Insurance by BogdanovaThe Encyclopedia of Insurance and Risk Science

RESEARCH DOSSIER

Climate Transition Risk in Insurer Investment Portfolios

Abstract. Climate transition risk enters an insurer's investment portfolio through changes in the prospective cash flows, default probabilities, recovery values, discount rates, liquidity and marketability of securities and other assets. Its financial significance is determined jointly by the issuer's capacity to adapt, the distribution of transition shocks across sectors and jurisdictions, the insurer's contractual investment exposures, and the nature and duration of insurance liabilities financed by those assets. This monograph develops a critical international synthesis of the financial transmission mechanisms for sovereign and corporate credit, listed equities, real assets, infrastructure, private credit and pooled funds. It compares direct carbon exposure metrics with issuer-level financial vulnerability, specifies a scenario-to-valuation workflow, and distinguishes asset-level losses from changes in regulatory solvency and accounting equity. Two transparent, hypothetical numerical exercises examine a corporate-bond spread shock and a multi-asset balance sheet shock. The research interprets EIOPA's insurer investment studies, IAIS investment and climate supervision principles, the NGFS scenario framework, selected peer-reviewed pricing evidence, and the relevant EU and IFRS architecture. Its principal conclusion is that decision-useful transition-risk analysis requires exposure transparency, issuer-specific cash-flow mapping, coherent repricing and liability treatment, horizon-sensitive liquidity testing, and disciplined validation of scenario limitations. The monograph makes an original analytical synthesis and contains no claim of newly developed theory, novel econometric identification, calibrated portfolio forecasts, or external peer review.

Keywords: Climate transition risk; insurers; investment portfolios; corporate bonds; equity risk; sovereign debt; asset-liability management; stranded assets; climate stress testing; credit spreads; liquidity; Solvency II; IAIS; NGFS scenarios; private credit; IFRS 9; IFRS 17; IFRS S2.

Introduction

Insurers finance future promises with present assets. The time between collecting premiums and paying claims makes the composition, valuation and availability of those assets central to insurance solvency. Changes in climate policy, production technology, consumer demand and financing conditions modify the economics of the securities insurers hold. For a fixed-income-heavy institution, even a modest repricing of particular issuers can matter substantially when measured against its excess of assets over liabilities. For an institution carrying participating or market-linked business, the consequences further depend on guarantees, policyholder sharing arrangements, options and the economic treatment of liabilities.

The central institutional foundation is the IAIS climate supervision application paper, published in April 2025. It identifies climate-related exposures in both insurance operations and investments and applies existing Insurance Core Principles rather than instituting a new self-standing global climate capital formula. The IAIS investment principle explicitly places the quality of an insurer's investments and their relationship with its liabilities within solvency supervision. These primary standards explain why an investment study must analyze the balance sheet as a connected financial system.

The research question is: under which identifiable assumptions does a climate transition affect the value and risk of insurer investment portfolios, and how do those effects change with the liabilities, liquidity obligations, accounting classifications and supervisory jurisdictions of the insurer? Subordinate questions concern the differentiation of issuer risk from sector labels, the timing of market repricing, the treatment of nominal and real rates, the role of sovereign exposures, the limitations of private-asset valuations, and the economic cost of apparently simple divestment responses.

An early empirical benchmark is EIOPA's transition sensitivity assessment of European insurers' equity and corporate-bond holdings. It illustrates the potential severity of losses in exposed holdings and the importance of distinguishing a sector-specific shock from a whole-portfolio aggregate. EIOPA's subsequent climate stress-testing methodology treats the exercise as an evolving supervisory and analytical capability. The EIOPA sustainability risk program places these sensitivity exercises alongside stress tests and work on asset-side climate exposures. The questions remain important because an aggregate ratio that looks small against total assets may imply a much larger percentage reduction in economic surplus.

This monograph therefore follows the risk through four separable layers: the economic transition, the issuer's balance sheet, the market or contractual value of the instrument, and the insurer's financial position. The same transition shock can have different effects at every layer. The objective is an economically traceable explanation, not a mechanically produced climate score.

Scientific Object, Scope, and Research Method

Scientific boundaries

The object of inquiry is asset-side transition risk in the investment portfolio of an insurance or reinsurance undertaking. Transition means changes arising from mitigation policies and their enforcement, energy and industrial technologies, changes in demand and business practices, and the market's anticipation or reassessment of those developments. The focus includes induced credit, market, concentration and liquidity risk. Physical hazards, climate-related insurance claims and climate litigation enter only where they change an issuer's financial condition or the insurer's asset-liability position. Their direct underwriting analysis belongs to separate scientific research objects.

The unit of observation may be a legal issuer, a security, an asset pool, a country, a fund or an insurer legal entity. The correct unit varies with the claim. A power generator's financing prospects require issuer-level analysis; a bond's immediate market loss requires instrument-level repricing; a reserve-backing portfolio needs legal-entity-level aggregation. Consolidated-group figures alone may obscure assets segregated for particular contracts or policyholder funds. A high-carbon label is therefore a screening variable whose financial meaning must be established by the specific claim, operating geography, maturity and capital structure.

The geographic scope is international. The legal discussion identifies EU Solvency II and IFRS standards with their own boundaries. Global IAIS ICPs are supervisory principles whose implementation varies by jurisdiction. IFRS S2 is a reporting standard effective from January 2024 under the standard's own text; legal application to any insurer depends on a jurisdiction's adoption and regulatory arrangements. The analysis does not generalize these requirements to all countries. Current EU amendments taking effect in January 2027 are distinguished from rules applicable during November 2026.

Source hierarchy and critical method

The inquiry applies comparative document analysis of official IAIS, EIOPA, NGFS and IFRS sources, complemented by peer-reviewed asset-pricing and credit-risk research. Primary institutional sources establish what supervisory standards and scenario documents actually say. Journal abstracts and publisher records support the scope and qualitative findings reported in those studies; they are not treated as independently replicated datasets or proof that a particular coefficient is universally stable. EIOPA's insurance climate stress-testing methodology provides an insurer-specific reference for organizing stressed assets and the resulting financial position.

A scenario is a conditional description of a possible future state. It becomes a portfolio stress test only after a mapping translates its drivers into a specified change in issuers' revenues, costs, financing conditions and asset values. The NGFS Phase V scenario documentation explicitly warns that the long-term pathways are not predictions. This distinction determines the evidentiary interpretation of every numerical stress outcome in this monograph.

Calculations below are synthetic and reproducible. They use stated positions, market-price conventions and shock assumptions to test accounting arithmetic and loss propagation. They do not estimate real insurer exposures or claim that shock magnitudes were calibrated by a regulator. Mathematical expressions describe conditional sensitivity, while interpretations identify the economic conditions under which an expression remains valid. Monetary illustrations use nominal currency units and deliberately exclude tax, management actions, hedging and second-round effects unless otherwise stated.

Valuation sequence

An auditable analysis proceeds from a position date, security identifier, legal owner and data provenance to a classification of transition drivers; then to issuer-specific cash-flow and default implications; to a shocked market or model valuation; to liability and capital adjustments; and finally to a management and liquidity assessment. The sequence is important. Applying a sector haircut after also using an issuer model that fully embeds the same shock creates double counting. Allowing a detailed credit-spread model to absorb transition policy risk, then imposing a second generic carbon charge, is equally vulnerable unless the two measurements cover explicitly different residual risks.

The distinction between information, assumption and inference is maintained throughout. An emissions inventory is information about present activity; a hypothetical future carbon price is a scenario assumption; the loss on a bond is an inference produced by a valuation method and the issuer's capacity to absorb that price. Quality control tests each connection.

Transition Risk Channels

Policies, technologies and demand

Policy transition shocks include changes to carbon pricing, emissions limits, subsidies, efficiency standards, procurement conditions, land-use rules, border measures and the enforcement of existing requirements. Different legal designs change who pays and when. A carbon price imposed on direct emissions enters an operating company's cost structure differently from a product standard that requires replacement capital expenditure. The value impact depends on cost pass-through, substitute availability, customer sensitivity, existing contracts, permits, taxes and the firm's financing capacity.

Technology transition includes improvements in clean generation, storage, industrial processes and energy efficiency, as well as obsolescence of existing capital equipment. Greater technological availability can lower future energy costs while reducing the value of incumbent assets. Demand shifts can accelerate depreciation before a formal rule changes. For some corporations, transition increases margins and expected investment opportunities; for others, it makes committed capital less productive. An economically balanced valuation must consider both favorable and adverse adjustments within a portfolio rather than presuming uniform losses across all energy-related exposures.

A formal income channel for issuer i is an assumed change in free cash flow: delta FCF(i,t) = delta revenue(i,t) - delta operating cost(i,t) - delta capital spending(i,t) - delta tax and working capital effects(i,t). A shock to emissions cost should be multiplied by the operational exposure and adjusted for ability to pass costs to buyers. If a change in expected terminal enterprise value follows, the effect on the company's equity is leveraged by existing debt and priority claims. If cash-flow coverage becomes fragile, a credit shock may occur before insolvency itself becomes likely.

Financing and market repricing

The holder of a bond is exposed to both expected cash shortfalls and changes in required yield. These are related but different quantities. An increase in perceived default probability or loss given default can widen the credit spread, while a liquidity premium may widen even without a comparable change in expected contractual loss. Equity responds to revisions in distributable future cash flows, discount rates and the valuation of embedded growth options. Real assets and infrastructure require project-level analysis of operating licenses, regulated revenues, service contracts, construction risk and residual values.

Market repricing can arrive discontinuously. A transition policy is commonly discussed before enactment, and securities may already incorporate part of its probability. A stress test that introduces the full policy as though it had been wholly unanticipated exaggerates the incremental surprise unless the assumed baseline is documented. Conversely, a delayed or abruptly implemented transition can create repricing not captured by smooth annual forecast curves. Investors should distinguish the cost of the transition that an issuer has already financed from the new information arriving at the stress date.

Battiston and colleagues' network-based climate stress test demonstrates the relevance of financial dependencies and propagation in a European bank application. Its specific banking sample is a methodological point of comparison, not a numerical insurer calibration. The conceptual implication for insurer research is that corporate-credit, sovereign and financial-counterparty exposures can be connected through the same economic transition and funding markets.

Scenario uncertainty and institutional assumptions

The NGFS short-term climate scenarios supplement long-horizon pathways with an approach better aligned to near-term macrofinancial shocks. Long- and short-term families answer different questions; a discounted multi-decade policy pathway cannot simply be treated as a one-year mark-to-market shock. Discount rates, expected inflation, commodity prices and monetary-policy response must follow a coherent economic narrative.

An essential source-quality issue arose in the NGFS Phase V material: its physical-damage methodology notice identifies the retraction of the Kotz et al. (2024) paper underlying some physical-loss estimates. NGFS states that outputs independent of that underlying study, including its short-term scenarios, are unaffected. The distinction permits a carefully scoped transition analysis but forbids treating the affected physical estimates as independently verified economic truth. In a mixed transition-physical test the damaged components require replacement, exclusion or a clearly disclosed sensitivity treatment.

Transmission into investor risk

The mechanisms can be expressed as a chain: policy, innovation or demand shock; issuer cash-flow adjustment; credit quality and valuation adjustment; portfolio aggregation; insurer economic surplus, liquidity and solvency response. Each step has competing explanations. A rise in credit spreads during a carbon-policy announcement may also reflect a general rate or recession shock. A firm with high emissions may be financially sheltered by pass-through mechanisms, while a lower-emission equipment supplier may have severe supply-chain concentration. Correlation with a sector proxy is therefore a starting hypothesis rather than proof of causation.

The practical research conclusion is that a portfolio's transition sensitivity is defined by its conditional financial response, not its emissions total alone. Portfolio carbon metrics provide useful screening and monitoring information, but require translation into issuer-level value drivers before they can justify a quantified loss claim.

Sectoral Credit and Equity Exposures

Corporate bonds and structured credit

Insurers often hold long-duration corporate fixed income to generate predictable cash flows for insurance obligations. Transition risk changes the issuing corporation's repayment capacity and the market discount rate for its obligations. A bond's coupon, remaining maturity, seniority, covenants, collateral and embedded options determine the magnitude of the value response. Price sensitivity to a spread shock tends to increase with duration; exposure to realized credit loss depends separately on default probability, recovery and the bond's position in the capital structure.

Ugolini, Reboredo and Ojea-Ferreiro's European CDS transition-risk study reports that a constructed transition factor affects CDS term structures more strongly for vulnerable firms. The result supports incorporating issuer vulnerability into a credit analysis. It does not establish that every corporate bond spread moves by the same amount or that CDS prices can be used without adjustments for liquidity and bond-specific terms. Dong and colleagues' 2026 corporate bond analysis offers further evidence from Chinese issuers, which is valuable for international comparison but cannot be generalized mechanically to regulated utilities, sovereigns or investment-grade portfolios worldwide.

A simplified bond-pricing equation for a fixed-coupon issue is P(y) = sum over t from 1 to n of C/(1+y)^t + F/(1+y)^n, with C the annual coupon, F the principal repayment and y an assumed flat annual yield. Real bond desks discount cash flows using a term structure of risk-free rates, issuer spreads and instrument characteristics; the flat-yield formula is used here solely for transparent arithmetic. The first-order change in value is approximately delta P/P = -D_mod delta y, where D_mod is modified duration. Convexity and discrete cash flows make the linear approximation less reliable for large shocks.

For a synthetic, default-free-at-baseline bond with face amount 1.0 billion currency units, a 4% annual coupon and seven remaining annual payments, a yield equal to 4% gives price 1.0 billion. Increasing its required yield to 5.5% while holding promised cash flows fixed gives approximately 0.914755 billion; the modeled loss is 0.085245 billion, or 8.52% of the starting price. The arithmetic follows the stated annuity discounting expression. It is a spread-and-yield valuation exercise, not an estimated climate default probability. If the market had partly priced the transition or if risk-free yields fell, the incremental effect would differ.

Climate exposures also enter securitized credit through borrowers, collateral, geographic industry dependence and servicer arrangements. Each tranche's expected loss and price depend on waterfall priority, attachment and detachment points, credit enhancement and underlying payment correlation. A simple weighted average of underlying carbon intensity does not determine senior-tranche impairment. Structured assets demand look-through that respects legal rights and cash-flow structure.

Listed equities

Equity absorbs the residual change in enterprise value after debt and other senior claims. Profitable producers of transition equipment or infrastructure may gain expected growth opportunities, while firms operating exposed legacy facilities may experience accelerated depreciation and costly replacement investment. These effects need not be proportional to current emissions. A company's chosen decarbonization pathway affects future capital requirements, transition credibility and the durability of its franchise; it cannot be inferred from a pledged target alone.

Bolton and Kacperczyk (2021), in Do Investors Care about Carbon Risk?, report that higher corporate emissions and changes in emissions are associated with higher stock returns in their U.S. sample, controlling for several conventional predictors. One interpretation is a carbon risk premium demanded by investors; the study does not imply that high-emitting shares necessarily lose value in every transition announcement. Ex ante expected returns, realized unexpected shocks and firm-level operating resilience are distinct. Selecting equities exclusively on historic emissions without valuation and business-model analysis risks replacing financial research with a sectoral proxy.

An insurer should assess both direct holdings and exposures through collective funds, external mandates, affiliated investment companies and derivative overlays. Investment manager policies matter because voting, financing and divestment can change transition exposure over time. For an insurance legal entity, ownership and asset availability must nevertheless be confirmed at the level relevant to the covered liabilities and regulatory restrictions.

Sovereign bonds

Sovereigns face transition effects through tax receipts, fiscal transfers, energy-import bills, export competitiveness, contingent liabilities and the cost of public investment. A hydrocarbon exporter can lose expected fiscal revenues as demand changes, while an importing economy can benefit from reduced fuel import exposure. The capacity to tax, borrow in domestic currency and implement policy substantially alters the relationship between carbon intensity and creditworthiness. Currency risk and monetary sovereignty further complicate any assumption of a universal sovereign climate spread.

EIOPA's sovereign bond portfolio assessment specifically examines European insurers' sovereign holdings under disorderly transition assumptions. Its existence reinforces the need to test this asset class separately. Assigning all government bonds a zero transition-loss assumption would exclude a documented supervisory research concern, while applying corporate-sector haircuts to government debt would disregard sovereign fiscal and currency mechanisms.

The choice of sovereign yield curve is fundamental to ALM. A fall in risk-free rates can raise both sovereign bond values and the present value of insurance liabilities. A widening sovereign credit premium can lower the value of sovereign holdings even as the benchmark risk-free yield falls. Stress tests must keep those components separate before aggregating portfolio and liability valuation effects.

Private credit, infrastructure and real assets

Private credit and privately placed securities expose insurers to information asymmetry and appraisal lag. Loan values may remain relatively stable in reports because they are measured through periodic models or contractual methods, while the economic value of borrower promises has changed. Contractual loan protections, amortization, financial covenants and sponsor support should be incorporated into loss analysis, and reference interest rates should be distinguished from credit spreads. A floating-rate instrument may protect against a rate move while remaining exposed to deteriorating debt-service capacity.

The IAIS 2025 global insurance market assessment identifies growing investments in private credit as a supervisory concern. It does not supply a universal transition-risk multiplier for such assets. Inference requires borrower and collateral look-through, the legal ranking of debt, concentration in asset managers or sponsors and plausible liquidation discounts. The same limitation applies to private equity, infrastructure funds and project financing.

Transition risk can be particularly nonlinear for infrastructure. A long-term power purchase agreement can stabilize near-term revenue but introduce concentrated counterparty risk. A regulated asset base can support cost recovery, subject to future allowed returns and regulatory decisions. A transport concession may be highly sensitive to changes in technology adoption and consumer travel. A real estate holding can experience capital expenditure demands from energy performance requirements even when physical hazard exposure is low. Those are separate economic channels, so a single label such as 'green infrastructure' has insufficient precision for valuation.

International and sectoral heterogeneity

Risk assessments must identify the legal and geographic setting in which an issuer operates. An industrial firm with operations in multiple markets may face divergent carbon prices, tax incentives, disclosure rules and import restrictions. Its revenues and financing may be in currencies different from those of its production costs. The financial effect on an insurer depends on whether those currency and commodity exposures are hedged, and whether hedges remain effective in severe market states.

The analytical result is that credit and equity classifications are complementary: sector and country identify where to examine a risk; the security's contractual and financial attributes determine its transmission into value. An insurer whose assets are apparently diversified by issuer count can remain concentrated in a shared policy or energy-price driver.

Portfolio Repricing Scenarios

Baseline definition and information set

A coherent scenario exercise begins with a valuation date and a baseline already reflected in market prices and issuer forecasts. Its first deliverable is a list of differences between the stressed state and that baseline. These differences might comprise a stricter policy trajectory, accelerated capital replacement, demand displacement, a higher financing premium or a changed default outlook. Without an explicit incremental surprise, a stressed valuation may confuse a projected structural transition with an unexpected market loss.

The insurer should keep separate an orderly pathway, a disorderly pathway, a fragmented pathway and a less ambitious policy pathway, along with appropriate macroeconomic and transition assumptions. The NGFS provides such scenario families as analytical inputs. Their model outputs need adjustment for jurisdiction, issuer sensitivity, time horizon and the valuation purpose. Use of the same scenario name in separate exercises does not guarantee identical exposures or pricing assumptions.

In the first stage, a vector of transition drivers is assigned by scenario and date: carbon cost, energy prices, sector demand, output, capital expenditure, inflation, interest rates and policy credibility. The second stage translates these drivers into issuer variables. The third stage maps issuer variables into instrument values and risk measures. A formal expression is V_i(s,t) = V_i(base,t) + Delta V_i(transition,s,t) + Delta V_i(macro,s,t) + interaction terms, with s a scenario. The interaction terms reflect the fact that transition expenditure, inflation and financing rates can change together. This is a valuation decomposition, not a separately proposed theoretical model.

Credit and equity consistency

The cost of capital is endogenous to credit quality and the financial environment. When a corporate issuer's cash flows deteriorate, a higher credit spread may be partly explained by the same forecast deterioration. A discounted cash-flow model can therefore double count the shock if it lowers earnings in full, increases default probabilities using those earnings, and adds an unrelated generic issuer haircut. Consistency requires either a joint valuation that explicitly captures these channels or a documented division of shocks with reconciled overlaps.

Equity repricing should reflect both cash-flow changes and discount-rate revisions. Bond and equity shocks from the same issuer should be linked through the firm's capital structure. A hypothetical 20% fall in a firm's equity value and a 10% bond loss cannot be treated as independent observations when both are driven by the same repricing shock. The insurer's aggregate scenario may still include both positions, but their joint distribution matters for uncertainty and concentration.

Multiple insurer funds may hold the same issuer via direct bonds, external funds and derivatives. Gross positions must be reconciled with derivative notionals, delta equivalents, collateral rights, netting terms, side pockets and any policyholder participation. Economic exposures and regulatory reporting categories should be bridged rather than silently equated.

Synthetic whole-balance-sheet experiment

Consider a synthetic insurer holding 100 billion currency units of invested assets and owing insurance liabilities with an economic present value of 92 billion. Initial economic surplus is 8 billion. Divide invested assets into corporate bonds of 60 billion, sovereign bonds of 20 billion, listed equity of 10 billion and private assets of 10 billion. This balance sheet is a mathematical demonstration; it represents no named insurer and is not calibrated to an EIOPA or IAIS sample.

For corporate bonds, assume 12 billion of transition-sensitive positions lose 8% and the remaining 48 billion lose 1%. The corporate-bond loss is 0.96 + 0.48 = 1.44 billion. Assume all 20 billion sovereign positions lose 2%, yielding 0.40 billion. Of the 10 billion equity portfolio, assume 4 billion loses 25% and 6 billion loses 7%, yielding 1.00 + 0.42 = 1.42 billion. Of the 10 billion private assets, assume 4 billion loses 12% and 6 billion loses 3%, yielding 0.48 + 0.18 = 0.66 billion. The total asset loss is 3.92 billion; stressed assets equal 96.08 billion.

Asset block Baseline value, billions Scenario loss, billions Stressed value, billions
Corporate bonds 60.00 1.44 58.56
Sovereign bonds 20.00 0.40 19.60
Listed equities 10.00 1.42 8.58
Private assets 10.00 0.66 9.34
Total 100.00 3.92 96.08

Suppose the same economic narrative increases the present value of insurance liabilities by 1.5%, or 1.38 billion, through the net effect of rates and contract cash flows. Economic liabilities become 93.38 billion. The stressed surplus is 96.08 minus 93.38, or 2.70 billion. Surplus has fallen by 5.30 billion, equivalent to 66.25% of the initial 8 billion, although the loss on invested assets alone is 3.92% of the starting asset value. This difference is the economic significance of balance-sheet leverage. The specified liability shock is an independent assumption, not an output deduced from the asset haircuts.

Under a different internally consistent liability path in which the present value of obligations falls by 1.38 billion, economic surplus becomes 5.46 billion rather than 2.70 billion. The same asset stress leads to different capital consequences. A solvency capital ratio cannot be calculated from these values alone because eligible own funds, the insurer's capital requirement, jurisdictional adjustments, policyholder loss-sharing, taxes and model recalibration have not been specified. Reporting the synthetic change in surplus as a regulatory SCR loss would be a category error.

Sensitivities, timing and feedback

For each stressed position, analysts should compare an instantaneous market shock with a multi-year cash-flow path. Insurance obligations can extend for decades even if a market-value stress is measured over a much shorter window. The same eventual transition expenditure will not necessarily produce the same immediate price change if the path is well anticipated, financed gradually or accompanied by a favorable adjustment in risk-free yields.

Sensitivity analysis should vary sector mapping, scenario ambition, policy timing, issuer pass-through, inflation, bond duration, credit spread, recovery assumptions, technology adoption, correlations and the liquidity discount on private assets. Reverse stress testing starts from a capital or liquidity threshold and asks what combinations of shocks would reach it. Those combinations may be implausible as probabilistic forecasts but useful as transparency tests for fragility.

A second-order effect arises when many insurers or other investors simultaneously rebalance similar sectors. The resulting liquidity demand can widen bid-ask spreads, increase financing costs and amplify reported losses. This is an economic possibility rather than a measured law of universal asset-market behavior. Its significance rises with crowded holdings, leverage in intermediated structures and the concentration of financing sources.

Climate Investment Concentration

Measuring concentration correctly

Issuer count, industry weight and portfolio emissions answer different questions. A portfolio can have thousands of bond lines yet be concentrated in a few utility or transport transition drivers. A high weighted-average carbon intensity may indicate direct emissions exposure without describing balance-sheet vulnerability. Conversely, a relatively low average can hide large tail losses from a small portfolio sleeve with significant leverage or impaired recoveries.

A practical concentration review therefore considers issuer and parent-group aggregation, high-carbon activities, geography, facility location, technology dependencies, energy import and export sensitivity, duration and credit rating, security seniority, and common ownership by asset managers. It should identify sector-to-sector financial dependencies. Utility companies, equipment manufacturers, transport operators and commodity-producing borrowers are linked through input costs and investment cycles. A single policy shock can be transmitted through several nominally distinct industries.

EIOPA's 2020 transition sensitivity work examines corporate bonds and equity linked to fossil-fuel extraction, energy-intensive industries, motor vehicles and power generation. This is a useful initial classification scheme, but the results depend on the chosen holdings, scenario assumptions and sector mapping. It cannot establish an internationally transferable haircut for every issuer in each sector.

Interpreting observed supervisory aggregates

In its October 2026 risk dashboard, EIOPA reports a median exposure to climate-relevant assets around 3.1% of total assets and characterizes ESG-related risk as medium with an increasing outlook. The dashboard also reports an indicator concerning insurers' green bond holdings. These data are a descriptive, date-specific supervisory snapshot. The definition of 'climate-relevant assets' and the reporting population are essential to interpretation; the median is not the amount held by a typical specific insurer, and a percentage of total assets is not a stress-loss percentage.

A research-oriented portfolio audit should therefore resist treating headline ratios as sufficient exposure estimates. A 3% sector concentration can be negligible or material depending on the loss severity, leverage, associated liability sensitivities and business concentration. A smaller private-credit position can be economically more significant if financed against a concentrated short-term liquidity requirement. The relevant denominator may be total assets for screening, economic surplus for solvency sensitivity, or available liquid assets for a cash-flow horizon test.

Sovereign and financial-system clustering

Sovereign exposures can create domestic clustering: an insurer may hold domestic government bonds, invest in companies that depend on the same economy and write policies whose claims and lapses rise when that economy deteriorates. Transition policy or an energy-price shock can therefore affect investments, demand for insurance and household or business financial conditions together. Holdings of banks and other financial intermediaries add a further link because lending and collateral values respond to issuer and macroeconomic shocks.

The ability to diversify climate transition risk internationally depends on correlations under severe scenarios, not simply on the number of countries represented. A global technology substitution shock may affect producers in several jurisdictions simultaneously; divergent policy regimes can also create offsetting returns. Currency hedging introduces collateral and liquidity requirements. The assessment needs a multi-dimensional concentration map rather than a single red-versus-green allocation figure.

The limits of divestment as a valuation response

Selling an exposed security reduces the insurer's ownership of that specific claim, but the effect on total insurer risk depends on the replacement asset, transaction cost, market depth, hedging and the ongoing trajectory of liabilities. It may cause a realized loss, transfer risk to another holder or increase concentration in an apparently safer sector. Engagement or financing credible investment changes can alter future issuer cash flows, though the benefits and timing remain uncertain. Portfolio optimization should treat transition objectives, liability matching, return, liquidity and capital resilience as simultaneous constraints.

This is a comparative financial conclusion, not a mandate for a specific investment policy. The issue addressed by prudential science is whether risks are understood and controlled with appropriate supporting evidence.

Asset-Liability Mismatch

Duration and rate sensitivity

The insurer's investment portfolio is held in service of policyholder and beneficiary claims. Long-term guarantees, annuities, participating contracts, short-tail protection and catastrophe reserves create different cash-flow profiles. Changes in market rates can affect both securities and the present value of obligations; changes in default risk usually affect invested credit assets more directly. An asset-only 'climate loss' is therefore incomplete as a measure of insurer financial resilience.

A simplified local approximation for economic surplus is delta S = delta A - delta L, where A is the market-consistent asset value and L is the economic value of liabilities. For a fixed-income block, delta A approximately equals -D_A A delta r -D_sp A delta spread, plus convexity, defaults and option effects. For discounted liability cash flows, the pure interest-rate term is approximately -D_L L delta r, subject to contract participation, guarantees, options and the valuation framework. The combined first-order risk-free sensitivity is approximately -(D_A A - D_L L) delta r. Spread shocks generally require a separate treatment because credit spreads of assets and discounting of insurance obligations are not mechanically identical.

The approximation makes explicit that a policy scenario linked to disinflation and lower interest rates can have a very different surplus effect from a disorderly scenario linked to inflation and rising benchmark rates. The latter may also modify claims inflation, lapse behavior, contract expenses and the present value of nominal promises. Even the sign of an apparent hedge can change when embedded options or guarantees become material.

Cash-flow adequacy and liquidation

Transition risk can transform an apparently manageable mark-to-market movement into a liquidity problem if cash outflows coincide with impaired market depth. Insurers need cash for claims, surrenders, collateral calls and operating expenses. A long-horizon investment may be economically viable yet poorly suited to urgent sales. A private-credit fund with redemption gates or a senior secured asset with limited buyers cannot be valued as immediately fungible cash.

The Solvency II prudent-person rule requires EU insurers to invest in assets whose risks they can properly identify, measure, monitor, manage, control and report. It directs attention to the security, quality, liquidity and profitability of the portfolio and to assets suitable for the nature and duration of liabilities. That legal requirement supports analysis of liquid resources, asset availability and obligations together rather than treating a favorable credit rating as proof of liquidity.

Liquidity stress needs a dated schedule of forecast claims, lapse and collateral cash flows; maturity and coupon inflows; possible asset sales and haircuts; derivative margin requirements; and legally encumbered assets. A portfolio-wide percentage haircut does not directly give a liquidity shortfall. A separate funding bridge is required, including the assumption that insurance contract behavior itself may be sensitive to the same transition-driven macroeconomic shock.

Contract and accounting distinctions

The treatment of a bond in financial statements differs from its economic market value depending on classification and recognition requirements. The IFRS Foundation's IFRS 9 classification overview identifies amortized cost, fair value through other comprehensive income, and fair value through profit or loss categories according to the relevant contractual cash-flow and business-model requirements. Expected credit loss considerations are not synonymous with instantaneous market-spread marks. A synthetic market shock should not be described as an accounting charge without tracing the exact classification and recognition mechanism.

Likewise IFRS 17 insurance contracts establishes measurement and presentation requirements for insurance obligations, including current estimates of fulfillment cash flows and the contractual service margin. Economic solvency and IFRS accounting performance can diverge because they answer different questions and use different constraints. For participating contracts, the treatment of policyholder sharing affects net shareholder exposure; for unit-linked business, investment risks borne by policyholders require a separate analysis of retained guarantees and fees.

Hedging, capital and management actions

Interest-rate hedges can offset risk-free duration mismatch but do not automatically hedge issuer carbon-transition credit spread or equity-value risk. Credit derivatives may introduce counterparty, basis, tenor and collateral risks. Inflation hedges may respond differently from sector earnings and investment costs. Hedge effectiveness must be assessed in the same joint scenario that produces the exposure loss.

Management actions such as rebalancing, increasing high-quality liquid assets, revising guarantees on new business or improving asset look-through should be modeled with realistic execution constraints. An assumed instantaneous sale at pre-stress market prices creates an unsupported benefit. An assumed capital injection needs an identified source and conditions. Where policyholder contract rights or regulator-approved matching arrangements constrain asset substitution, those restrictions belong in the scenario.

EIOPA's climate ORSA application guidance supplies practical insurer-oriented materiality and scenario examples, including illustrative life and non-life undertakings. It supports a forward-looking management horizon broader than an immediate annual ratio. The analytical objective is resilience of the institution's capacity to meet obligations, not the optimization of an isolated investment statistic.

Supervisory and Reporting Questions

IAIS investment and risk governance

The IAIS Insurance Core Principles online resource places investment quality, asset-liability interdependence and information about complex pooled instruments within prudential investment supervision. Its climate application guidance describes the relevance of physical and transition risks to credit, market and other prudential exposures. Supervisory responsibilities differ across jurisdictions, but the international scientific requirement remains comparable: claims about an insurer's resilience must specify what assets are held, how they are valued and which liabilities they support.

Governance questions include whether a board receives material scenario results; whether risk limits are defined by issuer, exposure and scenario; whether risk, investment, actuarial and finance functions use consistent assumptions; and how exceptions, model changes and data gaps are escalated. The board does not need a single risk score that claims to settle uncertainty; it needs evidence of material vulnerabilities and the available management responses.

EU Solvency II chronology

Article 132 is a central existing EU investment rule. The broader Solvency II revision under Directive (EU) 2025/2 has an application date of January 30, 2027, as confirmed in EIOPA's Solvency II regulatory overview. Thus a November 2026 legal analysis must not state that every revised 2027 requirement already applies to an insurer. The July 2026 EIOPA implementation announcement identifies guidelines and draft technical standards supporting the new framework. Their precise legal character and implementation dates need differentiation.

A portfolio stress report should present the current applicable position, the prospective requirements and the economic risk assessment as distinct categories. An international research dossier cannot determine an individual insurer's regulatory capital requirement without jurisdiction, authorization status, asset classification, model approval, liability features and the rules legally in force at the valuation date.

IFRS reporting and disclosure

The IFRS Foundation's IFRS S2 implementation materials locate climate-related disclosure within an established reporting architecture. The IFRS S2 standard overview addresses risks and opportunities reasonably expected to affect cash flows, access to finance or cost of capital, including physical and transition risks. Its effective date at the international standard level does not alone establish mandatory domestic adoption.

A useful disclosure links stated exposures and transition assumptions to material financial sensitivity. It identifies scenario horizon, limitations of emissions data, sources of valuation inputs, material concentrations and the boundaries of the reporting entity. It separates gross financed emissions or carbon intensity from estimated financial loss and explains changes in methodology. Disclosure that reports a smooth emissions trajectory without examining stressed cash-flow adequacy may satisfy a communication objective while leaving the investment risk problem unresolved.

The archived TCFD disclosure recommendations organize information around governance, strategy, risk management, and metrics and targets, and treat scenario analysis as exploration of hypothetical conditions. The TCFD was disbanded after its 2023 status report, with monitoring transferred to the IFRS Foundation. The recommendations remain historically influential but should not be described as an institution issuing current requirements in 2026.

Scholarly validation and model risk

A credible supervisory or academic analysis must be able to reproduce the mapping from securities to issuers and from scenario drivers to valuation outputs. At minimum, it should retain a position-date file, a source vintage, issuer classifications, a record of expert overrides, shocked cash flows or spreads, interest-rate treatment, asset and liability valuation changes, a liquidity bridge, and sensitivity results. Data gaps need labels: missing issuer emissions, inadequate fund look-through, stale private-market appraisal, absent recovery assumptions or missing contract-level liability information.

Validation should include reconciliation of total holdings to financial statements, tests of scenario arithmetic, comparisons against alternative sector mappings, price-convention checks and independent examination of complex exposures. A favorable backtest of ordinary volatility is not sufficient evidence for a structural climate transition with limited historical analogues. Where an issuer cash-flow adjustment is inferred from sector research, analysts should disclose the extrapolation and test whether its magnitude is consistent with known margins, operating expenditure and debt service.

Discrepancies across models are informative. An issuer-level discounted cash-flow analysis, a market-implied spread approach and a broad supervisory haircut method may disagree because they use different baseline information and stressed dynamics. Rather than averaging them without explanation, research should identify which component generates the difference. The objective is disciplined comparison of conditional results.

Evidence interpretation and comparative institutional limits

The IAIS 2025 climate paper, EIOPA stress testing and the NGFS scenarios have different institutional functions. The first describes supervisory application of risk principles, the second provides insurance-focused stress-test methodology, and the third supplies conditional macroeconomic and transition pathways. IFRS 9, IFRS 17 and IFRS S2 organize recognition, measurement and disclosure within applicable accounting or sustainability frameworks. They cannot replace one another.

The research must also differentiate statistical evidence from regulatory obligations. The carbon premium described by Bolton and Kacperczyk is an empirical finding in a particular stock-market dataset. It does not set the capital charge for an insurer. An EIOPA sensitivity estimate is a conditional result based on specified holdings and assumptions. It does not guarantee an investment return. An NGFS pathway is a conditional input. Only by preserving these boundaries can the study support reliable insurer decisions and scientific inference.

Integrated Findings and Research Limits

The synthesis establishes that the materiality of transition risk depends jointly on exposed cash flows, security-specific loss mechanisms and the insurer's balance-sheet structure. A change in company emissions is one dimension of economic adaptation, while the value of a bond or equity depends on the conversion of those changes into future obligations, profits and discount rates. Financial losses may be concentrated in small asset sleeves even when sector exposure is limited as a percentage of total assets.

It also establishes an ordering principle within ordinary analysis: first identify and document the relevant change in the economic baseline, next calculate consistent issuer and instrument values, then aggregate economically correlated exposures, and finally assess liability, liquidity and capital consequences. This is the logical sequencing of known valuation and insurance balance-sheet methods, not a newly introduced theoretical doctrine.

A third finding concerns data granularity. Asset-class allocations cannot substitute for issuer and security information, while issuer-level emissions do not substitute for a cash-flow model. Look-through and legal ownership are especially consequential for private credit, structured instruments and fund investments. Where complete data are unavailable, transparency about omitted exposures is more reliable than the assignment of artificially precise climate probabilities.

A fourth finding concerns temporal consistency. Multidecade transition pathways, immediate mark-to-market shocks, near-term liquidity needs and the duration of insurance claims operate on different clocks. A meaningful insurer scenario must connect these clocks through explicit dynamic assumptions. A fifth finding concerns institutional interpretation: IAIS, EU prudential law and IFRS reporting address different aspects of the problem, and the effective dates and national adoption of requirements cannot be collapsed into a universal legal statement.

The synthetic examples demonstrate economic leverage and valuation sensitivity under their stated assumptions. They do not estimate actual insurer losses, establish scenario probabilities, derive regulatory SCRs or attribute observed market moves causally to climate policy. Peer-reviewed pricing evidence is treated as supportive context with market, country and period limitations. Primary institutional pages and available publisher abstracts were examined; proprietary supervisory data and paywalled full-text empirical replications were not obtained. This limitation matters particularly for transferring empirical coefficients across jurisdictions.

Further research requires matched security-holding and contractual liability datasets, transparent issuer transition-cost models, alternative macrofinancial scenario assumptions, dynamic default and recovery specifications, realistic funding and collateral simulations and independent evidence on private-asset market values. Such work can evaluate model calibration and the realized performance of different risk-management responses. It also offers a path from conceptual scenario consistency to empirically testable insurer resilience.

Conclusion

Climate transition risk in insurer investment portfolios is a financially traceable interaction between structural economic change, issuer vulnerability, security valuation and insurance obligations. Policies, technologies and demand alter operating cash flows and financing conditions. Credit spreads, default expectations, equity valuations, sovereign risk, private-asset appraisals and liquidity then transmit those alterations into the insurer's holdings. The ultimate prudential effect depends on the liabilities supported, the timing of required cash payments and the rules under which capital and accounting values are measured.

The combined evidence supports a firm methodological conclusion. Exposure screening is necessary for discovering possible concentrations, while quantified risk analysis requires issuer-specific transmission, coherent scenarios, independently checked valuation assumptions and an explicit asset-liability reconciliation. Insurer-wide financial resilience cannot be inferred from carbon intensity, sector holdings or isolated asset haircuts. The conditional balance-sheet calculation shows how an investment loss of 3.92% of starting assets can accompany a 66.25% reduction in initial economic surplus under a separately stated liability shock. It also shows why the capital interpretation changes when the liability shock changes.

International standards reinforce this conclusion from complementary perspectives: IAIS supervisory principles address prudent investment and risk governance; EIOPA provides insurer-specific sensitivity and stress-testing research within the EU framework; NGFS supplies forward-looking pathways with explicit limitations; and IFRS standards govern particular accounting and disclosure questions in adopting jurisdictions. The result is an original, critical analytical synthesis of established research and standards, with scenario assumptions and evidentiary limits made explicit. It introduces no proprietary predictive claim or new authorial theory.

References