How we compute suburb affordability

Every figure on our suburb affordability pages and in the Deposit Treadmill research is computed, not estimated — and computed directly from primary government records via an open pipeline, not a paid aggregator's black box. (The five earliest pilot suburbs were originally cross-checked against the major data providers; within ~2% of the government figures, they have since been migrated to the same government-direct method — more below.) This page documents exactly how, and what the data can and can't tell you.

Last reviewed 2026-05-30.

Where the medians come from

Median sale prices are computed directly from the NSW Valuer-General's Property Sales Information (PSI) — the public record of every property sale registered in New South Wales, roughly 2.2 million transactions over the last six years. This is the authoritative source the major data providers themselves license. We process it through an open, reproducible pipeline (the script is in our public repository, scripts/vg-suburb-medians.mjs), so any figure we publish can be re-derived from the government records by anyone.

How each median is calculated

Window: the trailing 12 months (to February 2026 for the current set) — the standard period for a stable median.
Residences only: we keep arms-length residence sales and apply a $50,000 price floor to drop nominal/$1 transfers; we exclude vacant land and non-residential sales.
House vs unit, never blended: each dwelling type gets its own median. We classify by strata title — strata-titled dwellings are counted as units, Torrens-title houses as houses. (Caveat below.)
Reliability floor: a dwelling type is only published if it clears a minimum sale-count threshold; below it, we suppress the figure rather than show a noisy median off a handful of sales.
Sample sizes shown: every page displays the number of sales behind each median, so you can judge how solid it is.

One honest classification caveat. The strata-title split is the standard proxy, and it's clean in genuine apartment markets. In a few house-dominated suburbs the strata stock is mostly townhouses, which can lift the “unit” median above what most people picture as an apartment. Where that applies we either disclose it or hold the suburb back until we can split apartments out specifically — we don't publish a misleading unit figure.

Stamp duty, deposit and concessions

Stamp duty is computed from the current state transfer-duty rates — the same calculation engine behind our stamp-duty calculator, kept current with each financial year's indexed thresholds (the NSW FY2025-26 scale is verified against Revenue NSW's own figures). For each suburb we show duty at both the standard rate and the first-home-buyer rate, including the tapered concession band where it applies. Deposit figures are simply 5% / 10% / 20% of the median, with the 20% line flagged as the point above which you typically avoid Lenders Mortgage Insurance.

We do not assert scheme eligibility (the 5% Deposit Scheme, First Home Super Saver, etc.) from price alone — those depend on income, citizenship, new-vs-established status and current price caps. We point you to the relevant scheme guide and the state revenue office instead of guessing.

The “deposit treadmill” model

Our Deposit Treadmill research answers “how long to save a deposit” — but unlike the usual flat figure, it models the deposit target rising as prices grow while you save. We take each suburb's 12-month price growth (also computed from the government records, current vs prior-12-month median) and find the first point where cumulative savings (at a stated monthly rate) meet the rising target. Where the target grows faster than savings accumulate, there is no such point — the deposit is a receding target. The “years” figure is a model output under a stated savings assumption, not a forecast.

Refresh cadence + provenance

The underlying government dataset updates daily; we recompute and re-verify on each publication and stamp every page with the data window and the date it was last sourced. Every suburb is now computed directly from the government records. The five earliest pilot suburbs were originally cross-checked across CoreLogic, PropTrack and Domain as a consensus — the two methods agreed within about 2% — and have since been migrated to the same government-direct method, so every figure is now first-party and independently reproducible.

What this does not do

It is not financial advice, and a median is a snapshot of a market, not a valuation of any specific property.
12-month growth rates are volatile and not a prediction — recent rates may not continue; the treadmill shows what happens if a rate holds, not that it will.
Eligibility for first-home-buyer concessions and schemes is determined by the relevant state revenue office and your lender, not by us.
Thin markets are suppressed rather than published — absence of a figure means “too few sales to be reliable”, not “no data”.

Suburb affordability by suburb — the pages this method produces.
The Deposit Treadmill — the original research (CC BY 4.0).
FHB scheme comparison methodology — the finance-scheme side of our transparency layer.