Physical climate risk: the data behind capital decisions

In the same week, two Brazilian cities made the same bet: climate data sells. Not as a sustainability narrative. As an investment argument.

It is worth understanding what each one did, because the pattern they are setting will knock on the door of private asset operators sooner than it seems.

Porto Alegre: resilience data as a showcase for investors

In the Innovation District, the city's old Fourth District, the city government launched a platform that gathers real-time data on flood protection, urban projects, and business opportunities. The showcase numbers:

1. 258 hectares and about 18,000 local businesses in the area, with 636 points of interest mapped on the platform.

2. Flood protection and drainage works: of the 60 planned after the 2024 floods, 48 are complete, 8 underway, and 4 in design.

3. More than R$ 1 billion in projected investments to transform the area, across sustainable infrastructure, flood protection, and ecological corridors.

The IT coordinator of the city's environment department (Smamus), Matheus Marchiori, summed up the intent in one sentence to Exame magazine: "We want to bring more confidence to investors and show that we are prepared for the new climate reality."

Note the verb. It is not "inform". It is "bring confidence". Resilience data has become a piece of investor communication.

Balneário Camboriú: dense monitoring as positioning

In the same week, Balneário Camboriú joined an artificial intelligence climate monitoring network across the 11 municipalities of the AMFRI region, with 38 monitoring points and data updated every 4 hours, replacing the previous system's daily updates.

A tourism city, a hot real estate market, high exposure to coastal events. The choice to invest in dense monitoring is not chance: it is positioning. A city that monitors climate at fine granularity signals to the market that it knows how to manage the risk it sells. And the city's annual contract in the network, R$ 83,900, shows something else: the cost of the data is orders of magnitude smaller than the risk it helps manage.

The signal behind the two cases

Cities do not do this out of technical vanity. They do it because they noticed something financial markets already knew: climate resilience has stopped being an environmental department agenda and has become a capital allocation criterion.

Investors look at drainage infrastructure before committing. Lenders look at extreme event exposure before releasing credit. Insurers look at projected loss ratios before pricing a policy. A city with that data ready, structured, and public gets ahead in the race for capital.

Now replace "city" with "asset". Power plant, industrial site, substation, railway, insured asset portfolio. The logic is identical. Whoever finances, underwrites, or grants a concession on that asset is also asking: what is the physical climate exposure, asset by asset, over the coming years?

The asymmetry: investors ask about the asset, operators answer with the average

Here is the problem. The investor asks for data on the specific asset. The operator, in most cases, still answers with the regional average, or worse, with "force majeure" after the event has already happened.

While that asymmetry persists, physical climate risk does not disappear. It just moves: into the insurer's loss ratio, into the company's accounting provisions, into the cost of capital of the financing. Someone always pays the bill for a risk that was not measured before it happened.

The numbers show the size of that bill.

Globally, insured losses from natural catastrophes reached US$ 137 billion in 2024, 27% above the ten-year average of US$ 108 billion. The protection gap, the difference between total economic loss and insured loss, reached US$ 181 billion: of the US$ 318 billion in economic losses from disasters worldwide, 57% had no coverage at all.

In Brazil, according to the study by the insurance confederation CNseg with EY presented at COP30, climate-related losses reached R$ 184 billion between 2022 and 2024. Of that total, 91% of losses had no insurance coverage.

Ninety-one percent. That is not a one-off market failure. It is the absence of a data standard that allows measuring, pricing, and provisioning physical climate risk by asset, before the event happens.

For insurers, the asymmetry has a technical name: pricing by regional average errs on both sides. It overcharges the well-protected asset, which walks away, and undercharges the exposed asset, which stays in the portfolio.

Physical climate risk is not a weather forecast. It is an index per asset

The data that supports a capital decision is not the city's rain forecast. It is the translation of a climate hazard into an index calculated for the specific asset, crossing climate with the territory's characteristics: vegetation cover, soil, slope, what sits in the hazard's path.

The difference shows when the index meets the real event. Three examples from our own work, in three different sectors:

1. At a power utility, the tree-fall index over the grid crossed wind with land use and mapped risk every 3 km, across 73 municipalities of the concession.

2. On a pipeline network, the landslide risk index was validated against the real geotechnical events of 8 years of history: 3 of the 4 events happened on stretches the index already classified as high or very high risk, without those events being used to train the model.

3. On a critical highway corridor, the forecast anticipated the road's actual flood closure by 47 hours.

That kind of data, per asset and validated against real events, is what a credit committee, an underwriting desk, or a board can use. Regional averages do not fit a pricing model. An index per asset does.

The data standard that closes the gap

That is exactly the gap i4climate, by i4sea, was designed to close: quantifying climate exposure asset by asset, with IPCC scenarios (SSP2-4.5 and SSP3-7.0) projected to 2075, in a format native to the frameworks the market uses as reference, such as IFRS S1/S2 (ISSB) and TCFD.

On regulation, the region is moving at different speeds, and the right way to read it is as access-to-capital context, not as a deadline threat:

In Chile, reporting has been mandatory since 2021 (CMF, NCG 461, with NCG 519 bridging to IFRS S1/S2). Even so, the May 2026 measurement on official CMF data shows alignment of only 24.1% to IFRS S1 and 25.5% to IFRS S2. The obligation arrived before the data: whoever holds asset-level exposure stands out even inside the rule.

In Mexico, IFRS S1/S2 reporting is mandatory for all BMV issuers since 2026 (CNBV). The question is no longer whether to report, but with which data: regional average or asset-level exposure?

In Peru, the standard was approved in March 2026, with mandatory adoption from 2029. It is the clearest window in the region: three years to build the data series the market will demand, before everyone looks for it at the same time.

In all three countries the reading is the same: the data is not born from the regulator. It is born from price, planning, and provisioning. The regulatory report is one of the outputs, not the reason to exist.

This is possible because i4sea works at a resolution of 1 to 3 km, while public forecasts operate at roughly 25 km, and maintains more than 10 years of proprietary climate reanalysis, currently monitoring more than 100 critical assets across Latin America and Europe, with 18 hydrometeorological hazards covered.

The sentence that sums up the positioning is worth repeating: price, plan, and provision with your asset's risk, not with the regional average.

Next step

If you do not know, today, what your asset's physical climate exposure looks like through 2050, that is the question the next investor, lender, or insurer will ask.

i4sea offers a free climate risk diagnosis of 1 asset or region, in 7 days, with the same data standard used by operators that already manage climate as a business variable, not as a surprise.

Sources

1. Porto Alegre Innovation District platform: Porto Alegre City Hall, Aug 4, 2026 and POA 24 Horas, Aug 4, 2026

2. AI climate monitoring network in the AMFRI region (38 points, 11 municipalities, data every 4 hours, R$ 83,900/year contract): Notícia Já, Aug 21, 2026

3. Insured losses of US$ 137 billion in 2024, US$ 108 billion average, US$ 181 billion protection gap (57% of US$ 318 billion uncovered): i4sea, insurance sector page

4. R$ 184 billion in climate losses in Brazil (2022-2024) and 91% uncovered: CNseg/EY Radar, presented at COP30, 2025

5. Regulation by country: Chile, CMF NCG 461/519 and MINCORE measurement on CMF data (May 2026, 24.1%/25.5% IFRS S1/S2) · Mexico, CNBV/BMV, mandatory since 2026 · Peru, approved March 2026, mandatory from 2029