Everyone Says the Market Has Split in Two. Our Reading of the Indices Says It Split Into Four

A couple in their late thirties own a four-room flat they are perfectly happy in. Good floor, good town, mortgage comfortably serviced. Neither of them has ever thought of it as a mistake.

Then at a wedding dinner somebody says the markets are decoupling. Private is running away and HDB is stalling, and if you do not move across now you never will — your flat will be worth less relative to the condo you were going to buy every year you wait.

They go home and look at the headlines and the headlines agree. Three straight quarters of HDB decline. Seven straight quarters of private growth. And a decision that was never on the table becomes, over the course of one week, the only thing they can think about.

That is the frenzy The Straits Times is warning about, and it is worth taking seriously — because a fear of permanent exclusion is the single most reliable way to get people to buy something in a hurry.

We read URA's and HDB's official quarterly indices directly rather than re-derive them, so we can check a claim like this against the same source it came from.

HDB's own Resale Price Index stood at 202.8 in 2026Q2 — down 0.3% on the quarter. The run behind it:

Three consecutive declines from the 2025Q3 peak, and year-on-year the index is flat, at -0.0%.

URA's private Property Price Index sat at 219.4 in the same quarter, up 0.5% on the quarter and 2.9% on the year.

So the observation is correct. One index is going sideways-to-down, the other is going up. The columnist's numbers check out against ours.

The HDB medians say the same thing in dollars. Across the 2026 H2 half-year, every flat type is softer year-on-year: 2-room $370k (-1.3%), 3-room $436k (-3.0%), 4-room $628k (-0.3%), 5-room $728k (-1.6%), Executive $894k (-1.5%). That is not a collapse. It is a market that has stopped rising.

Here is where we would push back on the framing, and it is the reason this piece exists.

"Private" is not a market. URA publishes the private index by segment, and the segments are doing very different things.

Read that column top to bottom. The distance between landed at +7.0% and central-region condos at +0.5% is far larger than the distance between central-region condos at +0.5% and HDB resale at -0.0%.

Which means the story most people are being told — two markets, pulling apart — has the wrong cut. On a year-on-year basis an RCR or CCR condo owner has had more in common with an HDB owner than with a landed owner.

If there is a decoupling here, the seam is not between public and private housing. It is between landed and suburban condos on one side, and central condos and HDB flats on the other.

Zoom in from the year to the quarter and the picture gets more awkward for the simple version.

In the very quarter that produced the +0.5% headline for private property, the RCR fell 1.2% and the OCR was slightly negative. The rise was carried by the CCR, up 1.8%, and by landed, which URA reports islandwide.

So the quarter that is being used as evidence that private housing is pulling away from HDB is a quarter in which most non-landed private housing did not, in fact, go up.

That inverts the usual reading, and it deserves a caution of its own: one quarter of a regional index is a small, noisy thing, and the most recent quarter can be a flash estimate that URA revises later. We would not build a decision on it. But it is exactly as valid as the single-quarter comparison being used to argue the opposite.

The commentary's own framing gets at this: the resale HDB and private markets stay linked through overlapping buyer pools, households upgrading and downgrading, and the shared expectation that a home preserves wealth.

That is the part the frenzy logic misses. A permanent decoupling would require the flow of people between the two markets to stop. It has not. The HDB upgrader is the largest single source of demand for a suburban condo, and the money for the condo comes from selling the flat.

Which sets up the awkward loop inside the fear. If HDB resale prices soften, the upgrader has less to bring across. If enough upgraders rush anyway, they push suburban condo prices up and HDB resale supply up at the same time — widening exactly the gap they were rushing to get ahead of.

The fear, acted on at scale, is the thing that would make the fear true. That is not a reason to dismiss it. It is a reason to be very careful about acting on it in a hurry.

The fear assumes private property as a block is accelerating beyond reach. The segment breakdown does not support that for the homes most upgraders buy.

An HDB upgrader typically moves into a suburban or city-fringe condo — OCR or RCR. Over the past year, RCR is up 0.6% and OCR is up 3.9%, and in the latest quarter both were flat-to-negative. The 7.0% figure is landed housing, which is not the alternative to a four-room flat, and is largely closed to non-citizens besides.

So the honest version of the trade is much less dramatic than the headline: you would be selling into a softer HDB market to buy into a condo market that has risen modestly and, in the most recent quarter, not at all in two of three regions.

There are perfectly good reasons to upgrade — space, location, a school, wanting to. "Before the door closes forever" is not one this data supports, and a hurried purchase made on that basis carries costs — stamp duties, the transaction itself, a larger mortgage — that a few tenths of a percent of index divergence does not come close to covering.

A national index tells you what the aggregate did. It does not tell you what your flat did.

The index fell 0.3% in the latest quarter and is flat year-on-year. Underneath that, individual towns, flat types and lease positions move in different directions, and the composition of what happened to sell in a given quarter affects the reading.

The medians make the point about spread: $436k for a national median 3-room against $894k for an Executive, in the same half-year. Your flat's storey, its remaining lease, its town and its exact location within that town will each move it away from any national figure.

The useful question is not "what did the index do". It is what comparable flats in your block and your town have actually transacted at recently, and how your remaining lease compares to theirs.

The framing we are responding to — the decoupling talk, the risk of a buyer frenzy, and the point that the two markets remain linked through overlapping buyer pools, upgrading and downgrading, and the expectation that homes preserve wealth — is from The Straits Times' commentary, linked below. That is the columnist's argument, attributed rather than claimed as ours.

The figures are ours to the extent that we read them directly: the HDB Resale Price Index and national median resale prices by flat type from HDB's own published statistics, and the URA Property Price Index, its regional non-landed breakdown and the landed figure from URA's quarterly private residential statistics. We read the official indices; we do not re-derive them.

The caveats are set out in full in the reality-check section, and they are not decoration. These are indices rebased to 2009Q1 = 100, not dollar prices. HDB and URA publish separate indices that should be compared on direction, not level. The most recent quarter can be a flash estimate subject to revision. Regional figures are non-landed only; landed is reported islandwide. And the median resale prices are a dated half-year snapshot across all towns, gross of levies and fees, with the current half-year excluded until enough resales lodge.

The news:

• The Straits Times — Talk of private and resale HDB markets decoupling could spark buyer frenzy, 11 August 2026.

The data:

• HDB Resale Price Index and national median resale prices by flat type, HDB official statistics — 2026Q2 and 2026 H2 respectively.
• URA Property Price Index, regional non-landed price index and landed price index, URA quarterly private residential statistics — 2026Q2.

This is commentary by the PropKaki Editorial Desk, responding to a commentary published by The Straits Times. That argument belongs to that newsroom; the segment analysis, the framing and the opinions here are ours, and we have tried to disagree with the framing rather than the reporting.

It is opinion and general information, not financial, legal, tax or property advice. Nothing here is a recommendation to buy, sell, hold or upgrade, and index readings — particularly single-quarter ones, particularly provisional ones — are a poor basis for a decision of that size. Anyone weighing an upgrade should look at comparable transactions for their own flat and the specific homes they are considering, and take professional advice.

Published 12 August 2026.