Does the market really follow Bala’s Table? The lease-decay curve
- This is the foundation of a fair-value model: before you can price MRT, schools or size, you have to strip out lease decay — the biggest, most non-linear discount in the market.
- We estimated the curve from transactions and the answer is reassuring: the market prices 99-year leasehold very close to Bala’s Table, applying about 92% of the theoretical discount. A lease with ~90 years left holds ~93% of freehold value, ~81% at 60, ~77% at 50.
- The tilt that matters: the market is kinder than the table to young leasehold (85–94 years trades at ~95% of freehold, richer than Bala’s 93%) but turns harsher below ~55 years — buyers barely dock a long lease, then punish short ones as the policy cliffs loom.
- The trick is freehold: freehold ages but doesn’t decay, so it pins the pure building-age effect — letting us isolate lease decay instead of confusing it with an old building.
- Newer isn’t just longer-lease: a series of URA rule changes (bay windows out in 2009, balcony caps in 2019, floor-area harmonisation in 2023) mean a newer unit packs more usable space into the same quoted sq ft. The base layer is two halves — the lease-decay curve for the lease, a technical-efficiency adjustment for the build — combined in the calculator below.
- Best for: anyone weighing a leasehold buy, and as the base layer for spotting undervalued projects.
Why lease decay has to come first
To find undervalued projects you need a model of fair value — a weight for every attribute (MRT, schools, size, floor, location). But those weights are impossible to read cleanly while lease decay bleeds into them: a short-lease unit is also an old unit, often in an older part of town, so its low price mixes lease, age and location. Get lease decay wrong and every other weight is contaminated. So we solve it first — and we do it by leaning on a trusted prior, then letting the data correct it.
The curve: our data vs Bala’s Table
Everything here is a % of freehold value. The dotted line at the top is freehold — worth 100% by definition, the ceiling. The thick green line is the leasehold value our data implies at each remaining lease, and the shaded band between them is the discount the market applies for holding a lease instead of freehold — thin when the lease is long, wider as it runs down. The faint amber dashes are Bala’s Table; the green line barely clears it, so the market follows the table closely. Key leases are labelled; hover for any point.
What a leasehold is worth vs freehold, by remaining lease
| Remaining lease | Market (our data) | Bala’s Table | Gap (pts) | Caveats |
|---|---|---|---|---|
| 85–94 yrs | 95.0% | 92.8% | +2.3 | 18,395 |
| 75–84 | 93.3% | 88.8% | +4.6 | 16,770 |
| 65–74 | 85.9% | 84.5% | +1.4 | 9,726 |
| 55–64 | 86.8% | 80.0% | +6.8 | 1,116 * |
| <55 yrs | 64.1% | 75.0% | -10.9 | 652 * |
The thresholds that bite
The decline isn’t smooth. Two policy cliffs shape it: around 60 years remaining, CPF usage starts to taper (full CPF needs the lease to cover the youngest buyer to age 95); and around 30 years, bank financing dries up. Buyers price these ahead of the cliff, so value can erode before the lease actually reaches the threshold — the discount is about your exit lease, not just today’s.
Recalibrating leasehold: a newer build is a different product
Lease length isn’t the only thing separating a 2008 unit from a 2024 one. A series of URA rule changes reshaped what you actually get for a given quoted square foot — so a newer lease is often a better product, not just a longer one. This is the second half of the base layer: the lease-decay curve above prices the lease; a technical-efficiency adjustment prices the build. Together they say what one unit is fairly worth versus another.
| Build era (by approval) | Technical dead space in the quoted sq ft | Usable* |
|---|---|---|
| Bay-window era · pre-2009 | Bay windows + planter boxes + aircon ledge + voids | ~91% |
| Middle era · 2009–2023 | Aircon ledge + voids (no bay windows/planters) | ~95% |
| Harmonised · 2023 on | Aircon ledges & voids excluded from strata | ~99% |
It isn’t only those two rules. Several URA changes have re-shaped what sits inside a quoted square foot over the years — worth knowing when you compare a resale to a new launch:
| Effective | Rule change | What it did to the quoted sq ft |
|---|---|---|
| 2001 | Balcony Incentive Scheme — balconies, PES & roof terraces as bonus GFA up to 10% | Semi-outdoor space sold at $psf; balconies grew larger |
| 1 Jan 2009 | Bay windows & internal planter boxes lose GFA exemption | The old ~5% of “phantom” indoor space vanishes from new builds |
| 17 Jan 2019 | Balcony bonus cut 10→7% & capped at 15% of internal area; min average unit size 70→85 sqm | Smaller balconies; fewer shoeboxes (larger units, so a lower headline $psf) |
| 1 Jun 2023 | Floor-area harmonisation — aircon ledges, voids & PES/PRT counted as GFA | ~5% less saleable area; the quoted sq ft is now mostly usable |
Net effect: for the same quoted area, a post-2023 unit gives you meaningfully more usable floor than a pre-2009 one, and the years in between carry their own quirks. That’s a permanent, structural reason a newer 99-year lease can be worth more per quoted sq ft — separate from the lease, and separate from the (fading) newness premium.
Pay-up calculator: what one unit is really worth vs another
Set each property’s tenure and year — a 99-year lease from its commencement year (e.g. “99 years from 2010”), or a freehold by its year built. We work out the balance lease and the build vintage for you, then combine the two halves of the base layer — the lease-decay curve and the technical-efficiency adjustment — to estimate how much more, per quoted sq ft, one is genuinely worth than the other, before any newness premium.
(B vs A) per sq ft
Investor verdict
Young leasehold (roughly 85+ years) is barely penalised — it holds ~95% of freehold value and, over a 5–10 year hold, behaves much like freehold. That’s the market being kinder than Bala, not harsher. The danger is buying into the steep part of the curve: below ~65 years the discount accelerates, and it compounds because your buyer inherits an even shorter lease.
So the real question isn’t “freehold or leasehold?” — it’s “what lease will be left when I sell?” A 99-year unit with 90 years left, held 8 years, still has 82 — comfortably above the cliffs. One with 68 left, held 8 years, exits at 60 — right where the market starts to punish.
How to use this before you buy
Price the exit lease, not the entry lease
Subtract your holding period from today’s remaining lease and read the curve at that point. If it drops you below ~60 years, expect a steeper discount when you sell.
Compare like-for-like against freehold
A young leasehold at ~95% of freehold for meaningfully less $psf is often the better cashflow buy; a freehold premium only earns its keep on a very long hold. See the leasehold vs freehold growth study.
Check it against the model, not the myth
Value the specific unit in the Valuation tool (it accounts for lease) and read the project’s remaining lease and flags on its page before you rely on “leasehold is fine” or “freehold always wins.”
Treat below-curve pricing as a lead, not a verdict
A unit priced below the lease curve may be genuinely cheap — or have a real defect. It’s a shortlist signal for the fair-value model we’re building, to be confirmed by due diligence.