Credit conditions in 2026: what the data says
The cash rate rose a full percentage point in 2026, yet credit is still growing 8.4% a year, arrears are low and banks are well capitalised. The tightening shows in lenders' tests, modelled here for borrowing power and refinancing.
General information only, not advice. Read the full note
Next update. The Reserve Bank's next decision on the cash rate is due on 3 November 2026. We will update the rates and figures in this article after it.
Credit conditions come down to two things: what borrowing costs, and how easy it is to get. In 2026 the first has moved quickly. The Reserve Bank of Australia has raised the cash rate four times, by a full percentage point, and lenders have passed the rises on. The second has moved more quietly. Credit is still growing faster than a year ago, arrears remain low and the banks hold more capital than their regulator requires. The tightening is showing up somewhere less visible: in how much a lender will advance at today's rates, and in what happens when a loan made on yesterday's rates and values comes up for renewal.
This is Part 3 of The architecture of lending. It reads the data published by 12 October 2026 as a lender and an investor would, with two models of the mechanics behind the numbers.
Reported data
Credit conditions at a glance
- New housing loan commitments
- $97.6 billion
- June quarter 2026, down 5.2% on the March quarter and up 6.8% on a year earlier
- ABS5
The latest figure for each measure published by 12 October 2026. ADIs are authorised deposit-taking institutions: banks, building societies and credit unions. The trend line shows the cash rate target at the end of each month from December 2025 to September 2026.
The price of credit: four rises in 2026
The cash rate is the interest rate on overnight loans between banks, and the Reserve Bank's Monetary Policy Board sets a target for it. Lending rates sit above it and generally move with it, but the costs of banks' funding, from deposits and from wholesale markets, do not necessarily move by the same amount or at the same speed as the cash rate.9 Part 1 explains the chain from the cash rate to the rate on a loan.
The Board raised the target four times in 2026, by 25 basis points each time; a basis point is a hundredth of a percentage point. The rises took effect on 4 February, to 3.85%; on 18 March, to 4.10%; on 6 May, to 4.35%; and on 30 September, to 4.60%. In between, the Board left the rate unchanged at its June and August meetings.1 The cash rate is a full percentage point higher than at the start of the year. The three cuts of 2025 have been more than reversed, and the rate is at its highest since 2011.
The Board's statement on 29 September set out its reasons. "Inflation remains elevated and some of the upside risks flagged in August are materialising." "Short-term measures of inflation expectations remain elevated." And "growth in aggregate demand needs to remain subdued for a period", to ease the pressure on the economy's capacity. The Board noted that "housing prices have fallen in most capital cities and new housing loans have declined noticeably", and that labour market conditions had eased broadly as it expected. It would keep doing what it considers necessary to bring inflation back to target, "including increasing the cash rate target further if needed", and the decision was unanimous.2
In our reading, the words that matter most for borrowers are "for a period". The Board is not trying to stop the economy growing. It wants demand to run below what the economy can supply until inflation comes down, and credit is one of the channels through which it works. The Reserve Bank's October Financial Stability Review puts it in one line: the Board "has increased the cash rate by 100 basis points to help return inflation to target".8
The next decision is due on 3 November, at the end of the Board's two-day meeting,10 and new forecasts follow in the November Statement on Monetary Policy. We set out the causes of this year's inflation, and the longer history of the cash rate, in a separate article.
Reported data
The cash rate target, January 2024 to September 2026
RBA cash rate target at the end of each month, per cent
Show the figures
| Period | Cash rate target |
|---|---|
| Jan 2024 | 4.35% |
| Feb 2024 | 4.35% |
| Mar 2024 | 4.35% |
| Apr 2024 | 4.35% |
| May 2024 | 4.35% |
| Jun 2024 | 4.35% |
| Jul 2024 | 4.35% |
| Aug 2024 | 4.35% |
| Sep 2024 | 4.35% |
| Oct 2024 | 4.35% |
| Nov 2024 | 4.35% |
| Dec 2024 | 4.35% |
| Jan 2025 | 4.35% |
| Feb 2025 | 4.10% |
| Mar 2025 | 4.10% |
| Apr 2025 | 4.10% |
| May 2025 | 3.85% |
| Jun 2025 | 3.85% |
| Jul 2025 | 3.85% |
| Aug 2025 | 3.60% |
| Sep 2025 | 3.60% |
| Oct 2025 | 3.60% |
| Nov 2025 | 3.60% |
| Dec 2025 | 3.60% |
| Jan 2026 | 3.60% |
| Feb 2026 | 3.85% |
| Mar 2026 | 4.10% |
| Apr 2026 | 4.10% |
| May 2026 | 4.35% |
| Jun 2026 | 4.35% |
| Jul 2026 | 4.35% |
| Aug 2026 | 4.35% |
| Sep 2026 | 4.60% |
Reading the chart
Threefold Capital interpretation
- What it shows
- The cash rate held at 4.35% through 2024, fell to 3.60% with three cuts in 2025, and has risen four times in 2026 to 4.60%, above where it stood through 2024.
- What it means
- A borrower who took out or refinanced a variable loan in late 2025 will be paying a full percentage point more once the September rise reaches their repayments, if the lender passes every rise on; on the Reserve Bank's rule of thumb, most such borrowers' repayments at 12 October reflected three quarters of a point. In our view the larger change is in direction: the Board's guidance now points to further rises if inflation does not ease, not to cuts.
- How to respond
- Any plan that depends on borrowing is worth testing at a rate half a point above today's and at one a full point above. The models below show what those steps do to borrowing power and to a refinance.
How the rises reach borrowers
A change in the cash rate reaches borrowers through their lenders' funding costs, and this year the transmission has been quick and close to complete. By August, the Reserve Bank reported that "banks have passed on the three cash rate increases to deposit and lending rates". Variable mortgage rates rose by nearly 75 basis points between January and June, close to the 75 basis points of the three rises over those months, and variable business rates rose alongside the cash rate and bank bill swap rates.3 Bank bill swap rates, or BBSW, are a key market reference rate for banks' funding costs.9 The September rise came after these figures, and how fully it was passed on will show in data published later this year.
In our reading, the rises are travelling faster than they did in 2022 because almost every mortgage is now variable. The fixed share of outstanding housing loans fell to a historical low of less than 5% in 2025,11 against around 40% in early 2022,12 when fixed loans held the average rate on outstanding mortgages back for a year or more (Part 1 shows the pass-through of the 2022 rises). A rise in 2026 reaches nearly every mortgage holder.
It does not reach them at once. Cash rate increases "can take around three months to flow through to scheduled mortgage payments", the Reserve Bank notes, because of minimum notice periods for higher repayments and differences in borrowers' repayment cycles.3 On that rule of thumb, the September rise will reach most scheduled repayments around the end of the year, so at 12 October most variable-rate borrowers' repayments reflected three quarters of a point of this year's rises, not the full point. Even before the September rise, scheduled mortgage and consumer credit payments had risen to just under 12% of household disposable income in the June quarter, close to their 2024 peak.3
Smaller businesses feel the rises sooner than large companies. The Financial Stability Review expects pass-through to be faster for smaller businesses, because "many smaller businesses take out variable-rate business loans secured with a residential property mortgage".8 For the owner of such a business, the business loan and the family home carry the same rise at the same time.
One qualification runs the other way. Competition has narrowed the spread between mortgage rates and the cash rate since 2019, and the Reserve Bank concludes that changes in credit markets of this kind have made financial conditions less restrictive than they otherwise would be for a given level of the cash rate.13 A cash rate of 4.60% bites a little less today than it would have with the margins of 2019. Part 1 has the figures.
How much credit is flowing
Higher rates have not stopped credit from growing. Total credit, the lending of financial institutions operating in Australia, grew 8.4% over the year to August 2026, faster than the 7.2% of the year before. Housing credit grew 7.3%, business credit 10.9% and personal credit 5.1%, each faster over the year than a year earlier.4 The recent pace tells a different story for housing. The Reserve Bank's August Statement said that housing credit growth, measured over six months and annualised, "has eased and is expected to slow further in the period ahead", while business credit growth "remains close to its fastest pace since 2022".3
Reported data
Growth in credit over the year to August, 2025 and 2026
Growth in credit outstanding over the year, per cent. Seasonally adjusted and adjusted for breaks in the series. Business means lending to non-financial businesses
Reading the chart
Threefold Capital interpretation
- What it shows
- Every kind of credit grew faster over the year to August 2026 than over the year to August 2025. Business credit grew fastest, at 10.9%, and personal credit slowest, at 5.1%. Total credit grew 0.6% in each of July and August.
- What it means
- The stock of credit is still growing at about twice the rate of consumer prices, so this year's rate rises have not yet reduced the amount of borrowing. Credit outstanding is a slow measure; the flow of new lending, in the next section, turns first.
- How to respond
- Read aggregate credit growth as a description of the economy, not of any one borrower's access to credit. What matters for a particular loan is how its lender's tests move with rates, which the models below show.
Credit growth measures the stock of loans outstanding, and the stock turns slowly. We think it keeps rising for a while after rates go up because loans approved months earlier are still being drawn, construction loans draw down in stages as the building goes up, and businesses draw on limits they already have, sometimes because cash is tighter rather than because they are investing. Those are reasons for a slow turn, not for an acceleration, and business credit grew 1.0% in July and 0.9% in August,4 so new borrowing by businesses had not slowed by August; the flow of new housing lending, in the next section, is where the turn shows. Part of the growth is also prices: consumer prices rose 4.0% over the year to August,14 so the real increase in credit is well below the headline.
Business lending has also had help from outside the banks. Non-bank lenders have taken a growing share of business lending since 2022, especially smaller loans to small and medium businesses,13 and the Reserve Bank describes "ongoing strong competition between bank and non-bank lenders".3
New lending has turned down
The flow of new lending is where higher rates show first. The Australian Bureau of Statistics (ABS) now publishes its lending indicators each quarter, and the June quarter release of 14 August is the latest. It counts borrower-accepted commitments, firm offers of a loan that the borrower has accepted,15 and it leaves out refinancing.5
In the June quarter the value of new loan commitments for dwellings fell 5.2% to $97.6 billion, and their number fell 5.4% to 134,225, seasonally adjusted. Investors pulled back hardest: the value of their commitments fell 10.2% to $37.1 billion. Owner occupiers' commitments fell 1.9% to $60.5 billion, while first home buyers, who are counted among owner occupiers, took on 0.2% more, $18.4 billion. By value, every group was still borrowing more than a year earlier, and the total was 6.8% higher.5
Reported data
New housing loan commitments, June quarter 2026: change on the quarter and on a year earlier
Change in the value of new loan commitments for dwellings, per cent. Seasonally adjusted, excluding refinancing. First home buyers are part of owner occupiers
Reading the chart
Threefold Capital interpretation
- What it shows
- The value of new housing loan commitments fell 5.2% in the June quarter, led by a 10.2% fall for investors. Owner occupiers took on 1.9% less and first home buyers 0.2% more. Against the June quarter of 2025, every group was still borrowing more by value, from 6.0% more for owner occupiers to 10.0% more for first home buyers.
- What it means
- New lending turns before the stock of credit does, and it has turned. Investors are the most sensitive to rates, in our view, because the return on an investment property depends directly on the gap between its rent and its interest, and the Reserve Bank links weaker sentiment partly to changes to negative gearing and the capital gains tax discount. A fall concentrated among investors, with first home buyers steady, is consistent with a slowdown led by rates and tax settings rather than by job losses.
- How to respond
- For anyone planning to buy or refinance property, the useful signal is not the national total but what lenders now ask of borrowers like them, which the models below set out. The September quarter figures, due on 11 November, will show whether the fall continued.
The Reserve Bank's August Statement described "a sharp decline in new housing loan commitments over recent months, driven by investors".3 Housing prices have fallen too. The Financial Stability Review attributes the fall to weaker sentiment, "reflecting the somewhat restrictive stance of monetary policy and changes to negative gearing and capital gains tax discount policies".8
APRA's figures measure something different: loans actually funded in the quarter by the banks, building societies and credit unions it supervises, rather than commitments. On that measure, those lenders funded $200.5 billion of new housing loans in the June quarter, 6.8% more than a year earlier.6 The two series differ in coverage and definition and will not match, but they agree on the direction over the year.
Inside the lenders' property books
The Australian Prudential Regulation Authority (APRA), which supervises the banks, publishes quarterly figures on the property lending of authorised deposit-taking institutions (ADIs: banks, building societies and credit unions). The June 2026 figures, released on 17 September, show little change in the riskier kinds of lending, and low arrears.6
Reported data
ADIs' property lending, June 2025 and June 2026
Scroll the table sideways to see every column.
| Measure | June 2025 | June 2026 | Change |
|---|---|---|---|
| Residential loans outstanding, $ billion | 2,391.8 | 2,558.5 | +7.0% |
| Investor share of residential loans | 30.5% | 31.2% | +0.72 |
| Share of residential loans with a loan to value ratio of 80% or more | 17.6% | 16.7% | −0.9 |
| New residential loans funded in the quarter, $ billion | 187.6 | 200.5 | +6.8% |
| Investor share of new loans | 34.1% | 35.6% | +1.48 |
| Share of new loans with a loan to value ratio of 80% or more | 30.4% | 29.7% | −0.72 |
| Share of new loans at six times income or more | 5.5% | 5.6% | +0.1 |
| Share of new owner-occupier loans at six times income or more | 3.7% | 3.7% | −0.04 |
| Share of new investor loans at six times income or more | 8.7% | 8.9% | +0.13 |
| Share of residential loans 30 to 89 days past due | 0.66% | 0.54% | −0.12 |
| Share of residential loans non-performing | 1.04% | 1.01% | −0.03 |
| Commercial property exposures, $ billion | 463.3 | 501.6 | +8.3% |
| Commercial property exposure limits, $ billion | 498.2 | 541.0 | +8.6% |
ADIs are banks, building societies and credit unions. Shares are of residential loans outstanding at the end of June, or of new loans funded during the June quarter. A change in an amount is in per cent; a change in a share is in percentage points, as APRA publishes it.
Reading the table
Threefold Capital interpretation
- What it shows
- Residential lending by ADIs grew 7.0% over the year to June. Investors took a slightly larger share of new loans, a slightly smaller share of new lending was at high loan to value ratios, and lending at six times income or more held at 5.6%. Early arrears fell and non-performing loans were a little lower than a year earlier. Commercial property exposures grew 8.3%.
- What it means
- On these figures, lenders were not loosening their standards to keep volumes up, and the stress from higher rates had not shown up in arrears by June. Two things temper that reading. The figures end before the May rise had fully reached repayments and before the September rise, and commercial property exposures are growing faster than residential lending, in a market where the Reserve Bank sees strong competition for the better projects.
- How to respond
- Borrowers can expect lenders to apply the same tests as before, at today's higher rates. We read the competition for commercial property loans that the Reserve Bank describes as helping borrowers with strong tenants and moderate debt, and doing much less for the rest.
Lending at high multiples of income is well below the limit APRA introduced this year. Since 1 February 2026, ADIs may make no more than 20% of their new owner-occupier lending, and separately no more than 20% of their new investor lending, at a debt to income ratio of six times or more; loans to build or buy new homes, and owner-occupiers' bridging loans, are exempt.16 APRA sets limits like this under Attachment C of APS 220, which lets it limit particular types of lending.17 In the June quarter the shares were 3.7% and 8.9%,6 and the Reserve Bank describes the aggregate as "well below" the limits.8 The limits apply to each lender's own lending, so a low share for the system does not show whether a particular lender is close to its cap.
The serviceability buffer is unchanged. This is the margin above a loan's rate at which a lender must test whether the borrower can meet the repayments. APS 220 requires at least 3.0%, which APRA may vary between 2.0% and 5.0% (Attachment C, paragraph 4),17 and APRA left it at 3 percentage points in May.18 Because the buffer is added to the loan's rate, the test rate rises whenever rates do, which is the mechanism behind the first model below.
Arrears remain low, despite what the Reserve Bank calls "a recent pick-up".8 A loan is non-performing when the borrower is 90 days or more past due or the lender considers the borrower unlikely to pay in full (APS 220, paragraph 13(a)).17 Non-performing housing loans were 1.01% of residential loans in June, a little lower than a year earlier,6 although the Reserve Bank notes that housing non-performing loans "increased slightly in the June quarter".7 Its own measure of loans more than three months behind has "increased a little over the year to date but remains around pre-pandemic levels".8
Commercial property, or commercial real estate (CRE) in the Reserve Bank's shorthand, is the other large book. ADIs' commercial property exposures were $501.6 billion in June, up 8.3% in a year, against limits of $541.0 billion.6 The Reserve Bank notes that commercial property loans have risen as a share of banks' assets in recent years, driven by the major banks' lending to industrial property, and that the share classified as non-performing remains low. "Competition in CRE lending remains strong from both bank and non-bank lenders, particularly for higher-quality projects." Conditions are weaker for lower-grade offices and in areas with high vacancies, such as parts of Melbourne.8
Who is under pressure
Higher rates spread their weight unevenly.
- Share of household disposable income taken by scheduled mortgage and consumer credit payments in the June quarter 2026, close to the 2024 peak
- Under 12%
- RBA
- Variable-rate owner-occupiers whose income does not cover their repayments and essential expenses
- Around 2%
- RBA
- Companies entering external administration for the first time in 2025‑26
- 14,153
- ASIC
The aggregate figures say that households and businesses are coping. The pressure sits in particular groups.
Households. The Reserve Bank estimates that around 2% of variable-rate owner-occupier borrowers have a cash flow shortfall: their income does not cover their scheduled repayments and essential expenses. The share rose a little in the first half of 2026, and for borrowers in deficit the median shortfall is around 6% of income. Most mortgagors hold large savings in offset and redraw accounts, and the median borrower could cover more than a year of scheduled payments from them.8
The Reserve Bank projects the share in shortfall to stay around its current level, a little under 2%, for some time. One reason is that borrowers who took out loans in recent years "have faced loan servicing capacity requirements greater than those faced by pandemic-period borrowers and well above current interest rates".8 The buffer did its job: people who borrowed recently were tested at rates higher than the ones they now pay. The group we would watch most closely is recent buyers with small deposits. The share of lending at the highest loan to value ratios has increased since the Australian Government 5% Deposit Scheme was expanded in October 2025, the Reserve Bank says, although it remains low overall;8 APRA's broader measure, loans at 80% or more of value, was a little lower in the June quarter than a year earlier.6
Businesses. Total company insolvencies as a share of operating companies declined over the past year and are around their longer-run average, but they remain elevated in some industries, "particularly hospitality, construction and transport".8 The Australian Securities and Investments Commission (ASIC) counted 14,153 companies entering external administration for the first time in 2025-26, down from 14,722 in 2024-25 and equal to 0.38% of registered companies. Construction accounted for 24.5% of them, and accommodation and food services for 14.7%.19,20
External administration is not the same as closure. ASIC's count includes voluntary administrations, which can end in a deed of company arrangement that lets a company keep trading, and small business restructurings, in which the directors stay in control. It does not count companies wound up while solvent, or companies that are simply deregistered.21,22
Who carries the losses matters as much as how many companies fail. Around 75% of the companies that entered insolvency in 2024-25 had fewer than 20 full-time employees, and around 70% owed nothing to secured creditors.8 That, we think, is why a high count of external administrations can sit alongside low losses for lenders that hold security: most of the cost falls on suppliers and other unsecured creditors, and on owners and guarantors. The banks' exposure lies more in smaller businesses whose loans are secured on the owner's home, where rate rises pass through fastest.
Non-banks and private credit. Lenders outside the banking system are small relative to it. Non-bank lenders account for around 6% of financial system assets. Private credit, lending by investment funds rather than banks, accounts for around 10% of total business debt on the largest available estimates, and less than 2% of financial system assets. It is, though, "typically concentrated in real estate lending, including for construction and development", and the Reserve Bank refers to "the large number of funds with exposures to the recent insolvency of a larger property developer". The banks' own exposure is limited: the Reserve Bank's estimate of the upper bound on their drawn exposures to private credit funds is 2.8% of their assets, and it expects the actual figure to be much smaller.7 Estimates of size vary with what is measured. The Reserve Bank put private credit at around $50 billion of credit outstanding at December 2025, noting gaps in the data, against an industry estimate it cites of $224 billion in assets under management.13
Reported data
Where the strain shows, and how far it reaches
Scroll the table sideways to see every column.
| Measure | Latest | Period |
|---|---|---|
| Scheduled mortgage and consumer credit payments, share of household disposable income | Just under 12% | June quarter 2026 |
| Variable-rate owner-occupiers whose income does not cover repayments and essential expenses | Around 2% | First half of 2026 |
| Median shortfall of those borrowers, share of income | Around 6% | October 2026 estimate |
| Scheduled mortgage payments the median borrower could meet from offset and redraw balances | Over a year | October 2026 estimate |
| Housing loans 30 to 89 days past due, ADIs | 0.54% | June 2026 |
| Companies entering external administration for the first time | 14,153 | 2025-26 |
| Companies entering insolvency with fewer than 20 full-time employees | Around 75% | 2024-25 |
| Companies entering insolvency with no debt owed to secured creditors | Around 70% | 2024-25 |
The shortfall and savings buffer figures are the Reserve Bank's estimates. The insolvency profile is for 2024-25, the latest year the Reserve Bank reports.
Reading the table
Threefold Capital interpretation
- What it shows
- Repayments take just under 12% of household income, close to the 2024 peak, and around 2% of variable-rate owner-occupiers are short each month, by a median of around 6% of income. The median borrower holds more than a year of payments in offset and redraw. Of the companies that fail, most are small and owe nothing to secured lenders.
- What it means
- Stress is concentrated, not general. A small share of households is short each month, while a much larger share has savings to absorb further rises. In business, a high number of failures translates into modest losses for secured lenders, and most of the cost falls on unsecured creditors, owners and guarantors.
- How to respond
- For a household or a business, the number that matters is its own buffer: months of repayments held in cash or offset, and the cash left each month after debt service at a higher rate. In our view an owner whose business loan is secured on the family home has the most reason to plan conservatively, because one rise reaches both.
How resilient the lenders are
The question for the system is whether lenders can absorb losses and keep lending. On the Reserve Bank's evidence, they can. Its overall judgement is plain: "The Australian financial system has a good degree of resilience." It sees external factors as the most prominent threats to financial stability.23
Capital is the shareholders' money behind a bank's loans. It absorbs losses before depositors are affected. APRA measures it against risk-weighted assets, the bank's exposures weighted by their riskiness. The banking system's ratio of Common Equity Tier 1 capital, the highest-quality form, to risk-weighted assets was 12.4% in June 2026, "well above regulatory requirements".7 The requirement, with its buffers, is 8.00% to 10.25% of risk-weighted assets depending on the kind of bank;24 Part 1 sets out how it is built up, what capital does and how risk weights work.
Banks also hold provisions, money set aside for expected losses, of around 0.7% of total credit outstanding. More than 90% of housing non-performing loans, and more than 50% of business ones, were considered well secured in June, which limits the loss if they are not repaid. Liquid assets are well above the regulatory minimums, and the Reserve Bank concludes that banks "could absorb large losses from a very adverse shock while continuing to extend credit".7 The Council of Financial Regulators said in September that the banking system had "maintained strong capital and liquidity positions alongside prudent lending standards".25
The Reserve Bank also tests the system against severe scenarios. They are not forecasts; they ask what would happen if a great deal went wrong at once.
Reported data
The Reserve Bank's severe scenarios: what they assume, and what they show
Scroll the table sideways to see every column.
| Scenario | What it assumes | What the Reserve Bank estimates |
|---|---|---|
| A very adverse downturn | A shock, possibly from overseas, in which GDP growth falls 3 percentage points, housing and commercial property prices fall 20%, unemployment rises to 6.3%, inflation to 7% and the cash rate to 5.6% | The banking system's CET1 ratio falls from 12.4% to around 11.6%, with very few banks expected to use a substantial amount of their capital buffers; the share of mortgagors at higher risk of default rises to around 5%, a little higher than the peak in 2023 |
| A large fall in housing prices | A uniform fall of 20% from current levels, applied to the loans in the Reserve Bank's Securitisation System data | Around 5% of mortgages fall into negative equity, owing more than the home is worth |
The Reserve Bank's own scenario estimates, published in its Financial Stability Review. They test resilience; they are not forecasts. The downturn's assumptions are drawn from the Review's two chapters, which describe the one scenario; the negative equity estimate uses the Reserve Bank's data on securitised loans.
Reading the table
Threefold Capital interpretation
- What it shows
- In the Reserve Bank's very adverse downturn, the banks' CET1 ratio falls from 12.4% to around 11.6%, very few banks use much of their buffers, and the share of mortgagors at higher risk of default rises to around 5%, a little above the 2023 peak. A uniform 20% fall in housing prices would leave around 5% of mortgages in negative equity.
- What it means
- In the Reserve Bank's estimate, the banks' capital would absorb its very adverse downturn with room to spare, and most mortgages carry enough equity to withstand a large fall in prices. The scenario also points, we think, to where the real risk sits: unemployment, more than interest rates, is what turns a stretched borrower into one who defaults.
- How to respond
- Read these results as evidence about the system's capacity to keep lending through a downturn, not about any one lender's appetite. A well capitalised lender can still decline a loan that fails its tests, which is the subject of the next two sections.
Interest rates and borrowing power
Higher rates do two things to credit. They raise the cost of existing loans, which the figures above trace. They also shrink new loans, through a mechanism most borrowers never see.
Before approving a home loan, an ADI must check that the borrower could keep up the repayments at a rate at least 3 percentage points above the one they will pay, the buffer described above. In simplified form, the lender works out what is left of the borrower's income after tax once living expenses and other repayments are paid, and lends no more than that surplus would repay at the test rate. Part 2 explains how lenders measure income and expenses, and the rules behind the buffer.
The consequence is arithmetic. When rates rise, the test rate rises by the same amount, and the loan that a given surplus can repay falls. Nobody at the lender has to change a policy: the same rules, applied at a higher rate, produce a smaller loan. That is how credit can tighten without any lender tightening.
Our model takes a hypothetical household with $12,000 a month after tax, $5,000 a month of living expenses and no other debts, borrowing over 30 years at an assumed rate of 6.25%. The rate is an assumption, not a reported figure; the Reserve Bank publishes the average rates on new and outstanding housing loans each month in its Statistical Table F6, and the model's rate control can be set to the current one. Tested at 9.25%, the household's surplus of $7,000 a month supports a loan of $851,000, before the further margins lenders keep, such as a minimum surplus after the test repayment and an expense benchmark scaled to income. At 5.25%, as if the 2026 rises had not been passed on, the same household could borrow $932,000. If lenders pass the year's rises on in full, they cut its borrowing power by 8.7%. One more percentage point would take it to $781,000, a further cut of 8.2%.
The buffer itself accounts for $286,000. Tested at the loan's own rate, the household could borrow $1,137,000. The buffer leaves room: at the actual rate, the repayments on $851,000 would be $5,239 a month, 44% of income after tax.
Illustrative model
Model: what higher rates do to borrowing power
Choose a scenario, or set your own income, expenses and rate. The lender lends what the monthly surplus would repay at the test rate, the loan's rate plus the buffer.
Without JavaScript the model shows one scenario: Household, $12,000 a month.
Borrowing power $851,000tested at 9.25%
- What a lender might lend
- $851,000
- Monthly repayment at the loan's rate
- $5,239
- Against one point lower, before the 2026 rises
- −8.7%
- Repayment, share of income after tax
- 44%
The lender tests the repayments at 9.25%, 3.00 percentage points above the loan's rate of 6.25%. Without the buffer it could lend $286,000 more.
At this rate, one more percentage point on rates cuts borrowing power by 8.2%. One point less, as before the rises of 2026, would lift it to $932,000.
At the loan's actual rate, the repayments would take 44% of income after tax, $5,239 a month.
Interest rates and borrowing power
Illustrative model
What each change does to borrowing power in the household example
Scroll the table sideways to see every column.
| Change | Borrowing power |
|---|---|
| Loan rate a quarter of a point higher | −$18,000 |
| Loan rate one point higher | −$70,000 |
| Loan rate half a point lower | +$39,000 |
| Loan rate one point lower, as before the 2026 rises | +$81,000 |
| No buffer: tested at the loan's own rate | +$286,000 |
| A car loan costing $900 a month | −$109,000 |
| Living expenses $500 a month higher | −$61,000 |
| Income after tax $1,000 a month higher | +$122,000 |
Each line changes one setting of the household example, with income of $12,000 a month after tax, living expenses of $5,000 a month and a 30-year loan at 6.25% tested at 9.25%, and leaves the rest as they are.
Reading the model
Threefold Capital interpretation
- What it shows
- In the household example, the 2026 rises cut borrowing power by 8.7%, and one more percentage point cuts it by a further 8.2%. The buffer removes $286,000. Other debts matter more than most borrowers expect: a car loan costing $900 a month removes more borrowing power than a full point on rates.
- What it means
- Borrowing power falls automatically as rates rise, because the lender's test rate rises with them. Across the economy, that reduces how much new credit households can take on, without any change in lending rules, and it is one of the ways higher rates slow the housing market. It works in reverse too: a cut lifts borrowing power by the same arithmetic.
- How to respond
- A borrowing figure is only as good as the rate it was worked out at. Before relying on one, it is worth working it out at the test rate rather than the advertised rate, and again a point higher. In the example, clearing a car loan of $900 a month does more for borrowing power than a fall of a full point in rates, which no borrower controls.
The refinancing squeeze
A home loan usually runs for decades; many commercial property and business loans do not. The Commonwealth Bank (CommBank), for example, describes its Lease Doc loan for tenanted commercial investment property as running for up to five years or until the lease expires, whichever is shorter (page seen 12 October 2026),26 and ANZ's commercial broker team has described investment property funding with terms of up to five years, matched to the remaining lease (21 November 2025).27 Those are two lenders' terms for particular products, not market norms, but they show why commercial borrowers refinance far more often than home owners. At each renewal, the lender tests the loan again as if it were new.
Two tests do most of the work. The value test caps the loan at a share of the property's value today: the loan to value ratio, where a lender's published maximum can be as low as 65%.26 The income test caps it at the loan whose interest the property's net rent covers a set number of times: the interest cover ratio. The lender lends the lower of the two. Part 4 sets out lenders' published limits, and Part 5 how they test a business.
Rising rates tighten the income test directly, because the same rent covers less interest. Softer values tighten the value test. A borrower can meet every repayment, keep the tenant and collect the rent, and still owe more than a lender will now advance. The difference has to come from somewhere: new equity, a sale, or a lender whose limits differ.
Our model takes a hypothetical investor with a loan of $1,300,000 at 4%, secured on a commercial property that was worth $2,000,000 when the loan was made, and net rent of $120,000 a year. The property is now worth 10% less, and the loan is up for renewal at 7% with a lender that will lend up to 65% of today's value and wants the rent to cover the interest at least 1.5 times. The rates and limits are our assumptions, not any lender's. Lenders usually test the rent at a rate above the loan's (Macquarie's guide for brokers, dated May 2026, says its minimum interest cover for property investors includes a buffer, currently 1 percentage point)28 and often on principal and interest repayments rather than interest alone, so a real income test is tighter than ours; setting the model's new rate higher shows the effect.
At 4%, the rent covered the interest 2.31×. At 7%, it covers interest on the same balance only 1.32×, below the lender's minimum. The income test now allows a loan of $1,143,000, less than the $1,170,000 the value test would allow, so the investor needs $157,000 of new equity to refinance. Interest on the balance rises by $39,000 a year. On a lender's stricter test the gap would be larger: the model's figure is a floor, not an estimate.
The striking result is that the fall in value is not what creates the gap. With the value unchanged, the shortfall is the same $157,000: the rate alone does it. Value takes over only when it falls further. At 20% down, the value test binds and the gap widens to $260,000. A rent 10% higher helps the income test but not the value test, and leaves $130,000 to find. A loan paid down to $1,000,000 before the renewal passes both tests, with $143,000 to spare.
Illustrative model
Model: the refinancing squeeze
Choose a scenario, or change the property's value, the rent, the rates and the new lender's limits. The new lender lends the lower of what its value test and its income test allow.
Without JavaScript the model shows one scenario: Value down 10%.
New loan available $1,143,000shortfall $157,000
- Equity needed to refinance
- $157,000
- Most a new lender would lend
- $1,143,000
- Interest cover at the new rate
- 1.32×
- Loan, share of today's value
- 72%
The income test sets the limit: at 7.00%, net rent of $120,000 covers the interest 1.50× on a loan of no more than $1,143,000. The value test alone would allow $1,170,000.
To refinance the $1,300,000 owed, the owner must find $157,000 from other money, or a lender whose limits differ.
At the new rate, interest on the same balance rises by $39,000 a year, to $91,000.
The two tests a new lender applies
Illustrative model
The refinance under each scenario
Scroll the table sideways to see every column.
| Line | Value unchanged, rate up | Value down 10% | Value down 20% | Value down 10%, rent up 10% | Loan paid down to $1 million |
|---|---|---|---|---|---|
| Change in value since | 0% | −10% | −20% | −10% | −10% |
| Net rent a year | $120,000 | $120,000 | $120,000 | $132,000 | $120,000 |
| Loan balance now | $1,300,000 | $1,300,000 | $1,300,000 | $1,300,000 | $1,000,000 |
| Value today | $2,000,000 | $1,800,000 | $1,600,000 | $1,800,000 | $1,800,000 |
| Loan allowed by the value test | $1,300,000 | $1,170,000 | $1,040,000 | $1,170,000 | $1,170,000 |
| Loan allowed by the income test | $1,143,000 | $1,143,000 | $1,143,000 | $1,257,000 | $1,143,000 |
| Most a new lender would lend | $1,143,000 | $1,143,000 | $1,040,000 | $1,170,000 | $1,143,000 |
| Equity needed to refinance | $157,000 | $157,000 | $260,000 | $130,000 | $0 |
| Loan to value today | 65% | 72% | 81% | 72% | 56% |
| Interest cover at the old rate | 2.31× | 2.31× | 2.31× | 2.54× | 3.00× |
| Interest cover at the new rate | 1.32× | 1.32× | 1.32× | 1.45× | 1.71× |
| Extra interest a year | +$39,000 | +$39,000 | +$39,000 | +$39,000 | +$30,000 |
In every scenario the property was worth $2,000,000 when the loan was made, the old rate was 4% and the new rate is 7%, and the new lender lends up to 65% of today's value with minimum interest cover of 1.5 times.
Illustrative model
Equity needed to refinance, by the fall in value and the new interest rate
Scroll the table sideways to see every column.
| Change in value | Rate 6.00% | Rate 7.00% | Rate 8.00% |
|---|---|---|---|
| 0% | $0 | $157,000 | $300,000 |
| −10% | $130,000 | $157,000 | $300,000 |
| −20% | $260,000 | $260,000 | $300,000 |
| −30% | $390,000 | $390,000 | $390,000 |
The rest of the example is unchanged: a balance of $1,300,000, net rent of $120,000 a year, a maximum loan of 65% of today's value and minimum interest cover of 1.5 times. $0 means the loan passes both tests.
Reading the model
Threefold Capital interpretation
- What it shows
- At 7%, the example needs $157,000 of new equity whether its value has held or fallen 10%, because the income test binds first. A 20% fall in value makes the value test bind and lifts the gap to $260,000. In the grid, at 8% the income test sets the gap for any fall in value up to 20%, and at a 30% fall the value test sets it at every rate shown.
- What it means
- A refinance is a new credit decision made on old numbers. A borrower who has done nothing wrong can need new equity because the rate used to size the loan has risen and the asset has been revalued. The gap arrives as a lump sum on a known date if the borrower has to change lender; an existing lender more often agrees a staged paydown, a higher margin or extra security in exchange for rolling the facility, which is why the early conversation described below matters. Lower rates help only until the value test takes over: in the grid, a property that has lost 20% of its value needs the same equity at 6% as at 7%.
- How to respond
- For a loan due for renewal in the next year or two, it is worth running both tests at a rate above today's and a value below the last valuation, to see which one binds. The answer shows where effort counts: on rent and lease terms if the income test binds, on paying down debt if the value test does.
The squeeze is not universal. The Reserve Bank notes that strong competition among lenders has allowed some commercial property investors to refinance at favourable rates, partly offsetting higher interest rates.8 Where it lands, in our reading, is on loans that were sized on low rates, on weaker or shorter-leased assets, and on borrowers who arrive at the renewal date without the information a new lender needs.
How we read it
What follows is our interpretation of the data: where the pressure will show first, what lenders tend to do at this point in a cycle, and what the borrowers who refinance most easily tend to have in common.
Where the pressure will show first
The pressure from this year's rates will show first at refinancing dates, not in missed repayments. Arrears are the last stage of stress, and they are reported late. A refinance, a facility review or a covenant test is a scheduled moment at which a lender tests a loan again at today's rates and values, and in 2026 both have moved against loans written when rates were lower. The groups we would watch are below.
Threefold Capital interpretation
Where we expect the pressure to show first
Scroll the table sideways to see every column.
| Borrower | Why the pressure lands here | How it tends to show first | Data to watch |
|---|---|---|---|
| Commercial property investors with loans written at lower rates | Short loan terms, an income test at today's rates, and softer values for weaker assets | A smaller loan offered at renewal, or a request for a paydown or more security | APRA's commercial property exposures; the Reserve Bank's commentary on commercial property |
| Small businesses with variable loans secured on the owner's home | The fastest pass-through of rate rises, with the family home behind the business | An overdraft that no longer returns to zero, tax and super falling behind, requests to restructure | ASIC's external administrations; business non-performing loans |
| Recent home buyers with small deposits | The smallest savings buffers, and the least equity if prices fall | Hardship requests, and loans 30 to 89 days past due | APRA's early arrears; the Reserve Bank's shortfall estimates |
| Builders and developers | Fixed-price contracts, presales that can fall over, and funding from private credit at higher rates | Delayed settlements, extended construction loans, failures along the chain of subcontractors | ASIC's external administrations by industry; the Reserve Bank's reporting on private credit |
| Residential investors | Rent has to carry more of the interest, and the Reserve Bank links weaker sentiment partly to changes to negative gearing and the capital gains tax discount | Fewer new loans and more sales | ABS investor commitments; APRA's investor shares |
Reading the table
Threefold Capital interpretation
- What it shows
- Our reading of which borrowers feel this year's rates first, why, and where the evidence will appear.
- What it means
- Most of these pressures arrive on a date: a renewal, a review, a settlement. That makes them easier to see coming than a fall in income, and easier to prepare for.
- How to respond
- Know the date on which the lender will next look at the loan, and what it will test. Most of what improves the outcome takes months, not weeks.
What lenders tend to do in this phase
Lenders at this point in a cycle rarely announce that they are tightening, and rarely need to. The tests do it for them: the serviceability buffer sits on a higher rate, interest cover is measured at a higher rate, and valuations follow prices down. Beyond that, we would expect the following, roughly in this order.
- More attention to the existing book. Loans with renewals, reviews or covenant tests coming up are reviewed first, with earlier requests for financial statements, rent rolls and valuations.
- Terms before headline limits. Shorter terms, principal repayments in place of interest only, tighter covenants and more conditions tend to change before a lender's published maximum loan to value ratio does.
- Pricing for risk rather than withdrawal. With the strong competition the Reserve Bank describes for higher-quality commercial property loans, we would expect lenders to keep competing for strong borrowers and to charge weaker ones more, rather than to stop lending to a sector.
- Other lenders filling gaps, at a price. A borrower who falls outside one lender's tests may find a lender whose limits differ, usually at a higher rate and on shorter terms.
Bank pricing so far points, we think, to competition rather than caution. The spread between banks' lending rates and their funding costs has widened since early 2025, but mainly because funding costs fell relative to the cash rate, and it remains well below its pre-pandemic level; the spread over the cash rate itself is at its narrowest in almost two decades.11
What the borrowers who refinance well have in common
The borrowers who come through a refinance like the one modelled above most easily tend to have done a few things early.
- A year out. They know the expiry and review dates and the covenants in the facility agreement a year ahead, and have run the lender's tests on today's figures at a higher rate and a lower value.
- An early conversation, about the renewal. A lender that hears about a gap a year ahead has options it lacks a month before expiry, such as a longer term, a staged paydown or more security, and the borrowers who do this well treat it as a commercial conversation about the renewal rather than a request for hardship help. Separately, the banks that have adopted the Banking Code of Practice, called Code banks here, have obligations to small business customers in financial difficulty, which the Code defines as being unable to repay what is owed, expecting to be unable to pay upcoming repayments, or having difficulty meeting repayment obligations (paragraph 168). The Code asks such a customer, or their representative, to contact the bank as soon as possible (paragraph 167), and with the customer's co-operation a Code bank will work with them to help find a sustainable solution (paragraph 175).29
- Headroom built before it is needed. Retained cash, debt paid down and leases extended all improve the tests. In our model, a balance of $1,000,000 instead of $1,300,000 turns a shortfall into headroom.
- The information ready. Current financial statements and management accounts, a schedule of tenants and lease expiries, a plain account of any weak period, and a recent valuation if one exists, though a lender will usually instruct its own valuer and rely on that report. A credit officer who has to chase information tends to assume the worst about what is missing.
- The alternatives known in advance. A partial sale, an equity partner or a lender with different limits each has a cost, and the time to learn it is before the facility expires.
The outlook: three scenarios
A year ahead, the range of outcomes for credit is wide. We separate three things: what has been measured, what the Reserve Bank forecasts, and the scenarios we use to test plans.
What has been measured. Consumer prices rose 4.0% over the year to August, and the trimmed mean, which leaves out the largest price moves, 3.6%.14 Unemployment was 4.6% in August.30 The cash rate is 4.60%, credit is growing 8.4% a year, and new housing lending fell in the June quarter.
What the Reserve Bank forecasts. In August, the Reserve Bank expected inflation to return to the middle of its 2 to 3% target range by early 2028, and unemployment to rise to 4.8% by the end of 2028.31
Reported data
The Reserve Bank's August forecasts for inflation, unemployment and growth
Scroll the table sideways to see every column.
| Per cent | Dec 2026 | Jun 2027 | Dec 2027 | Jun 2028 | Dec 2028 |
|---|---|---|---|---|---|
| Headline inflation | 3.6 | 2.8 | 2.6 | 2.4 | 2.4 |
| Trimmed mean inflation | 3.3 | 3.0 | 2.6 | 2.4 | 2.4 |
| Unemployment rate | 4.5 | 4.6 | 4.7 | 4.8 | 4.8 |
| Growth in GDP | 1.4 | 1.5 | 1.6 | 1.6 | 1.8 |
Forecasts finalised on 5 August 2026, before the September rise, with the cash rate assumed to follow market pricing at that date, which included around half of one further rise by the end of 2026.3 Inflation and growth are year-ended changes; unemployment is the average rate in the quarter. New forecasts are due in the November Statement.
Reading the forecasts
Threefold Capital interpretation
- What it shows
- The Reserve Bank expected inflation to ease through 2027 and reach 2.4% by mid 2028, unemployment to drift up to 4.8%, and growth to stay below 2% throughout.
- What it means
- The forecasts describe a slow return to target with a gently weaker labour market. In our view that path, on its own, would mean arrears that rise, but slowly. The forecasts were finalised before the September rise, though the market pricing they assume already included about half of one further rise, and the Board said in September that some of the upside risks it had flagged in August were materialising.2
- How to respond
- Treat the forecast as the Reserve Bank's central case, not as a plan. A borrowing plan that works only on the forecast path has no margin for the risks the Board itself describes.
The scenarios we use. These are planning cases, not forecasts. They describe directions and mechanisms, not levels.
Forward-looking scenario
Three cases for credit conditions through 2027
Scroll the table sideways to see every column.
| Case | What would have to happen | What it would mean for borrowers | Signs it is happening |
|---|---|---|---|
| Rates on hold | Inflation eases broadly as the Reserve Bank forecast in August, and the Board holds the cash rate into 2027 | Repayments stop rising once the September rise has passed through, and borrowing power stays where it is. Loans written at lower rates still face the income test at today's rates when they are renewed, so refinancing pressure fades only as those loans roll over. Arrears drift up slowly as savings are used | Monthly inflation and the trimmed mean falling; the Board's statements no longer pointing to further rises |
| One more rise, or two | Inflation stays high, the upside risks the Board described in September keep materialising, and the cash rate rises again | Each rise cuts borrowing power again and raises the bar at refinance. More households move into shortfall, more refinances need new equity, and lenders tighten terms for weaker borrowers. The banks start from a strong position | Services prices and wages staying firm; short-term inflation expectations high; the November forecasts revised up |
| Cuts in 2027 | Inflation falls faster than forecast, or unemployment rises faster, and the Board cuts in 2027 | Borrowing power recovers by the same arithmetic that cut it. Refinancing eases through the income test before values recover. If the cuts come because unemployment is rising, arrears can rise even as rates fall | Inflation back in the target band; unemployment rising faster than forecast |
Planning cases, not forecasts. None of them assigns a level to the cash rate.
Reading the cases
Threefold Capital interpretation
- What it shows
- Three directions for rates through 2027, and what each would mean for borrowing power, refinancing and arrears.
- What it means
- The case that looks best for borrowers carries its own risk. Cuts that come because unemployment is rising would ease repayments for those still in work while raising arrears among those who are not. In our view the second case, at least one more rise, is the prudent one to plan around, not because we predict it but because a plan that survives it survives the others.
- How to respond
- Test any borrowing plan against the second case before relying on it, and decide in advance what would be done if it arrived: which costs would be cut, which assets could be sold, and which conversations with a lender would start first.
Our models give a sense of scale, as arithmetic rather than prediction. In the household example, a further quarter-point rise passed on in full would take borrowing power from $851,000 to $832,000, and a half-point cut would lift it to $890,000. In the refinancing example, a new rate of 7.5% rather than 7% would widen the gap from $157,000 to $233,000, while a rate of 6.5% would narrow it only to $130,000, because the value test then sets the limit.
What to watch. The September consumer price index, later in October; the Reserve Bank's financial aggregates for September, on 30 October;32 the Board's decision on 3 November,10 with new forecasts in the November Statement on Monetary Policy; and the ABS lending indicators for the September quarter, on 11 November.5 After those, APRA's September quarter statistics and the Reserve Bank's half-yearly Financial Stability Review32 will show whether arrears and commercial property lending have begun to move.
Conclusion
Credit in 2026 is dearer, still plentiful in total, and tighter at the edges. The cash rate has risen a full percentage point, lenders have passed the rises on quickly, and with almost every mortgage now variable they reach nearly every borrower. Credit is still growing at 8.4% a year, but new lending has turned down, led by investors. Arrears are low and rising only slightly, and the banks hold capital that would, on the Reserve Bank's estimates, absorb its very adverse downturn with room to spare.
The tightening that matters for most borrowers is not in the headlines. It is in the arithmetic of a lender's tests. In our household example, the year's rises take borrowing power from $932,000 to $851,000 without any lender changing a rule, and in our refinancing example an investor who has met every repayment needs $157,000 of new equity to keep the loan. In our view, the borrowers who come through this phase well are the ones who run those tests on themselves before a lender does, and who start the conversation while there is still time to act on the answer.
Method and sources
Cut-off. This article uses data published on or before 12 October 2026. It does not include the September consumer price index, the September financial aggregates (30 October), the Board's November decision and Statement on Monetary Policy, the September quarter lending indicators (11 November) or APRA's September quarter statistics.
Sources. Each figure comes from its original publisher: the Reserve Bank, APRA, the ABS, ASIC, the Council of Financial Regulators and the Australian Banking Association, and three lenders' own pages for the terms of their own products. Quotations are from the documents cited. No secondary sources are used.
Periods and measures. Periods differ, and each figure is given for its own. Credit growth is the change over a year in the stock of credit provided by financial institutions operating in Australia, seasonally adjusted and adjusted for breaks in the series; where the Reserve Bank's six-month-ended annualised measure is quoted, the text says so. The ABS figures are borrower-accepted commitments, seasonally adjusted and excluding refinancing, and first home buyers are part of owner occupiers. APRA's figures cover banks, building societies and credit unions and count loans funded, so they do not match the ABS series. The Reserve Bank's figures for shortfalls, savings buffers, negative equity and its scenario are its own estimates, and its forecasts were finalised on 5 August 2026. The cash rate chart and the dashboard's trend line show the target at the end of each month.
Our calculations. Month-end levels of the cash rate come from the Reserve Bank's table of changes, and the rise of 1.00 percentage point in 2026 is 4.60% less 3.60%. The rate was last above 4.60% before 2 November 2011, when it was cut from 4.75%, which is why we call it the highest since 2011. The 75 basis points of the first three rises are compared with the rise in variable mortgage rates from January to June, and the three quarters of a point reflected in most variable-rate borrowers' repayments at 12 October is those three rises, with the September rise still to reach scheduled repayments. Credit growth of about twice the rate of consumer prices compares 8.4% with 4.0%. The timing of the September rise in repayments applies the Reserve Bank's three-month rule of thumb to 30 September.
The models. Both are hypothetical, and every setting is our assumption. The borrowing power model lends the amount whose monthly repayments at the test rate, the loan's rate plus the buffer, equal income after tax less living expenses and other repayments. It leaves out expense benchmarks, discounts on some kinds of income, credit card limits, minimum surpluses and lenders' own policies, and its scenario before the 2026 rises assumes they were passed on in full. The refinancing model lends the lower of a maximum share of today's value and the loan whose interest the net rent covers a set number of times, at the new loan's own rate and on interest alone, which is gentler than most lenders' tests. It leaves out fees, repayments of principal, the tenant, the lease and how a lender values the property. Each model is written twice, once for the figures printed on the page and once to recalculate them as the controls move, and the two are tested against each other on every scenario and thousands of other settings. Every figure quoted from a model is filled in by the model itself.
Labels. Reported data: figures published by the source named, including the Reserve Bank's forecasts and scenario estimates, each in an exhibit of its own. Threefold Capital interpretation: our reading of the evidence. Illustrative model: hypothetical figures from our models. Forward-looking scenario: a planning case, not a forecast.
Limitations. Credit growth and arrears move slowly, and the data end before the September rise reached repayments. The share of high debt to income lending is for the whole system, not any one lender. Estimates of private credit vary with the definition used, and lenders' published terms change.
Corrections. If you see an error, please tell us at threefoldcapitalau@gmail.com. We correct published articles and record what changed at the top of the page.
Sources
- Reserve Bank of Australia, Cash rate targetThe cash rate target and the date each change took effect, including the three cuts of 2025 and the four rises of 2026
- Reserve Bank of Australia, Statement by the Monetary Policy Board: Monetary Policy Decision, 29 September 2026The rise to 4.60% and the Board's reasons, quoted
- Reserve Bank of Australia, Statement on Monetary Policy, August 2026, Financial Conditions, 11 August 2026Pass-through to deposit, mortgage and business rates; scheduled payments; the three-month lag; credit growth, including the six-month-ended annualised measure for housing; the decline in new housing loan commitments; competition between lenders; market pricing of around half a further rise by the end of 2026 at the time of the Statement
- Reserve Bank of Australia, Financial Aggregates, August 2026, 30 September 2026Growth in total, housing, business and personal credit over the year, and in July and August
- ABS, Lending indicators, June Quarter 2026, 14 August 2026New loan commitments for dwellings by value and number, seasonally adjusted and excluding refinancing; the next release on 11 November 2026
- Australian Prudential Regulation Authority, Quarterly authorised deposit-taking institution property exposure statistics, June 2026 highlights, 17 September 2026Residential and commercial property exposures, new loans funded, loan to value and debt to income shares and arrears of banks, building societies and credit unions
- Reserve Bank of Australia, Financial Stability Review, October 2026, Resilience of the Australian Financial System, 1 October 2026Bank capital and the very adverse scenario, with its assumptions in endnote 4; provisions; non-performing loans and how well they are secured; liquidity; non-bank lenders; private credit
- Reserve Bank of Australia, Financial Stability Review, October 2026, Resilience of Australian Households and Businesses, 1 October 2026The 100 basis points of rises; cash flow shortfalls, savings buffers and serviceability; negative equity, and the very adverse downturn's unemployment, inflation and cash rate; pass-through to smaller businesses; insolvencies and the industries where they remain elevated; housing prices; high loan to value lending; commercial property
- Reserve Bank of Australia, Banks' Funding Costs and Lending Rates (explainer)Funding costs do not necessarily move by the same amount or at the same speed as the cash rate
- Reserve Bank of Australia, Monetary Policy Board meeting schedulesThe 2026 meeting dates, including 2 and 3 November
- Reserve Bank of Australia, Bulletin, May 2026, Developments in Banks' Funding Costs and Lending Rates, 28 May 2026The fixed-rate share below 5% in 2025; the spread between lending rates and the cash rate, the spread between lending rates and funding costs, and why funding costs fell relative to the cash rate
- Reserve Bank of Australia, Bulletin, March 2022, Developments in Banks' Funding Costs and Lending RatesFixed-rate loans had risen to around 40% of housing credit by early 2022
- Reserve Bank of Australia, Bulletin, February 2026, Recent Changes in Credit Markets and Their Implications for Monetary Policy, 26 February 2026The narrower spread of mortgage rates over the cash rate since 2019; non-banks' share of business lending; estimates of the size of private credit
- ABS, Consumer Price Index, Australia, August 2026Inflation of 4.0% and trimmed mean inflation of 3.6% over the year to August 2026
- ABS, Lending indicators methodology, June Quarter 2026Borrower-accepted commitments, and the quarterly release
- Australian Prudential Regulation Authority, APRA to limit high debt-to-income home loans to constrain riskier lending, 27 November 2025The limit of 20% of new lending at a debt to income ratio of six or more, applied separately to owner-occupier and investor lending, from 1 February 2026, and the loans it excludes
- Australian Prudential Regulation Authority, Prudential Standard APS 220 Credit Risk Management, Banking (prudential standard) determination No. 14 of 2022Paragraph 13(a), non-performing exposures; Attachment C, paragraph 4, the serviceability buffer of at least 3.0%, which APRA may vary between 2.0% and 5.0%; Attachment C, APRA's power to limit particular types of lending
- Australian Prudential Regulation Authority, APRA maintains current macroprudential policy settings in highly uncertain environment, 28 May 2026The serviceability buffer kept at 3 percentage points in May 2026
- ASIC, Corporate Insolvency Update, Issue 41, October 202614,153 companies in 2025-26; 0.38% of registered companies; shares by industry
- ASIC, Corporate Insolvency Update, Issue 37, September 202514,722 companies in 2024-25
- ASIC, INFO 80, How to interpret ASIC's insolvency statisticsWhat the count of external administrations includes and excludes
- ASIC, INFO 41, Insolvency: a glossary of termsVoluntary administration, deeds of company arrangement and small business restructuring
- Reserve Bank of Australia, Financial Stability Review, October 2026, Financial Stability Assessment, 1 October 2026The overall assessment of the financial system's resilience and the main threats to it
- Australian Prudential Regulation Authority, Response to submissions: Finalising the bank capital reforms, November 2021Common Equity Tier 1 requirements including buffers of 8.00%, 9.25% and 10.25% of risk-weighted assets (Table 3)
- Council of Financial Regulators, Quarterly Statement by the Council of Financial Regulators, 3 September 2026Strong capital and liquidity positions alongside prudent lending standards
- Commonwealth Bank of Australia, Lease Doc loan, page seen 12 October 2026For this product: a term of up to five years or the lease expiry, whichever is shorter, and a loan to value ratio of up to 65%
- ANZ, Commercial property made simple with ANZ support, 21 November 2025Commercial property investment funding with terms of up to 5 years matched to remaining lease periods
- Macquarie Business Banking, Commercial Lending Guide (for brokers), dated May 2026For commercial property investors, minimum interest cover that includes a buffer, currently 1%
- Australian Banking Association, Banking Code of Practice, effective 28 February 2025Paragraphs 167, 168 and 175, on financial difficulty
- ABS, Labour Force, Australia, August 2026Unemployment of 4.6% in August 2026
- Reserve Bank of Australia, Statement on Monetary Policy, August 2026, Overview, 11 August 2026Forecasts for inflation, unemployment and growth, finalised on 5 August 2026
- Reserve Bank of Australia, Schedules and eventsThe Financial Aggregates release on 30 October 2026; the Financial Stability Review is published half-yearly
General information only, current at 12 October 2026. It is not financial, credit, tax or legal advice and does not take into account anyone's objectives, financial situation or needs. The rules described are simplified, and what applies in any particular case depends on the documents and the facts. Threefold Capital does not provide credit assistance and is not a lender. It is not an offer or an invitation to invest. The models are illustrations, not forecasts or quotes, and the views are the firm's own.