The work
Three studies, published with their limits.
Every figure here comes from matched buy-and-sell pairs, the same unit bought and later sold, so each row is one owner's real outcome rather than a point on an index. Sample sizes are printed. So is the part of each study that does not support what I would like it to.
Study 01
5.6% vs 2.9%
The entry price law
What decides whether a resale property grows. Price paid against the same district in the same year beats freehold, floor height and postcode, and it is not close.
13,732 private + 99,705 HDB pairs
Read the studyStudy 02
0 of 584
New launch against resale
Not one new launch in the record was ever priced more than 20 percent below its own district. Resale did it 658 times. The gap is a price story, not a product story.
2,777 matched pairs held over three years
Read the studyStudy 03
45.4%
What the land already cost
Land is about 45 cents of every dollar of a launch price. 2026 land is the dearest on record and none of it has launched yet.
39 government land sale sites matched by hand
Read the studyMethod
Why matched pairs, and not a price index.
An index tells you what the average price did. It cannot tell you what happened to a person, because the thing being averaged changes every quarter as different homes transact.
A matched pair is the same unit bought and later sold by the same owner. It gives a real holding period, a real entry price and a real outcome, including the ones that lost money. It is more work and a smaller sample, and it is the only version worth putting in front of someone who is about to commit.
Every chart carries its own source and pull date. URA restates caveats for weeks after a month closes, so a figure quoted today can move, and a stale chart is visible rather than hidden.
Standing rule
The limits go on the page, not in the footnotes.
Each study ends with what it cannot support. A sample of 56 that flatters the argument does not get used. A five year private record does not get blended with a thirty six year HDB one and called the same thing. A result about resale does not get quoted as a result about launches.
That section is not modesty. It is the reason the rest of the numbers are worth anything.
Where these numbers come from
Computed from URA caveat data, URA government land sale tender results and the HDB resale register, pulled 12 September 2026. Past transactions describe what happened to other owners and are not a forecast for any particular unit. Nothing here is financial advice.