing, because the honest reading demands it.

A low launch-weekend take-up is not automatically a failure. A developer selling only large family homes to families genuinely expects a slower, longer sales campaign, and prices accordingly. Families buying a $3m to $4.8m home do not decide in a showflat queue over a weekend. They bring parents. They walk the neighbourhood on a weekday. They check which primary schools are within a kilometre. They sell an existing property first. That process takes months, not forty-eight hours.

Marcus Chu of ERA Singapore described it precisely: “buyers of larger homes are taking more time to assess factors like long-term liveability”. Justin Quek of Realion noted that “demand was primarily driven by owner-occupiers and families drawn to the project’s spacious, family-centric designs”. Both of those are descriptions of a deliberate, deliberative buyer, not an absent one.

Which means the right question about Amberwood is not “why only 11%” but “what is the sell-through at month six and month twelve, and at what prices”. If a large-format project absorbs steadily over a year without discounting, that is a successful campaign with an unremarkable opening weekend. If it stalls and the developer starts repricing, that is a different story. The launch weekend cannot distinguish between the two, and anyone telling you otherwise in September was guessing.

And note the staged release here is consistent with exactly that plan. Seventy units out of 212 in phase one is what you do when you expect a measured absorption. The 33% phase figure is arguably the more honest description of how that weekend went against the developer’s own expectations.

For context, 2025 — and the date trap

These next two projects are 2025 launches, not 2026. I am putting them in a separate table on purpose, and they are deliberately kept out of my 2026 scoreboard, because mixing years is how bad analysis happens.

A little more on each. Zyon Grand sold 590 of 706 units, 84%, at an average S$3,050 psf, with 84% Singaporean and 14% PR buyers, and one of its two five-bedroom penthouses went for above S$10m. Springleaf Residence sold 870 of 941 units, 92%, with one-bedders from around $860,000 and five-bedroom units of 1,453 to 1,475 square feet from around $2.97m.

Springleaf is worth noting against my own argument, incidentally: a one-bedroom from roughly $860,000 is about as low as entry quantum gets in a private launch of that scale, and the project cleared 92%. That is consistent with the unit-mix reading.

What the new launch take-up rate should change about how you buy

Let me be useful rather than clever. Here is how I actually use this metric with my own clients, and where I tell them to ignore it.

1. Treat it as a pricing verdict, not a quality verdict

A high take-up rate tells you the market agreed with the developer’s price on the units released. That is it. It says nothing about construction quality, layout efficiency, the management corporation you will inherit, the view in five years when the next plot gets built out, or what the unit will fetch on resale. I have watched projects sell out in a weekend and then trade sideways for years. I have watched slow starters do perfectly well. Launch speed and long-run performance are two different questions.

2. Use it to estimate your negotiating room, and nothing else

This is where the number genuinely earns its keep. High project-level take-up means the developer has very little unsold stock and no reason to move on price. Low project-level take-up means there is inventory, and inventory eventually produces flexibility — in the form of better stacks still available, or, later in the campaign, incentives. That is a practical read you can act on.

3. Always get the figure for your unit type

By now the reason should be obvious. Lucerne Grand’s 61.4% was 87.5% if you wanted a standard 2-bedroom and 51% if you wanted an ordinary 4-bedroom. The project number was the wrong number for both of those buyers.

4. Compare against resale before you let urgency decide

Launch weekend is engineered to compress your decision time, and a published take-up rate is part of that engineering. The counterweight is to know what comparable completed stock costs in the same area. If you have not done that comparison, you cannot tell whether a fast-selling launch is well priced or merely well marketed. I set out how I run that comparison in my piece on resale condos versus new launch in 2026. Do that homework before the weekend, not during it.

5. Cross-check the numbers yourself where you can

You can follow unsold-inventory movement over time on the portals — 99.co and the usual listing sites will show you what is still being marketed months after launch, which is a far better test of absorption than any single weekend. And if you want to sanity-check a quantum against your own finances rather than against a brochure, run it through the money math tools first. Affordability is arithmetic. Launch-weekend atmosphere is not.

6. Know what your own property is worth before you commit to a new launch

Most new launch buyers in 2026 are upgraders. Justin Quek flagged exactly that at Lucerne Grand, where HDB upgraders formed part of the buyer pool. If you are one of them, the take-up rate of the project you want is a second-order concern. The first-order concern is what your current place will actually fetch and what that leaves you. Get a free valuation done before you queue for anything. I would rather you knew your number than the developer’s number.

And if you prefer to watch rather than read, I break down launch results and walk through projects on video at my video hub.

Who should chase a hot launch, and who should not

Frequently asked questions

What is a good new launch take-up rate in Singapore?

There is no single threshold, and anyone who gives you one is guessing. What I can give you is the 2026 distribution: published launch-weekend results ran from 11% to 98.8%, with most projects landing between roughly 55% and 92%. So anything above about 70% on a project-level denominator was an above-average weekend in 2026, and anything below 40% was clearly on the slow side. But the sample is small and the context matters more than the number. A 50% weekend on a large-format prime project and a 50% weekend on a mass-market project with plenty of two-bedders mean very different things.

Why do two reports give different take-up rates for the same project?

Almost always because they used different denominators. Units sold divided by total units in the project gives one answer. Units sold divided by units released in that phase gives another. Amberwood at Holland in September 2026 is the clean example: 23 sales was reported as 11% of the 212-unit project and, in the same article, as 33% of the 70 units released in phase one. Both are correct. Neither is wrong. They are simply answers to different questions, and the gap on that project was threefold.

Does a high take-up rate mean the condo is a good investment?

No, and this is the most common mistake I see. A high take-up rate means the market agreed with the developer’s pricing on the units that were released, in the week they were released. It says nothing about rental demand, completion-year supply in the area, layout efficiency, build quality, or resale pricing in five years. I have seen fast sell-outs trade flat for years and slow starters perform well. Use take-up to estimate your negotiating room, not to grade the asset.

What was the Lucerne Grand take-up rate on launch weekend?

Lucerne Grand sold 350 units over its launch weekend of 3 to 4 October 2026, reported as 61.4% of the 570 units, at an average of $2,480 psf. Buyers were 93% Singaporean and 7% Permanent Residents. The more interesting detail is the split by unit type: the standard two-bedroom type cleared 84 of 96 units, or 87.5%, while the four-bedroom category as a whole cleared 48 of 95, or 51%. The headline 61.4% is an average of outcomes that varied enormously inside the same building.

Why did Amberwood at Holland only sell 11 percent?

The clearest factor in the published data is that Amberwood has no two-bedroom units at all. The smallest configuration starts at a three-bedroom from about $2.516m, so the entry cheque is large regardless of the psf, which at roughly $3,019 was actually the lowest of the 2026 prime launches. PropNex chief executive Kelvin Fong put it as the project having limited entry-level options. There is also a fair counter-reading: large-format family homes are expected to sell over months rather than over a weekend, so a slow opening on that kind of project is not automatically a failure.

Is a 50 percent launch weekend take-up rate bad?

Not by itself. Half a project sold in forty-eight hours is a lot of transactions. Whether it is good depends on three things: the denominator used, the unit mix, and the price band. Fifty percent of a whole project, in a development with large units only, is a strong result. Fifty percent of a small released phase, in a project with plenty of affordable two-bedders, is a weaker signal. Ask which denominator produced the number before you judge it.

Where can I check take-up rate figures myself?

Launch-weekend figures are usually released by the developer and picked up by the property press — EdgeProp and Stacked Homes carried most of the 2026 numbers in this article, and they generally state the date, which many aggregator posts do not. For absorption after launch, the monthly developer sales data is the better tool, and watching what is still being marketed on portals such as 99.co months later tells you more about real absorption than any single weekend ever will. If you want the by-unit-type breakdown, ask an agent with access to the developer’s sales chart.

Should I wait after launch weekend to get a better price?

Sometimes, and it depends on what is left. If the project cleared most of its stock and your preferred unit type was among the fastest-moving, waiting costs you choice and gains you nothing — the developer has no pressure. If there is meaningful unsold inventory in your configuration, time is working in your favour, because inventory eventually produces flexibility on stack, floor or incentives. The deciding input is the project-level unsold count for your unit type, not the headline percentage.

Talk to me

If you are weighing a 2026 launch against resale, or you are an upgrader trying to sequence a sale and a purchase without getting caught in between, that is the conversation I have most days. Bring me the project and I will bring you the numbers behind the headline.