How Fast Does a 99-Year Lease Really Decay in Singapore?

The Official Table Says 88%. The Real Market Says 68%.

Here's the number that should worry anyone buying a leasehold condo or HDB flat with 67 years left on the lease: the official government reference (the SLA/Bala Curve) says that unit is still worth 88.2% of what it would be worth freehold. I pulled the actual resale transaction data and ran the numbers myself. The real market price for that same 67-year lease is closer to 67.7% of reference value — a 20 percentage point gap that the official table simply does not show you.

That gap is the whole point of this guide. I built my own lease-decay curve from real transactions — not land-premium theory — because I kept seeing buyers and sellers anchor on the SLA numbers and get blindsided by how leasehold units actually trade. If you're one of them, this is the correction.

Why the Official Curve and the Real Market Don't Agree

The SLA/Bala Curve was never built to predict resale prices. It was built for a specific government purpose: working out how much a leaseholder should pay to top up or extend a lease back toward 99 years. It's a land-premium formula, not a market-pricing model.

The open market doesn't price a shrinking lease the way that formula assumes. Buyers worry about financing (banks cut loan tenures hard once a lease gets short), about resale-ability down the line, and about CPF usage rules that kick in as a lease shortens. None of that psychology is in the SLA table. So I went and measured what actually happens to transaction prices as leases shrink, instead of relying on a formula designed for a different job.

To do that, I used hedonic regression — a statistical method that isolates the price effect of just ONE factor (in this case, years of lease remaining) while holding everything else steady, like location, floor level, unit size, and the time period the sale happened in. That way, what's left over is a clean read on what lease length alone is doing to price.

I ran this on real URA private resale transactions and HDB resale transactions from June 2021 to now — about five years of actual deals. Each market segment (CCR, RCR, OCR, and EC for private property; each town for HDB) gets its own fitted curve, using units with 90 to 99 years remaining as the 100% reference point, since that's as close to a fresh lease as the data allows.

How Fast Different Segments Actually Decay

Private property doesn't decay at one uniform rate. Prime districts hold up noticeably better than the city fringe. Here's the average decay rate per year of lease lost, measured from 90 years remaining down to 50:

Segment	Real decay rate (%/year)	What the official SLA table assumes
CCR (Core Central Region — Orchard, Marina)	0.95%/yr	Same curve for every segment, no market feedback
OCR (Outside Central Region — suburban condos)	1.01%/yr	Same curve for every segment, no market feedback
RCR (Rest of Central Region — city fringe)	1.10%/yr	Same curve for every segment, no market feedback

RCR decays the fastest of the three, CCR the slowest. Put simply: prime property holds its value better as the lease shortens. That's worth knowing if you're weighing a prime-district leasehold unit against a similarly priced city-fringe one — the lease clock is not ticking at the same speed.

It Gets More Specific Than That — Down to the District

Segment-level averages are useful, but they hide real variation underneath. I also fitted a separate curve for each of Singapore's 28 postal districts, wherever there was enough transaction data to trust the result. That worked for 24 of the 28 districts — the other four didn't have enough deals, so they fall back to their segment's average.

Two examples of what shows up once you go district-by-district: District 7 (part of CCR) decays at 1.06% a year — faster than the 0.95% CCR average. District 20 (part of RCR) decays at 1.16% a year — faster than the 1.10% RCR average. So two units in the same broad segment, even the same general part of town, can carry meaningfully different real decay rates depending on the specific district they sit in. If you're comparing two leasehold options in the same segment, the district-level number is the one to ask for, not the segment average.

HDB Isn't Immune Either — And the Town Matters

The same pattern holds for HDB resale. I fitted a separate curve for 23 of the 26 HDB towns that have resale data, and used the 5 official HDB planning regions (Central, East, North, North-East, West) as the fallback for towns with too little data of their own.

Two towns show how much this varies. Toa Payoh decays at 1.14% a year — noticeably faster than its region's average. Hougang decays at just 0.67% a year — noticeably slower. That's a genuinely useful finding for anyone hunting HDB resale: the specific town you're buying into matters more than which broad region it's in. Two flats with the same lease remaining, in the same region, can be losing value at very different speeds.

Does Flat Size or Bedroom Count Matter? Yes for Private, Not Clear Yet for HDB

I tested whether unit size changes the decay rate too. The private property result is real and worth knowing. The HDB result is genuinely inconclusive, and I'd rather tell you that plainly than dress it up as a finding.

Private property: smaller units hold their per-year value better, and the effect is bigger the closer you get to the city. Within OCR condos, 1-bedroom units decay at about 0.74% a year versus 1.00% a year for 3-bedroom units. Within RCR condos the gap widens a lot — 1-bedroom units decay at only about 0.41% a year, while 3- and 4-bedroom units decay at roughly 0.88–0.94% a year, more than double the rate. If you're buying a small leasehold unit in RCR as a long-term hold, this works in your favour more than the segment average would suggest.

HDB: the picture is mixed, and I tested only the three largest towns. In Woodlands, 3-Room, 4-Room and 5-Room flats decayed at nearly identical rates — a gap of about 0.1 percentage point, not meaningful. In Punggol, the spread across room types was much bigger. In Sengkang, I couldn't isolate a clean room-type effect at all, because different room types there are concentrated in different lease-age bands (older blocks tend to have a different room-type mix than newer ones), which muddies the comparison. My honest read: room type may matter for HDB decay in some towns, but the evidence isn't consistent enough yet to state it as a general rule. More research needed before I'd bet on this one.

This Curve Will Be Recalibrated — It's Not a Fixed Formula

Unlike the SLA table, which is a static formula, this curve is fitted from the last five years of actual transactions. That means it gets recalibrated as new data comes in, and it should — because buyer behavior genuinely shifts over time in ways that change how the market prices a shrinking lease. Pre-COVID buyers skewed more toward property as an investment vehicle; post-COVID there's been a much stronger shift toward buying for quality of life and owner-occupation. Those are different buyers with different risk tolerances for a short lease.

A data-driven curve like this captures the current pattern in how the market is actually behaving. It's a snapshot, not a guarantee of where things go next, because the underlying buyer preferences driving it keep evolving. I'll be re-running this as more transactions come in rather than treating today's numbers as permanent.

What This Model Can't Do Yet

Two honest limitations, because a model that hides its blind spots is more dangerous than one that names them.

First, it cannot yet cleanly separate pure lease decay from building age and en-bloc or tier-reset effects. A building's physical condition and aging amenities pull price down in a way that looks similar to lease decay but isn't the same thing, and a fresh 99-year lease from a recent Government Land Sales site resets the clock in a way this version doesn't fully untangle yet. Solving that needs transaction data reaching further back than 2021, which isn't in the dataset this curve is built from.

Second, it doesn't yet incorporate broader economic or demographic signals — things like consumer spending patterns or GDP per capita trends — that likely also shape how buyer preferences, and therefore lease-decay pricing, shift over time. That's a real direction for future work, deliberately left out of this version rather than bolted on half-finished.

What This Means If You're Actually Deciding

If you're buying, selling, or holding a leasehold property in Singapore, stop anchoring on the SLA/Bala table as if it tells you what the market will pay — it wasn't built for that job, and the gap between what it says and what buyers actually pay can run 20 percentage points wide at 67 years remaining. Use the segment, district, or town-level real decay rate instead, and if you're comparing a small unit against a large one in the same building, know that size changes the math too, especially in RCR. The lease is always shrinking. The only question worth asking is how fast the market you're in is actually pricing that in — and now you have a real answer instead of a formula built for something else.

Sources: URA private residential transaction data and HDB resale transaction data (via data.gov.sg / URA REALIS); Roncasa's own hedonic regression analysis of that transaction data, last fitted 25 Sep 2026.