São Paulo new-build flat snapshot — July 2026 methodology

I

ink.awake

Property investor
I’m deciding whether this is ready to publish as a July 2026 São Paulo snapshot or should remain a working note. Current indications for new-build flats are 88 days on market, asking-price movement of -5.9%, and visible financing sensitivity around R$5,572,000. These are not an official index. What sample definitions, completed-sale evidence, inventory changes and neighbourhood splits are needed before the figures are useful?
 
Keep it as a working note for now. The figures may be useful, but only after defining the listing universe, observation period and treatment of duplicate or relisted units. Otherwise 88 days and -5.9% look more precise than the underlying sample may justify.
 
At minimum, I’d put a short methodology box above the numbers: new-build only, which São Paulo neighbourhoods, active versus withdrawn listings, and the date the data were collected. Also say whether the asking-price movement compares identical units or changing monthly samples.
 
What period does -5.9% cover? Month on month, since first listing, or something else? That missing denominator changes the interpretation completely. It would also help to know whether reductions are measured per flat or from the aggregate asking-price mix.
 
The R$5,572,000 point needs similar clarification. Is it a sample boundary, a cluster where financing-dependent demand changes, or simply the price of one observed listing? I would not call it a market threshold without seeing how that conclusion was reached.
 
Agreed. A compact table could separate metric, definition, covered dates, sample count and last revision. That would answer Oscar’s period question and prevent the financing figure from being mistaken for a universal cutoff.
 
Completed sales will probably arrive on a different timetable from listing observations, so don’t force them into the same July cut if they are incomplete. Publish their completion dates separately and label any lag rather than treating asking and transacted prices as simultaneous.
 
I’m less convinced that publication must wait. A clearly labelled directional snapshot can still prompt useful contributions. The problem is not that it lacks completed sales today; it is whether readers can distinguish observed listing behaviour from conclusions that require sale evidence.
 
Fair. Then inventory should appear beside time on market. An 88-day figure means something different if available units are rising, falling or merely rotating through relistings. Even a statement that inventory change is not yet established would be better than leaving the relationship implicit.
 
“New-build flats” may still mix completed unsold units, off-plan releases and assignments. Those can have different marketing clocks and pricing practices. Can the sample be divided by development stage, or is that information unavailable?
 
Neighbourhood splits are essential before attaching this to São Paulo as a whole. I wouldn’t demand a figure for every district, but the summary should reveal whether the sample is broad or concentrated. A city-level number driven by a narrow cluster could mislead.
 
That concentration issue also affects my property-type question. If development stage cannot be separated, at least disclose the mix. I’d rather see an honest “mixed new-build sample” than neat subcategories built from too few observations.
 
Please preserve each revision date. If members later add completed-sale evidence, the original July 2026 listing snapshot should remain distinguishable from the amended version. Otherwise later information could appear to have been available at the initial publication date.
 
For any contributed link, note what claim it supports and when it was accessed. A link alone is not enough if it points to a changing listing page. Members can also state plainly when something is an on-the-ground observation rather than documentary evidence.
 
How is “days on market” handled when a unit disappears and returns? If the clock resets, 88 days could understate the marketing history. If it does not reset, there needs to be a reliable way to identify the same flat across listings.
 
Another complication: advertised reductions may not capture incentives that leave the headline price unchanged. That does not invalidate -5.9%, but it does mean the wording should remain “asking-price movement,” not buyer discount or achieved-price change.
 
Joana’s distinction should sit directly beside the figure, not in a footnote. Readers tend to carry a percentage forward without its qualifier. Something like “observed asking-price movement; completed-sale comparison pending” would keep the claim narrow.
 
There is also a calculation choice to disclose: is -5.9% the change in an aggregate figure, the average of unit-level changes, or a weighted result? I’m not arguing for one method, only against mixing them between revisions.
 
Duplicate developer, agent and portal listings could affect both inventory and time on market. If deduplication is uncertain, report listings rather than implying a count of unique flats. That wording change alone would make the limitations much clearer.
 
For completed sales, match as closely as possible on neighbourhood, development stage, unit size and timing. A citywide completed-sale figure placed beside a narrow high-price new-build listing sample could create an apples-to-oranges comparison.
 
You must log in or register to reply here.
Back
Top