GUIDES & INSIGHTS · MYTH BUSTER

“I gladly paid the ABSD.” Is that actually wise?

He was not complaining. That was the interesting part. A Singaporean buyer, second property, 20% Additional Buyer’s Stamp Duty written off without a flinch — because, he said, it is freehold landed, and freehold landed always wins. It is a common belief and it is not stupid. It is also only true under conditions most people never check.

Farhan Adenan · 7 Aug 2026 · 9 min read

ABSD rates are those in force since 27 April 2023 as published by IRAS and corroborated across independent guides. Rates change with cooling measures — verify before transacting. Break-even ranges are illustrative arithmetic on stated growth assumptions, not forecasts.

There is no freehold rate and no landed rate. ABSD is set by who you are and how many residential properties you already own. Tenure does not appear in the table. Neither does landed versus non-landed. A freehold bungalow and a 99-year studio attract exactly the same 20% for the same buyer.

Which means the sentence contains a hidden assumption worth surfacing. “I gladly paid the ABSD because it is freehold landed” only makes sense if the freehold landed part is expected to earn the 20% back. It is not a discount. It is a claim about future returns — and claims about future returns can be tested.

All-in = price + ABSD + Buyer’s Stamp Duty on the current residential scale (1% / 2% / 3% / 4% / 5% / 6%, top two bands from 15 Feb 2023). Legal fees, valuation and renovation excluded.

Here is the trap inside the logic. People reason about ABSD as though it were a fixed toll — pay it once, get access to the better asset. It is not fixed. It is a percentage, so choosing the more expensive, more prestigious, more freehold property increases the penalty in lockstep. The very quality being used to justify the cost is what makes the cost larger.

And it is dead capital in the strictest sense. You cannot borrow against it, it earns nothing, it is not deductible against rental income, and it does not come back when you sell. On a $4M house, $800,000 leaves your account and never appears in the asset. It is recovered only if the property appreciates enough to cover it — which is an entirely different proposition from “freehold landed holds its value.”

Five to eight years to get back to zero is the honest headline. That is not a reason to refuse — plenty of good decisions take a decade. It is a reason to be clear about what you have actually signed up for: this is a long-hold commitment, and anyone treating a second property as a medium-term trade has mispriced their own patience.

Add Seller’s Stamp Duty and the first years are worse still. Sell inside three years and SSD lands on top of an ABSD you have not yet recovered. The structure of the tax system is telling you, quite loudly, that this is a hold-for-a-decade decision. The buyer who says he gladly paid is right to be relaxed only if he genuinely means to stay that long.

Freehold landed is genuinely scarce, and scarcity is the one thing policy cannot manufacture. Landed housing is a small fraction of Singapore’s residential stock, restricted to Singaporeans for most of it, and effectively fixed in supply — the state is not zoning new landed estates. Everything else being built is high-rise. That is a real structural argument and it deserves respect.

No lease decay means no built-in downward escalator. Our own decay curve work shows what happens to leasehold exits as the remaining term shortens. Freehold sidesteps that entirely. Over a thirty-year hold, that difference compounds into something significant, and it is the strongest single point in the buyer’s favour.

And landed is the one segment where you own the land itself. Redevelopment optionality, the ability to rebuild — see our terrace rebuild guide — and the plain fact that they are not making more of it. If you are going to pay 20% for anything, an asset class with fixed supply and no expiry is a defensible thing to pay it for.

1. “Freehold landed always appreciates” hides enormous variance. Landed is not one market. It is dozens of micro-markets where road width, plot shape, level differences, drainage reserves and the road line plan can move value by six figures on otherwise identical plots. The segment average tells you almost nothing about the specific house. A good landed purchase is an outstanding asset; a poorly-chosen one is an illiquid, expensive, slow-selling problem.

2. Liquidity is the cost nobody prices. Landed transacts thinly. When you want out, the buyer pool is small, largely restricted to Singaporeans, and highly specific about what it wants. Compare that to a mainstream condo where a hundred near-identical units trade a year and the bank knows exactly what to value it at. Paying 20% to enter a thin market is a real risk, not a badge.

3. The comparison is almost never run against the alternative. The relevant question is not “is freehold landed good?” It is “is freehold landed plus $800,000 of dead capital better than what that same $4.8M does elsewhere?” — including a first-property upgrade with no ABSD at all, or decoupling done properly, or simply a smaller second property where 20% is a smaller absolute cheque. Most people who paid gladly never ran that comparison. That is the actual error, and it is not about tenure at all.

The instinct is sound. Freehold landed is scarce, does not decay, and has been a good long-term store of value in Singapore. If you are buying a second property to hold for twenty years and hand to your children, it is a rational thing to want.

The reasoning is not. ABSD is not a fee you pay for freehold landed — it is the same rate on anything, and it gets bigger precisely because landed costs more. Saying “I gladly paid it because it is freehold landed” is a reason to have chosen the asset, not a reason the tax was cheap. Those are different sentences and only one of them survives contact with a spreadsheet.

The test, in one question: if ABSD were abolished tomorrow, would you still buy this exact house at this exact price? If yes, you made an asset decision and the tax was a cost of entry you accepted with open eyes. If no — if the 20% was rationalised after the fact by the quality of the asset — then the tax made the decision, and it made it in the wrong direction.

If you own

If this is a home you will hold for decades, the argument is defensible

Freehold landed is genuinely scarce, it does not decay on a lease, and under the Residential Property Act the pool permitted to buy it is narrow by law. Over a long enough hold, an entry cost amortises and those characteristics do not go away.

The condition is the hold. At 3% growth the duty alone takes 7.4 years to earn back on a $4 million house. If you can say honestly that you are holding past that, the tax is a price of entry rather than a mistake.

If you invest

If you are buying it as an investment, run the comparison you have been avoiding

$979,600 of duty is unborrowable, earns nothing, is not deductible against rental income, and does not come back when you sell. The question is not whether landed is a good asset. It is whether landed-minus-the-duty beats the alternatives — including a first property in another name at 0% ABSD, if your household still has one available.

And check the exit before the entry. Seller’s Stamp Duty runs four years on anything acquired from 4 July 2025, at 16%, 12%, 8% and 4%. Sell in year two and you have paid the duty twice.

MARKET & POLICIES · INSIGHT 003

The decay curve: what your 99-year lease actually loses, decade by decade

The exit record by lease decade — from triple digits to 49% — and the freehold twist nobody expects.

GUIDES & INSIGHTS · GUIDE 001

Rebuilding your first terrace: what it actually costs, and how long it actually takes.

$400 psf to build, 25% more to finish, and 18–24 months of which only half is construction. The costs and rules before you commit to a plot.

Dataset — ABSD rates are those in force since 27 April 2023 as published by IRAS and corroborated across independent guides. Rates change with cooling measures — verify before transacting. Break-even ranges are illustrative arithmetic on stated growth assumptions, not forecasts.