RESEARCH DOSSIER
Climate Risk Governance and Insurer Strategic Decisions
Abstract. Climate risk governance is the organization-wide system through which an insurer receives information about physical and transition risks, determines their materiality, assigns authority, and makes accountable decisions about underwriting, investments, reinsurance, capital, operational preparedness, adaptation, and the continuity of policyholder protection. This scientific monograph investigates the connections between these decisions and the uncertainty of climate-related hazards, scenario modeling, and market responses. The method combines analysis of authoritative insurance supervisory texts, international scientific assessments, climate disclosure standards, financial scenarios, and selected empirical insurance market research, together with independently checked illustrative calculations. It compares frameworks issued by the International Association of Insurance Supervisors, the United Kingdom, European Union bodies, Canada, Australia, and the United States. The study distinguishes supervisory guidance from applicable national requirements and distinguishes climate hazard trends from portfolio losses, policyholder outcomes, and the effect of insurer governance. Its principal analytical finding is that a climate governance program is substantively operational when scenario information reaches decision-makers empowered to change the insurer's retained exposures and when executed choices can be traced and evaluated. Exposure quality, time-horizon consistency, adaptation effectiveness, reinsurance dependence, claims surge capacity, and social consequences of withdrawal are necessary considerations in that evaluation. The published evidence documents important insurance-market responses and regulatory expectations but does not identify a universal causal effect of governance committees on insurer solvency. Original findings are explanatory interpretations of existing research and institutions; no new proprietary theory or measured causal result is asserted.
Keywords: Climate risk governance; insurer strategy; enterprise risk management; board oversight; physical risk; transition risk; climate scenarios; underwriting; investments; risk appetite; catastrophe reinsurance; ORSA; solvency; adaptation; claims surge; insurance availability; catastrophe protection gap; insurance disclosures; insurer decision rights; comparative financial regulation; risk-based pricing; causal inference.
Introduction: Research Object, Questions, and Method
An insurer's business is a continuing commitment to finance uncertain future events. Underwriting determines the contingent promises, invested assets support the payment obligations, capital absorbs adverse experience, reinsurance transfers selected layers of risk, and claims operations convert financial resources into fulfilled contractual obligations. Climate-related change matters because it can alter both the distribution of events that cause insurance losses and the economic environment in which these commitments must be maintained.
Climate governance examines the institutional point at which that information becomes action. The board sets business strategy and acceptable risk, management conducts operations, actuarial and risk functions estimate exposures, and legal and supervisory constraints delimit the choices available. A map of wildfire hazard, a modeled future transition pathway, or a report on insurance take-up has no direct strategic effect until interpreted through insured property, investments, contractual terms, capital, and a named decision authority. This monograph studies that transformation.
Its first question concerns definitions and the division of responsibility: which insurance risks are climate-sensitive, where is authority allocated, and how are the assumptions reported? The second concerns decision mechanisms: how can assessment change underwriting, reinsurance, investment, adaptation, capital, claims preparedness, and product availability? The third concerns evidence: what do authoritative texts and empirical research show about these mechanisms, and what further evidence would be needed to identify the effect of climate governance on insurer performance or consumer outcomes?
The research method is critical documentary analysis, cross-jurisdictional comparison, institutional inference, and explicit hypothetical calculation. Primary supervisory sources include the IAIS 2025 Application Paper, the PRA's December 2025 statement, Canadian Guideline B-15, Australian practice guide CPG 229, and EIOPA's ORSA material. Physical science comes from the IPCC impacts assessment. Scenario narratives and methods are drawn from NGFS. Observations on insurer choices are drawn from government reporting and selected research on homeowners insurance.
Evidence strength depends on its object. An international application paper establishes how supervisors are advised to approach a subject; it does not establish the effect of the advice on claim costs. A scenario shows conditional model implications; it is not a probability forecast. A pricing study may identify a market response without observing the internal committee that approved it. An insurer's published climate policy describes governance arrangements without demonstrating every resulting decision. The analysis preserves these distinctions.
There is no original proprietary claims panel, board-minutes collection, complete insurer investment dataset, firm-specific climate-model implementation, or randomized governance reform in this dossier. Quantitative examples are explicitly hypothetical. The resulting conclusions concern institutional structure, decision logic, comparative scope, and identification of research questions; their evidentiary reach is limited accordingly.
Definitions, History, and Documentary Foundation
A financial definition of the climate-risk object
Climate-related physical risk concerns potential harm from acute meteorological and climatic extremes and chronic climate processes. Relevant insurer pathways include storm, flood, wildfire, heat, changing precipitation, and secondary or compound losses. Climate transition risk concerns financial changes related to decarbonization policies, technological transitions, changing consumer demand, legal developments, and repricing of assets and activities. Liability risk may arise where claims against insured parties or the insurer itself are affected by climate-related duties or representations. The IAIS climate supervision framework and Australian financial-risk guidance describe these pathways in their respective institutional contexts.
These drivers cross existing insurance risk categories. A change in rainfall intensity can affect property underwriting, casualty exposure indirectly, catastrophe reserves, reinsurance attachment, and claims surge operations. A shift in energy technology can change the value of fixed-income holdings, the risks of insured industrial projects, and opportunities for new coverage. The relevant risk unit is therefore a financial exposure, claim process, investment or contractual commitment, conditioned on a climate driver. The words physical and transition identify transmission mechanisms; they do not constitute a complete capital allocation.
Physical climate information is also geographically mediated. Two properties exposed to the same regional weather pattern can experience different loss because of construction quality, elevation, fire defenses, occupancy, deductibles, and claims costs. The IPCC account of interacting hazards and vulnerability characterizes climate risk as an interaction of hazard, exposure, and vulnerability. An insurer adds contract coverage, limits, insured values, claims settlement rules, and reinsurance to translate the resulting damage into liabilities.
How the regulatory discussion developed
Traditional insurer risk management developed through underwriting selection, reserves, investment discipline, pooling, catastrophe protection, and solvency supervision. Contemporary climate governance extends these functions to the possibility of changing hazard distributions, long-duration transition pathways, correlated effects on assets and liabilities, and wider market consequences.
The institutional record has a discernible sequence. EIOPA's 2021 opinion on climate ORSA scenarios emphasized integration of climate risk scenarios in own-risk assessments, followed by its 2022 materiality and scenario guidance explaining possible practice. The IAIS issued a more comprehensive Application Paper in April 2025. Its official replacement announcement explains the relationship to earlier 2021 guidance. In the United Kingdom, the updated 2025 PRA expectations replaced an older supervisory statement. A contemporary comparison needs the operative document date and scope rather than treating every archived publication as simultaneous.
There has also been a shift toward public disclosure of how boards oversee climate-related information. The IFRS S2 disclosure standard organizes climate-related disclosure around governance, strategy, risk management, and metrics and targets. It was issued in 2023 and subsequently amended. Adoption and effective application are dependent on local requirements. Disclosure creates a more observable institutional record; its reported governance processes remain distinct from demonstrated outcomes.
Normative hierarchy and legal scope
IAIS Insurance Core Principles provide an internationally recognized supervisory framework. The 2025 IAIS Application Paper is a guide for applying supervisory principles and does not introduce additional enforceable requirements of its own. Australia's CPG 229 describes prudent practice and acknowledges the preexisting standards it supports. An EIOPA opinion and an application guide have functions within European supervisory convergence, which differ from binding regulations or the requirements of national competent authorities. PRA and OSFI documents state expectations within defined domestic supervisory frameworks. The IAIS notice on the document's status gives explicit support for this distinction.
A scientifically meaningful comparison therefore identifies issuer, legal authority, instrument type, date, targeted entity, and decision topic. An internationally active group may use one central climate-risk vocabulary while applying different legal requirements to domestic insurers or branches. A global head office policy cannot by itself establish compliance in each licensed entity.
Institutional units and financial horizons
For a nonlife insurer, annual underwriting, catastrophe risk, insurance availability, and reinsurance renewals may dominate the near-term decision calendar. A life insurer may have long-duration assets and liabilities exposed to several macroeconomic and climate pathways. A reinsurer faces geographically connected catastrophe losses and changing retrocession conditions. Mutual insurers may be subject to member-related objectives and different accountability structures. These differences alter materiality, stakeholder consequences, and feasible governance instruments.
Climate horizons are similarly plural. A one-year solvency capital measure, an annual property renewal, a three-year financial plan, a ten-year geographic strategy, and a multi-decade investment exposure cannot be substituted for one another. Governance requires an explicit mapping between the duration of assumptions and the duration of decisions. The EIOPA ORSA scenario guidance treats longer climate horizons as relevant to materiality while presenting examples of insurer-specific use.
Research boundary
The science of climate hazards, actuarial catastrophe loss calibration, insurance contracts, and public disaster finance form the analytical inputs of this dossier. The subject here is the insurer's responsibility for interpreting those inputs and making strategic choices. A paper that identifies a hazard trend is supporting evidence for risk identification, not automatic evidence of governance effectiveness. A firm's selection of treaty limits, risk appetite, or portfolio concentration is evidence of governance only insofar as the decision authority, assumptions, and implementation can be examined.
Climate Oversight and Authority
Board responsibility and operational decision rights
Effective climate governance places climate-related financial drivers within the insurer's ordinary system of responsibility. The board determines the insurer's strategy and the tolerable level of risk in relation to commitments and resources. Executives implement business plans; underwriting, investment, actuarial, finance, reinsurance, claims, legal, and risk functions contribute distinct information and execution. The IAIS corporate-governance guidance and the OSFI risk governance framework support allocation of ownership to material climate-related risks within established enterprise risk arrangements.
A new committee can coordinate information but does not independently transfer legal authority to change product terms, capital, investments, or reinsurance. The relevant empirical question is which officer may approve a risk limit, halt a portfolio expansion, alter a reinsurance program, authorize a model override, or escalate concerns to the board. Governance can use dedicated climate structures or integrated existing committees when decision paths remain sufficiently clear.
A materiality assessment starts with actual insurance and investment exposures, then asks how physical or transition pathways could change their loss or cash-flow distributions. A large modeled climate impact on an industry that the insurer neither covers nor invests in might be less material than a smaller local hazard affecting most of its insured concentration. Materiality assessment therefore combines risk magnitude, exposure, duration, vulnerability, limits, and strategic dependency.
Risk appetite as a decision constraint
Risk appetite is the amount and type of risk an insurer is prepared to take in pursuing its objectives. Its translation into limits may involve gross and net catastrophe accumulation, premium adequacy, concentrated exposures, single-counterparty protection, investment credit quality, liquidity, capital coverage, or growth restrictions. The IAIS discussion of risk appetite connects material climate risks with enterprise risk management and appetite statements.
An underwriting portfolio may be compliant with a net annual catastrophe budget while exceeding a geographic accumulation limit under a particular model. The limit should identify the exposure unit, event metric, model version, return-period convention, deductible, reinsurance recoveries, and the treatment of model uncertainty. Without these details, management and the board cannot know whether a breach reflects portfolio growth, improved data, a new hazard estimate, or a genuine increase in retained risk.
A stated limit requires an operational consequence. For a proposed new line, the decision could be acceptance with a restriction, higher reinsurance, revised policy terms, an additional independent analysis, or a decline within applicable law. An exception should record the approving authority, duration, evidence, and monitoring trigger. Repeated undocumented exceptions would be relevant evidence of weak control even if the formal policy were detailed.
The path from data to a board decision
The climate-risk process links source data, hazard models, insured or invested exposures, policy terms, financial translation, aggregation, scenarios, options, and action. At each stage a different professional discipline contributes and may introduce uncertainty. A hazard return period is not an insured loss distribution; a portfolio's emissions intensity is not a credit loss; a climate pathway's economic damage estimate is not a near-term capital charge. Keeping the model's units visible is essential.
Consider a board request to continue writing coastal residential property. A useful decision memorandum identifies exposure values, modeled gross and net catastrophe losses, location quality, vulnerability assumptions, reinsurance conditions, affordability implications, and alternatives such as tighter limits or mitigation incentives. It should state what changes under alternate scenarios and who implements the approved option. The PRA expectations on data and controls and APRA's oversight guidance encourage comprehensive treatment of financial risks; the specific memorandum format is this monograph's analytical recommendation.
The risk function should have the authority to challenge whether assumptions are fit for the contemplated decision. It may identify model instability, poor geocoding, insufficient tail calibration, or stale policy terms. Senior management can decide among lawful options within its mandate while documenting the basis for accepting or rejecting challenges. The quality of that record is more informative than the mere existence of a risk dashboard.
Incentives, controls, and claims service
Underwriting premiums and market share are immediately observable; rare-event losses may take years to appear. Consequently, a strategy that increases near-term premium revenue while accumulating catastrophe exposure can look successful after a quiet season. The board should evaluate growth using risk-adjusted outcomes, exposure constraints, and alternative modeled events. Compensation and business budgets can reward expansion without equally reflecting capital or liquidity consequences, making independent challenge essential.
Climate governance also includes claims readiness. A severe event can generate simultaneous demands for adjustment staff, contractors, emergency accommodation, technology, legal interpretation, and reinsurance payments. The IAIS recommendations on conduct and claims discuss handling catastrophe claims and clear policyholder communication. The insurer's ability to pay promptly and fairly is a strategic output distinct from the modeled balance-sheet loss.
Oversight evidence versus disclosure evidence
The IFRS S2 categories of climate oversight encourage reporting of governance processes in disclosures to investors and other users. A disclosure of board oversight is an observable statement about internal arrangements. It does not prove that a specific investment was sold after a risk trigger, that an underwriting limit was changed, or that customers experienced faster settlements.
A more informative assessment combines public governance descriptions, minutes or audit records where available, approved policies, decisions taken, changes in financial exposures, and monitored outcomes. A researcher should distinguish an announced climate target from an implemented business decision, and distinguish the strategy's financial effect from unrelated changes in hazard, prices, competition, or regulation.
Professional authority and proportionality
A large multiline group may require sophisticated models, cross-border controls, and group-wide escalation. A smaller specialist may require proportionate data and a clearly defined board discussion around concentrated underwriting exposure. Supervisory proportionality does not determine that climate risk is always material; materiality must be established from an insurer's business. It instead concerns the intensity and form of assessment appropriate to the risk.
The Canadian climate management guideline and the Australian risk management guide allow a comparative reading of how national supervision approaches management processes. The common substantive problem is whether the institution can move from information to a decision within its authority before the risk has already crystallized.
Physical and Transition Scenarios
What a climate scenario establishes
A scenario is a structured account of possible future conditions and the mechanisms linking those conditions to financial exposures. The NGFS scenario portal presents alternative long-term and short-term climate-economic pathways, including differing degrees of physical risk and transition disorder. They are conditional modeling exercises whose interpretation depends on their assumptions; the number attached to a scenario does not, by itself, give the unconditional probability of that future.
The insurer needs to start from a decision question. For a one-year property renewal, it may examine current hazard, catastrophe tail, inflation, and reinsurance prices. For a decade-long underwriting strategy, it may assess concentration growth and migration, adaptation, construction, and market access. For a bond portfolio, it may examine credit and market repricing under differing policy and technology pathways. The EIOPA climate scenario guidance provides insurer-relevant materiality and scenario examples.
Financial translation of physical hazard
Conceptually, the claims loss in a particular modeled year can be expressed as a function of hazard intensity, values exposed, asset vulnerability, coverage, deductibles, limits, claims inflation, and adjustment costs. Expected insured loss is the sum or integral of this contract-adjusted outcome over plausible event states. Reinsurance then changes the retained result subject to attachment, limits, reinstatement, exclusions, and counterparty performance.
A hazard projection using a static book of insured properties can misrepresent a future insurance portfolio because exposure itself changes. Urban development, construction standards, building hardening, migration, coverage take-up, and premium changes may alter both the risk distribution and the social allocation of economic loss. The IPCC hazard-exposure-vulnerability framework provides the scientific basis for treating these components separately.
A historical increase in nominal insured claims may reflect inflation, growth in insured property values, coverage differences, catastrophe clustering, or weather. The attribution of the trend to climate change requires an appropriate event or long-run research design. Governance can use trend evidence for caution without representing a particular observed claims series as a clean measure of climate attribution.
Transition scenarios and assets
Climate transition can alter commodity use, electricity systems, transport infrastructure, technology investment, regulation, consumer demand, and corporate earnings. For insurers, these mechanisms enter the creditworthiness of bond issuers, asset market values, counterparties, property holdings, and the insured operations of clients. The OSFI account of transition risks explicitly identifies financial risk arising from the transition to a lower-carbon economy.
The investment function should distinguish an entity's emissions disclosures from its sensitivity to financial repricing. Two issuers can have similar emissions profiles but different leverage, technology options, policy exposures, and expected cash flows. A meaningful scenario must specify the macro-financial pathway and the assumed repricing, rather than converting emissions into credit loss through an unexplained multiplier.
Joint balance-sheet scenarios matter where a catastrophe simultaneously increases claims and decreases the value or liquidity of regional assets. Correlation may also affect reinsurance counterparties or asset sales. A single isolated underwriting shock cannot reveal all such joint financial vulnerabilities. The IAIS investment and solvency discussion recognizes material climate risk through existing prudential categories.
Hypothetical severe-loss illustration
Take a deliberately simplified two-state portfolio. In the ordinary state, annual gross insured catastrophe loss is $10 million; in the severe state, it is $450 million. Assign probabilities 98 percent and 2 percent for arithmetic instruction only. The expected gross annual loss is 0.98 × $10 million + 0.02 × $450 million = $18.8 million.
Suppose a single-event excess-of-loss reinsurance treaty attaches at $100 million and has a $250 million maximum payment. In the severe state the insurer receives $250 million and retains $200 million; in the ordinary state it receives nothing and retains $10 million. The expected ceded loss is 0.02 × $250 million = $5 million, while the expected retained loss is $13.8 million. This excludes the reinsurance premium, reinstatement, taxes, expenses, basis differences, credit risk, and recoverability timing.
If severe-state probability is instead set to 1 percent, with the ordinary-state probability 99 percent, the assumed expected gross loss is $14.4 million and expected ceded loss $2.5 million. The change follows from the artificial probability distribution. It supplies no measured estimate of a particular region's climate hazard. If the board defines a severe-state retained-loss tolerance of $175 million, the initial arrangement exceeds the tolerance by $25 million in that state despite delivering substantial reinsurance protection. This is the decision-relevant result: management must consider a change in retained exposure, treaty coverage, risk budget, or another lawful option.
The illustration shows that expected loss and tail exposure are distinct. Expected net loss alone cannot establish the ability to pay the remaining $200 million severe-state claim, the liquidity profile of reinsurance reimbursements, or the effect of concurrent investment losses. An appropriate board report would therefore include multiple loss measures, available capital, liquidity, and contractual protection.
Model versions and decisions under uncertainty
Scenario estimates can vary with model vintage, emissions assumption, financial transmission, damage function, hazard downscaling, exposure inputs, and market responses. The NGFS scenario methodology FAQ illustrates why version information is material to comparison. A changing modeled estimate should be decomposed into changing exposures, improved data, changes in the model itself, and new assumptions about future pathways.
A climate scenario can be useful without a precise likelihood. Sensitivity analysis tests whether management's decision remains satisfactory under alternative assumptions and which assumptions determine the outcome. A reverse stress exercise works backward from an unwanted result, such as inability to settle claims, to combinations of catastrophe events, market losses, reinsurance delays, and operational bottlenecks.
Backtesting also requires qualification. Observed historical claims may help evaluate geocoding, event footprints, contract logic, and claims translation. They cannot directly validate an unobserved multi-decade transition pathway. Methodological confidence should be allocated to the components actually tested. This discipline keeps scenario analysis within the limits of its evidence while making it useful for choices today.
Portfolio Steering and Adaptation
Underwriting strategies
Portfolio steering concerns decisions on which risks to accept, what coverage to promise, where to operate, which premiums and deductibles to apply, how fast to grow, and how much catastrophe exposure to retain. Climate-related tools include geographic accumulation limits, granular risk classification, vulnerability assessments, verified mitigation incentives, coverage design, stronger claims readiness, and reinsurance. Each choice must also respect relevant insurance contract and rating law.
The study by Boomhower, Fowlie, Gellman, and Plantinga on wildfire risk classification and pricing uses insurer and property-risk evidence to describe variation in the use of granular wildfire-risk information. Its relevance to governance is that the quality and granularity of information change the feasible decisions. A carrier able to distinguish individual exposure from broad regional averages may select, price, and mitigate differently. This research does not establish that any single classification system is optimal across markets.
Differentiating price by hazard can improve the relationship between promised protection and expected claim costs. It can also alter affordability and participation. Continued underwriting under inadequate prices can weaken the insurer's long-run ability to meet commitments, while immediate withdrawal can displace uncovered losses to households or public institutions. The appropriate strategy cannot be inferred from one financial ratio; it requires examining the insurance product, risk concentrations, consumer outcomes, and the scope of public intervention.
The Federal Insurance Office market study reports cost and availability developments from a large US homeowners insurance dataset. Its observation period and geographic aggregation define the inference. It is evidence that the insurance-market problem is substantive, rather than an estimate of the independent effect of an insurer's board committee.
Investments and integrated balance-sheet exposure
Insurers may hold assets exposed to geographic catastrophe, transition-sensitive industry changes, or climate-related regulatory developments. Asset and liability effects can occur together. For instance, a carrier writing property insurance and holding local municipal or real-estate exposures may confront higher claim payments and a decline in asset liquidity after the same event. Climate governance should therefore aggregate material connections across underwriting, investments, liquidity, and counterparties.
A liability-focused strategy with long investment duration must examine different time horizons than a short-tail property insurer. The IAIS materiality and investment guidance connects climate risk drivers to insurer investment, valuation, and solvency considerations. Portfolio decisions should state the financial transmission assumed: default, spread widening, asset repricing, revenue decline, or cash-flow interruption.
A published portfolio climate indicator may be informative for disclosure. It cannot replace cash-flow analysis where the decision is about future claims-paying assets. Metrics from different portfolios have different denominators and assumptions; combining them into a single score without a financial explanation can obscure concentration.
Reinsurance, capital, and strategic reliance
Reinsurance reduces net catastrophe risk under contractual terms and creates financial dependencies. Renewal prices, availability, exclusions, attachment points, reinstatements, collateral, and reinsurer creditworthiness all affect the amount and timing of effective protection. An insurer that meets an internal retained-loss limit only if recoveries are immediate and certain may have a liquidity problem even where its modeled net ultimate loss appears acceptable.
The severe-loss illustration demonstrates this distinction. A $250 million cession reduces a $450 million gross event to $200 million retained loss; it does not itself ensure the insurer can finance the remainder or withstand an additional market shock. The board should understand the treaty's layer structure and the consequences of adverse renewal terms. Reinsurance, exposure reduction, and adaptation investments are alternative strategic responses to be evaluated with consistent assumptions.
Adaptation and loss prevention
Risk can be influenced by building design, flood protection, land-use policy, fire hardening, maintenance, and emergency response. The IPCC assessment of adaptation options recognizes the role of vulnerability reduction and the limitations of available measures. Insurance terms can sometimes encourage verified adaptation through pricing, deductibles, coverage terms, or information, where law and product design permit.
Consider another explicitly hypothetical portfolio. Modeled expected annual insured loss is $32 million. A proposed mitigation program is assumed to reduce expected loss by 25 percent, to $24 million, and to cost $6 million annually. The modeled gross loss reduction is $8 million and the net annual difference before other factors is $2 million. The result follows only from these assumptions. If efficacy is lower or portfolio composition changes, the financial implication changes as well. A real approval requires credible efficacy data, capital and liquidity effects, costs, legal constraints, and a monitoring plan.
The insurer's control over adaptation is partial. It may alter contract incentives or its own claims infrastructure, while building standards, flood defenses, land use, and household investment frequently depend on other actors. Strategy records should separate actions the insurer can implement from public or private measures on which its risk assessment depends.
Protection gaps and geographic retreat
An insurance protection gap exists under a defined measure when economic losses or exposed risks are not effectively insured. Its magnitude can differ depending on whether the statistic measures historical losses, expected risk, insured value, or household participation. EIOPA's natural-catastrophe protection-gap dashboard contains modeled and historical perspectives that should be interpreted with their distinct denominators.
A higher price does not necessarily translate mechanically into lower insured loss for society. Some customers may change coverage, increase deductibles, leave the market, or implement mitigation. Others may be unable to obtain or afford protection. EIOPA's analysis of demand-side protection gaps highlights why insurance purchase behavior matters alongside insurer supply. The customer's decision is a separate causal mechanism from the insurer's pricing calculation.
A decision to withdraw from a location can reduce the carrier's near-term exposure while increasing uninsured loss borne elsewhere. Maintaining coverage at prices that cannot support foreseeable obligations can also generate broader losses if an insurer fails. Governance should document the financial and social consequences of plausible alternatives and distinguish the insurer's lawful duties from decisions reserved to public policy.
Property finance and regional feedback
The research by Keys and Mulder on insurance premiums and disaster risk uses mortgage escrow evidence to examine property insurance price changes and links to the housing market. Its findings concern the study's data, time periods, and identification. For insurer governance, they illustrate that a pricing decision can have consequences beyond premiums, potentially influencing affordability, collateral, and local development. This does not establish a single universal feedback coefficient or justify treating all premium increases as climate-attributable.
Strategic monitoring should therefore separate premium adequacy, portfolio profitability, customer renewal, policy nonrenewal, coverage availability, and claims results. A report that celebrates a reduction in catastrophe exposure without showing how it was achieved could miss an important change in policyholder protection. A report that counts newly insured policies without measuring the risk retained could similarly misrepresent sustainable growth.
Operational resilience and policyholder outcomes
Catastrophic weather can generate simultaneous demands for assessment staff, repairs, emergency accommodation, fraud safeguards, communications, legal interpretation, and reinsurance recoveries. Financial protection can be delayed by operational failure. The IAIS catastrophe claims recommendations support incorporating preparedness and conduct into oversight. Governance of this exposure includes business continuity tests, claims surge arrangements, partner capacity, reliable policy records, and post-event monitoring.
The decision to invest in claims technology or additional adjuster capacity belongs to the same strategic landscape as the decision to buy reinsurance. Both can affect whether the insurer delivers the promised contract performance after a severe event. Their outcomes require different evidence: a claims service measure is not equivalent to a modeled capital ratio.
Evidence Limits and Improvement
Regulatory texts, disclosed arrangements, and actual outcomes
Evidence about climate governance appears in several distinct forms. Climate assessments describe physical processes and vulnerabilities. Supervisory texts articulate expected risk management or reporting practices. Firm disclosures report governance arrangements. Insurance-market studies describe observed prices, selection, and financial consequences. Internal records, if available, show decisions, authority, and implementation. These sources cannot be collapsed into a single measure of governance effectiveness.
The IFRS S2 disclosure structure improves consistency of reported governance information where applied. Its stated coverage, however, is reporting. An observed disclosure that a board reviews climate exposure does not show whether the review changed a catastrophe limit. A claim about financial consequences needs actual insurer exposure or outcome evidence, not just a description of oversight.
The California reporting-year climate survey demonstrates a separately administered disclosure mechanism. Jurisdictional reporting populations and definitions differ, so counts of reported policies, committee names, or climate targets need normalization before comparison. The presence of a disclosure requirement can change reporting intensity without automatically changing insurer risk-taking.
Selection bias and causal identification
Suppose firms with stronger climate controls later report fewer catastrophe losses. Such a pattern could reflect the controls, but may also reflect geographic portfolio differences, reinsurance purchases, capitalization, exposure selection, luck in the observation period, or the fact that firms with greater prior risk invested in control upgrades. A causal estimate requires a clearly defined treatment and comparison, common outcome definitions, and a credible account of how governance adoption relates to preexisting exposure.
A possible research design would compare staggered implementation of a particular supervisory requirement across otherwise comparable firms, while studying pre-trends, anticipation, simultaneous rate regulation, catastrophe occurrence, and insurer composition. Another would exploit a documented internal change in limit approval authority and observe pre/post exposures with a justified control group. The governance variable should represent an actual intervention rather than merely counting climate committee meetings.
The Boomhower and colleagues' wildfire study and Keys and Mulder's escrow study offer valuable empirical evidence of property-insurance behavior. They do not, by themselves, measure a universal causal impact of board-level climate oversight. Their result populations, geographies, time periods, and data access determine the permissible inference.
Data provenance
Insurance risk estimates depend on addresses, construction, insured values, policy terms, portfolio dates, asset holdings, treaty terms, and loss adjustment assumptions. A regional hazard map can conceal substantial variation in property vulnerability. Claims inflation and reconstruction demand can alter nominal loss independently of hazard. Portfolio selection can change the insured exposure even when the physical environment is unchanged.
A replicable review should document data sources, observation windows, model versions, missing geocodes, imputation, vulnerability assumptions, claims development, policy coverage, and the bridge to financial metrics. The US homeowners market evidence gives a public example of extensive geographically organized insurance data; its aggregate reporting does not provide unrestricted firm-level microdata for all questions.
For a group whose catastrophe models produce substantially different results, the board should see the size and cause of differences and understand which decisions are robust to them. Where an answer depends on a disputed model assumption, a decision record can explain the sensitivity rather than conceal it behind a point estimate.
Scenario validation, uncertainty, and time-horizon errors
The NGFS methodology and vintage notes illustrate the consequences of changes to climate-economic model assumptions. Using a scenario value from one vintage with a different portfolio basis can produce apparent changes that reflect inconsistent data rather than genuine risk. A sound technical process versions each material scenario, its assumptions, models, and input data, and identifies the decisions that relied on it.
Some components can be backtested against claims or financial observations; others, especially multi-decade transition pathways, may have no direct historical analogue. An error metric for claims translation does not validate a future macroeconomic policy path. Validation should explicitly state its test target, outcome, sample period, and unsupported components. This protects against false precision.
A long-horizon cumulative physical risk cannot be inserted as a one-year loss without a conversion justified by time and probability. Conversely, a one-year risk-capital measure need not capture future loss of insurance availability or long-run changes in geographic concentration. The insurer should maintain distinct planning and financial horizons and a documented bridge between them.
Claim verification and bibliographic limitations
This monograph uses official regulatory and scientific documents and research working-paper landing materials. Not all linked publications' complete annexes, proprietary underlying datasets, or paywalled full texts were independently reproduced. The empirical papers are described at the level supported by their published abstracts and accessible research summaries. No claim is made that an original replication, comprehensive systematic review, or insurer-level causal estimation was conducted.
The source ledger identifies the checked claim and verification basis for each reference. The two numerical examples are authored illustrations, not source-estimated parameters. The probabilities, loss levels, reinsurance coverage, and adaptation efficacy were specified for demonstration and checked arithmetically. The source-to-claim distinction also requires identifying where a regulator describes a desirable practice versus where a government imposes a legal obligation.
An agenda for more discriminating evidence
Useful future data would tie material board or committee decisions to the legal entity, date, portfolio, model version, alternatives, sign-off, exception, and later financial or policyholder outcome. Analysis could examine whether observed changes improve risk-adjusted capital resilience, claims execution, reinsurance security, product availability, or the distribution of uninsured economic loss.
A narrow implementation study might trace one firm's decision to revise a geographic limit and ask whether new business composition changed as intended. A cross-firm study might test whether a supervisory change altered scenario use or the consistency between climate disclosure and underwriting. Both require carefully defined measurement and appropriate confidentiality safeguards. The existing research provides reasons to undertake these tests but does not preempt the answers.
Comparative Synthesis and Research Conclusions
Common supervisory questions and different legal instruments
Across the IAIS international supervision framework, PRA supervisory statement, OSFI climate guideline, APRA prudential practice guide, and EIOPA materiality guidance, the recurring questions concern board oversight, integration into enterprise risk management, material risk identification, scenarios, data, and actions. Regulatory convergence is therefore visible at the level of institutional concerns. It should not be inferred at the level of identical legal duties or enforcement intensity.
The comparison demonstrates why a multinational insurer needs consistent risk terms and local legal mapping. A policy may be simultaneously useful as internal best practice, relevant to a supervisory expectation, and material to public disclosure. Its legal basis and evidentiary standard may differ in each use. Statements about universal obligation would exceed the cited documents.
Heterogeneous insurer business models
Property carriers face claims concentration, catastrophe reinsurance, annual renewal, product redesign, and geographical access. Life insurers may be more exposed through long-duration invested assets, credit transitions, and long-horizon financial and health assumptions. Reinsurers face international accumulations and retrocession dependencies; mutual and shareholder-controlled firms may operate with different stakeholder objectives. A single unqualified climate-governance score would compress these differences.
A meaningful comparison asks whether each entity identifies its material climate pathways, has people with authority to act, translates data into risk-relevant financial units, records alternatives and decisions, and checks implementation. The sophistication of models can differ proportionately with exposure. The requirement for identifiable decision rights and documented accountability is conceptually constant.
Four evidence-based findings
First, climate governance extends across existing insurer financial functions. Hazard and transition drivers enter underwriting, investments, liquidity, reinsurance, operations, and financial resilience. The conclusion follows from the risk transmission discussed by IAIS and national supervisors. A standalone climate committee cannot by its existence demonstrate that all these effects have been integrated.
Second, the value of climate information depends on its route to an authorized decision. A scenario with no specified management question has indeterminate strategic meaning. A decision with no identified exposure basis, financial horizon, or version-controlled model is difficult to audit and revise. The analytically decisive unit is the documented interaction between information, alternatives, authority, execution, and monitoring.
Third, insurer resilience and the public protection gap are connected but analytically separate. Better pricing and exposure selection can protect a carrier's solvency while changing affordability or availability. The EIOPA catastrophe protection-gap dashboard, the EIOPA analysis of insurance demand, and the US homeowners insurance investigation supply distinct lenses on these consequences. Governance reporting needs to distinguish the insurer's loss and capital position from consumer participation and uninsured social loss.
Fourth, evidence of effective climate governance cannot be reduced to the number of committees, policies, scenarios, or disclosures. The published record describes expectations and certain market responses, but does not supply universal governance-to-solvency causal estimates. Decision-level data, credible comparison, and explicit confounding controls are required for that claim. The distinction between descriptive and causal evidence is a finding about the current research record, not a claim of new institutional theory.
Strategic action amid uncertainty
An insurer frequently has to act before uncertainty resolves. It can use alternative scenarios, ranges, risk appetite limits, policy renewal decisions, independent model challenge, and documented options. The NGFS scenario approach supplies possible pathways, while the EIOPA ORSA guidance shows how insurers can bring climate scenarios into existing assessment processes.
A decision's quality is assessed by the appropriateness of evidence to exposure, the legitimacy of authority, consideration of alternatives, and monitoring of results. More model precision alone cannot settle a legal or distributional choice. A smaller insurer may have a simpler but well-controlled process; a large institution may have many climate reports but weak links between them and executed portfolios.
Limits of inference and prospects for research
The scientific scope of this dossier is international but the empirical studies examined principally concern observable homeowners insurance outcomes in the United States. Public regulatory and disclosure sources document multiple jurisdictions but rarely reveal the confidential board record or complete insurer risk positions. This limits direct cross-country causal generalization. The recommended research agenda includes governance event records, insurer panels with exposure controls, and joint measurement of capital resilience and consumer outcomes.
The original synthesis establishes a precise interpretive conclusion: climate risk governance is measurable as the institution's accountable passage from risk information to constrained financial and operational decisions, followed by monitoring and adjustment. This definition organizes existing evidence without establishing an independent named doctrine or a new regulatory principle. Its scientific usefulness lies in locating where more complete evidence would change a conclusion.
Conclusion
Climate risk governance is a central part of insurer strategy because evolving physical hazards and economic transitions can alter underwriting loss distributions, claims settlement needs, investment returns, liquidity, reinsurance security, and the future availability of protection. Insurance supervisors increasingly articulate convergent expectations concerning oversight, materiality, enterprise risk integration, and scenario use, while national legal authority and instrument type remain distinct.
The comparative research identifies decision rights, coherent time horizons, traceable model assumptions, and feedback into executed portfolios as the substantive features that connect climate science with insurer management. Portfolio measures capable of preserving solvency may influence affordability and the insurance protection gap, making policyholder consequences a necessary additional observation. Public documents and market studies demonstrate real vulnerabilities and pricing responses, while firm-specific governance effects require further decision-level evidence and credible causal identification.
The original contribution is an integrated source-grounded interpretation of insurer climate-risk decisions, the derivation of two hypothetical financial examples, and a comparative identification of what documentary and empirical evidence can substantiate. The monograph reaches an explanatory scientific conclusion without asserting a newly invented theory, external peer review, or quantitative results that the available evidence cannot support.
References
International Association of Insurance Supervisors (IAIS). (2025). Application Paper on the Supervision of Climate-related Risks in the Insurance Sector. Basel: IAIS. https://www.iais.org/uploads/2025/04/Application-Paper-on-the-supervision-of-climate-related-risks-in-the-insurance-sector.pdf
Prudential Regulation Authority (PRA). (2025). Supervisory Statement SS5/25: Enhancing Banks' and Insurers' Approaches to Managing Climate-related Risks. London: Bank of England. https://www.bankofengland.co.uk/prudential-regulation/publication/2025/december/enhancing-banks-and-insurers-approaches-to-managing-climate-related-risks-ss
Australian Prudential Regulation Authority (APRA). (2021). Prudential Practice Guide CPG 229: Climate Change Financial Risks. Sydney: APRA. https://www.apra.gov.au/practice-guides/cpg-229
Office of the Superintendent of Financial Institutions (OSFI). (2025). Guideline B-15: Climate Risk Management. Ottawa: OSFI. https://www.osfi-bsif.gc.ca/en/guidance/guidance-library/climate-risk-management
European Insurance and Occupational Pensions Authority (EIOPA). (2022). Application Guidance on Climate Change Materiality Assessments and Climate Change Scenarios in ORSA. Frankfurt: EIOPA. https://www.eiopa.europa.eu/publications/application-guidance-climate-change-materiality-assessments-and-climate-change-scenarios-orsa_en
European Insurance and Occupational Pensions Authority (EIOPA). (2021). Opinion on the Supervision of the Use of Climate Change Risk Scenarios in ORSA. Frankfurt: EIOPA. https://www.eiopa.europa.eu/publications/opinion-supervision-use-climate-change-risk-scenarios-orsa_en?prefLang=et
International Sustainability Standards Board (ISSB). (2023, amended 2025). IFRS S2 Climate-related Disclosures. London: IFRS Foundation. https://www.ifrs.org/issued-standards/ifrs-sustainability-standards-navigator/ifrs-s2-climate-related-disclosures/
Network for Greening the Financial System (NGFS). (2024–2025). NGFS Scenarios Portal: Long-term Climate Scenarios and Short-term Scenario Narratives. Paris: NGFS. https://www.ngfs.net/ngfs-scenarios-portal/
Intergovernmental Panel on Climate Change (IPCC). (2022). Summary for Policymakers. In Climate Change 2022: Impacts, Adaptation and Vulnerability. Cambridge University Press. https://www.ipcc.ch/report/ar6/wg2/chapter/summary-for-policymakers/
U.S. Department of the Treasury, Federal Insurance Office (FIO). (2025). Homeowners Insurance Costs Rising, Availability Declining as Climate-related Events Take Their Toll. Press release, January 16. https://home.treasury.gov/news/press-releases/jy2791
California Department of Insurance. (2026). Climate Risk Disclosure Survey: Reporting Year 2025 Guidance. Sacramento: California Department of Insurance. https://www.insurance.ca.gov/0250-insurers/0300-insurers/0100-applications/ClimateSurvey/index.cfm
European Insurance and Occupational Pensions Authority (EIOPA). (2025). Dashboard on Insurance Protection Gap for Natural Catastrophes. Frankfurt: EIOPA. https://www.eiopa.europa.eu/tools-and-data/dashboard-insurance-protection-gap-natural-catastrophes_en
European Insurance and Occupational Pensions Authority (EIOPA). (2026). Climate Insurance Protection Gaps: A Demand-side Challenge. Eurofi Magazine contribution, March 25. https://www.eiopa.europa.eu/publications/climate-insurance-protection-gaps-demand-side-challenge_en?prefLang=el
Boomhower, J., Fowlie, M., Gellman, J., & Plantinga, A. (2024; revised 2025). How Are Insurance Markets Adapting to Climate Change? Risk Classification and Pricing in the Market for Homeowners Insurance. NBER Working Paper No. 32625. https://doi.org/10.3386/w32625. https://www.nber.org/papers/w32625
Keys, B. J., & Mulder, P. (2024; revised 2026). Property Insurance and Disaster Risk: New Evidence from Mortgage Escrow Data. NBER Working Paper No. 32579. https://doi.org/10.3386/w32579. https://www.nber.org/papers/w32579
International Association of Insurance Supervisors (IAIS). (2025). IAIS Publishes Comprehensive Application Paper on the Supervision of Climate-related Risks in the Insurance Sector. News release, April 16. https://www.iais.org/2025/04/iais-publishes-comprehensive-application-paper-on-the-supervision-of-climate-related-risks-in-the-insurance-sector/
Network for Greening the Financial System (NGFS). (2024–2025). NGFS Scenarios Portal: Frequently Asked Questions. Paris: NGFS. https://www.ngfs.net/ngfs-scenarios-portal/faq/