Million-dollar HDB sales hit a record as broader resale prices dip

A record 596 resale HDB flats changed hands for at least S$1 million in July–September 2026, according to ERA Singapore’s analysis of public transaction records. Yet HDB’s final figures released on October 23 put its Resale Price Index at 202.5, down 0.1% from the previous quarter—its third consecutive quarterly decline.

For someone searching for a home, those numbers can feel like two different accounts of the same market. The useful distinction is that one counts transactions above a dollar threshold; the other tracks broad price movements while accounting for the mix of flats sold. Million-dollar sales represented about 7.8% of the quarter’s resale transactions, according to ERA Singapore Key Executive Officer Eugene Lim. They were a record-setting minority, rather than the typical sale.

A record at the top, a small dip across the market

The quarterly index decline followed falls of 0.1% in the first quarter and 0.3% in the second. That makes three successive declines, but the latest movement was modest: a tenth of a percentage point across the market, not a statement that every flat became noticeably cheaper.

HDB’s index measures price movements across resale transactions, taking account of differences in the flats being sold. That matters because the homes changing hands are not an identical basket each quarter. A quarter with more expensive flats sold should not automatically be read as a quarter in which comparable homes all became more expensive.

The million-dollar tally asks a simpler question: how many recorded transactions reached or crossed S$1 million? It does not measure how much the same kinds of flats increased in value, nor how prices moved below that threshold.

Both results can therefore coexist. More transactions can clear S$1 million while the broader, mix-adjusted measure edges lower. The apparent contradiction comes from treating a count of high-priced deals as though it were a measure of price changes throughout the market.

What 7.8% looks like in everyday terms

ERA’s reported 7.8% share translates to roughly one in 13 resale transactions reaching S$1 million. Put another way, about 92 in every 100 transactions were below that mark. Those are rounded calculations from the reported share, not a separate transaction count.

That comparison gives the record some proportion without dismissing it. Nearly 600 purchases at that price are substantial. But a buyer whose search is below S$1 million is still looking within the price range where the large majority of that quarter’s transactions took place.

The share also tells us something the raw tally cannot. A count describes how many expensive transactions occurred; a share describes how common they were among all transactions. Reading the two together is more useful than treating “record” as a description of every buyer’s experience.

There is an important boundary here: below S$1 million is not the same as affordable. The tally sorts transactions by one price threshold. It does not measure what a particular household can comfortably pay, and it offers no breakdown of the financial commitments of the households involved.

The expensive deals were concentrated, not evenly spread

ERA’s earlier analysis, using data available on September 30, counted 597 million-dollar transactions, of which 519 were in mature estates. Within that snapshot, Toa Payoh recorded 93, Queenstown 86 and Bukit Merah 56.

Those three towns together accounted for 235 transactions—about 39% of that earlier tally. Mature estates accounted for approximately 87%. These calculations show how geographically concentrated the high-priced sales were; they should be read using ERA’s 597-deal snapshot, rather than mixed with the later 596 figure.

That concentration helps explain why a national headline may feel very different from a household’s own search. Someone considering a flat in one of the concentrations is looking at a different set of available homes from someone searching elsewhere. Even within a town, the transaction records distinguish flat type, floor area, storey range and remaining lease.

A town total is therefore a starting point, not a price tag for every flat within its boundaries. The useful next comparison is between homes with similar characteristics—not simply between a flat for sale and the most eye-catching transaction carrying the same town name. HDB’s public dataset itself cautions that individual prices depend on many factors.

Why the count is 596 in one account and 597 in another

The later reported tally was 596, while ERA’s October 1 commentary recorded 597, based on data available as at September 30. The difference is one transaction, or less than 0.2% of the earlier total. It leaves the broad scale—about 600 million-dollar deals—essentially unchanged.

Still, the distinction deserves to stay attached to the numbers. The 596 figure is ERA’s analysis of public transaction records, not a headline statistic published by HDB alongside its price index. ERA’s town breakdown belongs to its earlier data snapshot.

The underlying HDB resale dataset is updated daily. Different extraction dates can therefore matter when comparing tallies, although the supplied accounts do not establish exactly which transaction explains this one-deal difference. The sound approach is to identify the source and cut-off, rather than quietly blend the two versions.

This is also why the September 29 flash transaction total should not be used to independently reconstruct the later reported 7.8% share. A denominator from one reporting date and a high-price count from another need not describe precisely the same set of transactions.

What the two numbers can tell you about a particular flat

For a household weighing an actual purchase or sale, the index and the million-dollar count answer different preliminary questions. The index describes the broad direction of resale prices. The tally shows how often transactions crossed a conspicuous price threshold, with ERA’s breakdown revealing where many of those transactions occurred.

Neither replaces a comparison with recent transactions for similar homes. The public records allow a closer look at location, flat type, size, storey range and remaining lease—the details that disappear when thousands of transactions become one quarterly headline. HDB also warns that recorded prices are indicative because individual deals depend on multiple factors.

The clearest reading of the quarter is therefore quite specific: million-dollar HDB transactions reached a record, while broad resale prices edged down. For buyers and sellers, the remaining question is local: what have comparable flats actually sold for? That is where these national numbers become a useful conversation about a particular home.