The Same Three Letters Produced the Biggest Wins and the Biggest Losses. The Difference Is One Street.

A seller at Cliveden at Grange (River Valley Road, D9) bought their unit around 2007, held it for 17 years, and exited in November 2024. They lost $3,707,680.

Another seller at the same project held 19 years and lost $2.2 million. A third held 18 years and lost $2.17 million.

All three are CCR. All three were patient. None of it helped.

The conventional belief about CCR property is simple: buy in a prestige district, hold long enough, and the market eventually vindicates the price. This dataset — covering 25,271 transactions from Q3 2024 to mid-2026, in a market that rose 6.7% over the study period — tests that belief directly. What it finds is that the CCR label is doing almost no work. Your outcome depends on a far more specific question: which part of CCR did you buy in?

The same region label produced a +$4.84M average gain at Ardmore Park and a -$2.43M average loss at Marina Collection. The sellers in both cases held for roughly 15–17 years. The difference between them is geography.

Two tables. One region label. Completely different outcomes.

Here are the CCR projects that produced the strongest average gains and the steepest average losses over the study period. Same region. Different address.

CCR Winners — Orchard-Tanglin / Bukit Timah Corridor

CCR Losers — Marina Bay / Downtown Core / Sentosa–Southern Islands

What the tables show is not that CCR is good or bad. It is that CCR is two markets sharing a postcode. The Orchard-Tanglin-Bukit Timah corridor produced compounding gains across multiple projects and hold periods. The Marina Bay, Sentosa Cove, and Newton-Scotts cluster produced persistent losses, often across holds exceeding a decade.

OUE Twin Peaks deserves a note: at 43 transactions it is the most statistically reliable data point in the loss table, and the loss is relatively modest at -$332k on an average hold of 8.9 years. This is not the story of the patient seller who waited 17 years for nothing. But it is still a loss, in a rising market, in an Orchard-adjacent address — and the explanation is almost certainly unit size (typically 1,055 sqft at $2,241 PSF), which limits the buyer pool at exit.

The 17-year hold that still didn't work

The whole-market picture makes the CCR loser cluster even harder to explain away.

At 10 or more years, 95% of all transactions across the whole market came out profitable. That is an extraordinarily high success rate. It supports the instinct that patience rewards property owners.

But the 5% who lost money are not randomly distributed. Every single transaction in the 7-year-plus loss list is CCR. Zero are RCR. Zero are OCR. Here they are:

Thirteen transactions. Holds ranging from 12 to 19 years. All CCR. Losses ranging from $2.1M to $3.7M per transaction. In a market that is broadly up.

The geographic concentration is striking. Five are Sentosa/Southern Islands. Four are River Valley, all at Cliveden at Grange. Two are Newton, both at Helios Residences. The one Marina Bay entry — Marina Bay Residences — held 19 years and lost $3.58M. These are not scattered failures. They are clustered in the same sub-districts.

Why Sentosa and Marina Bay failed structurally

The location data tells the clearest part of the story.

Sentosa Cove is 3.0km from Harbourfront MRT. There are no malls within 1km, no hawker centres within 500m, no schools within 1km. A resident without a car has almost no access to the ordinary infrastructure of daily life. Marina Bay was conceived as a global business district — branded residences for international executives, adjacent to a financial hub.

When lifestyle premiums are the only thing supporting a price level, a shrinking buyer pool is catastrophic.

The policy dimension made it worse. From 2010 onward, Additional Buyer's Stamp Duty effectively priced out the foreign buyer base that Sentosa Cove and Marina Bay were conceived around. The 2011 and 2013 ABSD rounds were not foreseeable from the vantage point of a 2007 purchaser. By the time these tightening measures were in place, the entry price was locked in — and the buyer pool that could have grown into it had been cut significantly. This is a structural demand problem layered on top of a location problem. The two are not easily separated.

Ardmore Park works because a large and diverse pool of buyers — local professionals, families, expatriates — can access it without a car, within walking distance of one of Singapore's best-served transport and retail hubs. When the original seller is ready to exit, there is deep demand at the other end. That breadth of demand is what compounds capital over time.

What the entry price trap looks like in numbers

The data gives us exit prices and loss amounts. It does not give us confirmed launch prices — those figures are not in the dataset. But the arithmetic is straightforward and internally consistent.

One Cliveden at Grange seller exited in December 2025 at $3,202 PSF on a unit of roughly 2,500 sqft. The recorded loss was $2,169,510.

Work backwards: gross exit proceeds of approximately $8.0M. Add back the $2.17M loss and the implied entry price is around $10.2M — roughly $4,000 PSF on a 2,500 sqft unit.

This is not a confirmed figure from a launch database. It is reconstructed from exit price plus loss. But it is consistent with published CCR pricing history from the 2007–2008 pre-GFC peak, when prime district launches regularly cleared $3,500–$4,500 PSF. The seller paid peak-cycle pricing. The market has spent 17 years failing to recover to that level at this address.

The same arithmetic applies across the loser table. Marina Bay Residences exited at $2,680 PSF after 19 years with a $3.58M loss on a unit implying approximately $4,700–$5,000 PSF at entry. Marina Collection exits at $1,513 PSF with a $3.68M loss on a 3,272 sqft unit — implying entry around $2,600 PSF, which was not cheap in 2007 for a Southern Islands address without MRT access.

The pattern is consistent: buyers paid above what the location's structural demand could eventually support. The market moved up overall, but never enough at these specific addresses to close the gap.

The Orchard exception — and where it breaks down

Ardmore Park, Grange Residences, and Regency Park are the clearest cases of CCR compounding. All three have 10 or more transactions, all three are in the Orchard-Tanglin-Bukit Timah corridor, and all three produced gains well above $1.7M on average holds of 14–20 years.

The mechanism is not mystical. Broad buyer access, deep retail and transport infrastructure, and a location that appeals to a wide pool of eventual buyers at exit. These are the same fundamentals that make OCR volume leaders like Treasure at Tampines (Tampines Ave 10, 341 transactions over the study period) and Riverfront Residences (Hougang Ave 7, 163 transactions) structurally resilient — not spectacular, but not catastrophically wrong either. The OCR high-volume projects do not appear in either the top-20 profit table or the bottom-10 loss table. That absence is itself a finding: they are the boring middle, not the dangerous extreme.

But "Orchard-adjacent" is not a sufficient condition on its own. Two projects sitting on Scotts Road — some of the most recognisable real estate geography in Singapore — are in the loss column.

Scotts Square is a 47-storey tower on Scotts Road with 17 transactions averaging -$575,271 and a typical unit size of 947 sqft at $3,230 PSF. The Scotts Tower has 9 transactions averaging -$926,437 on an 850 sqft typical unit at $2,099 PSF.

Both are genuinely Orchard-adjacent. Both are losing money. The most likely explanation is unit size relative to price. A 947 sqft unit at $3,230 PSF represents an entry price above $3M for a flat that is smaller than many HDB 5-room units. The buyer pool for a $3M CCR unit under 1,000 sqft is extremely thin. Boutique tower units at ultra-high PSF are not a subcategory of "Orchard CCR that compounds" — they are their own category, with their own demand constraints.

The Scotts Tower and Scotts Square do not disprove the Orchard thesis. They refine it: location quality is necessary but not sufficient. The unit has to be a size and price point that a broad buyer pool can absorb at exit.

Leedon Residence: the case against waiting 20 years

Leedon Residence (Holland Road, Bukit Timah corridor, D10) has the most statistically