HomeResearch › Leasehold vs freehold

Leasehold vs freehold: which held value better?

“Freehold always holds its value” is the oldest rule in Singapore property. We put it to the test on the numbers. Taking every private condo completed at least five years ago (TOP 2021 or earlier) with enough resale caveats to measure a trend — 1,209 projects — we compared the 5-year resale $psf growth of freehold / 999-yr projects against 99-year leasehold, split by market region and controlled for location, size and age. Source: URA Data Service.

TL;DR — the takeaway

Over the last five years, leasehold quietly kept pace with freehold — and beat it in the suburbs. On matched pairs (same district, similar size and age), 99-yr leasehold appreciated on resale $psf about as fast as freehold in the prime core and city fringe, and faster in the OCR suburbs. The extra you pay for freehold did not buy faster growth in this window.

The five-year verdict: a dead heat, tilting to leasehold outside the core

Across all 196 matched freehold–leasehold pairs, the median difference in 5-year resale $psf growth was -0.8 percentage points a year in favour of leasehold — small enough to call a tie overall, but clearly leasehold-favouring in the suburbs. Freehold’s real edge — no lease decay, en-bloc land value, holding in perpetuity — is a long-hold and lease-tail story that a recent 5-year window, in a rising mass-market, simply doesn’t capture.

CCR · matched gap
dead heat
RCR · matched gap
leasehold +0.3pp
OCR · matched gap
leasehold +1.8pp
All pairs (median)
leasehold +0.8pp

Matched-pair gap = median (freehold minus leasehold) 5-yr resale $psf CAGR, pairing each freehold project with a leasehold one in the same district of similar size and vintage. A positive value favours freehold; negative favours leasehold.

The 5-year scorecard, by region

Median 5-year resale $psf growth (per year) for each tenure, in each URA market region. If freehold reliably “held value better,” the dark bars would sit clearly above the amber ones. They don’t — the two are close everywhere, and leasehold is ahead in the suburbs.

0%1%2%3%4%5%6%7%8%CCR freehold 3.2%/yrCCR leasehold 2.0%/yr3.2%2.0%CCRRCR freehold 5.2%/yrRCR leasehold 5.7%/yr5.2%5.7%RCROCR freehold 5.8%/yrOCR leasehold 7.3%/yr5.8%7.3%OCR5-yr resale $psf CAGR (per year)
Freehold / 999-yr99-yr leaseholdeach bar = median project CAGR in that region

Region by region

MetricCCR FHCCR LHRCR FHRCR LHOCR FHOCR LH
5-yr resale $psf CAGR3.2%2.0%5.2%5.7%5.8%7.3%
Current median $psf$2,225$1,918$1,668$1,726$1,554$1,345
Qualifying projects28168357112150241

Current $psf = median of resale caveats in the last ~2 years. Note the freehold price premium: buying freehold costs about +16% more per sq ft in the CCR and +16% in the OCR — yet that premium bought no extra 5-yr growth. In the RCR, freehold and leasehold $psf are almost level, partly because the leasehold set there includes many big, recently-completed launches.

Controlling for location, size & age

Region medians can be skewed by where each tenure clusters — freehold dominates some enclaves, leasehold others. The cleaner test is a matched pair: take a freehold project and the most similar leasehold project in the same district, of similar size and vintage, and compare their 5-year growth directly. That removes the location, size and age confounders so the remaining gap is closer to a pure tenure effect.

CCR
-0.1pp
40 pairs
RCR
-0.3pp
73 pairs
OCR
-1.8pp
83 pairs
All pairs
-0.8pp
196 pairs

Median (freehold − leasehold) 5-yr resale $psf CAGR within matched pairs. Positive = freehold grew faster. The matched result confirms the region medians: a near dead heat in the prime core and city fringe, and a clear 1.8pp/yr edge to leasehold in the suburbs.

Why the suburbs favour leasehold here. The OCR leasehold set is dominated by large, newer 99-yr launches (2016–2019 TOP) that appreciated hard off a lower entry base during the 2021–2024 mass-market surge, while their freehold neighbours were older and more fully-priced. That is a genuine finding for this window — but it is a base-effect and cycle story as much as a tenure one.

The projects behind the numbers

Five freehold and five leasehold projects in each region — the most-traded qualifying names, so the $psf and CAGR are well-supported. These are illustrative examples of each cohort, not the whole sample.

Core Central Region (CCR) — the prime core

D1/2/4/6/9/10/11. Freehold edges ahead here (3.2% vs 2.0%/yr) — the one region where the tenure premium partly earned its keep.

Freehold / 999-yr
ProjectDistUnitsTOP$psf5-yr CAGR
Rv ResidencesD102482015$2,332+2.7%
Valley ParkD107281997$2,187+3.2%
Nouvel 18D101562014$3,029-0.0%
Spottiswoode ResidencesD23512013$2,289+2.8%
Skyline ResidencesD42832015$2,202+1.9%
99-year leasehold
ProjectDistUnitsTOP$psf5-yr CAGR
D'LeedonD101,7032014$2,044+4.5%
Reflections At Keppel BayD41,1292011$1,722+0.9%
The Sail @ Marina BayD11,1112008$2,026+2.3%
The InterlaceD41,0402013$1,682+5.4%
Marina One ResidencesD11,0422017$1,993-5.2%

Rest of Central Region (RCR) — the city fringe

D3/5/7/8/12/13/14/15/20. Essentially level (5.2% vs 5.7%/yr).

Freehold / 999-yr
ProjectDistUnitsTOP$psf5-yr CAGR
Flamingo ValleyD153932014$1,825+3.7%
Vacanza @ EastD144732014$1,535+5.8%
City Square ResidencesD89102009$2,077+3.9%
The WatersideD155021992$2,198+5.4%
One AmberD155622010$2,276+6.3%
99-year leasehold
ProjectDistUnitsTOP$psf5-yr CAGR
Sims Urban OasisD141,0242017$1,903+4.9%
Parc RivieraD57522019$1,689+4.8%
Eight RiversuitesD128432016$1,858+5.6%
Commonwealth TowersD38452017$2,217+4.4%
The TrilinqD57552017$1,801+4.5%

Outside Central Region (OCR) — the suburbs

The mass-market heartland. Leasehold led (7.3% vs 5.8%/yr) — newer 99-yr launches appreciating off a lower base.

Freehold / 999-yr
ProjectDistUnitsTOP$psf5-yr CAGR
Ferraria Park CondominiumD174722009$1,240+6.0%
Edelweiss Park CondominiumD175172006$1,118+5.2%
Signature ParkD219281998$1,711+4.4%
The CascadiaD215362010$2,232+4.4%
Carissa Park CondominiumD175282001$1,200+6.7%
99-year leasehold
ProjectDistUnitsTOP$psf5-yr CAGR
Sol AcresD231,3272018$1,470+7.7%
High Park ResidencesD281,3762019$1,622+5.8%
Kingsford WaterbayD191,1572018$1,455+3.8%
The TopiaryD287002016$1,465+8.9%
The MintonD191,1452013$1,517+8.1%

Investor verdict

Over a recent 5-year hold, tenure was not the deciding factor in how fast your $psf grew. Matched like-for-like, freehold and 99-yr leasehold moved together in the CCR and RCR, and leasehold actually out-grew freehold in the OCR by ~1.8pp a year. If your holding period is short and you’re buying a younger leasehold, the “freehold holds value better” rule did not hold in this window.

But read the caveats before concluding freehold is overrated. Five years is a short window, and it coincided with a mass-market surge that flattered newer suburban leasehold off a low base. Lease decay barely bites a 99-yr project in its first two decades — the freehold advantage shows up later, once a leasehold’s remaining tenure falls toward 60 years and bank LTV / CPF rules tighten its future buyer pool, and it shows up as en-bloc land value and the ability to hold in perpetuity. This study can’t see those; it measures the near term only.

So how to use it: if you’re buying to hold 5–10 years and sell while the leasehold is still young, a well-located newer 99-yr project — bought ~16% cheaper per sq ft — is a perfectly rational, and lately better-performing, choice. Pay the freehold premium when you’re buying for the very long term, for a legacy / en-bloc play, or in the prime core, where scarcity is real and freehold’s edge is most defensible.

How is this worked out? — region map, filters, CAGR & matching
Universe
Every private condo completed in 2021 or earlier (at least 5 years old) with enough resale caveats to fit a trend — a project needs ≥8 resale transactions spanning ≥3 years within the trailing 5 years. New-sale and sub-sale caveats are excluded so launch pricing doesn’t contaminate the resale trend.
Tenure
Freehold bucket = Freehold, 999-yr and estate-in-fee-simple; Leasehold bucket = 99-year. Other lease lengths (e.g. 103-yr, 946-yr) are set aside to keep the contrast clean.
Region
URA market segments by district. CCR = D1,2,4,6,9,10,11; RCR = D3,5,7,8,12,13,14,15,20; OCR = all remaining districts (D16–19,21–28). This follows URA’s Core / Rest of / Outside Central Region convention; D6 has almost no qualifying stock, so CCR is effectively the prime D9/10/11 core plus the downtown/Sentosa districts.
CAGR
Each project’s 5-yr resale $psf growth is the slope of an ordinary-least-squares regression of log($psf) on transaction date over the trailing five years, expressed as an annual compound rate. A regression (rather than first-vs-last median) uses every caveat and is robust to thin or lumpy data. Region figures are the median project CAGR in that region×tenure cell.
Matched pairs
The primary result. Each freehold project is paired with the leasehold project in the same district that is closest in size (log units) and vintage (completion year); we report the median (freehold − leasehold) CAGR. This holds location, size and age roughly constant, so the residual gap approximates a pure tenure effect. We chose matched pairs over a naive 5+5 pick precisely because tenure correlates with district and vintage.
Note: five years is a short window that coincided with a mass-market boom; it flatters newer suburban leasehold bought off a low base and cannot capture lease decay (which bites mainly below ~60 years remaining) or freehold’s en-bloc / perpetual-hold advantages. Caveats also don’t adjust for floor, facing or unit size. Treat this as a near-term resale read, not a full-life tenure verdict.
For educational purposes only — not financial advice.

See also: Project sizing — does unit count move $psf? › · Amber Road freehold ›