Is Thomson Reserve a good investment?

Is Thomson Reserve a good investment?

On the evidence available in July 2026, Thomson Reserve reads as a sound long-hold family home in a location with a genuine track record, not as a high-return investment. Its two nearest tracked neighbours, Thomson Impressions and Thomson Three, delivered median annualised gains of 2.8% and 2.9% on profitable exits held five to seven years. The District 20 median over the same holding band was 3.87%, and the all-Singapore median 3.73%. The dollar profits quoted around this launch are real. They are also, divided by the years it took to earn them, unremarkable.

The 1,268-unit project at Bright Hill Drive, previewing in October 2026 (est.), arrives with a specific argument attached: neighbours have compounded, the school catchment is elite, the land was bought cheap. Each part is defensible. Each looks different once you widen the comparison set.

What we tested, and against what

Two datasets do the work. Our price trends compilation holds 32,414 profitable resale exits across 199 completed Singapore condominiums, for sales lodged between January 2018 and April 2026. Our land sales dataset holds 256 URA and HDB tenders and collective sales since 2016, and powers the land cost tracker.

One property of the first dataset governs everything below, so it belongs up front. It contains profitable exits only, a deliberate choice explained on Price Trends. Every number here describes the size of gains that were made, never the probability of making one. When a case-study slide shows three neighbouring projects and every transaction is a gain, that is partly a fact about the neighbourhood and partly a fact about which transactions get shown.

One note on the set: Bishan Point, often cited alongside the two Thomson projects, is not among our 199 developments. Where an older District 20 leasehold comparator helps we use The Gardens at Bishan (1997, 764 units), the closest match we hold.

What the neighbouring projects actually returned

Across every profitable exit in our window, Thomson Impressions produced a median gain of $198,000, or 18.8% gross, over a median 6.4-year hold. Thomson Three produced $346,000, or 26.9%, over 8.4 years. Both are healthy. Both sit, annualised, near the middle of the national distribution rather than above it.

District 20 as a whole beat the national median by about 60 basis points a year, which supports the broad locational argument. But the two projects nearest the site both sat below their own district's median, and Thomson Grand well below at 1.89%. The district is strong; these particular buildings are not the reason why.

Why the five-to-seven-year window is the weakest part of the case

The holding period commonly used to frame this launch is five to seven years, on the reasonable grounds that it clears Seller's Stamp Duty and matches a typical upgrade cycle. It is a sensible planning horizon. It is not, on our data, the horizon that produced the returns being quoted.

Filter to exits held between five and seven years and the picture dims. Thomson Impressions falls from 3.18% a year to 2.79% across 42 exits; Thomson Three from 3.54% to 2.88% across 21; Thomson Grand collapses to 0.64%. The District 20 benchmark for that band is 3.87% across 327 exits.

The Gardens at Bishan is the interesting row: a median 12.8-year hold and 7.63% a year, the strongest figure in the comparison. A near-30-year-old leasehold outperforming two much newer neighbours is not a story about the buildings but about entry price and time. Those owners bought before the estate matured and held through more than one cycle, and the lease decay that should have worked against them was outweighed by two decades of location repricing.

That is the honest version of the long-term argument for this address, and it is stronger than the five-to-seven-year version. It carries a condition: it needs a decade or more, which is a different commitment from an upgrade timed to the end of the SSD period. Whether you can hold that long is a financing question before it is a market question, so model it in the affordability calculator first.

Is $1,178 psf ppr really the lowest RCR land cost?

The land rate is correct and the arithmetic is sound. The former Thomson View Condominium, a 255-unit 1987 development on a 504,314 sq ft site with a 2.1 plot ratio, sold collectively for $810 million to a UOL, SingLand and CapitaLand Development joint venture. That is $1,178 psf ppr once the land betterment charge and the lease upgrading premium for a fresh 99-year term are included. The High Court granted the sale order on 1 July 2025, after a stop order earlier that year.

The claim built on top of that number needs a boundary. Among the six upcoming Rest of Central Region sites in our pipeline, $1,178 is indeed the cheapest, by a clear margin. Against the RCR market as it stands, including projects selling now, it ranks ninth of 28.

Artisan 8 deserves a second look: a 34-unit project in the same District 20, secured in April 2024 at $1,059 psf ppr and selling at an average of $2,361 psf. Thomson Reserve's land is 11% dearer and our indicative range sits above Artisan 8's current average. Whatever else $1,178 is, it is not a discount to the district.

The same care applies to the two Outside Central Region comparisons, Lentor Central (Plot 4) at $1,278 and Hougang Central at $1,179. Both really are dearer per square foot of plot ratio than Thomson Reserve, which is an unusual result for a city-fringe site. It is also a two-point sample of unlaunched GLS plots: Lentor Gardens Residences was awarded at $920 psf ppr in April 2025, as we set out in our analysis of Lentor's first profitable sub-sales. The pipeline is expensive; the market is not uniformly so.

What the land cost implies about launch pricing

Land cost is the largest single input into a launch price, which is why a cheap site matters, and also why a cheap site rarely becomes a cheap flat. Developers price off a multiple that absorbs construction, financing, marketing and margin. Across the 53 launched projects in our land cost tracker that multiple has a median of 2.12×; for the 22 RCR projects in the set it is 2.11×.

Apply 2.12× to $1,178 and you get roughly $2,500 psf. Our indicative range is $2,450 to $2,700 psf (est.), from the model in our pricing methodology. The land advantage, then, shows up mostly as a lower launch price than the neighbouring pipeline, not as headroom accruing to the buyer after purchase. That is the part that gets lost when a land-cost chart is presented as an upside chart.

None of this is developer-confirmed; pricing is expected at the October 2026 preview. And 1,268 units is a large release into one district, cutting both ways: negotiating room on the weaker stacks now, real resale supply around 2035 when the SSD-free cohort turns over. Every awarded site and indicative range we hold sits in the GLS pipeline tracker.

Does the Ai Tong School catchment add value?

Thomson Reserve sits about 0.57km from Ai Tong School by straight-line measure, well inside the 1km band. Ai Tong is a Special Assistance Plan school and one of 71 our dataset rates very-high-demand, the top of a four-tier scale computed from the last completed Primary 1 exercise. On the catchment claim, marketing and data agree.

The 2025 exercise is worth decoding rather than summarising, because the detail changes what the catchment is worth. Ai Tong balloted at Phases 2A, 2B and 2C. The Phase 2C ballot reached the Singapore Citizen band within 1km, with 76 applicants for 43 places; Phase 2B was tighter at 62 for 20, also balloting within 1km. Living inside 1km buys entry to a ballot that roughly four in ten within-1km 2C applicants lost. It does not buy a place, and treating the two as equivalent overstates a real advantage. Beyond 1km there was no practical route in at 2C that year. Two other very-high-demand SAP schools sit outside the ring: Catholic High at 1.56km and CHIJ St. Nicholas Girls' at 2.0km.

Then the price question, where our own research cuts against the sales argument. We mapped all 182 MOE primary schools against 107 tracked launches for the 1km question. In the OCR, projects within 1km of a very-high-demand school carry a median average price about 4.6% higher. In the RCR the relationship inverts, pricing about 4.3% lower. Thomson Reserve is an RCR project. What holds up across both regions is buyer mix, not price: a 7.2 percentage-point higher HDB upgrader share. A top catchment reliably changes who your neighbours are and how liquid the resale pool is; it does not reliably add a premium, and here the measured sign is negative. Check the rings and tiers for any tracked launch on the primary schools map.

Distances here and across the site are straight-line estimates from OneMap postal-code geocodes, not the official MOE home-school distance, and we make no guarantee of accuracy. Verify any address with the OneMap School Query service before relying on it.

What this analysis does not show

Four limits, all load-bearing.

The dataset only contains winners. We can tell you what profitable sellers at Thomson Impressions made; we cannot tell you what share of sellers were profitable, because loss-making caveats are excluded by design. Read every median here as "what worked", never as "what happens".

The project-level samples are small once you filter. Twenty-one five-to-seven-year exits at Thomson Three is a thin base for a median. It is the base the neighbourhood argument rests on too, which is rather the point, but the gap to the District 20 benchmark is a signal, not a settled fact. The 327-exit district figure is the sturdier number on that table.

All gains are gross. Buyer's stamp duty, mortgage interest, agent commission, legal fees, maintenance and any Seller's Stamp Duty are excluded. Over a five-year hold, interest alone typically eats a large share of a 2.8% gross return. Net them down before using them as a forecast, and note that purchases from 4 July 2025 carry the reinstated four-year SSD schedule of 16%, 12%, 8% and 4%.

Past performance next door is not a forecast. The owners in this dataset bought at 2013 to 2019 prices. A buyer at an indicative $2,450 to $2,700 psf (est.) enters above both neighbours' current averages of $1,988 and $2,287 psf. These returns describe a repricing that has largely already happened.

Where could this reading be wrong? The Cross Island Line. Bright Hill becomes a TEL and CRL interchange, and interchange status has repriced other estates in ways these caveats cannot capture, because they pre-date it. If the CRL lifts this corridor the way earlier lines lifted theirs, the 2.8% band understates what a 2026 buyer might see. That is a real argument for the project, and it is not in our numbers because it has not happened yet.

What it means for buyers

Buying to live there for a decade or more: the case is reasonable, and its best evidence is The Gardens at Bishan at 7.63% a year over a median 12.8 years, not the five-year slides. Mature amenity, three MRT stations and a tier-4 catchment are worth paying for if you will be there to use them.

An HDB upgrader in Bishan or Ang Mo Kio: 1,268 units is one of the largest single releases the district has seen, and it is aimed squarely at you. Sequencing matters more than the launch-day price. Bridging, ABSD remission timing and MOP alignment are worked through in our sell HDB, buy new launch guide, with borrowing limits in the TDSR guide.

Investing on a five-to-seven-year view: our data does not support the expectation that framing implies. The nearest comparables cleared about 2.8% a year gross in that band, before costs. If your model needs more, it needs a longer hold or a different project. Line the launch up against the rest of the pipeline in the new launch comparison tool.

Weighing resale instead: Thomson Impressions and Thomson Three average $1,988 and $2,287 psf in our dataset, and Artisan 8 sells at $2,361 psf on cheaper land in the same district. The new launch price dataset holds the bedroom-level detail. Our full project profile is at the Thomson Reserve deep dive; every launch we follow is under upcoming launches.

Frequently asked questions

Sources & methodology

Resale exit figures are computed from the PropertyInsider.sg price trends compilation: 32,414 profitable resale transactions across 199 completed Singapore condominiums, derived from URA caveat data for sales lodged between January 2018 and April 2026 (dataset generated 8 July 2026). It includes profitable exits only; the selection effect is disclosed on Price Trends. Medians are calculated column by column, so the annualised median is the median of annualised rates. The five-to-seven-year filter selects holding periods of 5.0 to 7.0 years inclusive.

Land rates, unit counts, tenure and status come from the PropertyInsider.sg land sales dataset of 256 URA and HDB tenders and collective sales since 2016, cross-checked against URA tender records, and from our project dataset (generated 24 July 2026). The 2.12× median land-to-launch multiple across 53 launched projects, and the 2.11× RCR sub-median, are computed from those sources and published in the land cost tracker. The $2,450 to $2,700 psf indicative range is a PropertyInsider estimate from pricing model v2, documented at propertyinsider.sg/research/pricing-methodology, and is not developer-confirmed.

Thomson View collective sale details, the $810 million price, the $1,178 psf ppr derivation inclusive of land betterment charge and lease upgrading premium, the 504,314 sq ft site area and 2.1 plot ratio, and the 1 July 2025 High Court sale order follow contemporaneous reporting and the developers' disclosures; the 1,268 unit count and October 2026 (est.) preview follow UOL Group's February 2026 results presentation as reported. Primary 1 figures come from the PropertyInsider.sg schools dataset (182 MOE P1-admitting schools, generated 19 July 2026), built from MOE General Information of Schools via data.gov.sg, MOE P1 vacancy and balloting releases for 2025, and OneMap geocodes. Seller's Stamp Duty rates follow the MND and MAS announcement of 3 July 2025.

Back to all research