Buying a 2 bedroom at launch beat buying the same condo later in 128 of 228 developments, and lost in most of the OCR

Buy the two-bedroom at the showflat, or wait and buy one in the same building after it is finished? This note compares the two buyers inside the same building. It covers 228 developments where a two-bedroom was bought in the launch year, another was bought one to four years later, and both have since been resold.

Two buyers appear throughout. The launch buyer bought a two-bedroom from the developer in the year the development was launched, paying in stages as the building went up. The later buyer bought a two-bedroom in the same development one to four years after that launch, either from the developer if units were still unsold, or from an earlier buyer as a sub-sale before completion or a resale after it. Both later resold, which is what makes the comparison possible.

The launch buyer came out ahead in most of them. In 128 of the 228 developments, 56%, the launch buyer made the larger gain when the unit was resold. In 126 of the 218 developments where the full costs can be modelled, the launch buyer earned the higher return on the money put in.

Return on the money put in means this. Count every dollar the buyer actually paid out, the deposits, the stamp duty, the instalments, the interest, the upkeep and the property tax, and every dollar that came back, the rent collected and the resale price less commission. The return is the annual rate that turns the money out into the money back. It is the measure an investor uses, because two buyers can make the same gain on paper while one of them had far more cash tied up for far longer.

On gain the two buyers were close. In the typical development, the launch buyer's gain was 2.6 percentage points higher than the later buyer's. On return on the money put in they were not close. The typical development gave the launch buyer 14.8% a year and the later buyer 11.9%. The launch buyer paid in stages while the building went up; the later buyer paid at once, for a unit that already cost more per square foot.

Where the building is decides it. In the CCR the launch buyer won 30 of 43 developments, made a typical gain of 32.6% against the later buyer's 24.1%, and earned the higher return however long the other buyer waited. In the OCR the later buyer won more developments than the launch buyer, 49 of 90, made a slightly larger gain, 9.9% against 8.7%, and earned a slightly higher return, 7.1% against 6.4%. The shoebox series found the launch buyer ahead in 41 of 48 developments, 85%; the same comparison for two-bedrooms gives 56%, and the difference is the OCR.

This note and the launch-versus-resale note ask different questions. That one compared a launch buyer with a resale buyer in a different building the same year, and the resale buyer won every year. This one keeps the building fixed. The later buyer is paying the finished building's price, which in the central regions is higher than its own launch price by more than the launch buyer's waiting cost.

Key takeaways

• The launch buyer beat the later buyer of the same building in 128 of 228 developments on gain, 56%, and in 126 of 218 on return on capital. For shoeboxes the figure was 85%. Development by development
• In the typical development the launch buyer's return on capital was 14.8% a year and the later buyer's 11.9%, a gap of 2.9 percentage points. Development by development
• In the CCR the launch buyer won 30 of 43 developments and earned a higher return than the later buyer at every length of wait, by 6 to 23 percentage points a year. In the OCR the later buyer won 49 of 90. By region
• A buyer who waited four years in the CCR earned minus 0.9% a year on the money put in; the launch buyer of the same buildings earned 21.2%. How long a wait matters

How the comparison is built

Every development where at least one two-bedroom was bought in the launch year and at least one was bought one to four years later, both later resold, with each buyer's own purchase price, resale price and hold. The comparison is made inside each development first, launch-year buyers against later buyers of the same building, and only then across developments, so a building with many later sales cannot outweigh one with few. The return on the money put in is computed the same way as in the launch-versus-resale note. The launch buyer pays in stages as the building goes up, the later buyer pays at once the stages already due, both start collecting rent two months after completion, and stamp duty, interest, upkeep, property tax and commission are all charged.

Development by development

	Launch-year buyer	Later buyer
Units resold	751	791
Typical gain at resale	17.5%	17.7%
Developments where this buyer's gain was larger	128	99
Typical return on capital at 2.5%	14.8%	11.9%
Developments where this buyer's return was larger (of 218)	126	92

Typical here means the median across the 228 developments. Half the developments gave a larger figure and half a smaller one.

Two ways of comparing the gain give different answers, and it matters which one a buyer reads. Taken across all developments, the typical launch buyer gained 17.5% and the typical later buyer 17.7%, which reads as if waiting did slightly better. But that compares a launch buyer in one building with a later buyer in a different one. Compared inside the same building, the launch buyer's gain was larger in 128 developments out of 228, and in the typical development it was larger by 2.6 percentage points. The later buyer's gain was larger in 99, and in 1 development the two medians were exactly equal, which is why the two counts fall one short of 228. The second comparison is the one that answers the question, because a buyer chooses between buying now and buying later in one building, not between the market's two averages. On return on capital both comparisons agree. The launch buyer's return was larger in 126 developments of 218, by 2.2 percentage points a year in the typical development and by 2.9 points between the two medians.

By region

• Launch-year buyer won
• Later buyer won

Region	Developments	Typical gain, launch	Typical gain, later	Launch won	Return at 2.5%, launch	Return at 2.5%, later
CCR	43	32.6%	24.1%	30 of 43	23.5%	14.9%
RCR	95	27.6%	24.7%	58 of 95	24.2%	16.7%
OCR	90	8.7%	9.9%	40 of 90	6.4%	7.1%

The central regions carry the whole of the launch buyer's advantage. In the CCR and RCR the launch buyer made the larger gain and earned the higher return in most buildings, and the gap on return was wide, 8.6 percentage points a year in the CCR and 7.4 in the RCR. In the OCR the later buyer was ahead on both measures, narrowly, and won 49 developments to the launch buyer's 40, with the remaining one exactly tied. The reason is what the later buyer paid. In the central regions a finished building sold for well above its own launch price, and the launch buyer kept that difference; in the OCR the difference was small, and the later buyer, who started collecting rent sooner on a building already finished, made it back.

How long a wait matters

Launch-year buyers against buyers who waited one, two, three or four years to buy the same building, by region. Gains and returns are the medians across developments; the return is at 2.5%.

CCR

Waited	Developments	Gain, launch	Gain, later	Launch won	Return, launch	Return, later
1 year	19	52.6%	33.1%	13 of 19	28.6%	20.4%
2 years	16	62.9%	21.4%	13 of 16	34.3%	14.4%
3 years	24	33.5%	28.8%	14 of 24	26.8%	20.5%
4 years	15	22.9%	2.9%	11 of 15	21.2%	−0.9%

RCR

Waited	Developments	Gain, launch	Gain, later	Launch won	Return, launch	Return, later
1 year	44	30.0%	25.0%	30 of 44	24.0%	13.8%
2 years	37	38.0%	34.7%	20 of 37	23.4%	21.8%
3 years	30	24.6%	26.4%	16 of 30	17.9%	21.5%
4 years	35	15.3%	13.3%	21 of 35	7.5%	9.7%

OCR

Waited	Developments	Gain, launch	Gain, later	Launch won	Return, launch	Return, later
1 year	46	13.8%	12.5%	23 of 46	9.1%	6.2%
2 years	42	11.8%	11.0%	14 of 42	8.1%	8.3%
3 years	18	7.3%	14.2%	9 of 18	8.0%	10.1%
4 years	32	8.1%	10.4%	13 of 32	6.1%	7.2%

In the CCR, waiting never paid, however long the wait. The launch buyer won most developments and earned the higher return at every wait. The buyer who waited four years, buying a finished building near its price peak, earned minus 0.9% a year on the money put in, while the launch buyer of the same buildings earned 21.2%. In the RCR the launch buyer made the larger gain at every wait but earned the lower return after three and four years, because by then the later buyer had been collecting rent for long enough to make up for a smaller price advantage. In the OCR the launch buyer was ahead on gain and on return only when the other buyer waited one year; at two, three and four years the later buyer was ahead on both. Two cells rest on few developments, the CCR's four-year wait on 15 and its two-year wait on 16, and show direction more reliably than size.

What waiting cost at entry

The later buyer paid more per square foot than the launch buyer of the same building, at every length of wait, though not steadily more the longer they waited. They paid 4.7% more after one year, 8.3% after two, 7.3% after three and 10.3% after four, across all developments. That is the launch discount. In the central regions it was most of the launch buyer's advantage; in the OCR it was small enough that the later buyer's earlier rent recovered it.

What this means at the showflat

For a two-bedroom in the CCR or RCR, buying at launch beat waiting for the same building in most developments and on most measures, and in the CCR the longer the wait the worse the outcome. For a two-bedroom in the OCR, waiting cost little at entry, and the buyer who waited, collecting rent sooner on a building already finished, came out slightly ahead more often than not. Neither is a rule about the market. Both come down to whether a region's finished buildings sell for much more than their own launch price, which the central regions' did and the OCR's did not. Read together with the launch-versus-resale note, the two findings fit. A launch buyer in a central-region building beats the buyer who waits for that building, and loses to the buyer who bought a resale unit in a different building the same year.

What this does not tell you

• The buildings that never resold both ways. Only developments with a resold launch-year purchase and a resold later purchase are in the data. That is 228, out of the far larger number launched in the period.
• Which unit. Within a building, floor, facing and stack move the price more than the year of purchase; these are medians across buyers, not a rule for one unit.
• Costs beyond the model. Additional buyer's stamp duty is not modelled, for the reason given in the launch-versus-resale note; renovation, vacancy and changes in interest rates are not modelled either.
• Today's launches. Launch prices in 2026 sit at a different level from the launches in this data, whose median launch year is 2006.

Method and data notes. Sale data to 7 May 2026.

Repeat sales from URA caveat data for private condominiums and apartments, two-bedrooms by resolved bedroom count, with the suspect class and two corrected units excluded, each pair a purchase and a later resale of the same unit. A development qualifies when it has at least one two-bedroom bought in its launch year and at least one bought one to four years later, from the developer or from an earlier buyer, both resold in the data; 228 qualify, with 751 launch-year and 791 later purchases. Gain is resale price less purchase price as a share of purchase price, measured per buyer, taken as the median within each development, then as the median across developments; "launch won" counts developments where the launch-year median gain exceeded the later-buyer median. Return on the money put in uses the model of the launch-versus-resale note with each buyer group's own median entry price, hold and floor area per development; the later buyer pays at once the stages already due at entry and the rest on schedule, rent begins for both two months after the model's completion, the market-wide 2025 two-bedroom rent per square foot is applied to the overall figures and each region's own rate to the regional tables, upkeep is proportioned to floor area, and property tax follows the registered non-owner-occupier schedule on an annual value equal to the modelled rent, and interest is at 2.5%, with 1.5%, 3.5% and 4.5% also run. This differs from the shoebox note in two ways. Entry is each group's measured median price rather than a fixed area times a median price per square foot, and property tax follows the registered schedule rather than a flat figure. Where two ways of aggregating give different answers, the figure compared inside each development is the one stated, and the pooled figure is shown beside it. Sale data to 7 May 2026.