The phrase doing the rounds this winter is that Sydney is handing back its COVID gains. It is a tidy line. It is also wrong by an order of magnitude, and the gap between the two numbers is where the real story sits.

Sydney peaked in January 2026. Values fell 1.2 per cent in June, putting the city 3.7 per cent below that peak, and a further 1.4 per cent in July. Call the drawdown five per cent. Over the June quarter Sydney led the capitals down at 3.2 per cent, and nationally July’s 0.7 per cent fall was the steepest single month since December 2022.

Now set that against the tide it is running out of. Australian home values have climbed roughly 47 per cent since March 2020 — about $280,000 added to the median dwelling. A five per cent retracement does not touch that. It barely dents the last eighteen months.

What we are watching is a conventional rate-driven correction sitting on top of an intact structural repricing. Those are different animals, and conflating them leads vendors to the wrong decision.

One market, two directions

The second thing worth naming is direction of travel within the market. This correction is running top-down. In the three months to July, upper-quartile values fell 3.2 per cent nationally while the lower quartile rose 0.3 per cent. Serviceability constraints are deflecting demand toward cheaper stock, and first-home buyer stimulus is meeting it there.

In our end of the market, that is not a footnote. That is the whole story.

New Zealand did what Sydney is being accused of doing

If you want to see a market genuinely hand back its pandemic gains, look across the Tasman. New Zealand prices are 18 per cent below their January 2022 peak and sitting at their lowest level since July 2023. The REINZ index rose 43 per cent in the eighteen months to November 2021, then fell 16 per cent down the other side — the largest of the five declines recorded since 1992.

In real terms it is deeper still. Adjusted for inflation, New Zealand house prices have fallen back to mid-2019 levels, 28 per cent below the 2021 peak, with Auckland down 35 per cent and Wellington 40 per cent in real terms.

Five years. A third of it in real terms. That is what a genuine repricing looks like, and Sydney has experienced nothing remotely comparable.

The affordability consequence is the part nobody in Sydney wants to hear. Auckland’s price-to-income multiple has fallen to 7.7 from 11.2 in 2022. Sydney’s is 13.8 — the second least affordable major market in the world. Brisbane, Perth and Melbourne now all rank as less affordable than Auckland.

Part of that is supply, not just price. Auckland has been building at volume since the Unitary Plan, and the improvement has held across a change of government and a full rate cycle.

Four markets, four settings

The Reserve Bank held the cash rate at 4.35 per cent on 11 August, after three consecutive hikes in February, March and May took it from 3.60 per cent. Inflation came in softer than the Bank’s own forecasts in the June quarter, which removed the immediate case for a fourth. The Bank has not removed the option.

Note the asymmetry. New Zealand’s cash rate sits 185 basis points beneath ours, and yet its mortgage rates are broadly comparable. BNZ expects floating rates above 6.5 per cent by year end and through 7 per cent in the first half of 2027. Cheap money did not save the New Zealand market, and it will not be what saves ours.

Sydney is second in the world, and closing

The median multiple is a blunt instrument. It is also the only measure that lets you stand Sydney next to Auckland, Toronto and London on the same axis, and on that axis Sydney sits second worldwide behind Hong Kong.

Nationally, servicing a new loan absorbs about 45 per cent of median household income, the deposit takes close to twelve years, and the dwelling value-to-income ratio has reached 8.2. Renters are committing a record 33.4 per cent of income.

Price is set by a formula now, not by sentiment

This is the part of the market that has genuinely changed, and it is where Australia has quietly followed New Zealand’s lead.

From 1 February 2026, APRA requires lenders to cap mortgages written at a debt-to-income ratio of six or above to 20 per cent of new lending, applied separately across owner-occupier and investor books. The serviceability buffer stays at three percentage points, which at current pricing means borrowers are assessed near nine per cent while paying around six.

New Zealand got there eighteen months earlier. The RBNZ has run a six-times cap for owner-occupiers and seven times for investors since July 2024, alongside loan-to-value restrictions. Ireland caps at three and a half to four times. Canada at four and a half.

Every major bank now has Sydney negative

Six weeks ago the institutional view was a soft landing. It is not any more, and the revisions have been violent.

ANZ downgraded on 11 August. It now expects capital city prices to fall 4.3 per cent this year and a further 3.4 per cent in 2027 — a peak-to-trough decline of 10.6 per cent across the capitals, with Sydney taking the steepest hit at 14.5 per cent from its January peak. In April the same bank had Sydney at 0.7 per cent lower for the year.

NAB moved a week earlier. Its August Housing Monitor has the combined capitals falling about 5 per cent across 2026, against the 2 per cent decline it forecast in June, with Sydney near 10 per cent and Melbourne 9 per cent.

A note of caution on any figure quoted this month. Commonwealth Bank’s widely circulated 15 per cent peak-to-trough call belongs to the 2022 cycle, not this one. The bank formally abandoned it in April 2023 and upgraded to a 3 per cent lift. CBA’s current position, published in June 2026, is prices broadly flat nationally with declines in Sydney and Melbourne, stabilising from early 2027.

The overlay nobody had priced in

The May 2026 Budget measures are now law. From 1 July 2027, negative gearing on established residential property ends for anything acquired after 7.30pm on 12 May 2026, and the 50 per cent capital gains tax discount is replaced with cost base indexation and a 30 per cent minimum tax rate.

Existing holdings are grandfathered. That cuts forced-sale risk, but it creates a lock-in effect — established investors now hold a tax reason not to sell. NAB named the reforms explicitly in its downgrade. Read together, the two effects pull in opposite directions: less investor demand for established stock, and less investor willingness to release it.

A listings problem, not a dwelling glut

Most commentary is conflating the two. They are not the same thing, and the distinction decides how you price a campaign.

SQM’s Louis Christopher on the July figures: national listing levels are almost 23 per cent higher than a year ago, an abnormal result when July usually records a winter lull. Both new and older listings are rising. More stock is coming to market and taking longer to sell, which he calls the typical read in a downturn.

Westpac’s Luci Ellis sets out the mechanism. Turnover has declined sharply, and while new listings are also easing they are falling more slowly than sales — so stock on market rises relative to the flow of sales, turning established housing into a buyer’s market.

Now hold that against the construction pipeline, because there is no physical oversupply anywhere in the country. June dwelling approvals rose 7.2 per cent to 18,328, yet 47,750 fewer homes were built over the year than required — a second consecutive year of Housing Accord shortfall. National vacancy is running near 1.2 per cent. Advertised rents are 5.9 per cent higher over the year, with unit rents up 7.9 per cent.

Values falling while rents accelerate and vacancy stays under two per cent. That is a credit event, not a supply event — and it tells you what reverses it.

Survey stress is climbing. Arrears are not.

These two numbers get quoted interchangeably and they measure entirely different things. The gap between them is the most misunderstood figure in the market right now.

Roy Morgan has 30.3 per cent of mortgage holders — about 1.61 million people — at risk of mortgage stress in the three months to June 2026, the fifth consecutive monthly rise, up from a three-year low of 23.9 per cent in January. Equifax reported mortgage accounts in financial hardship rising 5.3 per cent quarter on quarter.

Now the lender data. APRA has non-performing home loans at around one per cent, with early-stage arrears near 0.7 per cent. The RBA’s March 2026 Financial Stability Review put roughly one per cent of variable owner-occupier borrowers in cash-flow shortfall.

Households cut discretionary spending long before they miss a repayment. That is why survey stress can climb five months in a row while delinquency stays flat. The forced-sale wave the crash commentary requires has not arrived, and on this data is not close.

What the stress numbers do tell you is about borrowing capacity, not distress. A household spending thirty per cent of after-tax income on an existing loan is not a buyer for the next one.

The eastern suburbs have left the wage economy

Here is the figure that ends the argument. In Sydney’s eastern suburbs, an average wage earner would need 35 years to save a 20 per cent deposit on a median house. Having cleared it, servicing the mortgage would take one and a half times their income.

That is why the premium end is falling while the lower quartile holds. It is not a sentiment story. It is arithmetic. When the assessment rate moves from six to nine per cent, the buyer at $8 million loses more absolute capacity than the buyer at $1.2 million, and there are fewer of them to begin with.

Vendor discounting across the combined capitals has widened to 3.6 per cent, auction clearances are sitting near or below 50 per cent in Sydney, and the auction share of new listings has fallen from almost 45 per cent last November to just over 30 per cent in June. Sellers are moving to private treaty because the room is no longer in the auction.

Pricing to the last comparable is the error of this cycle. The last comparable was set under a different assessment rate.

There is one advantage still sitting on our side of the harbour, and it will not last. The listings surge has reached Sydney’s west, south west and Hills District. It has not yet reached the Eastern Suburbs, the Lower North Shore or the Inner West. Vendors here are still selling into a market that has not been asked to absorb a wave of competing stock.