The yield map is upside down. Cheap districts pay. Prime doesn't.
Ask a new investor where rents are strong and they'll say Orchard. The data says Orchard is where yields go to die — 2.9% — while Woodlands and the CBD fringe clear 4%. The map, the reason, and which side of it your money belongs on.
Farhan Adenan · 2 Aug 2026 · 9 min read
- Median gross yields by district run 2.45% (Sentosa's D06-adjacent core) to 4.26% (D02, Tanjong Pagar).
- The top of the map: D02 4.26 · D25 4.01 · D08 4.00 · D14 3.92 — fringe and heartland, not prime.
- The bottom: D10 2.62 · D20 2.76 · D21 2.77 · D09 2.91 — the prestige districts.
- The rule that falls out: rents follow people, prices follow prestige. The gap between those two is your yield.
Median gross yield per district: annualised rental contracts ÷ transacted values, matched at project level, URA data to 15 Jul 2026.
The full map
Twenty-seven districts, ranked by what they actually pay.
Exhibit 1. Median gross yield by district — annualised median rent over median price. The spread from top to bottom is about 1.6×, and it runs almost exactly opposite to how the same districts are ranked by prestige. Source: URA private residential caveats to 23 Jun 2026 (REALIS); POV matched-pair exit model. POV analysis.
| District | Median gross yield | Why |
|---|---|---|
| D02 Tanjong Pagar / CBD | 4.26% | Prices fell, rents didn't — see the paradox |
| D25 Woodlands | 4.01% | RTS corridor + int'l school tenants |
| D08 Little India / Farrer Pk | 4.00% | Medical cluster + city-fringe rents |
| D14 Geylang / Eunos | 3.92% | The address discount pays you monthly |
| D07 Bugis / Beach Rd | 3.71% | Compact CBD-fringe stock |
| D22 Jurong | 3.65% | Industrial + business-park tenancies |
| D27 Yishun / Sembawang | 3.56% | North value, real tenant depth |
| D04 Harbourfront | 3.55% | High rents rescue weak capital story |
| D18 Tampines / Pasir Ris | 3.53% | Heartland workhorse |
| D05 Clementi / West Coast | 3.51% | Universities + one-north |
| … | … | Mid-table: D12 3.35 · D19 3.31 · D23 3.28 · D15 3.06 |
| D09 Orchard | 2.91% | Prestige premium, tenant ceiling |
| D26 Upp Thomson | 2.83% | Low-density, few rentals |
| D11 Newton / Novena | 2.81% | Prime pricing, ordinary rents |
| D21 Upper Bt Timah | 2.77% | Owner-occupier land |
| D20 Bishan / AMK | 2.76% | School premium in price, not rent |
| D10 Holland / Bt Timah | 2.62% | The most expensive yield desert |
Gross, before maintenance/tax/vacancy. Net figures typically run 0.8–1.2 points lower.
The paradox at both ends
Both ends of the map are counter-intuitive.
Yield is a fraction, and Singapore forgets the denominator. D02 tops the map not because CBD rents boomed but because CBD prices deflated while rents held — our exits report card shows D02 sellers were profitable only 65% of the time. High yield there is the market's compensation for weak capital growth. Mirror-image: D10's 2.62% exists because prices carry a prestige premium tenants refuse to pay. A tenant rents the unit; a buyer buys the address. The gap between what those two people will pay is the whole map.
The genuinely rare squares are the double-positives — districts where yield AND exit records both rank top-ten: D25, D27, D18, D22, D05. Solid income (3.5–4.0%) sitting on 89–95% profitable exit histories. Nothing glamorous ever gets written about Woodlands, which is precisely why its numbers look like this.
Choosing your side
Income or prestige. Very few postcodes give you both.
If the money must pay you monthly — retirement income, mortgage offset — shop the top: D25/D08/D14/D22, plus D02 only if you accept its capital-flatness as the deal. Run net numbers (deduct ~1 point) and stress a vacancy month.
If the money is planted for a decade — growth first — the mid-table conveyor districts (D19, D18, D23, D27) offer 3.3–3.6% while you wait on the strongest exit records in the country. That combination, not Orchard, is historically Singapore's best total-return seat.
If you're set on prime — buy it knowing the rent won't carry it. Prime works when bought in downcycles and sold in upcycles; the yield map is simply the carrying cost of waiting. 2.6% gross means the property must appreciate ~2% a year just to match the boring districts' income head start.
The prestige subsidy, quantified
The prestige subsidy, quantified.
Exhibit 2. Gross annual rent on an identical $1.5 million outlay, at each district’s median yield. Choosing D10 over D02 costs $24,600 a year in forgone gross rent — roughly $2,050 a month, paid for the address. That is the subsidy, and it is a legitimate thing to buy as long as you know you are buying it. Source: URA private residential caveats to 23 Jun 2026 (REALIS); POV matched-pair exit model. POV analysis.
Run the numbers on the gap: D08 grosses 4%+ while D10 grosses 2.6%. On a $2 million property, that’s roughly $28,000 a year of income the D10 owner forgoes — an annual fee for the address, paid to no one, deductible from nothing. The tenant gets the Bukit Timah life; the landlord gets to mention it at dinner. Who’s subsidising whom?
The old defence was capital gains — and the regime took it out back. With optional capital taxed to the sidelines since April 2023, prime’s price engine idles, but its rents never rose to compensate. That leaves the fringe holding the only honest return in the market: cash, monthly, banked. Prices can tell stories. Rent clears.
Who this affects
Prestige and income are close to mutually exclusive here.
If you own
If you are buying to live in it, the yield map is your opportunity cost
Nobody chooses a family home on gross yield, and they should not. But the number tells you what the address is costing you annually, and $2,050 a month is a specific figure rather than a vague trade-off.
Knowing it does not change the decision for most households. It changes how honestly the decision is described, which matters when someone later calls the same purchase an investment.
If you invest
If you are buying for income, the map is upside down and that is the point
The districts that pay are the ones nobody puts on a brochure: D02 at 4.26%, D14 at 3.92%, D22 at 3.65%. The tenant pool there is deep, unglamorous and durable.
Two cautions. Yields this size usually come with small units, which means a thinner resale bid and a faster-moving tenant. And if this is a second property, 20% ABSD and Buyer’s Stamp Duty sit in the denominator whether you put them there or not — recompute the yield against what actually left your account, not the price on the listing.
Rents follow people. Prices follow prestige. Invest in the gap.
- Income buyers: D25 · D08 · D14 · D22 — and D02 with eyes open about why it pays.
- Total-return buyers: the conveyor belt — D19 · D18 · D27 — 3.3–3.6% yield on 94–95% exit records.
- Prime buyers: the 2.6% yield is your countdown timer — the appreciation has to show up, and soon.
Which square of the map fits your money?
Tell me the job the money has — monthly income, decade growth, or parking wealth — and I'll shortlist the three districts and five projects that actually do that job, with net numbers.
The 4% club: 99 projects where rental yields are actually real
Averages say two-something. The data says 99 projects clear 4% — and every high-yield cluster has a catch.
The 2.3× rule: developers price your condo years before it exists
44 launches, one formula — and what it already implies for Hougang, Bedok and Lentor.
How to check us: every number in this piece is computed from the primary record — URA caveats to 15 Jul 2026 — not from third-party estimates or hearsay. The links below are the official policy and news record behind the contextual claims.
- URA private residential transaction data (REALIS) — the caveat record every table in this article is computed from
- MAS · MOF · MND — "Measures for a Sustainable Property Market" (26 Apr 2023) — the official announcement lifting foreigner ABSD to 60%
Dataset — Median gross yield per district: annualised rental contracts ÷ transacted values, matched at project level, URA data to 15 Jul 2026.
Methodology & honesty notes. District medians computed from matched rental-contract and sale-caveat data per project (URA, to 15 Jul 2026), minimum sample floors per district; thin districts (D06 n=1, D26 n=10) are indicative. Gross yields exclude maintenance, property tax, agent fees and vacancy — net typically runs 0.8–1.2 points lower, more for small units. Cross-references: exit records from POV's matched-pair model (see "Who actually loses money"). POV Realty and Farhan Adenan are not the marketing agents for any project or listing referenced, and nothing here is financial advice — it's a starting point for your own due diligence, which is exactly how we'd use it.
Farhan Adenan · CEA Registration R068636D · Senior Associate Division Director, Huttons Asia Pte Ltd (Estate Agent Licence L3008899K).