Property Hotspot MethodologyWhat It Should Include — And What Most Lists Leave Out

Any hotspot methodology starts with this question. Are you trying to identify suburbs that willboom — before they do — or suburbs that are booming right now?

The instinct is to predict. Find the early signal, get in before the crowd. So we tested it. Infrastructure spending, population growth, and building approvals — the metrics that appear in most hotspot analyses — were used to build a prediction formula and backtested against real Australian suburb outcomes.

Fifty-five percent. Marginally better than a coin flip, and not by enough to act on. The reason isn’t that those signals are meaningless — infrastructure does affect property values. It’s that they operate at the wrong scale. Population shifts and infrastructure projects affect corridors over decades, not individual suburbs over the investment windows that matter.

Detection is a different question entirely. Instead of “will this suburb boom?” you ask “is this suburb booming right now?” That question has a measurable answer. And booms are long enough events that catching them 6–12 months after they start still captures 60–85% of total gains. With dramatically higher confidence.

BoomAU’s v2.3 score combines five components — Momentum (price growth acceleration), Growth Strength (annual growth against peers), Tightness (days on market plus current vacancy), Sustainability (rental yield plus vacancy trend), and Headroom (suburb median relative to the capital city median) — behind four hard filters that exclude stagnant or thin markets before any score is assigned. The full methodology is broken down step by step below.

Before a suburb is scored at all, it must pass four go/no-go filters. These are not scored components — they are hard gates. Fail any one and the suburb is not included.

These filters do two things. First, they exclude thin or stagnant markets where a score would be unreliable — a suburb that sells fewer than 30 homes a year produces a days-on-market figure that can swing wildly from one transaction to the next. Second, they focus the formula on exactly the segment of the market where backtesting found the strongest growth signals.

Of the 8,417 suburbs tracked nationally, 393 currently pass all four filters. Those are the only suburbs scored.

For the suburbs that pass the hard filters, five components combine into a detection score. Each component has a weight that reflects its contribution to boom detection accuracy in backtesting — not a judgment call.

Momentum carries the highest weight because price acceleration is the clearest observable sign of a boom already in progress — and the hardest to fake. Headroom carries the lowest formula weight (10%) as a scored component, but it shows up in the hard filters too and is the single strongest cross-suburb ranking signal in backtesting. Its influence is larger than the weight suggests.

The formula produces a score from 0 to 100. Four bands determine how a suburb is classified:

The thresholds were set by examining the score distribution across the 78-suburb backtest. The 80+ threshold cleanly separated genuine booms from non-booms — which is how the formula achieved a 20.2-point separation gap between the two groups.

Publishing a methodology without publishing backtest results is a claim without evidence. Here is the evidence.

The backtest used 78 suburbs — 28 that genuinely boomed and 50 controls that didn’t. The formula was applied to each suburb using only data available at the time of scoring, with no knowledge of the outcome. Each suburb was then classified as boom or no-boom and compared against what actually happened.

The 0% false positive rate is the result that matters most for investors. A formula that generates false signals — calling a suburb a boom when it isn’t — causes real harm. Capital deployed in a Pass-tier suburb underperforms by 6.4 percentage points per year on average. This formula generated no false signals across the backtest.

The 20.2-point separation gap confirms the formula isn’t just squeaking past a threshold. When it calls a suburb a boom, that suburb scored well above the threshold. When it didn’t call a boom, the suburb scored well below. Marginal calls are rare.

One important data quality constraint also came out of the backtest: suburbs with very low transaction volumes produce unreliable days-on-market figures. A suburb selling fewer than about 30 homes a year has a DOM median that reflects just a handful of sales — one unusually fast sale can drag it to 10 days, one slow listing can push it to 150. Below roughly 15 annual sales the DOM figure is not usable at all. This is partly why the hard filter exists, and partly why thin markets are excluded from scoring.

The 78-suburb backtest confirms the formula detects booms. A second backtest asks a harder question: do the four tiers (Strong Buy, Buy, Watch, Pass) actually predict which suburbs outperform over the next 12 months?

This used a walk-forward methodology across 12,360 postcode-months. Each suburb was scored using only data available at that point in time. Forward returns were measured 12 months later. No lookahead.

Excess return is measured against the market median — the average growth across all scored suburbs in the same period. A suburb that grew 12% when the market median grew 5% has an excess return of +7pp. This cancels the effect of broad market conditions and isolates suburb-level outperformance.

Walk-forward backtest, 12,360 postcode-months. No lookahead. Excess return = suburb 12-month growth minus market median growth. Full methodology →

The result is perfectly monotonic. Every tier outperforms the one below it — on excess return, on beat-market rate, and consistently across subsamples. Strong Buy suburbs beat the market 71% of the time with a +7.5pp excess return. Pass suburbs beat the market only 28% of the time with a −6.4pp excess return. That’s a 43-percentage-point spread in beat-market rate between the top and bottom tier — on the same asset class, in the same country, over the same time periods.

The monotonic result also means the tiers don’t need to be perfect. A methodology that reliably separates the top from the bottom — even if the middle tiers overlap somewhat — is actionable. This one separates cleanly across all four.

After running every available signal through backtesting, two things emerged that consistently predicted which suburbs would outperform after cancelling out the broad market tide.

What didn’t survive: past growth rate as a forward predictor (mean reversion dominates — past outperformers tend to underperform going forward), 5-year momentum, acceleration ratios, and anything that requires predicting rather than observing. These are not excluded because they sound wrong. They are excluded because they failed backtesting.

The two surviving signals are both checkable with free public data. You don’t need BoomAU to apply them to an individual suburb you’re already considering.

The hard part isn’t checking one suburb. It’s doing it across hundreds of suburbs, updated fortnightly, catching booms within weeks of starting rather than months after everyone already knows. That is what the formula automates — applied currently to 393 suburbs in three budget bands (under $400K, under $600K, under $800K).

The full backtest, the formula weights, the threshold derivation, and the walk-forward tier results are published on the proof page. No gating. Check the numbers yourself.