NextFin News - Only 35% of the 11,261 companies that disclosed environmental data in 2025 classified extreme weather as a material financial risk, yet the same group reported nearly US$3 billion in real losses from weather events last year alone. That gap between recognition and reality is the story: climate volatility has moved from a long-term scenario-planning exercise into the quarterly earnings line, and most corporate models are not built for it.
The Situation: Weather Is Already on the P&L
The numbers from CDP's May 2026 analysis are blunt. Of the companies that fully disclosed, extreme weather caused nearly US$3 billion in losses in 2025, driven mainly by higher direct costs of US$309 million and operational shutdowns of US$266 million. Heavy rain was the single largest driver, accounting for US$1.5 billion across disclosing firms. The forward-looking figures are an order of magnitude larger: companies project US$714 billion in future financial impacts, led by flooding at US$389 billion, cyclones at US$126 billion, and heavy rain at US$82 billion. Nearly half — 48% — of those risks are expected to materialize within the next two years, squarely inside current budget and investment cycles.
The losses are not confined to one sector or one type of asset. CDP expects them to spread through the systems businesses depend on: infrastructure, supply chains, insurance markets, and public services. Reduced production capacity alone accounts for a projected US$326 billion, with asset impairment or early retirement adding another US$122 billion.
Meanwhile, the insurance backstop is thinning. Swiss Re Institute estimates show global insured natural-catastrophe losses reached US$107 billion in 2025, the sixth consecutive year above US$100 billion. The first half of 2026 came in at US$42 billion, below the US$66 billion trend estimate, but the protection gap — the share of economic losses left uninsured — widened to US$424 billion in 2025 from US$395 billion a year earlier. When a disaster strikes, a growing share of the bill lands on corporate balance sheets rather than insurers'.
That is the tension the rest of this piece resolves: why are earnings models still treating weather as an outlier when the data says it is the baseline?
The Recognition Gap: Why 35% Matters More Than $3 Billion
The US$3 billion in realized losses is striking, but the more important number is the 35%. Only about a third of disclosing companies classify extreme weather as a material financial risk. That is not just a disclosure shortfall; it is a modeling failure. If a risk is not flagged as material, it does not get its own line in the forecast, it does not get capital allocated to mitigate it, and it does not get priced into the stock. The market can only reward companies for managing a risk that their own filings admit exists.
The cost of fixing that blind spot is surprisingly small. CDP's 2025 Disclosure Dividend report found the median cost of climate-related risks per company stood at US$39.4 million, compared with just US$3.1 million to mitigate them — nearly 13 times cheaper to act than to absorb the hit. Companies are, in effect, declining a 13-to-1 arbitrage by treating weather as a footnote.
The gap is even wider at the subnational level. Among 1,005 cities, states, and regions across 80 countries, 62% report already being significantly impacted by extreme weather, and close to a quarter specifically flag financial and insurance activities as highly exposed. Cities and companies are living in different realities: local governments are already adapting, while many corporate risk committees are still debating whether the hazard belongs in the annual report.
Extreme weather is already a financial risk. It has a dangerous domino effect, disrupting operations, reducing production and driving losses today, with far greater impacts lying ahead.
That assessment came from Amir Sokolowski, global director of climate at CDP, in the organization's May 2026 release. The domino metaphor is apt because it captures the transmission mechanism — and most models only price the first tile.
The Transmission Mechanism: How a Flood Becomes an EPS Miss
Weather does not hit earnings through a single channel. It propagates through at least four, and most consensus models capture only the first.
The direct channel is the obvious one: a factory floods, a mine shuts, a crop fails. CDP's data on operational shutdowns — US$266 million in 2025 — measures exactly this. But the second channel, supply-chain propagation, is where the damage multiplies. Marsh's 2026 supply-chain trends report puts the annual cost of global supply-chain disruptions at US$184 billion, with 65% of companies facing at least one bottleneck. Climate change is amplifying that: billion-dollar weather disasters now occur every three weeks, four times the frequency of the 1980s, when the interval was twelve weeks. A disruption no longer needs to hit your own asset to hit your earnings.
The third channel is the input-cost squeeze. When the Rhine and Danube run low, barges cannot carry full loads and freight rates spike. In August 2026, cargo ships on the Rhine were sailing at 30% capacity to avoid running aground, and European authorities resorted to blasting riverbeds to keep waterways open. Romania shut its sole working nuclear reactor, cooled by the Danube, for the first time, deploying naval forces to carry out underwater detonations to improve flow to the plant's cooling systems. Hungary's Paks plant, which supplies around 40% of the country's electricity, faced the same threat. Low water is not just a shipping problem; it is an energy problem, and energy is an input cost for every manufacturer on the continent. One German economist estimated the drought's hit to third-quarter value added at 1 billion to 2 billion euros.
The fourth channel is insurance repricing. As catastrophe losses mount, insurers withdraw from exposed markets, raise premiums, or narrow coverage. State Farm stopped selling new home policies in California in 2023, citing catastrophe exposure and inflationary pressures, and in 2025 moved to non-renew roughly 72,000 California policies — about 2% of its policy count in the state. The outright non-renewals were later reversed, but with fire coverage shifted to the California FAIR Plan, the state's insurer of last resort, which is more expensive and excludes liability. That is repricing in action: the risk did not disappear, it was pushed onto policyholders and, for commercial owners, onto corporate balance sheets.
The Second-Order Effect: Earnings Volatility Nobody Is Modeling
Here is the implication most consensus models miss. The problem is not only that weather lowers earnings in a bad quarter. It is that weather raises the variance of earnings, and variance is what discount rates punish.
Research on temperature-earnings relations quantifies the effect: a doubling in the frequency of 5% extreme temperature events implies a 37.4 basis point average change in earnings, with firms in negatively affected industries seeing a 28 basis point decrease. Against a quarterly EPS figure that may look small — until you recognize it as a second-moment effect. It widens the distribution of outcomes rather than merely shifting the mean, and a wider distribution is what drives multiple compression long before the average earnings line bends.
The utilities sector illustrates both sides of this. On the upside, heat waves drive demand: PJM's peak load hit 162.6 GW on July 2, 2026, about 2 GW above the prior record, and real-time prices averaged nearly US$350 per megawatt-hour during the early-July heat wave versus a year-to-date average of about US$64. The Dominion Load Zone saw US$467 per MWh against US$97 year-to-date. But that windfall is not pure profit. It arrives with grid stress, higher maintenance costs, and the risk that the next heat wave trips reliability constraints rather than revenue. NERC's 2026 Summer Reliability Assessment flagged three subregions at elevated risk under above-normal or extreme conditions. The same weather event that lifts one utility's quarterly revenue can force a capital-plan revision the next quarter.
This is why the cyclical-versus-structural question decides the conclusion. If weather volatility were cyclical, mean reversion would eventually restore the old earnings baseline and the market could look through it. It is not. The frequency of billion-dollar disasters has quadrupled since the 1980s; the protection gap is widening, not narrowing; and 48% of projected corporate impacts are expected within two years. This is a structural regime shift in the operating environment. A regime shift does not revert; it gets adapted to, or it gets priced.
The Counter-Thesis: Adaptation Is Real, So Is the Market Already Pricing It?
The strongest case against this reading is that corporate America is not asleep. Companies have been stockpiling inventory, diverting freight to rail and road, hardening assets, and buying business-interruption coverage. Felix Schmidt of Berenberg noted that German firms had likely not been caught by surprise by the Rhine's low water because climate change has made it a recurring challenge. If adaptation is real and widespread, the earnings impact should be contained, and the market — which is forward-looking — has already discounted the risk.
There is truth in that. Nearly US$3 billion in reported losses across 11,261 companies, against a global economy exceeding US$100 trillion, suggests the system is absorbing shocks without systemic failure. And the mitigation arbitrage — US$3.1 million to mitigate versus US$39.4 million in median risk cost — implies rational actors should already be acting.
But the adaptation argument has three weak points. First, adaptation is uneven: large firms with treasury functions can stockpile inventory and diversify suppliers; small and mid-cap firms cannot, and they sit in the same supply chains. Second, some hazards are not adaptable at the asset level. You cannot air-condition a river, and you cannot relocate a port. When the hazard is systemic infrastructure — waterways, grids, public services — firm-level adaptation hits a wall. Third, the insurance market, the clearest signal of priced risk, is still catching up: the protection gap grew by US$29 billion in a single year. If the world's most sophisticated risk-pricers are underpricing catastrophe exposure, equity analysts are almost certainly doing the same.
The falsifying signal is concrete. If, over the next two years, the share of companies identifying extreme weather as a material risk rises above 60% while realized weather-related losses stay flat as a percentage of revenue, the underpricing thesis is wrong — it would mean recognition has caught up and the residual risk is genuinely small. If instead recognition stays near today's 35% while losses climb toward the projected trajectory, earnings models remain exposed.
Outlook: Who Benefits, Who Is Exposed, and What to Watch
Cashed out, the mechanism points to three groups. The exposed are companies with concentrated physical assets in high-hazard zones, thin insurance coverage, and supply chains running through climate-stressed corridors — the Rhine, the Panama Canal, the U.S. Gulf and West Coast. Their earnings will show higher variance before they show lower averages, and that variance is what multiple compression attaches to. The beneficiaries are the enablers of adaptation: grid and dispatchable generation, water infrastructure, reinsurance and business-interruption capacity, and the engineering and construction firms that harden assets. Their revenue visibility improves precisely because hazard frequency is rising. And the arbitrage sits with the minority of companies that treat extreme weather as a system-level business risk today — they face a US$3.1 million mitigation bill against a US$39.4 million median risk cost, and the market has not yet rewarded them for it.
The forward look splits cleanly by horizon. In the short term — the next two quarters — earnings surprises will come from visible events already in motion: the European drought's hit to third-quarter value added, and the U.S. heat wave's effect on utility and data-center power pricing. In the medium term — the next two years, the window in which 48% of projected risks materialize — the signal to watch is the CDP disclosure rate: a move from 35% toward 60% would show recognition catching up. In the long term, the structural question is whether the protection gap narrows; if it keeps widening past US$424 billion, the residual risk stays on corporate balance sheets and the cost of capital for physical-asset-heavy businesses rises permanently.
Base case: weather-related earnings volatility becomes a standard analyst line item within two years, and companies that disclose and mitigate early trade at a premium. Upside case: a string of benign loss years and successful adaptation keep realized impacts near today's US$3 billion run rate, and the market never re-rates the risk. Downside case: a major catastrophe in a densely insured corridor pushes insured losses well past the 2025 level of US$107 billion, insurers retrench further, and the protection gap forces a broad repricing of physical-asset-heavy equities.
Climate volatility is not a risk that sits beside the earnings model. It is increasingly the variable that determines whether the model survives contact with reality.
Data as of August 31, 2026.
Explore more exclusive insights at nextfin.ai.
