When Physical Climate Risk Should Change a Decision

Key points

• Insurance and valuation already signal when physical climate risk should change a real asset decision. Most firms haven’t agreed what they’ll do when those signals move.
• Research on the Dutch market found climate risk tools shaped design and pricing far more often than whether a deal went ahead.
• Valuation standards and EU policy are pushing climate risk, and the value of resilience, into valuations and balance sheets. Undocumented adaptation is hard for insurers and valuers to credit.
• Five questions help a firm set the trigger in advance: Baseline, Signal, Action, Owner and Boundary.

Your insurer and your valuer are already telling you when physical risk should change a decision. Most firms haven’t agreed what they’ll do when they hear it.

When your premium last moved, could anyone in the room say whether it was about your building or your insurer’s book?

Most owners can’t. Renewal gets treated as procurement: a number to negotiate down and sign off. But insurance is often the earliest, clearest signal you’ll get about where an asset is heading. In The Legibility Problem Across Real Assets, Patrick Schmucki and I argue that an insurer pulling out of an area is saying something about long-term habitability that the valuer and the lender haven’t priced yet.

The information is there. What’s missing is a rule, agreed in advance, that says what it changes.

Plenty of data. Not many decisions.

Physical climate risk tools are everywhere now. What they change is much harder to pin down. A 2026 interview study of the Dutch market, Assessment Without Adjustment, found that assessments shaped design, conditions and pricing far more often than whether a deal went ahead. The developers interviewed had never turned down a development on climate risk alone. And the tools didn’t agree with each other: different numbers, different maps, same location.

So when the tools disagree and nothing in writing says which one wins, someone makes a judgement call. Buy another dataset and you’ve added another judgement call.

Two signals you’re probably already getting

Insurance. Adaptation can cut expected insured losses straight away. The trouble is the market often can’t see it. Annual renewal gives an insurer little reason to help fund a retrofit when you might switch next year. And if the catastrophe model can’t see your flood barrier, it can’t price it.

The bit you control is the paperwork. In a 2026 Aon survey with the G20 Global Infrastructure Facility, 40 per cent of the 25 insurers and reinsurers covered said they apply more conservative terms when a company can’t document its assessment and mitigation. Small sample, and it’s infrastructure rather than property, but the direction is hard to miss.

Don’t count on the discount either. The Actuarial Association of Europe worked through a residential example: a 10,000 euro measure on a 300,000 euro house, with a 25 per cent premium discount, takes 45 years to break even if you pay for it yourself. The bigger prize is staying insurable, keeping capacity at renewal and getting a better rebuild after a loss.

Valuation. The current RICS standard on ESG and sustainability in commercial property valuation names asset-level physical risk analysis, and the mitigation already in place, as evidence that may be relevant to value. Valuers are expected to document what they relied on, what they assumed and where the limits are. It doesn’t force a repricing. It does mean the reasoning behind a value is becoming visible.

Policy is heading the same way. The European Commission’s Framework for Climate Resilience doesn’t mince words: “Climate risks should be adequately reflected in asset valuations and financial accounts.” The Commission plans to work with European and international bodies to build material physical climate risks into valuation and accounting, so balance sheets reflect both the risk and the value of resilience measures.

That last part matters. If resilience is going to show up in value, someone has to be able to see it. Fund a measure without recording what it was meant to change, and what it did change, and you’ve given the valuer nothing to work with.

When repricing comes, it rarely comes gently. In flood-prone, wildfire and coastal markets it tends to follow a single event: an insurer pulling out, a rule change, one bad loss. By then you’re deciding under pressure.

Pilots deal with this on every landing. Before the approach they set a decision altitude, the height at which they go around if they can’t see the runway. They agree it on the ground, while there’s time to think.

Decide what each signal means before it arrives

We’ve boiled this down to five questions.

Baseline. What climate have you agreed to plan for?

Signal. What would tell you that’s wrong? Insurance terms, valuation evidence and lending conditions, not just hazard data.

Action. What would you do? Hold, invest, reprice, renegotiate or exit.

Owner. Who acts, and by when?

Boundary. What depends on someone else, like drainage, defences or the building next door?

Take insurance. Your rule might be: if the insurer reprices past an agreed level, cuts capacity or won’t renew, the asset goes to committee with four options costed. Fund the capex, negotiate, change carrier, or accept the terms. One trigger, one owner, one review window. You draw the line before the renewal notice lands, not after.

Lenders already do this. Loan covenants spell out in advance which ratios trigger action. We just haven’t done it for physical climate risk.

Start with what your own firm disagrees about

Here’s the cheapest finding most firms can get. Ask investment, asset management, valuation, risk and whoever looks after insurance the same factual questions about your baseline, separately. If the answers don’t match, that’s your finding. And it isn’t a consultant’s opinion. It’s your own people.

Patrick’s Edition 42 of The Adaptive Edge, Why Your Physical Risk Data Does Not Change a Decision, walks through how to build and test a rule.

We’ve built this into an offer called Ground Rules, a short question set returns a two-page diagnosis of where a firm stands with no confidential information required. Take it here.