Prices fell. Affordability got worse

Market analysis · September 2026

Prices fell. Affordability got worse.

Sydney values are 7.1 per cent below their February peak, and housing affordability has just hit the lowest level ever recorded. Both of those things are true at once, and the reason they are true explains the past five years better than any forecast will explain the next one.

---

Something happened this year that shouldn’t be possible on the ordinary reading of the property market. Sydney values fell 7.1 per cent from their February peak. Nationally, home values have now declined for five consecutive months. And in the middle of that decline, housing affordability hit the lowest level ever recorded.

A median-income household earning about $125,000 could afford 12 per cent of the homes sold across Australia in the last financial year. Mortgage repayments now consume 35.5 per cent of average household income — the highest share since 1989, and higher than at any point during the global financial crisis.

Falling prices did not help. That deserves an explanation, because the explanation is the whole story of the past five years.

Affordability is a ratio, not a price

The word gets used loosely. When most people say a house is unaffordable they mean it costs a lot. But every serious affordability measure — PropTrack’s, the RBA’s, the HIA’s — is a fraction. The cost of servicing the loan sits on top. Household income sits underneath.

Three things move that fraction, and they move at very different speeds. Interest rates move fastest and hit hardest, because they act on the whole loan every month for thirty years. Incomes move slowly and predictably. Prices move loudly, get all the coverage, and matter least of the three — which is the part almost nobody believes until they see the arithmetic.

Prices are the number everybody watches and the lever that matters least.

There is also a fourth constraint that doesn’t appear in the serviceability ratio at all: the deposit. Falling prices help a saver. Rising rates hurt a borrower. When both happen at once — as they did this year — the two hurdles move in opposite directions, and a household can find itself closer to the deposit and further from the loan on the same morning.

Five years, four turns

The sequence matters more than any single reading, because it shows which lever was doing the work each time the market changed direction.

Four turns in five years. Prices changed direction three times. Affordability tracked the cash rate almost the entire way.

The arithmetic of this year

Here is why a 7 per cent discount never reached the buyer. Take a $1.5 million purchase at 80 per cent lending, thirty-year principal and interest, and run it either side of this year’s rate moves.

The house is $105,000 cheaper and the repayment is higher. Three quarters of a per cent on the rate outweighed 7 per cent off the price, because rate movements compound across the whole loan while a discount only shrinks the principal once. Add income growth of roughly 3 per cent and the household is still marginally worse off than it was in January.

That is the record low, in one panel. Not a mystery — a compounding effect beating an arithmetic one.

What this looks like from the Eastern Suburbs

The obvious objection is that none of this applies here. A vendor in Bellevue Hill or Vaucluse is not selling to a median-income household on $125,000, and a buyer at $8 million is not stress-testing a serviceability ratio.

True — and beside the point.

Consider how a prestige sale actually completes. The buyer at $8 million is very often selling at $4 million to fund it. That buyer is selling a house at $2.4 million. And that buyer needs a bank. Every transaction at the top of this market rests on a chain of transactions beneath it, and the bottom link of that chain is made of credit. When borrowing capacity contracts, the bottom link doesn’t break loudly — it simply stops moving, and everything above it slows in sequence.

You can see it in the numbers. Nationally, sales volumes are running 15.5 per cent below a year ago and 11.5 per cent below the five-year average, while Sydney values have fallen 4.6 per cent over the same period. Volume has fallen more than three times as far as price. That gap is the signature of a market where willing parties exist at both ends and the chain between them won’t close.

Affordability doesn’t stop at the price point where it bites — it travels upward, and arrives at the top of the market as illiquidity rather than as price.

The other thing worth noting is where the falls started. Cotality’s data has premium markets leading the declines for most of this cycle, with the gap between upper and lower quartiles narrowing only recently, as 93 per cent of capital-city suburbs turned negative through winter. The top of the market moved first. It usually does, in both directions.

What would actually change it

Rates. Almost entirely rates, in the short run. A cut of 75 basis points would undo this year’s damage to serviceability faster than any other single lever, and market pricing is currently divided on whether the next move is up or down.

Incomes, slowly and permanently. This is the only lever that improves affordability without also making existing owners poorer, which is why it is the one everybody claims to want and nobody can deliver quickly.

Supply, structurally. Every economist quoted on this subject says the same thing, they are right, and it will take longer than most people’s remaining working lives to matter.

Tax settings, marginally and unevenly. From 1 July 2027, negative gearing on established residential property purchased after 12 May 2026 is abolished, and the 50 per cent capital gains discount is replaced by cost base indexation with a 30 per cent minimum rate. Existing holdings are grandfathered and new builds are exempt. Treasury’s own estimate is roughly 2 per cent a year off price growth and about $2 a week onto median rents. That changes who competes for a property. It does not change what the property costs to service.

And falling prices? On the evidence of the past five years — barely.

What I’d tell a vendor deciding whether to wait

Waiting for affordability to improve is not the same as waiting for prices to recover, and the past five years have shown they can move in opposite directions for years at a time. If your reason for waiting is that you expect rates to fall and buyers to return, that is a defensible position — it happened in 2025, and it may well happen again.

If your reason for waiting is that prices have fallen and you’d like them back, understand what you’re actually waiting for. Prices returning to February levels would require the very borrowing conditions that have just been withdrawn. The market rarely hands you both at once.