The six grades, tested instead of worshipped
The companion walkthrough gives six sites six letters. This written guide keeps those letters exactly as stated, then asks what the available evidence can actually carry. That is more useful than pretending a grade is a prediction. A letter cannot make a school run shorter, tell you which stack will face construction, or remove a condition attached to the final sales exercise.
Read the grade first, then the limit beside it. The market packet separates HDB four-room resale, private sales and private rents because they are different populations. It calls a named proxy a proxy, calls thin evidence thin, and does not manufacture a rental yield by dividing one unrelated average by another.
| Site | Exact grade | What this article can test | What it cannot promise |
|---|---|---|---|
| Toa Payoh / Caldecott | A | Mature-town context and named-comparable HDB evidence | A future block-level resale result |
| Bedok / Bayshore | B | Exact private-market observations and named HDB proxies | A water view or HDB premium |
| Tengah | B | The absence of HDB resale history and exact private observations | Proven BTO upside |
| Geylang / Mattar | C | Exact private-market cross-checks and named HDB proxies | A site result from a town trend |
| Yishun / Chencharu | D | The gap between Yishun and Canberra proxies | Exact Chencharu market history |
| Sembawang North | F | Near-area Sembawang/Canberra proxies | A verdict on a future precinct |
Four checks before you ballot
Use this as a document-checking list, not a scorecard. First, open the HDB project page and sales brochure: confirm the classification, flat mix, estimated completion and site plan for the exact launch. Second, trace the actual route for work, caregiving, school and the MRT or bus connection. “Near” on a map is not the same as a workable weekday. Third, check orientation and nearby construction for the stack you would accept. Fourth, treat past application rates as context, not as an odds calculator.
The evidence has three deliberately separate lanes: HDB four-room resale by named street (February 2025 to June 2026), positive-area private sales, and selected private rental strata (both January 2025 to June 2026). A named street can be EXACT, a NEAREST COMPARABLE, or NO DATA. Those labels stop a tempting but false shortcut: no HDB proxy becomes an exact site price, no private condominium becomes an HDB forecast, and no sale-and-rent pair becomes a yield claim.
For project rules and timing, read the official HDB Caldecott/Toa Payoh announcement, HDB conditions after buying a new flat, and LTA Jurong Region Line page. Final project pages and brochures remain the authority for classification, stack, completion and conditions.
Approximate locations only, not an official site plan. Map: OneMap, Singapore Land Authority.
Toa Payoh / Caldecott — Grade A
Toa Payoh / Caldecott is Grade A. The reason is the established-town case, not a claim that every future Caldecott flat will outperform. HDB has confirmed an October 2026 project next to Caldecott MRT, while the existing town evidence gives this location the strongest mature-context starting point in the six-site pack. That is a reason to read the documents carefully, not to skip them.
Toa Payoh / Caldecott: real neighbourhood map with nearest MRT and key amenities. Approximate site area only, not an official site plan. Map: OneMap, © Singapore Land Authority.
The fresh HDB packet is deliberately more awkward than a neat headline. Its nearest comparables are Bidadari Park Drive, Lorong 1 and 1A Toa Payoh, and Toa Payoh North—not Caldecott itself. Bidadari Park Drive has 120 four-room transactions, with a SGD 1,109,454 mean and 2020–2021 lease-commence stock; Toa Payoh North has 22 transactions, a SGD 505,722 mean, and 1973 stock. Those are named proxies with different lease cohorts, not one “Caldecott price.”
The exact-locality private check does not rescue that gap. Caldecott Close has no positive-area private-sale row in the selected January 2025 to June 2026 window. The rental check has only three selected-window records across incompatible detached-house area bands; its compact stratum is one 250–300 sqm detached rental at SGD 15,500 in January 2026. One record is not a local rent conclusion, and neither private rent nor private sale predicts an HDB BTO outcome.
The counterweight is the project rulebook. HDB says the final classification will be announced in the October exercise. Do not label this project Prime, Plus or Standard before that material exists. If Plus or Prime conditions apply, the household has to read the stated 10-year MOP and project-specific subsidy recovery rather than treating a central address as a free upgrade.
There is a second, less glamorous test: compare the option against the household’s holding period and cash-flow tolerance. A ten-year MOP, if applicable, is a lived constraint, not a footnote to a location. The evidence packet does not price that constraint for you, so a grade should not conceal it.
For a household, Grade A means “start here if mature-town access is your non-negotiable,” not “bid blindly because the chart looks strong.” This verdict would change if the final classification, flat mix, completion estimate, site layout, or the conditions attached to the project make the daily-life trade too expensive. The sales brochure and project page—not a town proxy—decide that question.
Bedok / Bayshore — Grade B
Bedok / Bayshore is Grade B. The grade recognises an East Coast context with real exact private-market observations, but it refuses the easy “waterfront equals guaranteed premium” story. Bedok’s town series in the earlier research stayed positive but slowed, and a town line cannot tell a household what a particular future stack will see, hear or pay for.
Bedok / Bayshore: real neighbourhood map with nearest MRT and key amenities. Approximate site area only, not an official site plan. Map: OneMap, © Singapore Land Authority.
The HDB evidence is still a proxy. Bedok South Avenue 2 has 29 four-room records with a SGD 485,213 mean, while Bedok South Road has 49 records with a SGD 690,417 mean; the latter mixes 1976–2022 lease cohorts. Neither named street is Bayshore, so neither number should be called a Bayshore HDB price. The point is variation inside a mature area, not a price target.
Private sale evidence is exact to Bayshore Road or Walk but remains private stock. Bayshore Park records 37 condominium sales at a SGD 1,415,447 mean, Costa Del Sol 52 at SGD 2,685,114, and Vela Bay 371 apartment sales at SGD 2,354,507 in its April–June 2026 window. The project types, sizes and sale windows differ. These figures describe specific private projects; they do not forecast an HDB BTO resale gain.
The rental observation is also narrow: Costa Del Sol’s non-landed 120–130 sqm, three-bedroom stratum has 68 rentals with a SGD 5,929 mean from January 2025 to June 2026. It is an exact Bayshore Road stratum, but it is not a yield input and not a promise about an HDB household’s future rent. A view story is incomplete until the actual site plan, orientation and nearby construction are visible.
That distinction matters because a coastal address can bundle several separate questions: access to the water, exposure, road noise, facing, construction around the site and daily transport. The figures here answer none of those stack-level questions. They simply show that there is enough exact private context to investigate without pretending it settles suitability.
For a household, Grade B means “worth investigating if East Coast access is the attraction, with the site plan doing the hard work.” It would change if the final brochure shows an orientation, surrounding development, flat mix or conditions that do not fit your routine. The useful next move is to trace the unit facing and daily route, not to convert a condo average into a ballot instruction.
Tengah — Grade B
Tengah is Grade B. This is a developing-town grade, not an established-resale grade. The packet finds no TENGAH HDB four-room resale rows at all. That absence is the honest finding: a new town has not yet built the normal post-MOP HDB resale history that would let this article test a conventional local HDB track record.
Tengah: real neighbourhood map with nearest MRT and key amenities. Approximate site area only, not an official site plan. Map: OneMap, © Singapore Land Authority.
Exact private evidence exists, but it answers a different question. Tengah Garden Residences has 676 positive-area apartment sales from April to June 2026, with a SGD 1,946,102 mean and SGD 22,800 mean per sqm. That is a substantial project-specific private observation, not evidence that a future HDB flat will follow the same path. Private and HDB ownership, eligibility and supply conditions are not interchangeable.
The exact rental signal is too thin to lean on. The selected Tengah Garden Walk Copen Grand stratum has one 140–150 sqm, five-bedroom EC rental at SGD 6,000 in December 2025. One rental record cannot describe a rental market, and combining it with the sale average to suggest yield would be false precision.
Transport timing deserves the same discipline. LTA lists Jurong Region Line Stage 2 for 2028 and a later Tengah in-fill station in the mid-2030s. That is useful context, but the final launch material must identify the station and access relevant to the actual project. “Near a future line” is not a commute plan until you map the walk, the opening date and the rest of the journey.
The missing HDB history also changes how to read rosy comparisons with mature towns. There is no like-for-like resale series in this packet that can confirm or reject the future outcome. A household can still choose Tengah for its own reasons, but it should name that choice honestly as a bet on fit and time, rather than evidence already earned.
For a household, Grade B means “a credible developing-town choice if you accept a longer evidence gap.” It would change if the final completion estimate, access plan, flat mix or your own timing makes that wait unworkable. The right question is whether you want this proposition despite missing HDB resale history—not whether missing history can be renamed proven upside.
Geylang / Mattar — Grade C
Geylang / Mattar is Grade C. It is neither a write-off nor a secret bargain hidden inside a single number. The earlier pack gives Mattar a mature-neighbourhood and Downtown Line context, while Geylang’s earlier town-level resale comparison was weaker than several other sites. A weaker town comparison is a prompt to investigate, not a verdict on a home that has not been built.
Geylang / Mattar: real neighbourhood map with nearest MRT and key amenities. Approximate site area only, not an official site plan. Map: OneMap, © Singapore Land Authority.
The HDB rows are named proxies, not Mattar. Circuit Road has 74 four-room transactions with a SGD 986,768 mean and a mixed 1967–2020 lease cohort; Pine Close has 13 at a SGD 818,069 mean with 1981–2000 stock. The difference itself shows why a town label is too blunt. Neither record establishes the future price of a Mattar BTO.
Mattar Road does have an exact private-sale cross-check: The Antares recorded 26 apartment sales from February 2025 to May 2026, with a SGD 1,508,453 mean, SGD 1,426,000 median and SGD 22,157 mean per sqm. It is evidence about that private project and its selected window, not a hidden comparable for public housing.
The exact rental stratum is similarly specific. The Antares, non-landed 60–70 sqm two-bedroom segment, has 62 rentals with a SGD 4,318 mean and SGD 4,300 median from January 2025 to June 2026. It cannot be paired with the sale row into a yield claim because the rows describe different observations and do not price an HDB BTO.
There is useful friction in that answer. A location can have a recognisable rail connection and still require a household to decide whether the transfer, walking route, caregiving route and flat layout work every week. Neither a town trend nor a condo stratum carries those facts. That is why the grade stays an interrogation, not a conclusion.
For a household, Grade C means “put the daily route, layout and final conditions ahead of a generic Geylang story.” It would change if the sales material and your own commute check show a genuinely better or worse fit than the town proxy suggests. The practical task is to test the exact journey and site plan, then decide whether the mature-area trade works for your household.
Yishun / Chencharu — Grade D
Yishun / Chencharu is Grade D. The grade should be read with its geographic warning attached. Historical nearby or precinct application signals were softer than stronger-demand comparators, which may matter to a household whose first problem is ballot chance. It does not turn historical application context into a future probability, and it does not make Chencharu the same place as every available proxy.
Yishun / Chencharu: real neighbourhood map with nearest MRT and key amenities. Approximate site area only, not an official site plan. Map: OneMap, © Singapore Land Authority.
The HDB rows are nearest comparables in Yishun: Yishun Street 81 has 47 four-room transactions with a SGD 578,518 mean and 1987–1988 lease-commence stock, with a SGD 593,000 median running higher than the mean — a reminder that even one named street’s 47 transactions do not cluster around a single number. That is a named Yishun proxy, not a Chencharu price. It can help a reader see the kind of established stock in the wider discussion, but it cannot stand in for a new project’s location, lease or layout.
The private evidence is even more constrained. The packet offers Canberra Crescent Residences and The Watergardens at Canberra as nearest comparables, and it says plainly that Canberra is administratively Sembawang, not Chencharu or Yishun. Canberra Crescent Residences records a SGD 1,739,009 mean and SGD 1,630,200 median across 307 sales; The Watergardens at Canberra averages SGD 1,591,388 with a SGD 1,590,000 median across its 74. Those apartment-sale observations are therefore not exact Chencharu evidence. Calling them Chencharu sales would be laundering a cross-town proxy.
The rental rows retain the same limit: Canberra Drive’s Visionaire three-bedroom EC stratum has 30 rentals at a SGD 4,043 mean (SGD 4,000 median), while Watergardens’ 60–70 sqm two-bedroom stratum has 46 at a SGD 3,257 mean (SGD 3,300 median). Those are Canberra observations. They do not establish Chencharu rent, yield or appreciation.
This is exactly where a label can become unfair. A softer historical application signal may be relevant to the ballot conversation, but it does not measure future schools, household routines, amenities or a particular stack. The packet offers no clean historical Chencharu private/HDB match, so the right response is an explicit evidence gap—not an invented story about demand.
For a household, Grade D means “examine the ballot and liveability trade with more scepticism, not with contempt.” It would change if the final brochure, amenity timeline, completion estimate or your actual route makes the project fit better than the proxy picture suggests. Use the official material to test the new place on its own terms; do not let an administratively different neighbourhood answer for it.
Sembawang North — Grade F
Sembawang North is Grade F. That is the exact grade from the companion work, and it needs the most careful reading because an F can easily sound like an order. The earlier research records softer historical nearby or precinct application signals than stronger-demand comparators. That may create a different ballot conversation; it is not evidence that a future home will be bad, cheap, inaccessible or impossible to resell.
Sembawang North: real neighbourhood map with key nearby amenities — no MRT station falls within this map's frame. Approximate site area only, not an official site plan. Map: OneMap, © Singapore Land Authority.
The HDB side is near-area evidence. Canberra Crescent has 75 four-room transactions with a SGD 739,615 mean (SGD 736,800 median) and 2018–2020 lease-commence stock; Admiralty Drive has 44 at a SGD 558,808 mean (SGD 557,500 median) with 2001–2002 stock. Both are named Sembawang-area comparables, not a historical Sembawang North BTO street. The two lease-commence bands sit nearly two decades apart, which is why their differing cohorts must not become a single promised average.
Private sales are also nearest comparables. Canberra Crescent Residences has 307 apartment observations at a SGD 1,739,009 mean (SGD 21,363 mean per sqm), collected from August 2025 to June 2026, and The Watergardens at Canberra has 74 at a SGD 1,591,388 mean (SGD 18,538 per sqm) over the fuller January 2025 to June 2026 window. These are Canberra project figures, not exact Sembawang North evidence. They are context only, with no claim that private prices predict public-housing performance.
The rental cross-check stays narrow: Visionaire’s selected three-bedroom EC stratum has 30 rentals averaging SGD 4,043 (SGD 4,000 median), and Watergardens’ selected two-bedroom stratum has 46 averaging SGD 3,257 (SGD 3,300 median). These cannot be made into a Sembawang North yield, because they are project-specific Canberra rents and the plan has no matching like-for-like price denominator.
The hard question is whether the grade is reacting to evidence or merely to unfamiliarity. Here, it is a cautious demand signal plus incomplete comparables, not a claim that the future precinct lacks value. The evidence cannot see your workplace, family support, preferred flat type or daily travel pattern. Those can outweigh a broad market label.
For a household, Grade F means “challenge the proposition hardest before you commit”—especially the bus/MRT connection, amenities sequence, site layout and your own family schedule. It would change if those final, project-specific facts deliver a fit that the broad proxies cannot see. An F is a demand for more checking, not permission to make a sweeping claim about people who choose the area.