Freehold vs Leasehold: Which Actually Made More Money on Resale?

The belief is simple and everywhere: freehold protects your money, leasehold slowly loses it. Matched like-for-like, the resale record says otherwise.

Compare the two the only fair way — same unit size, condos and apartments only, no landed — and leasehold matched or beat freehold on the odds of profit in all five size bands. Freehold led in none. Among sub-600 sqft units, leasehold resales profited 91.1% of the time against freehold's 83.9%; even in the largest band the two are level (freehold 80.4%, leasehold 81.1%). The much-promised freehold resale premium simply doesn't show up.

Now the guardrail, because this is exactly the kind of finding that gets over-read. It does not mean "buy leasehold." It measures one thing — whether a matched resale sold for more than it cost — and it deliberately says nothing about price stability in a downturn, downside risk, absolute dollar quantum, or the value of owning the land outright. Freehold can genuinely win on those, and they're real reasons people pay up. The honest headline is not "leasehold is better" — it's "freehold did not command the resale-profit premium the myth claims." For the full six-factor view see how to tell if your property will be profitable; to test a specific unit, use the Profitability model.

Before any number means anything, you have to know why the obvious comparison is broken — because most "freehold vs leasehold" claims you'll read are built on exactly the broken version.

If you simply split every private resale into freehold and leasehold and compare, you are not testing tenure. You're testing the mix of homes that happen to sit under each label:

• Every executive condominium is leasehold — and ECs are the single best-performing property type in this data, thanks to their subsidised entry price. So "leasehold" is partly an EC in disguise, and ECs drag the leasehold average up for a reason that has nothing to do with tenure.
• Landed skews heavily freehold — detached, semi-detached and terrace homes are mostly freehold, and landed carries lower profit odds. So "freehold" is partly a landed proxy, dragged down by the same confound in the opposite direction.

Put those together and a naive average flatters leasehold for the wrong reason — it's comparing cheap-entry ECs against big-quantum landed and calling the gap "tenure." That's why this study throws out the crude split entirely. The fix is to compare only like with like: same size band, condos and apartments only, no landed — so the freehold and leasehold homes in each row are genuinely comparable, and the one thing left varying is the tenure. Everything below rests on that control. When two groups differ in more than one way, a single average can't tell you which difference did the work.

A finding that overturns a popular belief has to be airtight on method, so here is exactly what sits behind every number.

• Matched pairs, not an index. We take private homes with a recorded purchase and a later sale and pair them, comparing the same unit's buy price to its own sell price. An index averages different homes selling in different periods; matched pairs remove that noise.
• Size-matched, non-landed only. Freehold is compared against leasehold within the same size band, across condos, apartments and executive condominiums — landed is excluded on purpose, because landed skews heavily freehold and its plots are wildly heterogeneous, so including it would smuggle the confound straight back in.
• Both tenures need real samples. A size band only appears when both freehold and leasehold have at least 30 matched pairs in it, so no row rests on a thin count. The smallest band still leans on 9,374 freehold and 10,461 leasehold pairs.
• Gross, not net. "Profit" means the sale price beat the purchase price — full stop. It does not deduct agent commission, buyer's stamp duty, any Seller's Stamp Duty, legal fees or loan interest.
• Current to June 2026, drawn from URA private caveats matched into consecutive buy-then-sell pairs.

One honest limit throughout: a pair needs a prior recorded purchase, so first-sale-only and very old units are under-represented. Read every figure as "what usually happened to homes like this," not a guarantee for yours.

Here is the like-for-like comparison in full — freehold against leasehold, matched within each unit-size band, condos and apartments only.

Read down the columns and the pattern is one-directional. Leasehold's profit odds are higher in every single band — by 7.2 points for the smallest units, narrowing to 0.7 points for the largest. On the typical gain, leasehold is also ahead in every band (for example +28.5% vs +25.4% in the 900–1,199 sqft band). Tally it up and it's leasehold ahead in 5 of 5 bands on odds, freehold ahead in 0.

Notice the shape of it. The leasehold advantage is largest among small units and fades as units get bigger — by the 1,600+ band the two tenures are effectively level. So the correct reading isn't "leasehold wins by a mile." It's that freehold, the tenure everyone assumes is the safer resale bet, did not once out-profit its leasehold twin — and at best drew level. That's the myth failing to appear, band by band.

The intellectually honest move is to not overclaim in either direction — including your own finding.

Where the edge is several points — +7.2 for sub-600 sqft, +5.2 for 600–899 sqft — it's consistent and sits on tens of thousands of pairs, so it's a real, repeatable pattern, not noise. But where the edge is a point or two — and especially the +0.7 at 1,600+ sqft — the sensible reading is "about the same," not "leasehold won." A gap that small, even on a big sample, is not something to build a decision on; treat those bands as ties.

That's why the verdict is framed the way it is. The claim that survives every band, from a 7-point gap to a rounding-error one, is the negative one: freehold did not command a resale-profit premium over a comparable leasehold unit. It never pulled ahead, and often fell a little behind. The positive spin — "leasehold is the better investment" — is a stronger claim than the data supports, because in the bands where it would matter most for a big-ticket purchase (the larger units), the two are level. The finding isn't that leasehold wins; it's that freehold doesn't win the way you were told it would.

This is the caveat that keeps the whole piece honest, so it gets its own space: the finding is narrow, and "leasehold is the better buy" does not follow from it.

What this study measures is one specific thing — did a matched resale sell for more than it cost, gross of fees — within comparable, non-landed units. Here is what it deliberately does not measure, each a genuine reason freehold can be worth more:

• Price stability in a downturn. Whether freehold holds up better when the market falls is a different question, about volatility and drawdown, not average resale profit. This study doesn't test it.
• Downside risk and lease decay over the long run. A leasehold's value behaviour deep into its tail — as the lease runs down toward the last decades — isn't captured by a base rate over historical pairs. The ageing-lease problem is real and freehold sidesteps it entirely.
• Absolute quantum. "Percentage profitable" is not dollars. A freehold home can be a larger, pricier asset whose absolute gain dwarfs a small leasehold unit's, even at similar odds.
• Owning the land outright. Freehold carries the option value of the land itself — en-bloc and redevelopment potential, and never having to top up a lease. That's a genuine asset this profit cut puts no number on.

So hold two things at once. On the specific question "which made more on resale, like-for-like," freehold did not earn its reputation. On the broader question "which is the better property to own," this study is one input, not the verdict — and the qualitative trade-offs live in freehold vs leasehold condo and 999-year lease vs freehold. Resale-profit odds are one lens on tenure, not the whole picture.

Turn the finding into a decision without overreaching in either direction.

First, drop the assumption that freehold automatically resells for more. Like-for-like, it didn't — so a freehold premium priced purely as "better resale value" is paying for something the matched pairs don't show. If two comparable units differ mainly by tenure and the freehold one costs meaningfully more, the resale record alone doesn't justify the gap.

Second, weigh the premium against what freehold genuinely gives you, because those things are real: no lease to top up, the land and its redevelopment or en-bloc optionality, and steadier value as a leasehold ages into its final decades. For a very long hold, an intergenerational hand-down, or a redevelopment play, that can be worth a premium — pay it with eyes open. For a shorter hold aimed at resale gain, the levers that actually moved profit were when you bought, what you paid, where and what type — not the tenure stamp. And note the tenure that quietly topped this dataset overall was leasehold-only by definition: the executive condominium, on subsidised entry — a thread picked up in are executive condos the most profitable property. For the whole framework, the profitability guide; for your specific unit, the Profitability model. Buy freehold for what it actually gives you — not for a resale premium the data doesn't confirm.

So should I just buy leasehold, then? No — that's not what this shows. The finding is narrow: compared like-for-like, freehold did not out-profit leasehold on resale, and leasehold edged it in all five size bands. But this measures resale-profit odds only, not price stability in a downturn, downside as a lease ages, absolute dollar quantum, or the value of owning the land — all of which can favour freehold. Read it as "freehold's resale-profit premium isn't supported," not "leasehold is the better buy."

Doesn't freehold hold its value better? Not on matched-pair resale profitability. Controlling for unit size and excluding landed, leasehold matched or beat freehold on the odds of profit in 5 of 5 bands — for example 91.1% vs 83.9% for sub-600 sqft units. "Holds value better" may still hold in a falling market or across a lease's final decades, but those are different questions this cut doesn't measure. On the specific question of resale profit, the premium the belief promises isn't in the data.

Why exclude landed from the comparison? Because including it would re-introduce the exact confound the study removes. Landed skews heavily freehold and its plots are highly varied, so folding it in would let "freehold" stand in for "landed" and distort the tenure comparison. Restricting to condos, apartments and ECs and matching by size keeps the two groups genuinely comparable — the only fair way to isolate tenure.

What we counted. Freehold versus leasehold profit computed within each unit-size band, over matched private resale pairs from PropKaki's resale_pairs_mv (URA private caveats matched into consecutive buy-then-sell pairs), current to June 2026, via the profit_by_tenure_size_mv matview. A "pair" is one unit with a recorded purchase and a later sale, so we compare the same home's buy and sell price. The comparison is restricted to condos, apartments and executive condominiums — no landed — and a size band appears only when both tenures have at least 30 matched pairs in it. Sub-sale flips under 180 days are excluded as noise.

Why non-landed only. A raw freehold-vs-leasehold split is a composition artifact: all executive condominiums are leasehold (the best-performing type) and landed skews heavily freehold (with lower odds), so an uncontrolled average measures property-type mix, not tenure. Restricting to non-landed and matching by size removes that confound; including landed (which skews heavily freehold, with heterogeneous plots) would re-introduce it.

What we did not claim. That leasehold is the better buy — this measures resale profitability of matched pairs only. That it captures price stability in a downturn, downside risk as a lease ages, absolute dollar quantum, or the value of owning the land outright and its redevelopment potential — all of which can favour freehold and are out of scope. That a small odds edge (a point or two, e.g. +0.7 at 1,600+ sqft) is a win for either tenure — read those as ties. That gross equals net — every gain is before agent commission, buyer/seller stamp duties, any Seller's Stamp Duty, legal fees and loan interest. And that a base rate guarantees any specific home; a pair needs a prior recorded purchase, so first-sale-only and very old units are under-represented. This is general information, not financial advice. To assess a specific unit, use the Profitability model.