What we found about shoebox units in Singapore

Shoebox units in Singapore no longer sell at a higher price per square foot than family-sized units in the same development. In 2017 they sold at 20.9% above family-sized units, and in 2025 at 3.1% below. Many guides published this year still describe the small-unit premium as fact.

A shoebox here is a private apartment of 538 sqft or less, the size set by URA's 50 square metre threshold. This page answers the questions people ask about shoebox units, with the answer in the first line of each section and a link to the note behind it.

What are you deciding?

• Buying one to live in
• Buying one to rent out
• Selling one you own
• Renting one as a tenant
• Renting out one you own

What the record shows

Do shoebox units still sell at a price premium?

No. Compared inside the same development, shoebox units sold at 20.9% above family-sized units per square foot in 2017 and 3.1% below them in 2025. In part-year 2026 data the discount is 11.5%.

Shoebox prices did not fall over that period. Shoebox psf rose from $1,368 to $1,699, a gain of 24%. Family-sized units rose from $1,066 to $1,676, a gain of 57%. The premium closed because family-sized units rose faster.

The shoebox premium inside the same developments fell from around 21% to below zero. Shoebox prices did not fall; they rose more slowly than family-sized prices.

Are the rental yields really higher?

Yes, and the gap has widened. Gross shoebox rental yield rose from 3.56% in 2017 to 4.30% in 2025. Family-sized rental yield fell slightly, 2.99% to 2.87%. The gap between them more than doubled.

The rent premium is the one part of the common view that the data supports. Shoebox rent per square foot was 52.6% above family-sized units in 2017 and 51.9% above in 2025. It was higher in 2022 and 2023, the years rents rose sharply, and it held through the pandemic and two rounds of cooling measures.

The market-wide figure combines three regions with different premiums. Split by region, the shoebox rent premium is about 47% in the core central region, 51% in the rest of central and 58% outside it, and the ordering has held in every year we measured without once crossing.

Building age moves it as well, and in the newer stock it moves sharply: the shoebox rent premium is 42.8% in buildings completed in 2009–2012 and 44.0% in 2013–2016, then falls to 32.9% in 2017–2020 and 24.1% in 2021–2024. The compression is on the family side, where rents rise with newness about twice as fast as shoebox rents do.

Shoebox rental yield rose because shoebox rents increased faster than shoebox prices.

Shoebox gross rental yield rose from 3.56% to 4.30% while family-sized rental yield fell slightly.

Did shoebox owners actually make money?

88.8% of them did. Matching the same unit bought and later resold, 6,680 of 7,523 shoebox resales since 2020 sold above their purchase price, with a median gain of $92,000 over a median seven-year hold.

Annualised, the median gain ranges from 0.8% to 3.5% a year depending on when the owner bought, before duties, commission, interest and holding costs.

The losses concentrate sharply by location: 44.2% of resales in the core central region (CCR) lost money against 6.0% in the outside central region (OCR). Among those that lost, the median loss was 5.2%.

Nearly half of core central region shoebox resales lost money, against one in sixteen outside central.

Is resale liquidity a problem?

No, and it has improved. Shoebox resale turnover nearly doubled between 2017 and 2025, from 2.4% of units in the transaction record changing hands in a year to 4.5%. The step change came in 2021 and has held since.

The concern is that a development with many small units can produce many sellers at once. The transaction record does not show sales drying up: turnover rose.

What does a shoebox actually cost?

$848,000 for a resale in 2025, and $1,268,007 for a new launch. The resale median was $680,000 in each year from 2017 to 2019 and rose by about a quarter by 2025. The new launch premium over resale increased from 13% in 2017 to 50% in 2025, and it holds in every region.

The cheaper half of the resale market has no equivalent in new-launch stock.

Does freehold protect a small unit?

No. Within freehold stock, shoeboxes went from an 18.7% premium over family units in 2017 to a 14.6% discount in 2025. Within leasehold stock they still hold a 5.1% premium.

Tenure makes almost no difference to rent. Shoebox rents are $6.00 psf freehold against $6.17 leasehold, 17 cents higher for leasehold.

The small-unit premium went negative in freehold and has never gone negative in leasehold. Because freehold sells for more and rents for the same, leasehold has the higher rental yield.

Does buying at launch beat buying resale?

It depends on the year of purchase. Matched so both groups bought and sold in the same months and held the same length of time, and measured as return on the cash invested at the mortgage rate used in the note's main table, resale buyers earned more in 2017, 2019, 2020 and 2021, by 12 percentage points in 2021. Launch buyers earned more only in 2018.

The reason is entry price. Median launch entry for a shoebox rose from $750,000 to $953,000 over those five years while resale entry stayed near $680,000.

Within a single development the picture is different again: buying in a project's launch year beat buying later in 41 of the 48 developments where both are measurable, because later buyers paid 8% to 24% more per square foot for the same building.

On return on the cash invested, resale won in four of the five purchase years tested, and the exception is 2018. Even buyers who waited one year after launch usually did worse than launch-year buyers.

Does building age or size explain any of it?

No. Grouped by completion year, the price premium does not move with age, until the newest stock, where 2021–2024 buildings show shoeboxes 16.4% below family units. The structural reason age cannot explain much is that shoebox stock has almost no age variation: fewer than 200 resale transactions in nine years come from buildings completed before 2009.

Region explains more of the change than building age does. A CCR shoebox is worth 4.5% less per square foot than in 2017. An OCR one gained 25.9%.

The shoebox premium shows no pattern with building age, except in the newest stock. The region has mattered more than the unit's size.

What it means for you

Buying for own stay

If you will live in the unit, the price-growth findings matter less to you than to an investor. One thing from them is directly useful at the negotiating table.

"The psf is high because it's small" is no longer, by itself, a reason a price is fair. Ask what family-sized units in that same development have actually transacted at, and treat a small-unit premium as something to be justified rather than assumed.

Buying for investment

Shoebox units have become better at producing income and worse at producing capital gains.

On income, in 2025 they have a rental yield of 4.30% a year, against 2.87% for a family-sized unit. That gap is the widest in nine years. On capital, they gained 24% over eight years while family-sized units gained 57%.

Which of those matters more depends on how long you plan to hold.

What matters more than the unit

In the two worked examples below, the same kind of purchase produced very different returns depending on financing, stamp duty and whether the unit was rented out.

We followed two purchases from start to finish, with every cost included.

A resale bought in the rest of central region (RCR) in 2021 returned as much as 5.9% a year, or lost 6.7% a year. The difference came down to two things: the mortgage rate, and whether the buyer paid additional stamp duty.

A new launch bought in 2018 returned 7.0% a year if the owner rented it out after it was built, and 3.1% if they left it empty until they sold. That one decision was worth $44,600.

The same resale unit, held the same five years, made money at a low mortgage rate and lost money at a high one. Whether a launch buyer rented the unit out made the biggest difference to what they earned.

Those gaps are wider than anything we found about shoebox units as a group.

One thing we cannot explain

An asset with a rental yield of 4.30% sitting next to one at 2.87% should attract buyers until the gap narrows. It has not, for four years running.

Either the extra stamp duties are keeping investors away and will continue to, or buyers see a risk that has not yet appeared in rents or prices. We do not know which.

Looking to sell

The key finding for a seller is that your unit's price per square foot has risen, while its premium over family-sized units in the same development has turned into a small discount.

Shoebox psf went from $1,368 in 2017 to $1,699 in 2025. That is a real gain. What disappeared is the uplift small units used to command over larger ones in the same building.

The practical consequence is about how you price. Applying the old small-unit premium to what family-sized units in your building currently fetch will put you above the market. The comparable that matters is what other small units in your development have transacted at recently.

If you own in the CCR, the position is harder: CCR shoebox psf is 4.5% below its 2017 level. If you own in the OCR, it is 25.9% above.

On the wider question of whether sellers have been making money, they mostly have. Across resales completing since 2020, 88.8% sold above their purchase price, and the median loss among the rest was 5.2%. The CCR is the exception, where 44.2% sold at a loss and the median loss was 7.7%.

Looking to rent

This data can help you judge whether a rent is in line with the market.

In 2025 a shoebox rented for about 52% more per square foot than a family-sized unit. That premium is smallest in the CCR at about 47% and largest in the OCR at 58%, and it is smaller in newer buildings: 24.1% in stock completed since 2021, against 42.8% in 2009–2012 stock.

Whether the unit is freehold or leasehold makes almost no difference to what you pay.

Looking to rent out

Your side of the finding is the one that held up. The shoebox rent premium was about the same in 2025 as in 2017, 51.9% against 52.6%.

Rents rose steeply in 2022 and 2023 for both unit sizes and then flattened. The premium was back at 51.9% by 2025.

Two things shape what you can charge. Location: the premium is largest in the OCR and smallest in the CCR, though CCR rents are the highest in dollars, at $7.02 psf against $5.60 in the OCR. Building age: the premium falls with newness, because family rents rise about twice as fast with newness as shoebox rents do. A new family unit commands proportionally much more than an older one; a new shoebox commands only a little more.

Tenure makes almost no difference to what you can charge. Shoebox rents are $6.00 freehold against $6.17 leasehold, a gap of 17 cents that runs in leasehold's favour.

What we tested, and what survived

The premium collapse could have had several explanations. We tested each one and reported the result whether or not it helped.

Composition: that different buildings were simply transacting. Failed: across 49 developments with sales in both periods, the premium compressed in 43.

Floor effects: that larger units sit higher. Failed: holding the five-storey floor band constant, it compressed in 33 of 35 development and floor band pairs.

A taste for space after 2020. Failed: HDB, the one housing market with essentially no investors, moved the opposite way over the same years.

Supply restrictions: that URA's shoebox caps made small units scarcer. Failed on direction: restricting new supply should raise the price of existing stock. It fell.

Tenure: that freehold and leasehold were driving it. Failed: within each region the two tenures moved together, while regions differed sharply.

Building age. Failed structurally: there is almost no old shoebox stock for age to explain anything with.

What remains is harder to name. One explanation is that investors withdrew, so the buyers who paid the premium stopped buying. Additional buyer's stamp duty for a citizen's second property rose from 12% to 20% between December 2021 and April 2023, and the foreign-buyer rate went from 20% to 60%. That is paid at purchase, on the whole price, regardless of what the unit returns in rent afterwards. The timing fits that explanation.

The sales volumes do not. If buyers had left, resale shoeboxes would change hands less often, because that stock is fixed. They change hands more often. Resale shoebox sales went from 465 in 2019 to 1,122 in 2025, and their share of all resale sales rose from 6.2% to 8.9%. What collapsed is new supply: the shoebox share of new launch sales fell from 16.6% to 6.3% over the same years. The data is consistent with a fall in new supply while buyers kept buying existing shoeboxes: more of the existing stock changed hands, at a price that softened relative to family-sized units.

That also explains why the supply test above failed on direction. Developers stopped selling shoeboxes years before they stopped delivering them: units sold in 2019 and 2020 were still completing in 2022 and 2023, so the stock available to buy kept growing while new sales of them dried up.

The investor explanation rested on elimination and timing, and the sales volumes do not support it.

The share of new-launch buyers giving an HDB address fell less for shoebox units than for 3 bedroom units. In 2017, 60.3% of new-launch shoebox buyers whose address type is recorded gave an HDB address, and by 2025 that was still 43.7%, a smaller fall than the one on 3 bedroom over the same nine years.

What we still cannot tell you

• Who is buying. Caveat records do not identify buyers or sellers, and no dataset we hold does. We can see how many shoeboxes change hands and at what price, but not who is on either side.
• What happens next. The record covers nine years, and nothing here is a forecast.
• Your specific unit. Every figure is a median across dozens or hundreds of projects. Buildings, stacks, floors and lease terms vary widely.
• Net returns. Rental yields here are gross. Maintenance, property tax, agent commission, vacancy and income tax all come out first.

Method and data notes. Sale data to 4 August 2026, rental data to February 2026.

Everything above is measured from URA caveat data for private non-landed resale transactions and URA rental contracts, with executive condominiums and collective sales excluded. Shoebox is 400–538 sqft on the sale side and URA's 40–50 sqm band on the rental side, both terminating at URA's 50 sqm threshold; family-sized is 1,000–1,500 sqft and the 90–130 sqm bands respectively. Comparisons are made within the same development wherever the data allows, and medians are taken across projects rather than pooled across transactions: pooling would let a change in the mix of units sold appear as a change in price. Each article states its own panel construction and limits. Sale data to 4 August 2026, rental data to February 2026. The launch-timing study uses the full transaction history rather than the recent-exit window, and says so.

30 October 2026. This note placed homes in market segments by postal district. URA places them by planning area, which moves about one in seven private sales between regions. The figures by region here use the earlier segments and differ from what the same records give now. How regions are set: How we measure prices and sales.