NAB Housing Monitor August 2026: Forecast Cut to −5% — Sydney and Melbourne Face 10% Peak-to-Trough Falls

At a Glance — NAB Housing Monitor, August 2026

Source: NAB Group Economics, Housing Monitor August 2026 (released 4 August 2026). Price data to July 2026; NAB's price figures are built on Cotality data.

When a major bank's economics team moves its housing forecast twice in two months, the revisions matter less as predictions than as admissions. NAB's June Monitor saw the eight capitals averaging a 2% decline across calendar 2026, with Sydney and Melbourne falling 6–7% over the year. The August edition, released on 4 August with data to July, replaces that with a 5% capital-city decline this year, led by falls of around 10% in Sydney and Melbourne — and its own commentary concedes the point plainly: recent data had left “the risk skewed firmly to the downside” of the earlier numbers, so the outlook was “formally revised down.”

The trigger was a July dataset that broke the last remaining bullish argument. National values fell 0.7% for a second consecutive month once revisions are counted, and the falls stopped being a two-city story. Perth, Brisbane and Adelaide, the markets that carried national growth through 2025, printed flat to lower. On Cotality's July index, regional Australia turned negative in the same month. And the revisions themselves keep landing in one direction: NAB flags Perth as “especially susceptible” to downward restatement — the same pattern as the 120-basis-point June revision we examined in our Cotality HVI July analysis.

For investors, a bank forecast is an input, not an instruction. The more durable value in this Monitor is its liquidity and rental data — turnover at 4.1%, days on market at a five-year high, rents still compounding at nearly 6% annualised — because those are the numbers that describe the market you'll actually transact in this spring. Let's take the release section by section.

What Did the August Housing Monitor Actually Say?

The Monitor's summary page carries five load-bearing claims, each worth separating from the headline.

First, the fall is broadening, not deepening city-by-city. Sydney (−1.4%) and Melbourne (−1.2%) continue to lead, and both now sit about 5% below their recent peaks — 5.1% and 5.3% respectively, measured peak to trough (from each city's price peak to its current level). But July's distinguishing feature is the mid-sized capitals joining: Brisbane, Perth and Adelaide are “flat to lower,” sitting 0.4–0.7% below their own peaks. A downturn confined to the two southern giants is a divergence story; one that includes the 2025 boom markets is a cycle.

Second, the revisions keep cutting one way. NAB notes that “revisions continue to show starker price declines than initially reported,” with Perth the standout case. June's national fall — first published at 0.4% — now reads 0.7%, the same as July's initial print. Investors should internalise the pattern: in a falling market with thinning settlement volumes, the first print is the optimistic print, and July's −0.7% may itself deepen.

Third, the forecast cut is structural, not cosmetic. The new numbers — capitals −5% in 2026, Sydney and Melbourne ~−10% peak to trough, mid-caps −2% to −4%, recovery “late 2027” — deepen the June edition's Sydney/Melbourne call and pull the whole capital-city complex into the red. One basis note: June's 6–7% was a calendar-2026 forecast while the new ~10% is peak to trough, so the like-for-like revision is more modest than the raw numbers suggest — the direction and breadth, not a doubling, are the story. Press coverage rounded Melbourne to −9%; the PDF's language is “~10% declines in Sydney and Melbourne.” Either way, NAB now sits clearly below the shallower 2026 corrections most forecasters carried into winter.

Fourth, the market is illiquid before it is cheap. The seasonally adjusted turnover rate fell to 4.1% of dwelling stock in July, from 5.2% a year ago — though NAB notes it remains above the ~3.5% trough of the 2018–19 slowdown. Median days on market hit 34, a five-year high. Prices are 0.4–5.3% off their peaks depending on the city; transaction activity is off more than 20% year on year. That gap is the defining feature of this phase.

Fifth, the rental market refuses to follow prices down. Vacancy remains near record lows, and advertised rents are growing at a 5.8% six-month annualised pace. NAB's own framing of why this downturn differs from 2018–19 rests on this point, and it deserves its own section.

Why NAB Says This Downturn Is Not 2018–19

The 2018–19 comparison is everywhere in the commentary around this cycle — similar peak-to-trough magnitudes forecast for Sydney and Melbourne, a similar credit-driven trigger, even a similar turnover trough. NAB's Monitor makes the case for why the parallel breaks, and it's the most analytically useful passage in the release.

In 2018–19, the price falls followed years of elevated apartment completions on the east coast. By the time values were declining, “improved supply conditions [had] been feeding through into cooler rents growth for some time” — landlords were cutting asking rents in Sydney even as prices fell, so the investor case deteriorated on both sides of the ledger simultaneously. Yields barely expanded; only the 2019 rate cuts and APRA's serviceability changes turned the market. Other conditions differ this cycle too — lending standards are tighter, migration is running harder, and tax settings are mid-reform — but the rental inversion is the difference NAB itself singles out, and it's the one that changes the holding math.

The 2026 setup inverts the rental side. Vacancy remains low — SQM's national measure sat at 1.3% in June, near its historic floor (we track it in our vacancy rate tracker) — and the Monitor has advertised rents growing at a 5.8% six-month annualised pace (the last six months' growth expressed as a yearly rate), running well ahead of the CPI's lagged 3.6% rents read. On the supply side, apartment and townhouse starts still outpace completions, detached pipelines have only normalised in NSW and Victoria, and Q2 input costs jumped 2.1% on materials, freight and fuel — none of which points to a supply wave arriving to cool rents in 2027.

The investor arithmetic follows mechanically. When prices fall while rents rise, gross yields expand from both ends: Cotality's combined-capital gross yield has already recovered to 3.50% from December's 3.34% low, and every month of the NAB forecast path — prices −5%, rents +5.8% annualised — widens it further. The expansion is uneven, though: it's strongest in the falling capitals' house markets and tight-vacancy suburbs, and weaker where strata costs or pockets of localised oversupply absorb the rent gains. Our analysis: this is the single most important difference between holding through this downturn and holding through 2018–19. Then, an investor's cashflow worsened while values fell. Now, for most portfolios, cashflow improves each quarter the downturn runs — which shortens the queue of forced sellers and is a large part of why NAB can forecast a 10% Sydney fall and a late-2027 recovery in the same document.

The Forecast Cut in Context: How the Banks Now Line Up

The revision history tells the story more honestly than any single forecast:

Source: NAB Group Economics, Housing Monitor June and August 2026 editions. Mid-sized capitals: August edition forecasts 2–4% peak-to-trough declines with recovery late 2027. Note the basis change across editions: June's Sydney/Melbourne figures were calendar-2026 forecasts; August's ~10% is peak to trough.

Three observations from our side.

The cut follows the data rather than leading it. Sydney is already 5.1% below peak and falling at more than 1% a month; a ~10% trough forecast requires only that the current run-rate persists into early 2027 and then decelerates. In that sense the new numbers are less a prediction than an extrapolation with a decay curve. The June edition, by contrast, required the falls to halve almost immediately — which is why it lasted two months.

Behind both revisions sits the same demand shock working alongside rates: NAB's June downgrade explicitly cited the enacted negative gearing and CGT reforms, whose 12 May 2026 purchase cut-off means investor demand for established dwellings is repricing well ahead of the rules' 1 July 2027 start. We covered the mechanics in our negative gearing and CGT reform guide; the point here is that this brake does not release when the RBA eventually cuts.

The mid-cap numbers are the genuinely contentious part. Forecasting only 2–4% peak-to-trough declines for Brisbane, Adelaide and Perth embeds a view that the 2025 boom markets correct gently despite stretched affordability and the same borrowing-capacity constraints. The alternative view — that the markets which rose 10–20% in a year have the most air underneath them — is at least as defensible. The tier data cuts both ways: cheaper segments are holding nationally (lower-quartile values +0.3% over three months), but Perth's repeated downward revisions suggest the boom markets' data is the least reliable in the set.

A late-2027 recovery is doing quiet work in the headline. “Capitals −5% in 2026” reads apocalyptic against 2025's gains, but the full shape — a 12-to-18-month decline, a floor, then recovery from late 2027 — is an ordinary cyclical correction, not a structural derating. The forecast is consistent with NAB's published rate view — the cash rate held at 4.35% through 2026 with easing in 2027 — and the recovery leg is essentially a rate-cut story. If the cuts slip, so does the recovery.

For how this squares with the index providers themselves, our Cotality HVI July deep-dive and PropTrack July analysis cover the underlying releases — NAB's price data is itself built on Cotality's index, so the Monitor is best read as Cotality's numbers plus NAB's lending book and forecast overlay.

Turnover at 4.1% and 34 Days on Market: The Liquidity Story

Price indexes get the headlines, but the Monitor's turnover data describes the market participants actually experience.

A seasonally adjusted turnover rate of 4.1% means roughly one in twenty-four dwellings changes hands annually, down from about one in nineteen a year ago. NAB notes this remains above the ~3.5% floor of the 2018–19 slowdown — there is still a functioning market — but the direction and pace matter. Combined with a median 34 days on market, the picture is of a market where vendors who can wait are withdrawing rather than discounting, and vendors who can't wait are doing the price discovery for everyone.

Three practical consequences for investors:

1. The negotiating window is measurable. A five-year high in days on market is the statistical version of what our negotiation guide documents anecdotally: listings aging past 45 days with a price reduction are where 8–12% discounts to guidance live. The Monitor says the median listing is more than halfway there.
2. Exit assumptions need a haircut. Any hold-period model that assumes a 30-day sale at appraisal value is now optimistic by both dimensions. If your strategy depends on a quick exit — bridging, flips, short settlements — the liquidity data is more important to you than the price forecast.
3. Trust the trend, discount the print. Thin turnover is precisely why June was revised from −0.4% to −0.7%: fewer settlements mean noisier first estimates. In a low-turnover downturn, treat every initial monthly print as provisional and every revision as more likely to subtract than add.

Lending, Arrears and the Construction Pipeline

The Monitor's housing-lending section is a preview: the ABS Q2 lending data lands next week (around 13 August — we covered the March-quarter release), and NAB expects it to show falling new commitments, with the investor retreat “more apparent in Q3.” That sequencing matches Reuters' late-July reporting of NAB's own application data — home-loan applications down 15% in the June quarter — and it matters for reading spring: the buyer pool is shrinking ahead of the listings that spring traditionally brings.

The stress indicators, by contrast, stay benign. Housing loan arrears sit around 1% of outstanding loans, higher in low-doc lending but stable overall. Only a small share of new lending is written at high debt-to-income or high loan-to-valuation ratios — though NAB flags that owner-occupier loans at LVRs of 90%+ ticked up late last year alongside the 5% deposit First Home Guarantee scheme. That is the program supporting the lower price tier, and it carries a caveat worth naming: the buyers holding up the cheapest quartile are also the most thinly capitalised against further falls. For deposit and LVR mechanics in this environment, see our deposit guide.

On the supply side, the Monitor reads as a slow squeeze rather than a wave. Dwelling approvals to June “held up” (our building approvals analysis covers the trend), apartment and townhouse starts continue to outpace completions, and pipelines remain elevated outside Sydney and Melbourne — but NAB expects new building demand to cool, and the Q2 Producer Price Index showed construction input costs jumping 2.1% in the quarter on raw materials, freight and fuel. Rising build costs into falling end values compress developer feasibility from both ends; the medium-term implication is less new supply into a rental market already at 1.3% vacancy, which is the structural case for rents staying firm through the price downturn.

What Would Make NAB's Forecast Wrong?

A forecast revised twice in two months deserves its own error bars. Our read of what breaks it, in both directions:

Shallower than −5%:

• Earlier or faster rate cuts. NAB's recovery leg is a rate story; if inflation keeps undershooting and easing arrives in early 2027 rather than late, the trough comes forward and the falls stop short.
• Migration and population pressure. Vacancy at 1.3% into strong population growth is the mechanism that ended previous downturns early in the cheaper tiers — the lower-quartile resilience (+0.3% over three months) is this force already visible.
• Government intervention. The 5% deposit First Home Guarantee is already supporting the lower tier; any expansion of demand-side support into a thin, low-turnover market would move prices quickly.

Deeper than −5%:

• The tax-reform brake doesn't release with rate cuts. If investor demand for established stock stays repriced through 2027, the recovery leg loses its main engine.
• The mid-cap call fails. Brisbane, Perth and Adelaide corrected 10–20% of the way through a boom's unwind in past cycles; 2–4% assumes an orderly landing in the least reliable data of the set.
• The revision pattern continues. Every recent month has been restated lower; if July's −0.7% deepens the way June's did, the −5% path is already behind the data.

What the August Monitor Means for Investors

If you're buying in Sydney or Melbourne: NAB's path implies roughly another 5% of downside from July levels before a late-2027 recovery. That is not an argument against buying — it's an underwriting parameter. Offers made 5–8% below current comparable sales are simply pricing the forecast; with the median listing at 34 days, a meaningful share of vendors will engage. The upper quartile, already falling 3.2% a quarter on Cotality's tiers, is where forecast-consistent discounts are most available.

If you're buying in Brisbane, Adelaide or Perth: the Monitor forecasts shallow 2–4% declines, which — if right — means the correction here is a repricing pause, not an opportunity set. Two cautions temper that: Perth's data keeps revising down, and these markets' affordability is stretched against local incomes after 2025. We'd treat mid-cap “bargains” skeptically and let individual vendor circumstances, not the index, define value.

If you're holding: the falling-price/rising-rent combination means most portfolios are cashflow-stronger each quarter even as paper values decline. The discipline is on refinancing and equity release — bank valuations will track the indexes down, so any equity-dependent plan should execute earlier rather than later in the forecast window, stress-tested at today's valuation less 5% rather than the peak appraisal, with a holding-cost buffer parked in offset. Arrears at 1% say your fellow borrowers aren't being forced out; nothing in this Monitor argues for selling into a 4.1%-turnover market.

If you run an SMSF: the calendar matters more than the forecast this week — from Monday 10 August, new limited recourse borrowing arrangements for residential property are off the table, so leveraged residential strategies inside super needed contracts exchanged by Sunday 9 August. For funds buying ungeared, the Monitor's combination — prices forecast to fall 5% while advertised rents compound near 6% annualised — improves the cash-purchase arithmetic each quarter. Our LRBA ban checklist covers the pivot options.

If you're selling: the liquidity data is your brief. Thirty-four days is the median, not the maximum; pricing at the front of the comparable range — not the top — is what separates a five-week sale from a five-month listing that follows the index down. Every month of delay costs ~0.5–0.7% on NAB's national path, before holding costs.

How We Read This Report

The Bottom Line

The August Housing Monitor is a major bank telling you, in its own numbers, that the 2026 downturn is real, broadening and roughly half-run in its two largest markets. The forecast cut — capitals from −2% to −5%, Sydney and Melbourne to ~10% peak to trough — mostly formalises what the July indexes had already shown, and the genuinely new information sits in the second-order data: turnover at 4.1%, days on market at a five-year high of 34, revisions consistently landing downward, and advertised rents still compounding at 5.8% annualised into near-record-low vacancy.

That last pairing is the strategic heart of the report. This is a downturn in which the income side of property keeps strengthening while the price side corrects — the reverse of 2018–19 — and it's why NAB can pencil a Sydney fall of 10% and a late-2027 recovery into the same forecast. For investors, the Monitor's practical message is not “wait” but “underwrite”: price the forecast into your offers, size your buffers against the revision pattern, and let expanding yields carry the hold.

We track every major index release in our Home Value Index tracker and will test NAB's −5% path against the August prints when Cotality and PropTrack report at the start of September.