Climate risk checklist for insurers: what's missing from the report

The compliance team has already read Circular 666 and Public Consultation 05/2026. It knows what IFRS S2 requires in governance, strategy, metrics and scenarios. What's missing is not understanding of the standard: it's physical data per asset. Hazard, coordinate, horizon, scenario and the evidence that supports each line of the report. Without that, the paragraph gets filled in and the conclusion stays empty.

This text does not explain the standard again. It shows what closes the gap: a 21-item checklist and six criteria for evaluating who supplies that data.

What Susep requires today and what Public Consultation 05/2026 adds

Susep Circular 666/2022 has been in force since August 1, 2022. It requires management of sustainability risks (environmental, social and climate), a formal policy and an annual report disclosed by April 30, with the GVR, EST and GER tables mandatory (MEM and OPO optional).

In September 2026, Susep opened Public Consultation 05/2026, which proposes revoking 666 and replacing it with a regulation aligned with IFRS S1/S2 and with CBPS 01 and 02 Pronouncements. The proposal expands governance and strategy, with climate scenario analysis and resilience. It is still a consultation, not a standard in force. The direction is set: Susep will demand conclusion, not spreadsheet.

The six fields every report requires and most don't have filled in

There are six physical data fields per asset that support reporting today, under 666, and that remain valid under CP 05/2026. Without them, each block of the report depends on aggregated estimate, and aggregated estimate does not survive the auditor's first question:

• Geographic identification of the asset: coordinate, not municipality. Address, not administrative region.
• Hazard per asset: which hydrometeorological hazard dominates at that point (hail, convective rain, wind, flood).
• Time horizon: exposure today and projected for 2030 and 2050.
• Climate scenario: which emissions trajectory supports the projection, in the format that IFRS S2 requires.
• Historical series: how many years of data support the stated probability, with no regional average in its place.
• Auditable evidence: source, methodology and model version documented for audit.

How to evaluate a climate data provider per asset

Evaluating a climate data provider per asset means asking six questions before signing a contract. All have objective answers, and the answer separates a nice-looking report from an auditable basis. Ask in writing, with a deadline to respond.

Spatial resolution: ask for the grid in kilometers and compare it with the size of the asset. For mapping historical susceptibility and current exposure, a grid of 1 to 3 km (against the ~25 km of global public models) changes the conclusion about hail or convective rain at a specific point.

Granularity per asset: demand data at the address level. Municipality hides risk concentration.

Hazard coverage: ask which of the main modeled hydrometeorological hazards relevant to the portfolio actually have reanalysis series calibrated with auditable surface data.

History: how many years of reanalysis support the data and whether the series is proprietary or inherited from a repackaged public model.

Scenarios: confirm whether there is a projection per horizon (2030, 2050) in the IFRS S2 format, using robust downscaling techniques of climate models (GCMs) to assess regionalized risk worsening trends, and not a generic climate change report or unrealistic promises of micro-local precision decades into the future.

Traceability: documented methodology, model version and update date per asset, in a format the auditor accepts without reservation.

These six questions separate two different things: consulting that delivers one report per year and data infrastructure that updates with every event. The first solves this year's reporting. The second solves the reporting and the next claim.

Where market checklists stop

The disclosure checklists available cover the paragraphs of the standard: governance, strategy, metrics, report structure. They do not cover the data that supports each paragraph. An IFRS S2 compliance checklist without physical data per asset becomes well-written prose with a borrowed number.

The bill shows up when the asset is specific: a warehouse in a hail zone, a substation on a slope. The municipality average does not answer. Measuring hazard at the asset's address is the missing layer.

This is where i4sea operates: resolution of 1 to 3 km, coverage with calibrated reanalysis for the main hydrometeorological hazards, more than ten years of proprietary reanalysis and more than 100 assets monitored in Latin America and Europe.

The cost of reporting with borrowed data

Reporting with aggregated data costs at underwriting, not at the report. When the portfolio is priced by the municipality average, the asset concentrated in a risk zone enters cheap and the claim arrives bigger than the premium paid.

In the first half of 2026, losses from climate events in Brazil totaled R$ 51.9 billion, an increase of approximately 11% over the R$ 46.7 billion recorded throughout all of 2025, according to CNseg's Radar. Drought accounts for 93% of these losses, and only 0.4% of them have insurance coverage. CNseg already launched a climate intelligence hub at COP30, in 2025, to organize this data at sector scale.

Whoever measures per asset decides before the event. Calculating the auditable number costs less than reacting to it, and parametric climate risk insurance depends on this same data so the trigger does not fire wrong.

Download the climate risk checklist for insurers

The form opens on this page: name, company and email, and the PDF arrives right away. The checklist is in Portuguese, which is the language of the document you download.

Frequently asked questions

References