Two per cent measures cash flow. A housing downturn hits the balance sheet
The Reserve Bank says fewer than two mortgaged households in a hundred cannot cover repayments and essentials. That is a flow number, and it says almost nothing about what falling prices do to the same borrowers.
The Reserve Bank of Australia sets the country's official interest rate and watches for dangers to the financial system. Twice a year it publishes a health check called the Financial Stability Review. The October edition says most mortgage holders can cope even if house prices fall sharply, and that the bigger threats come from overseas.
The Reserve Bank's reassuring figure counts who cannot pay this month. It says nothing about what happens when a stressed borrower can no longer sell the house for more than the loan.
The number doing all the work
Fewer than two borrowers in a hundred cannot pay. That is the figure carrying the October Financial Stability Review, the Reserve Bank of Australia’s twice-yearly health check on the financial system, and it is a flow measure wearing the clothes of a solvency one.
Read what it actually counts: variable-rate owner-occupiers whose income is insufficient to cover scheduled mortgage repayments and essential expenses. Everything above essential is already gone before you qualify. A household that has cancelled the insurance, stopped the extra super contributions, run the car into the ground and pulled the kids out of everything that costs money is, on this definition, not in shortfall. It is a measure of the last line, not of the distance to it. The Bank’s own projection — up to around two per cent, or a little above, as the cash rate has moved to 4.6 per cent — is conditioned on a rate path that the chief executive of the Commonwealth Bank described this week as live for another increase.
Flow and stock are not the same risk
A cash-flow shortfall tells you who misses a payment. It tells you nothing about what happens next, and what happens next is entirely governed by prices.
Australia’s mortgage book has been remarkably benign through two decades of rate cycles for a reason that has little to do with household budgeting discipline. When a borrower hit trouble, they sold. Prices had risen since origination, the sale cleared the debt with something left over, and a potential credit loss became an ordinary transaction with a real estate agent attached. Arrears converted into turnover. The loss never materialised.
That mechanism runs on rising prices. Reverse them and the same two per cent produces a different result: forced sales that do not clear the loan, lenders carrying arrears for longer because foreclosing crystallises a loss, and borrowers trapped in houses they cannot afford and cannot exit. None of that shows up in a shortfall count, because a borrower with no equity and a borrower with forty per cent equity look identical on a cash-flow line.
The Review’s media release joins the two with a conditional: households are well placed “even if housing prices were to fall sharply”. That is the sentence I would push on. The downturn is not a scenario the Bank is stress-testing against. It is the environment the Bank is reporting from.
Aggregate buffers belong to the wrong people
The second plank is buffers — the money sitting in offset accounts and available in redraw. Most borrowers have large ones. That is true and it is reassuring in aggregate, and aggregate is the problem.
Buffers accumulate in households with spare income. Cash-flow shortfalls occur in households without spare income. The two populations are close to mutually exclusive by construction, which is precisely why a system-wide buffer total is a poor guide to whether the stressed tail can absorb three more months of this. What matters is the buffer held by the people in the tail, measured in months of scheduled repayments, and that is a distribution, not a total.
The case against this argument
The strongest defence of the Bank’s position is simple: housing cycles do not break banks. Unemployment does. Australian lending is full recourse, loss rates through previous price falls have been trivial, and the Review is right that the serious threats on the horizon are offshore — the circular financing arrangements building around the artificial intelligence boom, where the same capital appears on several balance sheets at once, and a sharp move in government bond yields. A central bank that spent its stability review relitigating domestic mortgage arrears would be fighting the last war while the next one assembled in global credit markets.
I accept most of that. What I do not accept is that the two are separable. An offshore shock that lifts sovereign bond yields arrives in Australia as higher bank funding costs, which arrive as higher mortgage rates, and the employment effects arrive as lost income. The transmission channel from the global chapter to the Australian financial system runs directly through the mortgage book the domestic chapter has just declared sound. Structuring the document as offshore-risk-here and households-fine-there invites the reader to miss the join. The circular financing worry is a good one precisely because it is about hidden correlation. The Review should apply the same suspicion at home.
What would settle it
One table. The share of mortgaged households in cash-flow shortfall, cross-tabulated by equity band and by buffer size measured in months of scheduled repayments. If that cell is tiny, the Bank’s confidence is vindicated and the critics, including this one, are answered with data rather than reassurance. If it is not tiny, the country deserves to know before the next quarterly inflation print decides the November meeting rather than after.
The Bank has the loan-level data. It built the two per cent figure out of it. Publishing the overlap is not a new collection exercise. It is a decision about what to show.
The Reserve Bank should publish, in the April 2027 Financial Stability Review and every one after it, the share of mortgaged households that are in cash-flow shortfall broken down by equity band and by buffer size in months of repayments — not the two measures side by side in separate charts, but crossed. It should also state the house price path assumed in its household resilience work, so readers can see whether the assumption matches the market it is describing.
The Australian Prudential Regulation Authority, which supervises the banks, should pair that with a regular published figure for the share of outstanding housing loans where the debt exceeds the value of the property. Resilience claims that cannot be checked against equity are claims about last month's budget, not about the financial system.
What would change this view
Show me the overlap and I will drop the argument. If the Reserve Bank publishes, in its next Review, the share of mortgaged households that are simultaneously in cash-flow shortfall and hold little or no equity, and that share is a small fraction of the two per cent — and if the major banks' reported ninety-day housing arrears and realised housing losses stay flat through their 2027 results while prices keep falling — then the flow measure was an adequate proxy for the stock risk all along.
If that happens, the right conclusion is that recourse lending and a holding labour market do the work people claim for them, and the Review's structure is fine as it stands. If instead arrears rise while the published shortfall number sits near two per cent, the measure is tracking the wrong variable and the Bank will have told the country it was safe using a statistic that could not have detected the danger.
Worked from
- RBA relaxed about housing downturn, deeply worried by AI and bonds
ABC News · 2026-10-01