How we form a view

An APD View is a structured research object attached to a market — a suburb, city, state, rental market, supply market or the national market. Each view states a direction, a confidence level, an evidence strength, the indicators behind it, the evidence that argues against it, the events we are watching next and the measurable conditions that would change the conclusion.

Views are computed from structured monthly observations. A model can interpret, summarise and challenge a view; it never sets the direction and it cannot publish one. An editor approves every published view.

Direction describes where underlying conditions are moving. It is not a rating, a recommendation or a signal to transact — a market can be improving while remaining expensive, low-yielding or supply-exposed.

• Improving rapidly
• Improving
• Stable
• Weakening
• Weakening rapidly

Each indicator is compared with its own history, oriented so that a positive contribution means underlying conditions are improving, weighted, and rolled into analytical dimensions — demand, rental, price, supply and affordability. The composite score maps to a fixed band, so the same data always produces the same direction.

Confidence reflects the strength of the evidence rather than the strength of the conclusion: how many independent indicators agree, the quality tier of the sources, how fresh the latest reading is, how much history exists and how consistent the indicators are with each other. A view can be improving with low confidence, and we say so.

Strong, moderate or limited, based on indicator count, independent source count, months of history and data age. We publish weak evidence as weak evidence rather than hiding it.

A formal APD View requires, at minimum:

• 5 independent indicators
• 1 or more independent sources
• 24 months of comparable history
• data no older than 150 days
• a completed counter-thesis
• editor approval

Where those requirements are not met, the market is marked no formal APD View with the reason shown. We do not manufacture a conclusion because a page expects one.

• Fact — Objectively sourced data.
• Estimate — Modelled or externally estimated — not an observed transaction.
• Calculation — Derived mathematically from known or user-provided inputs.
• Market signal — An interpretation derived from observed data.
• Scenario — A hypothetical assumption, not a forecast.
• Editorial analysis — Australia Property Data interpretation of the available evidence.
• Forecast — A forecast, clearly labelled and not a guarantee.

• Tier 1 — Official government, regulator or primary dataset (ABS, RBA, APRA, state planning and revenue offices).
• Tier 2 — Established licensed property datasets and professional research.
• Tier 3 — Major reputable media and research houses.
• Tier 4 — Industry commentary and company-supplied data.
• Tier 5 — Unverified secondary sources.

View generation favours higher-tier sources. We do not treat every source as equal.

Every indicator is shown against its own historical average and, where enough history exists, its percentile within the available record. A sixth-percentile vacancy reading means vacancy has been lower than roughly 94 per cent of observations we hold. A percentile describes how unusual a reading is — it is not a measure of investment quality.

Scenario outputs use assumptions you choose and are labelled as scenarios, not forecasts. Modelled values are labelled as estimates. Calculations derived from your inputs are labelled as calculations. Where a forecast appears it is attributed and labelled.

Published views carry machine-readable watch conditions. When one is crossed, the system raises an internal review event — it never flips public research automatically. Every published view is archived with its evidence, counter-thesis and watch conditions, and a direction change requires a written explanation of what changed our view.

Market observations are monthly series and lag the market. Benchmark rates and official releases update on their publishers' schedules. Thin transaction markets produce unstable medians. Suburb-level readings are less reliable than city or state readings. Nothing here is personal financial or credit advice, and no research output takes your circumstances into account.