Insurers see climate risk first. The housing market should be listening.

Insurability is becoming a leading indicator of property value. In England, 1 in 9 new homes built in 2022 to 2024 sit in medium or high flood risk areas, and homes built since 2009 are excluded from Flood Re. This article sets out why insurers see the risk first, why adaptation and resilience need a shared definition, and five ways the market could reward resilience.

“Let’s stop building in dumb places.”

That was the headline of an OpEd I came across in my Google alert feed last August. And clicking through to the article, I was surprised to read the byline. This wasn’t written by a politician or a member of the public, but instead by Sascha Cowlrick, the executive general manager, Business, at Suncorp New Zealand. Suncorp NZ is one of a handful of insurers covering property in that country.

It stayed with me because it is direct, and because it is right. It is also the simplest version of the argument I made in a keynote to the Association of British Insurers, From risk to resilience. We are still building climate risk into the housing market. Insurers are the first to see it. And the question is what does the rest of the market does with that signal.

Insurers see it first

Climate risk reaches property values in a sequence:

Insurers sit at the forefront because they reprice every year against fresh loss data. Valuers rely on transaction evidence, which lags and thins out just as risk becomes apparent. In the UK, we are somewhere between the second and third steps.

That matters because of what insurance does. It turns catastrophic loss into a predictable cost. When it retreats, the cost doesn’t disappear. It moves to households, lenders and the state.

We are still building the risk in

One in nine new homes built in England between 2022 and 2024 sits in an area of medium or high flood risk. Homes built since 2009 are excluded from Flood Re, a rule meant to deter exactly this. On current patterns, another 115,000 homes will go up in England’s flood zones over the next decade.

The costs are already showing. Domestic flood claims rose 38% in 2025, to £312m. And it isn’t only flood. The average subsidence claim hit a record £20,000 in the second quarter of 2026.

Globally, Verisk models average annual disaster losses at $450bn, and 62% of that is uninsured. 2025 was the sixth year running that insured losses topped $100bn. Insurers are protecting margins by narrowing appetite. The costs keep rising.

Other markets show where this leads. In Australia, 1 in 7 households are in home insurance affordability stress. On average, those households would need 9.6 weeks of gross income to buy adequate cover. In the US, premiums rose by $648 per household between 2021 and 2024, and Florida now averages $9,462 a year. When cover becomes unaffordable, the risk lands on household balance sheets.

Adaptation and resilience are not the same thing

We tend to use the two words interchangeably. They are not the same, and the difference matters for who pays.

Adaptation responds to a permanently shifted baseline. It sits with the asset: where and how we build. Its benefits can be attributed to one owner.

Resilience responds to shocks that still happen, more often and more severely. It sits with the system: drainage, defences, catchments. Its benefits are shared and diffuse, and hard for any one party to capture.

In a recent paper with Patrick Schmucki, The Legibility Problem Across Real Assets, we found that three major taxonomies draw the line between the two in different places. If we can’t agree what we’re buying, no one can price it. And if no one can price it, no one will reward it.

Resilience works when the price rewards it

Alabama shows what happens when the pieces line up. Homes with FORTIFIED roofs had 73% fewer claims after Hurricane Sally. The model has four parts: a building standard, a grant to meet it, a mandated premium discount, and published claims evidence. Take any one away and the others weaken. Similar approaches are now being used for wildfire in California and to begin to address bushfire, and cyclone, risk in Australia.

The UK has some of the pieces. One in three eligible flood claims through Flood Re included Build Back Better provisions for adaptations for the home to reduce future risks. Natural flood management returns around £10 for every £1 invested over 30 years, once carbon, health and habitat benefits are counted). Flood Performance Certificates for new builds are due to be piloted by the end of 2026.

What’s missing is the link between them. A standard without a price signal is a recommendation. A discount without evidence is a cost.

Everyone has a role. No one has the obligation.

Government sets where and how we build but insurability is not tested as part of this process. Insurers price the risk and hold the data but have no duty to reward resilience. Households can be left carrying the loss, with no duty to adapt, or indeed may have limited means to do so. Developers on the other hand have a role in creating the climate exposure with no liability for where they build.

That gives us five places to break the loop:

None of these works alone. Each makes the others more likely to stick.

Flood Re ends in 2039. Who pays then?

That is a political question, not an actuarial one. It also reaches beyond flood, to subsidence, heat and whatever comes next. Insurers should be calling for a public answer.

In the meantime, the first step costs nothing. Let’s stop building in dumb places.