Freehold shoeboxes now sell for less per sqft than the family units beside them

A freehold shoebox now sells for 13.5% less per square foot than a freehold family unit. A leasehold shoebox still sells for more than a leasehold family unit — 5.9% more, down from 31% in 2017. The small-unit premium has collapsed in both, but it went negative in freehold four years ago and has not gone negative in leasehold yet.

We have been tracking the collapse of the shoebox premium — across the market as a whole, then by region, where it inverted in the core central region in 2021 and the rest of central in 2024. Tenure is the next obvious cut, and the one buyers ask about most.

One thing to know before the tables. Each tenure is measured against itself: freehold shoeboxes against freehold family units, leasehold against leasehold. We are not comparing freehold prices to leasehold prices — those differ for reasons of location and age that have nothing to do with tenure. What we are comparing is the premium inside each, which is a ratio and therefore unaffected by one tenure sitting at a higher absolute price than the other.

Freehold

Median resale psf, across the same developments at both ends of the period. Shoebox is 400–600 sqft, family is 1,000–1,500 sqft.

Year	Shoebox	Family	Premium	Developments (S / F)
2017	$1,511	$1,276	+18.4%	45 / 160
2018	$1,493	$1,421	+5.1%	51 / 143
2019	$1,550	$1,334	+16.2%	23 / 110
2020	$1,400	$1,340	+4.5%	26 / 97
2021	$1,477	$1,519	−2.8%	47 / 149
2022	$1,518	$1,641	−7.5%	36 / 109
2023	$1,684	$1,784	−5.6%	40 / 110
2024	$1,770	$1,888	−6.3%	45 / 127
2025	$1,756	$2,029	−13.5%	33 / 132

The premium is the shoebox figure over the family figure: how much more, per square foot, a small freehold unit fetched than a large one. A negative number means it fetched less.

Freehold crossed into negative territory in 2021 and has stayed there, reaching −13.5% in 2025 — the widest discount in the series. The freehold shoebox line itself has gone almost nowhere since 2017: $1,511 to $1,756, while the freehold family line moved $1,276 to $2,029.

The early years bounce — +18.4%, then +5.1%, then +16.2% — which is worth treating as noise around a level rather than as movement. The direction from 2020 onward is not noisy.

Leasehold

Year	Shoebox	Family	Premium	Developments (S / F)
2017	$1,318	$1,006	+31.0%	41 / 238
2018	$1,400	$1,078	+29.9%	49 / 236
2019	$1,372	$1,093	+25.5%	40 / 224
2020	$1,330	$1,093	+21.7%	46 / 236
2021	$1,426	$1,171	+21.8%	60 / 261
2022	$1,487	$1,278	+16.4%	60 / 242
2023	$1,595	$1,400	+13.9%	56 / 240
2024	$1,667	$1,466	+13.7%	53 / 247
2025	$1,644	$1,552	+5.9%	50 / 246

Leasehold falls in eight of the nine year-on-year steps, from +31.0% to +5.9%, and has not yet crossed zero. On the current slope it would cross within a year or two, but that is an extrapolation and not a measurement.

What the two tables say

Freehold started with the smaller premium — +18.4% against leasehold's +31.0% — and lost it faster. It inverted in 2021. Leasehold, four years later, is still positive.

Put plainly: the belief that a freehold small unit is the safer store of value does not survive contact with the last nine years. Freehold shoeboxes are the only cell in any cut we have run that now trades at a double-digit discount to its larger counterpart.

One important complication. Freehold stock concentrates in the central regions and leasehold in the suburbs. Our regional article found the premium inverting in the core central region in 2021 — the same year freehold inverted here. These may not be two separate findings so much as one fact seen through two labels. We cannot separate them with this data, and we are not claiming tenure is the cause.

A reading from practice, which we cannot test

What follows is interpretation from working in this market, not something the transaction record can confirm. Caveat data records the sale, never the buyer or the reason.

Freehold purchases skew toward buyers thinking in terms of legacy — an asset held across generations rather than a position to exit. Buyers in that frame are typically less budget-constrained, and a unit intended to be lived in or handed down tends to be a larger one. If the freehold buyer pool leans that way, demand concentrates in family-sized stock and the small-unit end of freehold loses its natural bidder.

Leasehold small units serve a different pool: buyers and investors working to a tighter budget, for whom a smaller unit may be the only entry point at all. That floor under demand does not disappear when sentiment shifts, which would explain a premium that erodes steadily rather than inverting.

We find this reading plausible and it fits the shape of the data. It is not evidence. Nothing in the caveat record identifies who bought, why, or with what intent, and a different explanation consistent with the same numbers would be equally unfalsifiable here.

What this does not tell you

• Whether tenure is doing the work. Freehold and central-region stock overlap heavily. This data cannot separate them.
• Anything about lease decay. Most leasehold stock in this sample still has seventy or more years remaining. What happens to a lease at forty years, or thirty, is a different question and out of reach here.
• Your development. These are medians across dozens of projects. Individual buildings diverge widely.
• What happens next. Nine years of history, not a forecast.

Method

Private non-landed resale caveats, condominium and apartment. Executive condominiums excluded, collective sales excluded.

Each tenure-and-size line runs on its own fixed panel: developments with at least three transactions in that band, in that tenure, in both the 2017–2019 and 2023–2025 windows — so each column describes the same buildings at both ends. Medians are taken across developments rather than pooled across transactions, which would let the changing mix of what sold masquerade as a change in price.

Because the shoebox and family panels are separate sets of buildings, the premium here is a comparison across developments rather than within a single one. Our first article in this series measured the premium inside each building, which is a stricter comparison but excludes any development that does not contain both sizes. The figures are therefore not directly comparable between the two pieces; the direction and the timing are.

2026 is excluded: the partial year leaves too few developments per cell. Data to 17 July 2026.