The standard insurance industry narrative about climate change focuses on exposure: rising frequency and severity of weather events, coastal flood zones requiring new pricing, heat stress impacts on agricultural yields. This is accurate as far as it goes. But there is a second-order problem that receives less attention: the historical data that underwrites the existing model is becoming progressively less useful for pricing future risk.
This is a model recalibration problem, and it is structurally different from an exposure problem. An exposure problem — more claims per year than before — can in principle be addressed by adjusting loss ratios and repricing accordingly. A model recalibration problem is harder: the actuarial models that price weather risk are built on statistical distributions derived from historical observation. If the underlying process generating weather events is non-stationary — if the distribution itself is shifting — then the historical parameters you are fitting to are no longer reliable guides to future frequency and severity.
Parametric contract structures handle this problem in a way that traditional indemnity insurance does not.
The Stationarity Assumption and Where It Breaks
Classical actuarial pricing assumes statistical stationarity in the underlying loss process: that the distribution of events in the future resembles the distribution of events in the historical sample, modulo known trends. For motor claims frequency, this assumption is reasonable — the distribution shifts slowly and predictably with macroeconomic conditions, road infrastructure, and vehicle technology. Catastrophe weather risk is a different story.
European flood event frequency has measurably shifted in the past 15 years. The Rhine basin and Meuse basin, the Po valley in Northern Italy, and the river basins of Central Germany have all experienced extreme precipitation events at frequencies that exceed the historical 100-year return period as calibrated on pre-2000 data. That does not mean the 100-year return period is wrong as a concept. It means the parameters of the distribution require updating, and those updates need to happen faster than the actuarial update cycle that traditional carriers typically follow.
Traditional indemnity insurance responds to this by adjusting rates at renewal — increasing premiums for flood-exposed properties, narrowing coverage in high-risk zones, or exiting markets where the loss potential is not priceable. All of these responses leave policyholders either underinsured or uninsured in the locations most exposed to climate change. This is the insurability problem that is receiving increasing regulatory attention from bodies including EIOPA and national supervisors across Europe.
How Parametric Structures Handle Non-Stationarity
Parametric contracts are structured around independently observable trigger events — meteorological measurements, satellite-derived indices, seismic station readings — rather than around historical loss distributions. The pricing problem shifts from "estimate the distribution of future losses" to "estimate the distribution of future trigger events." This is not a simpler problem. But it is a different problem, and in some respects a more tractable one.
Climate scientists produce regularly updated projections of future precipitation extremes, temperature distributions, and sea level trends. These projections — while uncertain in their precise magnitudes — can be integrated into parametric trigger pricing more directly than into traditional loss modelling, because the trigger event (rainfall exceeding X millimeters at monitoring station Y) is closer to the directly modelled physical process than the loss event (policyholder property flood damage).
A parametric drought product for European agricultural risk, priced using current-generation climate projections rather than historical rainfall statistics, can incorporate the non-stationarity of the underlying process at each annual renewal. The trigger parameters can be adjusted based on updated climate projections, without requiring a full actuarial re-rate of the underlying loss model. The product adapts with the science.
The Role of Earth Observation Data
The technical development that is enabling the current generation of climate-parametric products is the accessibility of high-resolution Earth observation data. Satellite radar, multispectral imaging, and ocean-atmosphere sensor networks now provide near-real-time measurements of the physical variables that drive weather-related insurance losses: soil moisture, snowpack depth, sea surface temperature, vegetation index, flood inundation extent.
This data infrastructure reduces the basis risk in parametric products — the gap between the measured index and the policyholder's actual loss experience — by allowing trigger definitions based on spatially granular physical measurements rather than point observations at distant weather stations. A flood insurance product triggered by satellite-derived inundation extent at the policyholder's specific location has materially lower basis risk than one triggered by river gauge level at the nearest monitoring point 30 kilometres away.
We invested in Descartes Underwriting specifically because the team understood that the technical challenge in climate-parametric insurance is the index design layer — building robust, actuarially defensible trigger mechanisms from the available Earth observation data. The modelling and the product structure are secondary to getting the index construction right. Teams that start with a product concept and then try to find a trigger that fits it build inferior products to teams that start with the physical data and design the trigger from the signal that is actually there.
What Parametric Climate Insurance Cannot Do
We are not arguing that parametric structures solve the insurability problem across all climate-affected risks. There are classes of risk where the basis risk inherent in any parametric design is too wide to provide meaningful coverage: complex urban flooding that combines pluvial and fluvial sources in ways that do not correlate cleanly with any single measurable index; wildfire behaviour in complex terrain that is sensitive to local wind patterns that are not captured by regional meteorological stations; crop damage from combinations of heat stress and moisture deficit that interact in ways that individual indices do not capture well.
For these complex cases, the answer may be hybrid structures: parametric triggers that provide immediate liquidity for the first tranche of loss, with an indemnity top-up layer that handles the quantum settlement through traditional adjustment. This two-layer approach solves the liquidity problem that drives the most acute policyholder distress in catastrophe events — the immediate inability to fund business continuity — while retaining the accuracy of indemnity adjustment for the total loss quantum.
The product innovation space here is real and largely unexplored at seed stage. The teams that are building the infrastructure — data pipelines, index pricing APIs, carrier fronting relationships — that would enable these hybrid structures to be deployed at scale are building something with a decade-long tailwind from climate risk repricing dynamics.
A Note on Regulatory Timing
European insurance regulation is beginning to engage with parametric products more seriously. EIOPA's 2022 consultation on natural catastrophe insurance gaps specifically acknowledged parametric structures as a tool for addressing coverage gaps in high-risk zones. National supervisors in Germany, France, and Spain have been in dialogue with carriers about parametric product authorisation in the wake of recent extreme weather events.
The regulatory window is opening. The window between first-mover regulatory approvals and market standardization — typically a 3–5 year period in European insurance markets — is when the companies with established products, working distribution relationships, and proven index methodologies compound their advantage. The companies building in this space now are well positioned for that window, if they survive the extended carrier procurement and regulatory approval timeline. That is the timing tension that seed capital needs to bridge.