How Long Will This Downturn Last? The Honest Framework

Clearance improved for a fourth consecutive week — 56.5% on 1,309 auctions (949 reported: 537 cleared, 412 uncleared) — and for the first time in this run the improvement came from outside the big two. Volumes eased 5.7% on last week's early-spring bump, and the Domain/Wilson read firmed to 51.8%, a second straight week above the halfway line, with auction markets 'meandering higher as spring approaches'. The same weekend a year ago cleared roughly 73.5% — the level remains deep buyer's-market territory even as the direction improves.

The auction floor: week ending 16 August

• Melbourne: 57.4% on 596 scheduled (474 reported; 82 sold prior, 187 under the hammer, 3 after, 128 passed in, 74 withdrawn) — eased off last week's post-April high of 60.8%, but a month of prints between 57% and 61% at four-years-cheaper prices reads as a demand floor, not a bounce fading.
• Sydney: 55.6% on 425 scheduled (313 reported; 98 sold prior, 73 under the hammer, 3 after, 42 passed in) — eased 1.4 points, and the tell was withdrawals: back up from 68 to 97, or 22.8% of scheduled auctions. This is the third freeze-thaw cycle of the winter — vendors who won't meet the market keep pulling out rather than passing in, which flatters the clearance rate and thins real supply at auction.
• Brisbane: 51.9% on 79 reported — the week's genuine surprise: a 13.8-point jump and Brisbane's first print above the halfway line in the six-week trend window, with only 10 withdrawals. One week doesn't remake a soft market, but it was the broadening the combined number needed.
• Adelaide: 61.9% — the best major-capital print of the week, though on a thin 42-report sample from 88 scheduled (the Wilson read had 52.9%, down from 58.2%). The series keeps oscillating with reporting lags; the range says Adelaide clears somewhere in the 50s consistently.
• Canberra: 60.0% on 35 reported — the firmest Canberra print of the winter, a second straight improvement. Perth cleared 3 of 6 reported — a private-sale market whose price index tells the real story.

No Tasmanian auctions were captured — the combined figure is exactly the six-city sum.

The price backdrop: no new index print, a sharper scenario map

The official July HVI still frames the price story — national values -0.7% in July (the steepest month since December 2022, median $928,421), Sydney -1.4%/-4.0% for the quarter, Melbourne -1.2%/-3.4%, with the August print due about 1 September. What's new is the framing: Cotality's August chart pack opens with downturn scenarios — what -5%, -10%, -15% and -20% from each city's cyclical peak would mean — and Sydney and Melbourne are already more than 5% below theirs. Vendor discounting has widened to 3.9% across the capitals (4.2% in Sydney — roughly $52,000 of negotiating room on the median), days on market have stretched to 33, and gross yields have rebuilt to a three-year high. The same pack recorded investors at a decade-high 40.3% of new lending in the final quarter before the negative-gearing reforms — the high-water mark of the investor era that the tax change ends.

Supply: Perth joins the listings story

National listings entered spring 23% above year-ago levels, and this week the surge reached the last tight market: REIWA counted Perth listings above 6,000, up 111% year-on-year — while rentals still lease in a median 16 days. Perth's scarcity premium is fading from the sales side even as its rental market stays the tightest of the majors; that combination — normalising prices, intact yields — is the late-cycle shape the sub-$700K corridors in our August Top 10 are built on.

The RBA held at 4.35% on 11 August — unanimous, a second straight pause, and hawkish in the details. The statement keeps headline inflation 'still too high', notes the trimmed mean is elevated and little changed, and explicitly retains the option of 'increasing the cash rate target further if upside risks materialise'. The Board doesn't expect inflation back around the target midpoint until late 2027, and no major forecaster has a cut before 2027. The next meeting is 28–29 September; the plateau is the planning assumption.

The RBA's own August chart pack fills in why the Board can afford patience, and why the economy can absorb it. Growth is modest but positive — +0.3% for the March quarter, +2.5% annually — though GDP per capita is contracting: population growth is doing all the aggregate work, which is the productivity concern behind the Board's inflation caution. The labour market is loosening slowly rather than breaking: unemployment 4.4%, participation still high at 66.9%, job vacancies down 2.1% from February but at 329,500 still historically elevated. Household spending actually rose 0.8% in June.

Two chart-pack numbers matter most for property investors. First, housing credit is leading prices down: housing loan values fell 3.8% from their December 2025 peak to the March quarter — owner-occupier lending -4.3%, investor lending -3.0% — and credit contractions of this kind have historically preceded and paced price declines, which is why the 2017–19 template (a credit-tightening downturn) fits this cycle better than the 2022–23 rate-shock one. Second, the system has no crash mechanism: Australia's residential stock is valued around $12.6 trillion against $2.6 trillion of debt — a system-wide loan-to-value ratio near 21% — and half of homeowners have no mortgage at all. Meanwhile first-home buyers took 29.0% of owner-occupier lending (against a 27.6% decade average), strongest in the ACT (37%) and NT (36.6%) — the demand rotation the reforms intended is visible in the credit data before the reforms have even taken price effect.

The practical positions stand: this is the top of the cycle, not the start of the descent; borrowing capacity stays roughly 12–20% below its 2024 peak with serviceability assessed near 9.4%; and falling prices are doing the affordability work that rates won't. Negotiate on price, not on hoped-for rate relief.

SQM's July print landed this week and held the national vacancy rate at 1.3% for a third month (40,771 vacant dwellings) — but the national number is hiding a two-speed market that keeps widening. Sydney (1.7%, +0.2 points year-on-year), Melbourne (1.7%) and Canberra (1.8%, the loosest capital) are gradually loosening; Adelaide tightened to 0.6% (from 0.7%), Hobart to 0.6%, and Darwin at 0.3% remains the tightest rental market in the country. Five of eight capitals sit below 1% vacancy — far below any definition of balance.

Rents keep repricing accordingly. July asking rents rose in five of eight capitals for houses through mid-winter — Canberra +2.1% for the month ($745), Melbourne +1.6% ($620), Brisbane +1.3% ($709), Adelaide +1.1% ($663), Sydney +0.3% to a national-high $878 — and the unit story remains the sharper one: Sydney units jumped 3.4% in a single month to $845 (+12.7% for the year), and Darwin is the fastest rental market in the country (houses +13.3%, units +18.2% annually). National asking rents are up about 7.2% over the year.

The investment arithmetic writes itself: rents rising 7% against prices falling is how entry yields repair, and combined-capital gross yields have now rebuilt to a three-year high. The forward supply picture extends it — investor lending is contracting (-3.0% from the December peak on the RBA's data) and the tax reform pulls more investors out from July 2027, which damages future rental supply rather than current occupancy. The two-speed split does carry an underwriting instruction: in the loosening capitals (Sydney, Melbourne, Canberra), stop assuming rent growth above wage growth — the vacancy direction says tenants are finding alternatives; in the sub-1% capitals, the floor under rents remains structural. (Vacancy: SQM, July; asking rents: July capital-city report.)

How long will this downturn last? The honest framework

Every downturn eventually reduces to one question, and this one reached it in August: how long? Four months in, the national index is down roughly 1.9% from its peak, the combined capitals about 2.5%, and Sydney and Melbourne are already more than 5% below their highs. Cotality's own chart pack now opens with downturn scenarios — the data house that spent two years publishing record highs is publishing fall maps. Here is the framework we'd actually use, rather than a single number.

What history says

Australia has never recorded a double-digit national price fall. The deepest correction on record — 2017–19, a credit-tightening episode like this one — took national values down about 9% over roughly two years. The 2022–23 rate shock was faster and shallower: values reversed inside a year once the rate cycle turned. Those two templates bound the plausible range. The bad news for anyone hoping for the 2022 template: that recovery was rescued by imminent rate cuts, and this time no cut is priced before 2027. The structure of this cycle — three hikes, a tax reform pulling investors from premium stock, housing credit down 3.8% from its December 2025 peak — looks much more like 2017–19 than 2022–23. Plan on duration, not a V.

What the forecasters say — and why the spread is the message

Domain sees about -3% over 15 months, the mildest call. NAB has capitals -5% in 2026, with Sydney around -10% and Melbourne -9% peak-to-trough. CBA sits near -7% through 2026. ANZ is deepest and longest at -10.6% by end-2027. Note what even the bear case is: every one of those numbers is a correction, not a crash — ANZ's floor would return capital-city values roughly to early-2024 levels. The crash mechanism is genuinely absent: half of homeowners have no mortgage, system-wide debt is about $2.6 trillion against $12.6 trillion of housing stock, and 4.4% unemployment isn't producing forced sellers. Downturns without forced sellers are slow repricings, not liquidations.

What ends it — three forces, in order of power

1. Rates. The 2019 and 2023 recoveries both started when the rate cycle turned. Current pricing says 2027 — the calendar's hard constraint on any recovery date. A faster inflation fall drags that forward; the RBA's retained tightening bias pushes it back.

2. Sentiment. Consumer confidence in housing has collapsed from 62% positive in late 2024 to around 24%. Sentiment troughs historically precede price troughs — and sentiment can turn on a single credible signal (the first cut, or three months of stabilising prints) far faster than fundamentals can.

3. The spring supply test. Listings entered spring 23% above year-ago levels, with Perth's now up 111%. The next three months tell us whether buyer depth can absorb an early, heavy selling season. This is the single most informative window of the cycle: if clearance holds its four-week improvement against the listings wave, the mild end of the forecast range wins; if it buckles, ANZ's 2027 trough becomes the base case.

The early signals lean stabilisation — but aren't a bottom

Clearance has now improved four consecutive weeks, and this week the lift broadened beyond Sydney and Melbourne. The pace of asking-price declines in the two big capitals has decelerated from 2–3% a month at the worst to under 1.2%. Falling faster → falling slower is how bottoms start; it is not yet a bottom. Watch the same dashboard we do: weekly clearance against the spring listings build, the ~1 September August HVI (do the mid-sized capitals follow Sydney and Melbourne toward the -5% line?), and the monthly credit prints.

Our read — and what to do with it

Base case: falls continue through spring and into 2027, front-loaded in Sydney and Melbourne (plausibly past the halfway point of their peak-to-trough already), shallower and later in Brisbane, Adelaide and Perth, while the affordable regional tier that never joined the downturn keeps grinding higher on yield.

For buyers: engage on price now rather than waiting for a bell. Vendor discounting is the widest in three years (3.9% capitals, 4.2% Sydney — roughly $52,000 on the median), so negotiate off quarterly trajectories and let the duration question be the vendor's problem. For holders: the rate plateau plus rents rising 7% annually means the carry improves every quarter — duration is your friend on the income side. For everyone: don't try to time the exact bottom. In 2019 and 2023 alike, the market repriced upward faster than sidelined buyers could re-engage — the cost of being six months early, at today's discounts, is far lower than the cost of being six months late.