Housing bubble · updated September 2026
Our call · Our lean
Not a national bubble — a localised correction. The froth capitals (Adelaide, Brisbane) are exposed; the majors are past the worst.

Is Australia in a housing bubble, and will prices crash?

It's not one bubble, and that's the key insight. We separate two questions people blur: is it expensive? (versus what households can borrow — yes, every capital is stretched) and is it running hot? (versus what it costs to build — only some cities are). A uniform national crash call misreads the market.

On our model the fresh over-running is concentrated in the smaller capitals that led the recent boom, while the big, expensive majors have already stalled — and in the latest data are gently falling. So the correction risk is real but localised: it's a froth unwind in the cities that ran hardest, not a synchronised collapse.

Which Australian cities are most at risk of a house-price correction?

The ones that ran hardest and now sit furthest above cost-adjusted fair value — the smaller capitals that led the boom. Counter-intuitively, the 'expensive' majors that have stalled are more defensive, because they've already given back their froth. The precise ranking and stress-test downside for each city are in our report.

How does MPLC know the difference between expensive and a bubble?

By adding replacement cost — what it actually costs to build — to the usual affordability lens. That's the piece most doom models miss: once you account for how much dearer building got, a permanent-sounding 'overvaluation' shrinks to a credible, cyclical range. Calibrated blind on pre-2020 data, the model flagged the 2021 boom and the 2023 correction — so its bubble-versus-expensive call has an out-of-sample track record.

The precise numbers behind this — fair value, the exact gap, five-year forecasts and the downside scenarios — are in the model. Why our model is different →  ·  Get the free weekly brief →