Building age doesn't explain the shoebox premium. There's barely any old shoebox stock.

We expected the age of a building to explain something. It does not. Grouping condominiums by the year they were completed, the shoebox premium shows no pattern with age at all — except in the newest stock, where it is sharply negative.

Across this series we have found the shoebox psf premium collapsing inside the same buildings, by region, and by tenure. Building age was the obvious next candidate: prime-district stock is often older, freehold shoeboxes are often conversions in older blocks, and both of those findings could plausibly have been age wearing a different label.

They are not. And the reason is worth stating before the numbers.

Key takeaways

• Grouped by completion year, the shoebox premium runs −3.6%, +3.6%, −1.6%, −16.4%. There is no gradient with age. The four cohorts
• Age cannot explain the earlier regional and tenure findings, because shoebox stock has almost no age variation — fewer than 200 resale transactions in nine years come from buildings completed before 2009. Why there is no old stock
• The exception is the newest cohort: in buildings completed 2021–2024, shoeboxes trade 16.4% below family-sized units. That is driven by family psf rising faster, not shoebox psf falling. The newest stock
• That cohort is entirely early exits, but both size bands sell at the same point in the cycle — 2.3 years against 2.2 — so the gap is not an artefact of who is selling. The newest stock

There is almost no old shoebox stock

Shoebox units in Singapore are a recent format. Across nine years of resale transactions in units of 400–538 sqft:

Completion era	Shoebox transactions	Projects
Before 2000	49	10
2000–2008	143	29
2009–2013	1,564	149
2014–2018	5,363	234
2019 onward	1,214	77

Fewer than 200 shoebox resale transactions in nine years come from buildings completed before 2009 — against nearly seven thousand from buildings completed after. Family-sized stock is the opposite: 543 projects completed before 2000 alone.

This is itself the answer to one question. Age cannot explain why prime-district or freehold shoeboxes underperformed, because there is no meaningful variation in shoebox age to do the explaining. Nearly all of it was built in one fifteen-year window.

It also constrains what can be measured. A building completed in 2023 has no resale history in 2017, so the paired-window method we use elsewhere — comparing each building against its own earlier self — cannot reach newer cohorts at all. What follows is a snapshot instead: where each completion cohort sits today.

Four completion cohorts, priced today

Resale transactions from 2023 to 2025, grouped by the year each project was completed.

Completed	Shoebox psf	Family psf	Premium	Projects (S / F)
2009–2012	$1,721	$1,785	−3.6%	40 / 127
2013–2016	$1,667	$1,609	+3.6%	153 / 143
2017–2020	$1,854	$1,885	−1.6%	49 / 57
2021–2024	$1,924	$2,300	−16.4%	34 / 40

Read down the premium column: −3.6, +3.6, −1.6, −16.4. There is no gradient. The premium does not decay as buildings age, and it does not strengthen. Three of the four cohorts sit within four points of zero, in no order.

The one cohort that stands apart is the newest — and it stands apart by a wide margin.

The newest stock

In buildings completed between 2021 and 2024, a shoebox trades 16.4% below a family-sized unit in the same cohort. That is a wider discount than any regional or tenure cut in this series has produced.

Both lines are rising with newness, as you would expect — newer buildings command more per square foot. Shoebox psf runs $1,721, $1,667, $1,854, $1,924 across the four cohorts. The difference is that the family line rises much faster in the newest cohort: $1,785, $1,609, $1,885, $2,300.

So this is not shoeboxes falling. It is family-sized units in recent projects being priced, and transacting, far above them.

One thing we checked before publishing this. Resales from the 2021–2024 cohort are almost all early exits — 90% within three years of completion. That is a selected group: people who sell soon after taking keys are not a random sample of owners. If shoebox owners were exiting earlier than family owners, the gap could be an artefact of who happens to be selling.

They are not. The hold periods are nearly identical: shoebox resales average 2.3 years since completion with 90.3% inside three years; family resales average 2.2 years with 89.2% inside three years. Both bands are selling at the same point in the cycle, so the gap between them is not explained by one selling sooner than the other.

What remains true is that the whole cohort is early exits. Owners who bought in 2022 and intend to hold for a decade are not in this data, and cannot be until they sell.

What this tells you, and what it does not

Age is not the mechanism. This was run as a test of whether our earlier regional and tenure findings were really age effects in disguise. They are not, for a structural reason: shoebox stock has almost no age variation to begin with. That closes a question rather than opening one.

A null result is still a result. We ran this expecting to find something and found an absence. The absence is informative — it removes one explanation from a list we have been narrowing across four articles.

The newest cohort is a live question, not a settled one. A 16.4% discount in 2021–2024 stock is a large number on 34 shoebox projects, drawn entirely from early exits. It is consistent with recent launches pricing family units aggressively, and it is consistent with the same investor withdrawal we have seen elsewhere reaching the newest product first. This data cannot separate those.

What it does not tell you: how any cohort's premium has moved, since the newer cohorts have no earlier history to compare against; anything about buildings completed before 2009, where shoebox stock barely exists; and anything about owners who have held rather than sold.

Method

Private non-landed resale caveats, condominium and apartment. Executive condominiums excluded, collective sales excluded. Shoebox is 400–538 sqft, matching URA's 50 sqm threshold; family-sized is 1,000–1,500 sqft.

Cohorts are defined by project completion year. The cross-sectional table covers resale transactions from 2023 to 2025 only — a snapshot, not a trend — because completion cohorts after 2020 have no resale history in the earlier window this series usually compares against. A project qualifies for a cell when it has at least three transactions in that size band over the window. Medians are taken across projects rather than pooled across transactions, which would let the changing mix of what sold masquerade as a change in price.

Because each size band is measured on its own set of projects, the premium here compares across buildings within a cohort rather than inside a single building. Our first article measures within buildings, which is stricter but cannot reach cohorts that lack an earlier window.

Hold period is measured as transaction year minus completion year, so it is an upper bound on ownership length for anyone who was not the original buyer.

Data to 17 July 2026.