Yishun/Sembawang Are The Top Performing Districts For Private Homes But That's Only One Half Of The Picture. Here's Why

In a previous article, I wrote that Yishun/Sembawang was one of the top performing neighbourhoods in terms of annualised returns on private homes.

That, is still true but what we found next was that a large chunk of the transactions from that finding were actually from ECs - 66% to be exact. And the problem with that is ECs come with a income ceiling, a 5yr MOP and eligibility rules that exclude a large share of private buyers so that got me thinking:

What if we removed all EC transactions from the mix, does Yishun/Sembawang still sit on top of that list?

That is the number that prompted re-running the full district ranking with ECs removed. What came back was not a minor adjustment. Three districts dropped significantly. One district — Bishan/AMK/Thomson, D20 — rose to first place. And the results were now grounded in transactions that any private buyer, at any income level, could have made.

Three Districts That Were Partly Running on EC Returns

The EC distortion is sharpest in three districts. Here is what the numbers look like before and after ECs are stripped out.

The drops are not cosmetic. D27 was the original top-ranked district at 5.3% per year. Remove ECs, and it falls to 3.4% — below seven other private districts. D25 drops from second to mid-table. D23 falls to match D27 at 3.4%.

The reason the distortion is so large is that ECs in these areas genuinely returned outstanding numbers. Parc Life in Sembawang: 7.2% per year across 114 transactions. Sol Acres in Choa Chu Kang: 8.0% per year across 158 transactions. The Vales in Sengkang: 8.3% per year across 68 transactions. When nearly half or two-thirds of all transactions in a district are running at those rates, the district average is no longer describing what a private buyer actually achieved.

ECs in this period returned 6.6% per year on average — with named projects reaching 8.3%. They earned those returns. The income ceiling and five-year MOP are the reason they belong in a separate column, not a flaw in the products.

The Corrected Ranking: What Private Buyers Actually Achieved

Here is the ex-EC ranking across the seven OCR and RCR districts with the most meaningful data.

D20 is now first. D26 is second — but with only 120 transactions, treat that as directional rather than definitive. D18 at third has 1,682 transactions and a 100% win rate; that is one of the deepest samples in the data. The three Jurong/Seletar/Bukit Timah districts are bunched at 4.1%. D19, which has the single largest transaction sample at 2,508, comes in at 4.0% — the most battle-tested number in the table, and still a strong result.

One sentence on the other end: the worst-performing condo resale projects in Singapore are concentrated in the Core Central Region, where some long-hold sellers still lost money. The full picture is in this dedicated piece on the CCR loss cluster.

Is D20 a Jadescape Story, or a District Story?

The first fair challenge to D20's new top ranking: Jadescape is a large, high-profile project that completed in 2022 and has been one of the most active resale condos in Singapore.

176 of D20's 744 transactions belong to Jadescape and Jadescape is returning 5.4% per year, is D20 really #1 — or is one flagship project doing all the work?

The answer requires looking at what else is in the district.

Jadescape leads on volume and return rate. But Bishan Loft, an older project from 2003 with larger units and a much longer hold period, is returning 5.5% per year — matching Jadescape despite being nearly two decades older. These are different products, different buyer cohorts, different price points, and they are telling the same story about the district.

Thomson Impressions at 3.8% per year is the weaker performer, which pulls the district average down from what Jadescape alone would suggest. That is exactly how a district average should behave — it blends the strong with the moderate, and D20's 4.7% survives that blend.

The Jadescape concern does not hold up.

D18 Is the Boring One That Earns It

Tampines and Pasir Ris are not exciting districts to write about. No unusual story, no counterintuitive geography, no flagship neighbourhood narrative. What they have is 1,682 transactions and a 100% win rate — the second-deepest private condo sample in the data.

Treasure At Tampines anchors the picture. Across 341 transactions — the most active single project in the entire dataset — it returned 5.6% per year on typical holds of about 4.6 years, at a PSF of roughly $1,753 on units around 915 sqft. Typical profit: around $300k. Those are consistent, repeatable numbers across a large sample. The Tapestry, completed in 2021, is a quieter performer at 3.6% per year across 144 transactions — which is what keeps D18's district average at 4.3% rather than the 5%-plus that Treasure At Tampines alone would suggest.

The mix matters. D18's 4.3% is not one outlier project; it is a district average that includes a genuine high performer (Treasure At Tampines) and a newer project still early in its return cycle (The Tapestry). If The Tapestry's hold periods lengthen, D18's average may rise. For now, 4.3% with 1,682 transactions and a 100% win rate is among the most defensible numbers in any private condo district.

D19 Is the Largest Sample — and It Shows Internal Variance

The single biggest number in the data belongs to D19: 2,508 transactions across Hougang and Punggol. That sample size should give the strongest signal of any district. The return it produces — 4.0% per year, 99% win rate — is solid. But the internal spread is wide.

Riverfront Residences in Hougang (163 transactions, 99-yr, completed 2023) returned 4.5% per year — above the district average. Botanique At Bartley returned 4.8% per year across 109 transactions. These two projects suggest the district has genuine upside in the right pockets.

Kingsford Waterbay, also in Hougang (143 transactions, 99-yr, completed 2018), returned 2.4% per year — with a 94.4% win rate that is the lowest of any project named in this analysis. There is a full explanation for why Kingsford Waterbay lags its neighbours, and it involves more than just location. The point here is that D19's 4.0% district average contains both a 4.8% performer and a 2.4% drag. A buyer who picks the right project in D19 can do meaningfully better than the district average suggests; one who picks Kingsford Waterbay will not.

This is the honest limitation of district-level averages. They tell you the central tendency. They do not protect you from the projects at the bottom of the range.

The D26 Number Is Interesting — and Thin

Upper Thomson and Mandai sit at 4.6% per year, just below D20. The two named projects — The Calrose at 5.75% annualised across 20 transactions and Meadows @ Peirce at 4.1% across 24 — are directionally consistent with the district figure. Seasons Park, older stock from 1997, also came in at 4.2%.

But 120 transactions across the entire district is genuinely thin.

A single unusual transaction cluster could move that figure materially. D26 belongs on the list as a signal worth monitoring but not as a confirmed second-place result. If D20 and D18 are the districts with strong evidence behind them, D26 is the one to watch as more transactions accumulate.

What the Re-Run Actually Shows

The original heartland-beats-prime finding was not wrong. The top private districts in Singapore — once you look at what private buyers can actually access — are still OCR and RCR. Bishan straddles both. Tampines, Hougang, and Sengkang are OCR. The geography of strong returns has not changed.

What has changed is the identity of the leader. Woodlands and Sembawang, which topped the original ranking, were running partly on EC performance that most private buyers could never access. Strip that out, and both drop to mid-table at 3.4–3.7% per year. The district that rises to first — Bishan/AMK/Thomson at 4.7% — never had a significant EC presence. Its result was always a private-buyer number. The ranking did not change because ECs are bad investments. They are not. It changed because the original ranking was, in those three districts, measuring something with a different buyer pool, a different product category, and a different return profile.

For the private buyer who read the original piece and had Woodlands or Sembawang on their shortlist: the case for those districts was partly built on numbers you were not eligible to produce. For Bishan, it was not.