Short-Form Memo
Climate-Risk Intelligence
From Hazard Models to Decision Infrastructure
Yanni Bills
Sector Investment Memo
May 2026
Abstract
Physical climate risk is moving from a disclosure topic to a priced operating constraint. Insurers, lenders, asset owners, infrastructure operators, and exposed corporates increasingly need to know how floods, fires, heat, drought, storms, and compound hazards affect underwriting, credit, asset value, downtime, resilience CapEx, and risk transfer. The venture opportunity is not in another climate dashboard. It is in companies that translate uncertain physical hazards into recurring, budgeted decisions. The winners will combine credible hazard science, asset-level exposure data, vulnerability and loss translation, uncertainty propagation, and workflow integration. The losers will sell polished exposure scores that look authoritative but do not change what a customer does.
1 Investment View
| Question | View |
|---|---|
| Core thesis | Back companies that convert probabilistic physical-risk intelligence into a recurring decision: underwriting, pricing, lending, capital allocation, resilience planning, operational continuity, or risk transfer. |
| Best wedges | Insurance and reinsurance analytics; asset-level risk intelligence for real estate and infrastructure; credit-risk tools for banks and lenders; operational adaptation platforms for exposed industries. |
| What to avoid | Static hazard maps, generic climate scores, compliance-only dashboards, services-heavy reporting businesses, and weather products that do not translate forecasts into economic action. |
| Why now | Losses are rising, corporate pain is operational rather than theoretical, formal risk-management capacity lags exposure, and financial institutions are beginning to embed physical risk into priced workflows. |
| Defensibility test | The company must own a hard-to-recreate layer: proprietary data, validated vulnerability/loss models, regulatory or institutional trust, embedded distribution, or a workflow that becomes painful to remove. |
| Investment standard | The product should change the price of risk, the allocation of capital, or the design of adaptation. Anything less is likely a feature, not a venture-scale company. |
2 Investment Thesis
Climate-risk intelligence refers to model-, data-, or AI-driven software that quantifies, forecasts, prices, and communicates exposure to physical climate hazards. The category includes products that estimate asset-level risk, operational disruption, or financial loss; translate probabilistic hazard information into decision-support outputs; and inform adaptation, risk-transfer, capital-planning, underwriting, lending, or operational-resilience measures.
The investable claim is simple:
Physical climate risk is becoming financially unavoidable, but it remains difficult to price, underwrite, and manage. Venture-scale companies can emerge where probabilistic hazard information is converted into a recurring, budgeted decision.
This memo excludes physical adaptation hardware, generic weather forecasting, pure emissions-accounting software, carbon-credit platforms, consulting reports without scalable software, and post-event disaster-response tools. The focus is external revenue generation through physical-risk analytics: the software and data layer that helps customers decide what to do before a loss occurs.
A strong company owns the relevant translation chain, going beyond answering whether a flood, fire, or storm could occur:
physical hazard → local exposure → asset vulnerability → expected loss → decision.
Many can produce a climate-risk output. Few, however, can answer the diligence question of whether the output is trusted enough, specific enough, and close enough to a budget owner that it changes a live decision.
3 Why Now
The timing is driven by four converging forces.
- Physical losses are rising. SVB reports that inflation-adjusted costs from billion-dollar U.S. weather and climate disasters have increased across decades, with the last ten years alone producing roughly $1.5 trillion in direct damages to property and crops [1]. Swiss Re reports that 2025 marked the sixth consecutive year in which insured natural catastrophe losses exceeded $100 billion [24].
- Corporate recognition has shifted from awareness to operating pain. MSCI's 2025 corporate resilience survey found that more than 80% of surveyed high-exposure companies said extreme weather had inhibited operations or increased costs since 2020; 99% assessed risks from extreme weather, and 85% estimated potential losses [17].
- Exposure is outpacing formal risk management. MSCI and Swiss Re's analysis of asset-owner portfolios found that 89% of assets faced multiple overlapping hazards, 55% of companies were severely exposed to physical risk, and only 30% of exposed firms formally disclosed integration of physical-risk management [19].
- Insurers and financial institutions are converting physical risk into priced workflows. MSCI's 2026 insurer survey found that 91% of surveyed property and casualty insurers saw opportunities in physical-risk management advisory services, while integration of physical risk into overall risk management remained uneven across regions [18].
Technical progress is also pulling the category forward. AI-native and hybrid weather models such as GraphCast, Pangu-Weather, GenCast, Aurora, and NeuralGCM have lowered the cost of short- and medium-range forecasting and expanded the feasible model class [15, 3, 21, 4, 14]. Lower inference cost can increase forecast frequency, scenario coverage, and localization. But AI weather prediction alone is not the product as enterprise buyers rarely want a standalone forecast. What they really want is to know how a changing hazard distribution affects things like assets, losses, or downtime.
That is why the category is shifting from "climate analytics" toward risk infrastructure.
4 Buyer Map
Climate-risk intelligence is not bought by a generic enterprise customer, as the strongest buyers are those for whom physical risk directly affects pricing, asset value, regulatory scrutiny, or operational continuity.
| Buyer | Budgeted pain | High-value use case |
|---|---|---|
| Insurers / reinsurers | Catastrophe losses, portfolio accumulation, model uncertainty, risk selection | Underwriting, pricing, reinsurance, parametric product design, customer resilience advisory |
| Asset managers / owners | Physical risk affects portfolio value, stewardship, fiduciary scrutiny, and disclosure | Portfolio screening, scenario analysis, capital allocation, divestment, resilience CapEx |
| Banks / lenders | Collateral impairment and borrower cash-flow exposure to local hazards | Loan-book stress testing, credit underwriting, mortgage / CRE risk, loss-given-default analysis |
| Real estate / infrastructure owners | Asset damage, insurance cost, maintenance planning, exit valuation | Acquisition diligence, underwriting assumptions, retrofit prioritization, asset-hardening plans |
| Utilities / energy / industrials | Asset downtime, grid disruption, worker safety, supply continuity | Resilience planning, outage mitigation, vegetation / asset prioritization, operational scheduling |
| Supply-chain-heavy corporates | Supplier, logistics, facility, and inventory disruption | Supplier screening, contingency planning, inventory positioning, operational resilience |
| Public sector / municipalities | Infrastructure exposure, community resilience, capital-project prioritization | Hazard mapping, adaptation planning, bond / grant support, emergency-preparedness planning |
The cleanest near-term wedge is insurance and reinsurance because physical climate risk already sits inside underwriting, pricing, and portfolio risk decisions. Asset managers and real estate/infrastructure owners are attractive because climate risk increasingly affects asset values and CapEx decisions. Banks and lenders are promising where physical hazards can be mapped into collateral impairment, borrower cash-flow stress, probability of default, or loss given default.
The common feature is budget proximity. Customers pay when the output changes a recurring decision with financial consequences.
5 Market Opportunity
The market is difficult to size because climate-risk intelligence is nested inside adaptation, resilience, catastrophe modeling, geospatial analytics, weather intelligence, enterprise risk, and financial-data software. Top-down estimates should therefore be treated as directional evidence, not as a clean TAM.
The broader adaptation and resilience economy is large. GIC and Bain estimate that selected adaptation-solution revenues could grow from roughly $1 trillion today to $4 trillion by 2050 [11]. McKinsey estimates that climate-resilience technologies could represent $600 billion to $1 trillion in addressable markets by 2030 [16]. BCG and Temasek similarly frame climate adaptation and resilience as a trillion-dollar private-market opportunity, with climate-intelligence solutions among the fastest-growing categories [31].
The narrower pure-play market remains smaller but is growing quickly. Fortune Business Insights estimates the climate-risk analytics market at roughly $1.8 billion in 2025, growing to $7.7 billion by 2034, while Technavio estimates climate-risk analytics platforms at roughly $905 million in 2025 with a projected 17.8% CAGR through 2030 [10, 25]. These estimates are imperfect because category definitions vary. The venture-relevant conclusion is narrower: the sector is early enough to be unsettled, but large enough to support meaningful software companies if products attach to recurring, high-value workflows.
A credible investment view should therefore be wedge-first rather than TAM-first. The best companies will win because a specific buyer cannot underwrite, lend, allocate capital, insure, operate, or adapt without their analytical layer.
6 Technical Architecture
Most products share a common stack:
hazard modeling → exposure mapping → vulnerability modeling → loss translation.
| Layer | What it must prove | Common failure mode |
|---|---|---|
| Hazard modeling | What may occur, where, with what intensity, over what time horizon, and with what uncertainty. Inputs may include NWP, climate-model ensembles, reanalysis data, catastrophe models, hydrology, fire-weather indices, satellite observations, or AI/ML forecasts. | Optimizing for average forecast skill while missing the tail events that drive loss. |
| Exposure mapping | Which assets, facilities, insured exposures, infrastructure nodes, suppliers, or portfolio holdings sit in the path of the hazard. | Treating geospatial proximity as commercial exposure without enough asset context. |
| Vulnerability modeling | How a specific asset or operation responds to the same hazard intensity, given construction, maintenance, redundancy, adaptation, and local context. | Confusing exposure with damage; producing precise-looking scores from generic vulnerability assumptions. |
| Loss translation | How physical vulnerability becomes expected loss, downtime, insurance cost, probability of default, asset impairment, resilience CapEx, or avoided loss. | Ending at a dashboard rather than a decision. |
The core technical challenge is an uncertainty-preserving translation. A strong product carries uncertainty through the stack rather than compressing it prematurely into a decorative score. Short-term weather products can be validated against frequent realizations. Long-horizon climate-risk products must rely more heavily on hindcasts, ensemble comparison, physical consistency, stress tests, and sparse data on tail events. This is especially important because the rare event is often the event for which the product is purchased.
Model class is not the same as commercial value. Physics-based numerical models offer explicit physical structure; AI models offer statistical expressivity and computational efficiency; hybrid models attempt to combine both. The principal diligence question is whether the architecture fits the hazard, time horizon, geography, tail-risk requirement, and customer decision. Recent evidence that AI weather models may underperform physics-based models for record-breaking events reinforces the need to evaluate tail behavior, not only average forecast skill [30, 22].
The technical bar therefore changes by horizon. Near-term weather intelligence must be operationally accurate and fast enough to alter a schedule, route, dispatch, or risk-control action. Subseasonal-to-seasonal intelligence must express anomalies and uncertainty clearly enough to inform planning before short-term forecasts arrive. Long-horizon climate-risk analytics must reason under nonstationarity, sparse validation data, and scenario uncertainty while still producing outputs that can survive model-risk review.
7 Product Archetypes
The sector can be divided into five partially overlapping archetypes.
- Climate-model-forward platforms. These companies emphasize long-horizon physical-risk analytics, scenario analysis, downscaling, and asset-level exposure. First Street, Jupiter Intelligence, Climate X, and XDI sit near this cluster [2].
- Weather-intelligence platforms. These companies emphasize short-term and subseasonal prediction for operational decisions. Tomorrow.io and Salient Predictions are representative examples [26, 23].
- Insurance and catastrophe-risk platforms. These companies emphasize vulnerability, loss translation, underwriting, and risk transfer. ZestyAI, Kettle, FloodFlash, Demex, Moody's RMS, and Verisk occupy parts of this landscape [29, 13, 9, 6, 20, 27].
- Asset and infrastructure-intelligence platforms. These products use satellite, sensor, geospatial, or asset-state data to inform infrastructure and operational resilience. They become climate-risk intelligence companies when monitoring is translated into forward-looking risk, avoided loss, or adaptation decisions.
- Workflow-forward planning platforms. These companies compete less on model novelty and more on embedding climate-risk outputs into a buyer's actual planning, underwriting, lending, investment, or resilience process.
The best companies will usually straddle multiple archetypes. A climate-model-forward company needs workflow integration. A weather-intelligence company needs sector-specific thresholds. An insurance-risk company needs validated vulnerability and loss models. An infrastructure platform needs asset-specific data and operational embed. Rather than rewarding long hazard menus, venture investors should favor products whose outputs are trusted when capital is actually allocated.
8 Business Models
The business model is the commercial shadow of the technical architecture. Revenue is captured where uncertainty becomes budgetary.
| Model | Revenue logic | What to diligence |
|---|---|---|
| Enterprise software licensing | Customers pay for a platform that ingests assets, runs hazard and scenario analysis, and generates portfolio or asset-level risk outputs. Pricing can scale by seats, assets, geographies, hazards, modules, or analytical depth. | Does the product enter a live workflow, or is it an annual disclosure dashboard? |
| Data, API, and embedded distribution | Risk information is embedded inside another system: mortgage platforms, financial terminals, insurer systems, real-estate software, or enterprise-risk platforms. First Street's API and data offerings illustrate this model [7]. ClimateCheck's partnership with Black Knight illustrates the appeal of inserting climate-risk data into existing mortgage workflows [32]. | Is the company the system of engagement, or an input whose pricing power may be capped by the distributor? |
| Risk-transfer and insurance-linked models | The company monetizes the movement of risk itself. Parametric insurance is the clearest example: FloodFlash pays out when a pre-agreed flood-depth trigger is met [9]. Demex similarly structures climate risk-transfer products with insurers, reinsurers, and brokers [6]. | Does the model price risk accurately enough to support underwriting, capital, and claims logic? |
| Advisory-led or services-augmented models | Services build trust, help ingest messy customer data, and reveal which workflows should be productized. | Are services a wedge into repeatable software, or are bespoke reports the actual product? |
Venture-scale outcomes are most likely where software recurrence meets workflow inevitability. A company that translates physical uncertainty into a decision can monetize that decision. A company that stops at hazard visualization must monetize attention.
9 Competitive Landscape and Moats
The competitive landscape is a contest over who controls the most decision-relevant layer of the stack.
First Street has moved from property-level climate-risk communication toward broader climate-risk financial modeling, with API, portfolio, and infrastructure-oriented products [7, 8]. Jupiter Intelligence emphasizes enterprise physical-risk analysis, stress testing, financial translation, and auditable methods [12]. Climate X competes through asset-level and portfolio-level financial translation for financial institutions [5]. XDI emphasizes quantification of extreme-weather and climate-change impacts to physical assets globally [28]. These companies compete on hazard science, geospatial coverage, vulnerability translation, and institutional trust.
Weather-intelligence companies such as Tomorrow.io and Salient Predictions compete through operational immediacy and forecast frequency. Their moat depends on whether improved forecasts actually change logistics, aviation, construction, agriculture, energy, or infrastructure decisions [26, 23].
Insurance-facing companies such as ZestyAI and Kettle face a sharper validation environment because underwriting models eventually meet claims, regulators, and renewal cycles [29, 13]. Moody's RMS and Verisk bring incumbent gravity from catastrophe modeling, insurance distribution, and institutional trust [20, 27]. If climate-risk intelligence remains a generic portfolio score or compliance feature, incumbents can absorb it. Startups need sharper wedges: better peril-specific science, proprietary asset or claims data, faster workflow integration, or underserved buyer segments.
The moat hierarchy is:
- Weak: attractive dashboard, generic map, or broad score.
- Better: proprietary or hard-to-recreate data.
- Stronger: validated model tied to claims, losses, asset performance, or operational outcomes.
- Strongest: recurring decision workflow whose removal would impair underwriting, lending, capital planning, risk transfer, or operations.
The right competitive question is therefore not "who has the most hazards?" It is "who owns the highest-value decision with the least replaceable evidence?"
10 Key Risks
| Risk | Why it matters |
|---|---|
| Model risk | A model may perform well in average conditions but fail in the tails, where customer losses are most convex. Rare events, compound hazards, and nonstationary extremes are the core test. |
| Data and validation risk | The satellite-era record is short relative to many hazard recurrence intervals. Local data are sparse, and long-horizon validation often depends on hindcasts, stress tests, physical consistency, and ensemble comparison. |
| Vulnerability translation risk | Hazard exposure is not loss. If vulnerability functions are generic or poorly calibrated, final risk scores may look precise while remaining economically fragile. |
| Decision-translation risk | A product can make risk intelligible without making it actionable. The output must alter underwriting, lending, acquisition, divestment, resilience CapEx, or operational planning. |
| Buyer and budget risk | Climate risk often touches sustainability, risk, operations, finance, and compliance teams. Diffuse ownership can slow procurement. Strong companies attach to a clear budget owner. |
| Commoditization risk | Hazard maps and dashboards are increasingly easy to imitate. Defensibility must come from data, validation, regulation, distribution, or workflow lock-in. |
| Trust and liability risk | Climate-risk outputs can influence property values, insurance availability, credit decisions, and investment behavior. The more consequential the output, the more it must survive model-risk review and external scrutiny. |
| Timing risk | The market may be inevitable but still develop unevenly. Some customers will buy for disclosure before embedding analytics into live decisions. That can be too slow for venture-backed timelines. |
The best companies turn these risks into moats. Model risk becomes a moat if the company validates better than peers. Data risk becomes a moat if the company obtains proprietary asset or claims data. Translation risk becomes a moat if the company maps hazards into losses with unusual fidelity. Buyer risk becomes a moat if the product embeds itself inside a recurring workflow.
11 Diligence Questions
A serious diligence process should separate scientific credibility from product theater.
11.1 Technical diligence
- Which hazards and time horizons does the product model?
- Does the company use numerical models, AI/ML models, hybrid models, catastrophe models, third-party feeds, or proprietary data?
- Does the product produce deterministic scores, probabilistic distributions, or scenario-conditioned outputs?
- How does the model treat tail events, compound hazards, spatial correlation, and nonstationarity?
- What has been validated: hazard occurrence, asset damage, financial loss, operational disruption, or customer decision quality?
- Where does uncertainty enter, and how is it propagated through the stack?
11.2 Commercial diligence
- Which recurring decision does the product change?
- Who owns the budget, and what system or process does the product enter?
- Does the product integrate into existing workflows, or does it require the buyer to create a new one?
- Does pricing scale with seats, assets, hazards, workflows, risk transferred, or economic value created?
- What proof exists that customers use the output for underwriting, lending, CapEx, allocation, risk transfer, or operations rather than disclosure alone?
11.3 Moat diligence
- What part of the stack is difficult for an incumbent or competitor to replicate?
- Does the company have proprietary asset, claims, loss, exposure, or operational data?
- Are models accepted by regulators, carriers, banks, ratings agencies, or model-risk committees?
- Does performance improve with more customers, more assets, more claims, or deeper workflow usage?
- If a large catastrophe-modeling, financial-data, or enterprise-risk incumbent copied the visible product, what would still remain defensible?
12 Conclusion
Climate-risk intelligence is compelling because physical risk is becoming economically unavoidable while remaining technically difficult to price. The sector should not be pursued through generic dashboards, static exposure maps, or broad climate scores. The most attractive companies convert uncertain physical hazard information into recurring, budgeted decisions.
The highest-conviction wedges are insurance and reinsurance analytics, asset-level risk intelligence for real estate and infrastructure, credit-risk tools for banks and lenders, and operational adaptation platforms for exposed industries. Each contains customers with repeated decisions and material downside.
The transcendental standard is that a climate-risk intelligence company is worth pursuing when it makes physical uncertainty more usable than its competitors can. Beyond visualizing risk, the winner changes the price of risk, the allocation of capital, or the design of adaptation.
References
[1] Baldi, D., et al., "The Future of Climate Tech 2026: At the Nexus of AI, Energy and Decarbonization", Silicon Valley Bank Signature Research, April 2026.
[2] Barrick, J., "A Market Map of Tech Solutions for Climate Adaptation and Resilience", SJF Ventures News & Insights, July 17, 2024.
[3] Bi, K., Xie, L., Zhang, H., Chen, X., Gu, X., and Tian, Q., "Accurate Medium-Range Global Weather Forecasting with 3D Neural Networks", Nature, Vol. 619, pp. 533-538, 2023.
[4] Bodnar, C., et al., "A Foundation Model for the Earth System", Nature, Vol. 641, pp. 1180-1187, 2025.
[5] Climate X, "Physical Climate Risk Analysis | Spectra by Climate X", Climate X, accessed May 2026.
[6] The Demex Group, "About", The Demex Group, accessed May 2026.
[7] First Street, "First Street API Documentation", First Street Documentation, accessed May 2026.
[8] First Street Technology, Inc., "First Street Expands Climate Risk Coverage from Real Estate to Companies and Infrastructure", PR Newswire, April 27, 2026.
[9] FloodFlash, "Parametric Underwriting at FloodFlash", FloodFlash, June 20, 2024.
[10] Fortune Business Insights, "Climate Risk Analytics Market Size, Share & Industry Analysis", Fortune Business Insights Report, May 4, 2026.
[11] Wong, D. R., and Kim, K. B., "Sizing the Inevitable Investment Opportunity: Climate Adaptation", GIC ThinkSpace, May 2, 2025.
[12] Jupiter Intelligence, "Climate Resilience Experts", Jupiter Intelligence, accessed May 2026.
[13] Kettle, "Balancing Risk in a Changing Climate", Kettle, accessed May 2026.
[14] Kochkov, D., et al., "Neural General Circulation Models for Weather and Climate", Nature, Vol. 632, pp. 1060-1066, 2024.
[15] Lam, R., Sanchez-Gonzalez, A., Willson, M., et al., "Learning Skillful Medium-Range Global Weather Forecasting", Science, Vol. 382, No. 6677, pp. 1416-1421, 2023.
[16] Trittipo, A., Ahlawat, H., Mysore, M., Kalavar, S., and Kane, S., "Climate Resilience Technology: An Inflection Point for New Investment", McKinsey & Company, September 29, 2025.
[17] Lee, L.-E., Towey, K., and Ashfaq, U., "What the Market Thinks: Findings from Our Corporate Resilience Survey", MSCI Institute, October 29, 2025.
[18] Mahmood, R., Palena, P., and Carlin, D., "What the Market Thinks: How Global Insurers Are Responding to Rising Physical Risk", MSCI Institute, February 26, 2026.
[19] Lehmann, J., Bowdrey, R., Wang, X., and Eichler, L., "Hidden in Plain Sight: Physical Risk in Asset Owners' Portfolios", MSCI Research Paper, November 3, 2025.
[20] Moody's RMS, "Introduction to Climate on Demand", Moody's RMS Developer Documentation, accessed May 2026.
[21] Price, I., Sanchez-Gonzalez, A., Alet, F., et al., "Probabilistic Weather Forecasting with Machine Learning", Nature, Vol. 637, pp. 84-90, 2025.
[22] Rasp, S., Hoyer, S., Merose, A., et al., "WeatherBench 2: A Benchmark for the Next Generation of Data-Driven Global Weather Models", Journal of Advances in Modeling Earth Systems, Vol. 16, No. 6, 2024.
[23] Salient Predictions, "AI-Powered Weather Intelligence from 1 Day to 1 Year Ahead", Salient Predictions, accessed May 2026.
[24] Swiss Re Institute, "2025 Marks Sixth Year Insured Natural Catastrophe Losses Exceed USD 100 Billion, Finds Swiss Re Institute", Swiss Re Group Press Release, December 16, 2025.
[25] Technavio, "Climate Risk Analytics Platforms Market Analysis, Size, and Forecast 2026-2030", Technavio Market Research Report, May 2026.
[26] Tomorrow.io, "The World's Weather Intelligence & Resilience Platform", Tomorrow.io, accessed May 2026.
[27] Verisk, "Catastrophe and Risk Solutions", Verisk, accessed May 2026.
[28] XDI, "The Physical Climate Risk Experts", XDI, accessed May 2026.
[29] ZestyAI, "Resources", ZestyAI, accessed May 2026.
[30] Zhang, Z., Chantry, M., Magnusson, L., et al., "Physics-Based Models Outperform AI Weather Forecasts of Record-Breaking Events", Science Advances, Vol. 12, No. 17, 2026.
[31] Oehling, D., Fischer, G., Sivaprasad, D., Nanji, T., Shandal, V., Sheridan, B., Chua, H. L., Zimmermann, F., Teo, J., Teo, M., and Lim, H., "The Private Equity Opportunity in Climate Adaptation and Resilience", Boston Consulting Group, May 6, 2025.
[32] Freedman, A., "First Look: ClimateCheck Inks Deal with Black Knight for Climate Risk Analytics", Axios, December 5, 2022.