The Win-Rate Column Is Lying to You. Here Is the $4.5 Million Proof.

Here is the full ranking. Win rate first, then what it is hiding.

All profit figures are gross — exit price minus entry price only. They exclude stamp duties, agent fees, renovation costs, and mortgage interest. At the bottom of this table, the real financial outcome for sellers is materially worse than the gross figure shown.

Two projects. Same win rate. $4.5 million apart.

Look at row 1 and row 14 in the table above. Both show 100%.

Ardmore Park (Ardmore Park Road, off Orchard, CCR D10): 12 transactions, average gross profit $4,841,612, average hold 15.7 years.

The Glyndebourne (Tyersall Avenue, near Botanic Gardens, CCR D11): 19 transactions, average gross profit $371,777, average hold 11.4 years. Freehold. Also CCR. Also 100%.

The gap between them is $4,469,835.

The win-rate column shows the same number for both. That is the trap. A 100% win rate tells you that every seller made something. It tells you nothing about whether they made $370,000 or $4.8 million. It cannot tell you whether the project next door is quietly producing losses on 96% of its transactions. It is a binary — profitable or not — and that binary hides everything important about the distribution underneath it.

The Glyndebourne and Ardmore Park both show 100% in the win-rate column. The average seller at Ardmore Park walked away $4.47 million more than the average seller at The Glyndebourne. The column cannot tell you which one you are in.

The Glyndebourne is not a bad outcome — $372,000 gross on an 11-year hold at a CCR freehold address is a positive return. The point is not that it failed. The point is that it looks identical to one of the best-performing luxury addresses in Singapore when you are reading the win-rate column. That is a reading problem, not a project problem.

Seven years in, still $293,000 down: the Marina Bay problem

Now look at the bottom of the table. Five projects. All CCR. Combined, they account for 124 resale transactions. Not one of them has a win rate above 10%. Most are at zero.

What connects them is not location alone, and not tenure alone. It is when they were built and what buyers paid to get in.

Marina One Residences (Marina Boulevard, Downtown Core, District 1): completed 2017, 56 transactions, win rate 4%, average gross loss $292,539, average hold 7.0 years. Only 2 of 56 sellers made money. The other 54 sold at a loss, on average nearly $300,000 below what they paid — before fees, taxes, and renovation.

Marina Bay Suites (Marina Boulevard, District 1): completed 2013, 20 transactions, win rate 10%, average gross loss $648,914, average hold 13.6 years.

OUE Twin Peaks (Leonie Hill, River Valley, District 9): completed 2015, 43 transactions, win rate 7%, average gross loss $331,651, average hold 7.9 years.

Cliveden At Grange (Grange Road, River Valley, District 10): completed 2011, 8 transactions, win rate 0%, average gross loss $1,920,380, average hold 13.9 years. Freehold.

Helios Residences (Cairnhill Circle, Newton, District 9): completed 2011, 10 transactions, win rate 0%, average gross loss $1,173,443, average hold 12.6 years. Freehold.

These five projects were sold off the plan between roughly 2007 and 2014 — a stretch that captured both the pre-global financial crisis peak and the sharp post-2009 CCR recovery. Launch-era buyers paid CCR prices set at or near the top of that cycle. Since then, the URA non-landed private price index has moved from around 197 in mid-2024 to around 211 in early 2026 — a recovery of about 6–7%. That recovery has not been large enough to bring this cohort back to breakeven. Current resale PSF at Marina One Residences is approximately $1,999. At Marina Bay Suites, approximately $1,936. Sellers who bought at launch-era pricing are still underwater at those levels.

This is the structural diagnosis: the problem is not that these owners did not wait long enough. At Cliveden At Grange, the average hold is already 13.9 years. At Marina Collection, 15.2 years. The problem is that the entry price was set above where the market is today, and a 6–7% broad recovery has not bridged that gap.

The leasehold explanation does not hold

A reader looking at Marina One Residences (99yr), Marina Bay Suites (99yr), and OUE Twin Peaks (99yr) will immediately reach for the obvious answer: leasehold tenure. But that explanation collapses the moment you look at Cliveden At Grange and Helios Residences. Both are freehold. Both show 0% win rates. Cliveden At Grange sellers are losing nearly $2 million on average after almost 14 years.

Leasehold amplifies the problem for some projects — a 99-year lease running from 2013 has shed meaningful residual value since launch, and that compounds against an already-high entry price. But leasehold is not the cause. Freehold status has not rescued a single seller at either Cliveden At Grange or Helios Residences. The cause is entry price relative to where the market is now.

The same test applies to unit size. Marina One Residences has a typical unit of 1,044 sqft. OUE Twin Peaks is 1,055 sqft. A reader might argue that compact CCR units simply do not appreciate well regardless of when they were bought. But The Imperial (CCR D9, freehold, 100% win rate, $1.93M average profit) has typical units of 1,443 sqft — larger, but not dramatically so. The Marbella (CCR D10, freehold, 100% win rate, $1.78M average profit) is 1,582 sqft. Size matters at the margins, but neither tenure nor unit size is the determinative variable here. Both can be ruled out by the data. Entry price is the consistent thread.

Hold longer and you almost always win — except here

Here is the broader dataset: 23,859 resale transactions across all hold periods.

The pattern is clean. Hold for two years and you win 97% of the time. Hold for a decade and you make over $700,000 on average. Patience is a genuine strategy for Singapore private property in aggregate.

But the Marina Bay cluster is sitting outside this pattern entirely. Marina Bay Suites sellers are averaging 13.6 years and still losing $648,000. Marina Collection sellers are averaging 15.2 years and losing $2.4 million. They are not in the 10+ year row above — or rather, they are in it numerically, but their outcomes look nothing like the 95% win rate the row suggests. The aggregate hold-period table is pulled strongly by the large volume of profitable long-hold freehold projects. The Downtown Core cohort is the exception that the aggregate hides.

The one that did not need 20 years

Before leaving the 100% zone entirely: Leedon Residence (Farrer Road, Bukit Timah, CCR D10, freehold, completed 2015) is worth a single paragraph because it punctures a different assumption — that meaningful CCR gains require a 20-year hold.

28 transactions. 96.4% win rate (1 seller lost money). Average gross profit $1,700,028. Average hold: 8.3 years. Typical unit size: 2,131 sqft. The large floor plate matters here — a 2,131 sqft unit absorbs a meaningful dollar gain even from moderate PSF appreciation. But the combination of volume (28 transactions is the largest sample in the top-profit cluster outside the Downtown Core loss group), profit magnitude, and an 8.3-year average hold makes this the clearest counterexample to "CCR means waiting forever." The entry price was right. The product matched genuine demand — 4-bedroom units here rent for around $18,000 a month, and 3-bedrooms at approximately $13,500, reflecting sustained institutional tenant demand from the Bukit Timah corridor.

Sentosa Cove: the market that left

Three projects form their own sub-cluster at the bottom of the table: Marina Collection (0%, –$2.43M), Seascape (0%, –$937k), The Coast At Sentosa Cove (16%, –$465k). All are 99-year leasehold. All completed between 2009 and 2011. All are located on Sentosa Cove island, subject to additional regulatory constraints on foreign ownership and specific use designations that have historically suppressed the resale buyer pool.

Marina Collection sellers have held for an average of 15.2 years and are still losing $2.4 million on average. Seascape, despite a shorter average hold of 6.1 years, shows 0% win rate — meaning sellers who got in and out quickly lost just as badly as those who stayed.

Caribbean At Keppel Bay (Keppel Bay Drive, Bukit Merah, RCR D4) is the contrast that makes the geography argument complicated: also 99-year leasehold, also waterfront, also adjacent to the southern coast. But Caribbean completed in 2004, is not under the Sentosa Cove designation, and sits in RCR rather than CCR. Result: 77 transactions, 96.1% win rate, average profit +$709,000. Same broad waterfront address. Completely different outcome. The Sentosa Cove regulatory and demand environment, combined with 2009–2011 completion, explains the divergence more than leasehold tenure does.

What the ranking is actually telling you

The URA private price index has recovered about 6–7% since mid-2024. Broad private property in Singapore is not in distress. The 97–98% win rates for 2–10 year holds confirm that most sellers, most of the time, are coming out ahead.

But five projects completed in the 2010s — some freehold, some not, some in the Marina Bay financial district, some in CCR Orchard and Newton — are still producing losses for the majority of sellers after holds of seven to eighteen years. They share a launch vintage and an entry price set above where the current market trades. The broad recovery has not been large enough to close that gap.

The win-rate column cannot show you this. It collapses a $4.5 million difference between two freehold CCR projects into the same single percentage. It shows you Marina One Residences at 4% without telling you that 54 of 56 sellers walked away with less than they paid, gross, before fees. And it shows The Glyndebourne at 100% without telling you that the average profit there is closer to what a good HDB resale year looks like than what the rest of the CCR top table delivers.

The data on entry price is the more useful number. So is the completion year. The projects at the top of this table share freehold tenure, large floor plates, and entry prices set before or well below the 2012–2014 CCR peak. The projects at the bottom share the opposite.

For more context on how Singapore's most expensive districts handle profit and loss across different property types, see D10's $3.16M Average Is a Detached-House Story. Semi-Detached Buyers Are Reading the Wrong Number. and Singapore's Most Expensive Planning Area Is Also Its Worst for Condo Resale Profits.

Market Intelligence — Resale Condo Profit Ranking

Scope: Cross-market private resale | Source period: Q3 2024 – Q2 2026 | Total transactions analysed: 25,271 across 1,295 projects

Key metrics — projects in this ranking:

• Highest average profit (100% win rate): Ardmore Park — $4,841,612 gross average on 12 transactions
• Lowest average profit (100% win rate): The Glyndebourne — $371,777 gross average on 19 transactions
• Largest sample in loss cluster: Marina One Residences — 56 transactions, 4% win rate, –$292,539 average
• Deepest average loss: Marina Collection — 7 transactions, 0% win rate, –$2,434,336 average
• Strongest mid-table return: Leedon Residence — 28 transactions, 96.4% win rate, +$1,700,028 average, 8.3-year hold

Notable transactions — top and bottom of the ranking:

All figures gross. Source: Coalwood resale dataset, Q3 2024–Q2 2026.