Gary Lim · ERA Senior Division Director · CEA R009877B · BuyCondo Team
Published 17 September 2026 · 17+ years, 500+ transactions

The short answer on HDB resale volume by town

A seller asked me this in July, in a Serangoon kitchen, and I did not have a good answer. She wanted to know whether her town was a busy market or a dead one. I gave her the reply every agent gives — “demand is healthy, don’t worry” — and I was annoyed with myself in the car afterwards, because it was a feeling, not a number.

So I went and got the numbers. This article is the complete picture of HDB resale volume by town: every resale transaction Singapore has registered since January 2017, 238,162 of them, sorted into all 26 HDB towns, divided by how many flats each town actually contains.

The headline is easy. Sengkang sells the most flats. Bukit Timah sells the fewest. If you have read a property blog this year, you already knew something like that.

The useful finding is the one that took me three days to see, and it is this: that ranking tells you almost nothing. Sengkang sells 33 times more flats than Bukit Timah because Sengkang has 27 times more flats. Once you divide by stock, the entire country collapses into a band 1.3 percentage points wide, and 23 of the 26 towns sit inside it.

Which means the honest answer to my Serangoon seller was: your town is not the problem. Very few towns are. 新加坡买房，就找对的团队 — and the right team tells you when the data says your worry is misplaced.

How I built these numbers (and why nobody else has)

Everything here comes from HDB’s own resale registration records, pulled record by record from the government’s resale flat prices registry — the same database that produces every median price HDB publishes. I did not take a single transaction count from a news article or a property portal.

Three decisions shaped the result, and you should know all three before you trust the tables.

1. A 116-month window, not twelve

Most volume articles use one year. A year of data in a town like Bukit Timah is 56 transactions — one quiet quarter swings it 20%. So I used January 2017 to August 2026: 116 complete months, 9.67 years. September 2026 is still being lodged, so I pulled each town’s September count separately and subtracted it. Nothing in these tables includes a partial month.

2. The denominator is sold flats, not all flats

To turn volume into a rate you need to divide by stock — and this is where everyone else stops, because the denominator is genuinely awkward to get. I used HDB’s own dwelling-unit table by town and flat type, and I took sold units only, excluding rental flats. A rental flat cannot be resold, so counting it as part of the market would quietly punish towns with large rental estates — Bukit Merah, Central Area, Ang Mo Kio. Nationally that is 1,032,699 sold units.

That table’s most recent financial year is 2021. Normally I would call that stale. Here it is close to ideal: FY2021 sits almost exactly at the midpoint of a January 2017 – August 2026 window, so it is the right average stock level for the period I am measuring. Where it does distort, it distorts in one direction — the fast-growing towns (Punggol, Sengkang, Tengah) have added flats since, so their true turnover is a little lower than I show. I flag that again where it matters.

3. Three cross-checks that had to pass

I do not publish a computed figure until something independent agrees with it.

• My registry count for August 2026 is 2,521 transactions. The SRX/HDB flash estimate reported that month at 2,524. Three deals apart.
• My August 2026 figure is +14.4% year-on-year. The published flash figure was +14.1%.
• My Central Area total since 2017 reproduced an earlier independent pull of the same town exactly, to the transaction.

Good enough. Let’s look at the table.

The full 26-town HDB resale volume by town table

Sorted busiest to quietest. Per year is the 116-month average. Turnover is annual transactions as a share of sold stock, and 1 in N is the same number in plain English: how many flats you’d have to walk past before one of them sold this year.

Source: HDB resale registrations via data.gov.sg, Jan 2017 – Aug 2026, computed by the BuyCondo Team. Stock: HDB dwelling units by town, sold units only, FY2021. Blocks: HDB property information register, residential blocks only.

The top six towns — Sengkang, Woodlands, Punggol, Tampines, Yishun, Jurong West — account for 42.7% of every HDB resale transaction in Singapore. The bottom six account for 7.1%. If you sell in Bukit Timah, Marine Parade or the Central Area, you are operating in a market perhaps one-thirtieth the size of the one your Sengkang colleague works in.

Now here is where the standard article ends and the actual analysis starts.

The busiest-town ranking is a population ranking

Put each town’s annual resale volume next to its flat count and the relationship is almost a straight line. The correlation is r = +0.966, which means 93.2% of the variation in “which town sells the most flats” is explained by nothing more exotic than how many flats the town has.

That is not a market insight. That is arithmetic.

Look at the two extremes with this in mind. Sengkang does 1,975 resales a year and Bukit Timah does 60 — a 33-fold gap that sounds enormous. Sengkang has 69,113 sold flats and Bukit Timah has 2,554 — a 27-fold gap. Almost the entire difference is accounted for before you have said one word about demand, prices, MRT lines or school catchments.

I am not being clever at anyone’s expense here. I would have written the same article a year ago. The raw ranking is genuinely interesting if what you want to know is where the work is — and for an agent deciding which town to farm, 1,975 deals a year versus 60 is the whole business case. But it is the wrong number for a homeowner asking “will my flat sell?”

The number that actually matters: turnover

If you own a flat, you do not care how many flats sold in your town. You care what share of flats like yours sold — because that is your competition, and that is your buyer pool.

So: annual transactions divided by sold flats. Nationally that is 2.39%. One in every 42 HDB flats in Singapore changes hands in a given year.

Sit with that for a second, because it reframes a lot of seller anxiety. In a 100-unit block, roughly two to three flats sell in a year. Not two or three a month. Two or three a year. When someone tells me their flat has been listed six weeks with three viewings and something must be wrong, the base rate says: probably nothing is wrong. You are in a market where your block produces a couple of transactions annually, and you are trying to be one of them.

Turnover league table: 1 in 30 to 1 in 50

The same 26 towns, re-sorted by the rate rather than the count. I have added the share of each town’s blocks with under 60 years of lease remaining, because it turns out to be the column that explains the table.

Lease column: share of each town’s residential blocks with under 60 years remaining, computed from HDB’s block register. Full methodology is in our companion piece on HDB lease decay by town (hdb-lease-decay-by-town).

The ranking has been turned inside out. Tampines drops from 4th busiest to 9th. Bedok falls from 7th to 19th. Queenstown — a town nobody would call a dead market — sits dead last. And Bukit Timah, which is by a distance the quietest resale market in Singapore in absolute terms, comes in 10th, comfortably ahead of Bedok, Toa Payoh, Clementi and Ang Mo Kio.

Bukit Timah is not illiquid. It is small. Those are completely different problems and only one of them is yours to solve.

The 1.3-point problem: 23 of 26 towns are the same market

Here is the finding I did not expect, and the reason this article exists.

The entire national range of turnover runs from 1.99% (Queenstown) to 3.29% (Punggol). That is a spread of 1.30 percentage points, top to bottom, across every HDB town in the country. Twenty-three of the twenty-six sit between 2.0% and 2.7%.

Volume varies by a factor of 33. Turnover varies by a factor of 1.66 — and if you strip out Punggol and Queenstown at the two ends, by about 1.3.

I want to be careful about what I am and am not claiming. Turnover is not the same as speed. This data tells you the rate at which a town’s flats transact; it cannot tell you that a Punggol flat sells in 30 days and a Queenstown one in 90. HDB does not publish days-on-market and nobody can compute it from the registry, because a registration records the sale, not the listing. Anyone who quotes you a days-on-market figure by town is quoting an agency’s internal listing data, which covers their own listings only.

What turnover does tell you is how thick the market is around your flat. And on that measure, Singapore is remarkably uniform.

What actually moves turnover: how old the town is

If it is not size and it is not desirability, what produces the 1.3 points of spread that do exist?

Lease age. Set each town’s turnover against the share of its blocks with under 60 years of lease left and the correlation is r = −0.659 — town age explains 43% of the variation in turnover. Nothing else I tested came close.

The top five towns by turnover are Punggol, Sengkang, Woodlands, Yishun and Sembawang. Three of those five — Punggol, Sengkang, Sembawang — have zero blocks under 60 years of lease. Not a low share. Zero.

The bottom four are Queenstown (57.9% of blocks under 60 years), Serangoon (54.3%), Ang Mo Kio (85.5%) and Bukit Merah (67.6%).

Two mechanisms, pulling the same way

On the seller side, it is the MOP pipeline. A young town has thousands of flats crossing their five-year Minimum Occupation Period in a cluster, and a meaningful slice of those owners sell almost immediately — upgrading, relocating, cashing out. That is a supply engine an old town simply does not have. Punggol’s 3.29% is not a sign that Punggol is desperate; it is a sign that Punggol has been manufacturing first-time sellers on an industrial schedule.

On the buyer side, it is financing. CPF usage and the loan-to-value limit both depend on the remaining lease covering the youngest buyer to age 95. As a town’s leases shorten, the pool of buyers who can finance a purchase there without extra cash narrows — and it narrows from the young end first, because a 35-year-old needs 60 years of lease while a 55-year-old needs only 40. Fewer eligible buyers, fewer transactions.

Old towns lose supply and demand at the same time, which is why the effect shows up cleanly despite everything else going on.

Deals per block: publishing a metric that doesn’t work

There is a third way to measure HDB resale volume by town, and on paper it is the most seller-friendly of the lot: transactions per residential block per year. Your block is your real competition, after all. I computed it for all 26 towns — it is the last column of the main table — and then I had to conclude it is the weakest of the three.

Here is why. Blocks are not the same size.

A Queenstown block holds 2.1 times as many flats as a Pasir Ris block. So deals-per-block rewards towns with big slab blocks and punishes towns built as small point blocks, regardless of how liquid either actually is. Queenstown has the worst turnover in Singapore and the third-best deals-per-block, purely because its blocks are enormous.

The statistical tell is clean. Turnover correlates with town lease age at r = −0.659. Deals-per-block correlates with the same thing at r = −0.05 — which is to say, not at all. The block-size noise swamps the signal.

I am leaving the column in the table because it is a real number and some of you will want it, and because I would rather show you a metric I rejected than quietly bin it. But turnover is the one to read.

Pasir Ris: the town that breaks both metrics

Pasir Ris is the sharpest illustration of everything above, so it earns its own section.

• By volume it is 17th of 26 — 718 resales a year. Middling.
• By turnover it is 6th of 26 at 2.44%. Comfortably above the national rate.
• By deals per block it is 26th of 26 at 1.63. Dead last, by a wide margin.

All three are correct. They disagree because Pasir Ris has the smallest blocks in Singapore — 66.7 sold flats each — and an unusually heavy weighting toward big flats. Nearly a quarter of its resale volume is executive stock, a share no other town comes close to. Small blocks full of large flats.

The practical read for a Pasir Ris owner: your town’s market is healthier than the raw count suggests, but there are genuinely few transactions in your block in any given year, so you will have less recent comparable evidence than a Tampines seller when it comes time to price. That is a pricing problem, not a demand problem, and it is fixable — you widen the comparable set to neighbouring blocks of the same vintage and flat type rather than insisting on your own address. You can run your own numbers with the free calculators at listings.sg/tools before you take anyone’s word for a valuation, including mine.

The market nobody noticed is running at exactly its ten-year average

One more thing fell out of the data, and it is worth a section because it contradicts the mood.

The long-run average across my whole 116-month window is 24,637 resale transactions a year. The most recent twelve months — September 2025 through August 2026 — came in at 24,591.

That is a gap of 0.2%.

After a year of headlines about a softening index, a first quarterly decline since 2019 and prices drifting, the actual number of Singaporean families buying and selling resale flats is sitting within a rounding error of where it has sat for a decade. August 2026 was up 14.4% on August 2025. Prices and volumes are telling different stories, and volume is the one that determines whether your flat finds a buyer.

I have written before about how the price side of the 2026 HDB resale market is behaving, and about the record run of million-dollar flats happening at the top of the same market. They are not in conflict. A flat market with a healthy transaction count and a widening spread between the best and worst stock is exactly what you’d expect at this point in the cycle.

What this means if you’re selling

Five things I would now tell the seller in that Serangoon kitchen.

1. Stop treating your town as the variable

Across 26 towns the liquidity range is 1 in 30 to 1 in 50 flats a year. Your town moves the needle by maybe 30% at the extremes and by almost nothing in the middle 23. Your asking price moves it far more than that.

2. Recalibrate what “slow” means

In a 100-unit block, two to three flats sell a year. If you have been listed eight weeks, you have not been rejected by the market — you have barely met it. Panic-cutting at week six is the single most expensive mistake I see, and it is usually a base-rate error rather than a pricing one.

3. In an older town, price for the buyer’s financing, not the buyer’s taste

If you own in Ang Mo Kio, Bukit Merah, Queenstown or the Central Area, your constraint is that a chunk of the buyer pool cannot use full CPF and full LTV on your lease. That is a cash-gap problem for them. The fix is to market to the buyers for whom the age-95 test is not binding — older upgraders, downsizers, cash-strong households — rather than to fight for the 32-year-old who structurally cannot afford you.

4. In a young town, your competition is your neighbours, not the town

Punggol and Sengkang turn over fast because MOP cohorts arrive together. That cuts both ways: when your block hits MOP, twenty of your neighbours list in the same quarter with near-identical layouts. Being first out with clean photography and a realistic number matters enormously in a homogeneous town and barely at all in a heterogeneous one.

5. Thin block history is a pricing problem with a known fix

Pasir Ris, Bukit Timah, Marine Parade and the Central Area sellers all face the same thing — not enough recent deals in the block to anchor a price. Widen the comparable set by vintage and flat type across neighbouring blocks, and be explicit with your buyer about why you did. Buyers accept a well-argued comparable set. They do not accept a number with nothing behind it.

If you are weighing whether to sell at all, our pieces on the HDB-to-condo price gap and on what executive maisonettes are really fetching cover the upgrade maths. For town-level detail, we have full deep dives on Hougang, Bedok and Bukit Merah, with the rest of the 26-town series publishing through the autumn.

Who should care about town liquidity — and who shouldn’t

Frequently asked questions about HDB resale volume by town

Which HDB town has the highest resale volume?

Sengkang. It averaged 1,975 resale transactions a year between January 2017 and August 2026, ahead of Woodlands (1,765) and Punggol (1,762). But that ranking mostly reflects town size: Sengkang has 69,113 sold flats. Measured as a share of stock, Punggol is the most active town, not Sengkang.

Which HDB town is the quietest for resale?

Bukit Timah, at roughly 60 resale transactions a year across the whole town – the thinnest HDB resale market in Singapore. Marine Parade (150 a year) and the Central Area (193) are next. All three are quiet because they are small, not because their flats are hard to sell; Bukit Timah’s turnover rate of 2.35% is actually above the national average.

What is the average HDB resale turnover rate in Singapore?

About 2.39% a year, which works out to roughly 1 in every 42 HDB flats changing hands annually. Across all 26 towns the range is narrow: from 1.99% in Queenstown to 3.29% in Punggol, with 23 of the 26 towns sitting between 2.0% and 2.7%.

Does living in a quiet HDB town make my flat harder to sell?

Almost certainly not. Once you divide transactions by the number of flats each town contains, the entire country falls into a band 1.3 percentage points wide. The difference between the best and worst town is 1 flat in 30 versus 1 flat in 50 per year. Asking price, flat condition and marketing move your outcome far more than your town does.

Why do Punggol and Sengkang have the highest HDB resale turnover?

Because they are young towns with large cohorts of flats crossing the five-year Minimum Occupation Period at the same time, and a meaningful share of those owners sell immediately. Punggol, Sengkang and Sembawang all have zero blocks with under 60 years of lease remaining. Town age explains about 43% of the variation in turnover across the 26 towns.

How long does it take to sell an HDB flat in Singapore?

There is no reliable published figure, and be sceptical of anyone who quotes one by town. HDB’s registry records the sale date, not the listing date, so days-on-market cannot be computed from public data. What the data does show is the base rate: in a typical 100-unit block only two to three flats sell in an entire year, so a listing that has run six to eight weeks is not behind schedule.

Is HDB resale volume in 2026 lower than usual?

No. The twelve months to August 2026 recorded 24,591 resale transactions against a long-run average of 24,637 a year since January 2017 – a difference of 0.2%. August 2026 alone was up 14.4% year-on-year. Prices have softened while transaction counts have held at their decade average.

Should I choose an HDB town based on how easily I can resell later?

Town matters much less than lease profile. The strongest predictor of how actively a town’s flats trade is how old its leases are, because CPF usage and loan-to-value limits both tighten as remaining lease falls below what covers the youngest buyer to age 95. Buy the lease you can exit rather than the town name.

Disclaimer. This article is general information, not financial, legal or investment advice, and it does not take account of your personal circumstances. Transaction figures are computed from HDB resale registration records published via data.gov.sg for January 2017 to August 2026 and are accurate as at 17 September 2026; HDB continues to lodge late registrations, so historical counts may shift marginally. Dwelling-unit stock figures are HDB’s published FY2021 town table, sold units only. Turnover rates are our own computation and are not an HDB statistic. Lease-remaining shares are computed from HDB’s block register using year of completion, which produces a conservative floor. Property values can fall as well as rise. Always obtain independent advice before a transaction.

Gary Lim · CEA Registration No. R009877B · ERA Realty Network Pte Ltd · CEA Licence No. L3002382K