Nine Mutual Banks Modelled Climate Risk. The Risk They Found Was Borrower Income — and the Insurance Line on Your Files

Climate stories usually arrive in the broker channel wearing the wrong clothes. They come as sustainability commitments, green loan discounts, or long-range warnings about coastal postcodes in 2050 — the sort of thing you file under “interesting, not urgent” and move on from, because your next appointment is in eleven minutes.

The work nine of Australia’s customer-owned banks published this month is a different kind of document, and it is worth eleven minutes. They did not model damaged houses. They modelled the borrower. And the answer they landed on — that the leading exposure is household income and employment disruption, with insurance costs sitting inside that — points at a line item that lives on your fact-find, not on a lender’s stress-testing spreadsheet.

What the nine banks actually did

On 11 August 2026, nine customer-owned banks released Shared climate scenarios for the mutual banking sector, described as an industry first. The work was built with the actuarial firm Finity and Climate KIC Australia, and it gave the participating institutions a common set of assumptions to test their lending books against rather than each mutual building its own.

Two scenarios were developed, characterised as shared, plausible and challenging versions of the future. Across both, the analysis identified household income and employment disruption as the leading risk area — shaped by climate shifts, insurance costs and broader economic transition.

Stephanie Elliott, chief impact officer at the Customer Owned Banking Association, framed the motivation in terms of proximity to the customer base: “Customer-owned banks are closely connected to the communities they serve and are already seeing how climate pressures can affect household finances.”

Sharanjit Paddam, a principal at Finity, put the transmission mechanism plainly: “Climate change is already reshaping the risks facing Australian households, and customer-owned banks feel that through impacts on their customers.” Chris Lee, chief executive of Climate KIC Australia, added that “climate impacts cut across communities, markets and institutions, so resilience can’t be built in isolation” — which is a reasonable description of why nine competitors pooled the modelling in the first place.

Aaron Newman, chief executive of Queensland Country Bank, tied it back to the member base: “As a member-owned bank, our focus has always been on the wellbeing and resilience of the communities we serve.”

The number that makes this a broker problem

The mutuals’ exercise is qualitative in what has been published so far. The quantitative case has already been made elsewhere, by the same actuary.

The Actuaries Institute’s research paper Home Insurance Affordability and Home Loans at Risk, on which Sharanjit Paddam is lead author, estimates that 5% of Australian households with mortgages are experiencing insurance affordability stress — “representing $57b of loan balances and 3% of all home loan assets.”

That is the sentence to sit with. Not because $57 billion is a systemic number — at 3% of home loan assets, it plainly is not — but because of where those loans are. They are not evenly sprinkled across the country. They are concentrated in exactly the regional, coastal and northern markets where a large share of broker-written volume and a large share of customer-owned banking activity happens to sit.

The same body of research sets the wider context. The proportion of affordability-stressed households rose to 15%, or 1.61 million households, in the year to March 2024 — up from 12% in 2023 and 10% in 2022. Those households spend an average of 9.6 weeks of gross income on home insurance, roughly seven times what a non-stressed household spends.

Two caveats worth carrying. First, that data runs to March 2024; premiums have moved since, and in one direction. Second, in reporting the mutual banks’ work, Australian Broker cited a non-insurance rate of one in five in northern Australia against 11% nationally. We have not been able to trace that figure to a primary release, so treat it as that outlet’s reporting rather than an established number — though it is directionally consistent with the affordability research.

The fastest-moving line on the form

Here is why this belongs in a broker publication rather than a risk-management one.

Compare the Market data, reported by Mortgage Professional Australia in July 2026, put the rise in home and contents quotes across Australia’s five largest capital cities at an average of $373.93 — 14.78% — over the 12 months to June 2026, roughly double the market’s long-run growth rate. Earlier Canstar research had the average home and contents premium climbing 14% across 2025, from $2,452 to $2,795, with New South Wales up 18% year on year.

The drivers are not mysterious. The Insurance Council of Australia reports building material costs up 40% since 2022. Extreme weather produced $4.8 billion in insured losses in 2025 across 294,000 claims from declared events, with more than $4.1 billion of that originating in Queensland. Rebuild costs and catastrophe experience both feed the premium.

Now put that next to your process. On a typical file, home insurance is a small declared expense — a few hundred dollars a month at most, often folded into a general “insurance” bucket, frequently benchmarked away entirely. It rarely changes an outcome on its own. But it is, right now, moving faster than almost any other line on the form. A household paying around $2,800 a year and absorbing a 14% increase is finding roughly $400 a year it did not budget for. Do that twice and it is a meaningful share of the buffer you assumed at approval.

Where it shows up in your process

The transmission from premium to credit outcome is slower and less dramatic than a flood, and it runs entirely through the household budget:

1. The renewal notice arrives higher than last year.
2. The household absorbs it — or quietly trims cover to hold the premium flat. Higher excess, contents cover reduced, flood cover dropped.
3. Either disposable income falls, or the security becomes underinsured. Often both, in sequence.
4. Nothing appears anywhere in a lender’s data until either arrears move or an event happens.

For brokers, that means the signal reaches you before it reaches your lender — and it reaches you at three specific points:

At application. The declared expense figure. Whether it is a real number from a real document or a rounded recollection from a client who has not opened the renewal notice.

At product and lender selection. Where the security sits, what cover is in place, and whether the lender’s settlement requirements around evidence of insurance are likely to create friction on this particular property.

At refinance and repricing. The annual review conversation, where a client’s position has shifted for a reason that never appeared in a rate change. If you have last year’s premium on file, that conversation is short and you look sharp. If you do not, you are reconstructing it live.

The best interests duty question

It is tempting to jump straight to compliance framing here, and it is worth being careful about how far that goes.

ASIC’s Regulatory Guide 273, Mortgage brokers: Best interests duty, was issued on 24 June 2020 and guides Part 3-5A of the National Consumer Credit Protection Act 2009. It says a broker should consider “the consumer’s personal circumstances and financial situation, to the extent that they could affect the suitability of different products” (RG 273.48(c)), and also “reasonably foreseeable changes to the consumer’s personal circumstances and financial situation” (RG 273.48(d)). On cost, it notes that “the cost of a credit product can significantly affect the outcome the consumer achieves” (RG 273.51). And it is explicit that the two obligations are not the same thing: there are circumstances “where you might satisfy the responsible lending obligations but fall short of complying with the best interests duty” (RG 273.110).

What RG 273 does not do is tell you that a rising insurance premium is a reasonably foreseeable change you are obliged to model. That is a judgement about your process, and it belongs with your licensee and your aggregator’s compliance team — not with a news article, and not with an assumption you make on your own.

The practical point sits below the legal one, and it is much simpler. A file that records the actual premium, the source it came from, and what it was the year before is a better file than one that records a benchmarked estimate. That is true whether or not anyone ever asks. It takes about ninety seconds.

Where your panel could diverge

Worth being clear that what follows is analysis, not reported lender policy.

Customer-owned banks are, by construction, less geographically diversified than the majors. A regional mutual’s book is concentrated in the communities it was built to serve — which is precisely the structural feature that makes shared scenario work attractive and individual scenario work expensive. That concentration is the most likely reason nine of them pooled the exercise rather than each running their own.

If any of this eventually surfaces in credit settings, the plausible early form is not blanket postcode exclusions. It is quieter than that: more specific commentary in valuations on catastrophe-exposed properties, firmer evidence-of-insurance requirements at settlement, or a narrower appetite at the margins in particular locations. Those are the changes that arrive as a note in a lender bulletin, not a headline.

The broker response to that is not to pre-empt it. It is to notice which of your lenders serve concentrated regional books, and to make sure the insurance position on a catastrophe-exposed security is something you have asked about before you submit rather than something that surfaces at settlement.

The five-minute insurance check

This is the entire practical payload of the article. It is short on purpose.

Nothing in that list requires a new system, a new subscription, or a conversation with your aggregator. It requires one extra question and one extra line in a note.

A worked example

Consider a standalone house in regional Queensland, joint applicants, purchasing. The client’s renewal notice shows $2,900 for the year, up from $2,450 — an 18.4% increase.

Take that as arithmetic rather than prophecy: if the same rate of increase repeated for three more years, the premium would be roughly $4,800 — about $1,900 a year more than today, or around $160 a month out of the same household’s disposable income. That is not a forecast. Premiums may well moderate; the point is only that the arithmetic of a compounding expense is unforgiving, and that a benchmarked estimate would never have shown you the starting delta at all.

What changes in your file because of that? Possibly nothing about the recommendation. But you now know the client’s buffer is being eaten by something other than rates, you have a documented reason to talk to them again in twelve months, and you can raise the underinsurance question — a genuinely useful conversation for a client on a catastrophe-exposed property — before it becomes urgent. Whether they act on it is their decision and their insurer’s; whether they were told is yours.

What to watch next

• Whether individual customer-owned banks publish their own outcomes from the shared scenarios, rather than the sector-level framing released so far.
• The next Actuaries Institute affordability update, which will refresh the $57 billion estimate against post-2024 premium movements.
• Any tightening in evidence-of-insurance requirements at settlement, which would be the first place a change of appetite becomes visible to brokers.
• Whether the majors follow the mutuals into published scenario work, or continue to treat it as an internal exercise.

The takeaway

Most of what is written about climate and lending asks brokers to think in decades. This does not. The mutual banks’ own conclusion pushes the question back to something immediate and mundane: can the household still pay, and is one of its fastest-growing costs being captured accurately anywhere in the process?

On most files, the honest answer is no — not because anyone is careless, but because insurance has always been a small, boring number that behaved itself. It has stopped behaving itself. The fix costs ninety seconds per file, and it improves the quality of your advice, the quality of your notes, and the quality of your next annual review conversation at the same time.

Ask for the renewal notice. Ask what it was last year. Write both down.

Sources: Shared climate scenarios for the mutual banking sector (Customer Owned Banking Association, Finity and Climate KIC Australia, released 11 August 2026); Home Insurance Affordability and Home Loans at Risk (Actuaries Institute, lead author Sharanjit Paddam); ASIC Regulatory Guide 273, Mortgage brokers: Best interests duty, issued 24 June 2020; Compare the Market data as reported by Mortgage Professional Australia, July 2026; Canstar and Insurance Council of Australia data as reported in the same article; Australian Broker, 19 August 2026.