Growth signal research:
20 thresholds, 25 years.
Every metric SuburbIQ shows has been tested against historical 12-month forward growth across the Australian transactional dataset. Here's how each signal earns its threshold — what counts as green, amber, or red — and how those thresholds map to our BUY / REVIEW / WALK verdict.
How a threshold is built.
Take vacancy rate as an example. We look at every suburb-month in the HTAG dataset (roughly 4.7 million observations). For each month, we record the vacancy rate and the next 12 months of price growth. Group the observations by vacancy decile. The lowest deciles (vacancy under 1%) correlate with average forward growth of around +4.8%. The highest deciles (vacancy above 3%) correlate with average forward growth of around -0.6%. We set the BUY threshold at the median of the top three deciles — roughly 1.5% vacancy. Above that level, we flag amber. Above 2.5%, red.
The 20 tested signals.
Stock on market, days on market, vacancy rate, gross rental yield, rent growth, search index, growth annualised over 1Y, growth annualised over 3Y, hold period, clearance rate, IRSAD decile, GRC (Growth Risk Composite), GPD (Growth Potential Differential), GSP (Growth Slope), RCS (Recovery Cycle Score), DDI (Demand Depth Index), STGS (Short-Term Growth Signal), BBT (Bull-Bear Tide), MAD index (Market Activity Differential), and EDI (Economic Demand Index). Each has a published threshold.
How they combine.
The verdict (BUY / REVIEW / WALK) isn't a single rule. It's a hierarchy. First, the cycle stage must be Recovery or Mid for BUY. Late-cycle peak is automatic REVIEW or WALK regardless of other signals. Next, structural depth (RCS, DDI) must clear the BUY threshold. Then short-term signals (STGS, BBT) confirm or downgrade. Two or more red signals in different categories downgrades from BUY to REVIEW. Three red signals downgrades to WALK.
Where the thresholds fail.
Suburbs with fewer than 30 sales per year have unreliable signal readings — high variance, prone to one-transaction skew. We flag these as MISSING in the report rather than score them. Prestige markets (median above $3M) have different vacancy and yield dynamics than mass-market suburbs; we apply a separate calibrated threshold band. Brand-new estates without three years of sales history are scored on price-band peers rather than direct signals.
Why this matters.
The threshold is the difference between data and information. A raw vacancy figure of 1.8% is data. The same number labelled "amber — historically associated with 1.2% forward growth, sample 412,000 observations" is information. The whole point of SuburbIQ is the second one.