RESEARCH DOSSIER
Pension Portfolio Sustainability Risk and Long-Horizon Investment
Abstract. Pension portfolio sustainability risk is the possibility that long-run environmental, social, governance, economic, demographic, or institutional changes impair the financial capacity to deliver retirement benefits when due. The decisive unit of analysis is the pension entitlement and its risk-bearing institution. Defined-benefit promises depend on assets, discounted liabilities, contributions, sponsor strength, and possible protection arrangements. Defined-contribution outcomes depend more directly on investment accounts, contribution histories, inflation, withdrawals, and retirement timing. This monograph integrates primary international pension statistics and standards, jurisdictional pension rules, corporate disclosure frameworks, physical and transition climate evidence, supervisory scenario exercises, and observed liquidity crises. Its analytical contribution is an explicit connection between financially material sustainability exposures, long-horizon security cash flows, pension liabilities, short-dated settlement constraints, and member-level retirement outcomes. Simplified sensitivity derivations demonstrate how improving reported funding can coexist with declining asset values and urgent collateral calls. Source criticism identifies a material retraction affecting selected chronic physical-damage estimates in NGFS Phase V scenarios. The result is an evidence-based comparative study with documented assumptions and unresolved empirical questions, without creation of a proprietary theory or doctrine.
Keywords: Pension portfolio sustainability; long-horizon investment; financial materiality; defined-benefit pensions; defined-contribution pensions; asset-liability management; longevity; sponsor covenant; climate physical risk; transition risk; fiduciary duty; investment governance; ESG disclosure; duration; liquidity; liability-driven investment; pension funding; retirement income.
Introduction
Pension finance turns current contributions and accumulated capital into future retirement income. The OECD's 2026 pension-market account estimates OECD pension assets at USD 78 trillion at year-end 2025, including approximately USD 7.3 trillion in public pay-as-you-go reserve assets. The aggregate reflects heterogeneous institutions and accounting boundaries, so its magnitude is evidence of market scale rather than a measure of retirement security, climate vulnerability, or available liquid collateral.
The OECD analysis of sustainability integration reports some sustainability-risk integration requirements for pension or insurance investors in 41 of the 53 jurisdictions surveyed. This is an inventory of different obligations, not evidence that identical laws apply in each jurisdiction. Interview-based observations identify investment practices and perceived obstacles; they do not independently estimate causality between sustainability integration and realized pension investment performance.
The research questions are threefold. Which sustainability-related risks are financially material to diversified pension portfolios supporting multi-decade obligations? How do mandates, data gaps, liabilities, sponsor support, and liquidity transform a security-level hazard into retirement-income risk? Which official legal, financial, and empirical records justify conclusions across differing jurisdictions and pension designs, and how stable are those conclusions under alternative assumptions? Primary institutional sources, a peer-reviewed asset-pricing study, actuarial identities, and historical case evidence are assessed within their own evidentiary limits.
Research Foundations and Institutional Genealogy
Financial promises define the relevant risk
A defined-benefit pension arrangement promises benefits specified by a governing instrument, subject to its own service, indexation, survivor, vesting, amendment, and insolvency rules. The related obligations require projection of benefit cash flows and selection of mortality, inflation, and discounting assumptions. IAS 19 Employee Benefits distinguishes defined-benefit and defined-contribution obligations by their economic substance, but employer accounting measurements can differ from regulatory funding valuations, pension-trust valuation rules, and insurer buyout prices.
A defined-contribution arrangement generally accumulates contributions and returns in member-related accounts, while the final retirement benefit depends on savings history, fees, investment outcomes, access rules, annuitization, and withdrawals. The OECD recommendation on DC design treats default strategies, contribution incentives, and payout arrangements as interacting design considerations. A guaranteed minimum benefit, collective risk-sharing provision, or acquired insurance annuity can shift exposures to other institutions; each risk allocation must be traced to the actual contract.
The pension sponsor is a separate economic risk bearer whose funding capacity may be correlated with investment assets. A pension-protection institution may offer specified statutory security; an insurer may accept a contractual annuity obligation after a buyout; a trustee and external asset manager carry governance and operational responsibilities. The presence of these parties does not make their assets or obligations interchangeable. The economic meaning of portfolio sustainability depends on which entity bears the loss and whether that loss can prevent timely benefits.
Historical evolution of sustainability integration
The OECD Core Principles of Private Pension Regulation establish international benchmarks for governance, benefit security, prudence, and appropriate risk diversification. The 2019 IOPS ESG guidelines provide nonbinding guidance to pension supervisors on financially relevant environmental, social, and governance factors. Their 2025 implementation note makes financial materiality and proportional costs central: obtaining information has value when it can change a material investment or risk-management decision.
Under Article 19 of Directive (EU) 2016/2341, occupational pension institutions within the directive's scope invest in the long-term interests of beneficiaries under the prudent-person principle; consideration of long-term ESG effects is permitted. Applicable national transposition, pension governance, and investment law determine operative requirements. In the United Kingdom, occupational-scheme climate guidance details climate governance and reporting duties for specified schemes beginning in October 2021. These instruments establish different legal obligations, and neither can be projected unchanged onto every global pension plan.
Three meanings that require distinct measures
Portfolio financial sustainability describes the continuing capacity to fund a specified retirement objective. Sustainability-related financial risks arise when environmental, social, or governance conditions change cash-flow expectations, asset prices, financing costs, or liabilities. Sustainability-targeted investment identifies an investor's choice of exclusions, allocations, engagement, or impact objectives. The concepts interact, but a security's sustainability label does not by itself demonstrate a favorable risk-adjusted return, an adequate pension contribution, or a reliable source of collateral.
The OECD's 2026 investor research distinguishes financial integration, thematic investment, exclusions, and preferences-driven choices. A transition project can offer new cash flows while exposing a fund to technology, financing, and concentration losses. A conventional security can have credible but underreported exposure to physical damage or governance failures. Classification is an input to valuation, not a substitute for it.
Sustainable Investment Risk Identification
Physical exposure and financial transmission
The IPCC AR6 Synthesis Report documents climate hazards, their physical consequences, and uncertainty about future developments. For a pension fund, exposure becomes financially material through specific issuers and instruments: a flood may impair a factory, reduce earnings, alter insurance costs, decrease tax revenues, or damage collateral supporting a bond. Adaptation measures, geographical diversification, pricing power, and insurance coverage modify these transmissions. A hazard classification is therefore not directly a bond default probability or an investment-return forecast.
Transition shocks emerge from regulation, energy prices, changing technology, consumer demand, investment requirements, and repricing. The IEA World Energy Outlook 2025 compares distinct policy and energy-system pathways with different assumptions. Such pathways are conditional analytical scenarios. The financial shock to an asset depends on unexpected changes relative to market prices, rather than on emissions or policy exposure considered independently of valuation.
Social and governance failures can produce litigation, impaired financial statements, labor disruption, conflicts of interest, and weaker issuer repayment capacity. They also affect pension schemes internally when delegation, sponsor oversight, custody, and incentive systems fail. A climate transition can weaken a sponsor's profitability at the same time that its pension assets lose value. Separate exposures can thus become correlated precisely when supplementary contributions are most necessary.
Asset-level identification and double counting
A defensible risk record identifies issuer, security, parent-company structure, operating sites, sector, currency, capital structure, geographic and contractual dependencies, and data date. It then states the modeled pathway from hazard to net operating cash flow, discount rate, default distribution, or terminal value. Dietz and colleagues' climate value-at-risk study illustrates a global asset valuation approach under model assumptions. It supplies no observed return forecast for a specified pension trust without an independent holdings map and calibration.
Expected climate losses may already be partially embedded in market yields, credit spreads, property values, and insurance costs. Adding a full unconditional model loss on top of an already adjusted price could double count risk. Conversely, backward-looking market parameters can omit new regulatory or physical hazards and changes in dependence across issuers. Exposure analysis must separate expected fundamentals, existing market pricing, and incremental information shock.
Multiple time horizons
A fund can face collateral payment within days, benefit payments each month, bond maturities over decades, and obligations to retirees whose expected lifetimes stretch further. Let C_t be expected cash outflow at date t and d_t its purpose-consistent discount factor. The simplified liability is L_0=Σ_t E[C_t]d_t. This identity is conditional on the valuation measure, discounting convention, survivor probabilities, and contractual benefit rules. Asset cash flows X_{i,t} and their liquidity determine whether those payments can be met.
The NGFS short-term climate scenarios explore channels extending to 2030, while its long-term scenarios describe alternative pathways spanning decades. Combining horizons can reveal both immediate repricing and accumulated physical losses. The models are scenarios rather than probability-calibrated forecasts. Long investment horizons can make cumulative hazards significant while also expanding uncertainty and the possibility of unmodeled events.
Portfolio Mandates and Data Gaps
Risk appetite, benchmark, and fiduciary design
The investment mandate limits admissible assets, concentration, tracking error, currency exposure, duration, liquidity, and delegation. A sustainability-led strategy must be compared with a feasible alternative under the same retirement objective, liability profile, benchmark, and fees. Excluding a high-emissions sector may alter country and sector weights, factor returns, and risk diversification. The OECD's portfolio construction findings explicitly discuss concentrated national markets, exclusion constraints, benchmark effects, and different investment policies. These observations motivate careful attribution, not a presumption that any one approach consistently outperforms.
A trustee's decision should state the financial mechanism addressed, the investment alternatives, expected risk reduction, transaction and data costs, benchmark consequences, implementation authority, and a rule for revisiting the decision. Delegated managers rewarded on short evaluation periods may pursue strategies misaligned with decades-long benefit horizons. Conversely, invoking long horizons without sufficient collateral procedures can create acute vulnerabilities. Governance quality depends on converting evidence into an implementable response.
Disclosure does not eliminate uncertainty
IFRS S1 disclosure requirements address sustainability-related matters expected to affect a reporting company's prospects; IFRS S2 climate disclosures address climate-specific information. Their global promulgation differs from binding jurisdictional adoption, issuer eligibility, actual compliance, and external assurance. Pension asset owners cannot infer universal data coverage merely because an international issuer standard exists.
Unlisted companies, sovereign assets, pooled vehicles, private infrastructure, and securities with complex ownership structures may have sparse emissions, asset-location, and transition data. The ECB's 2025 portfolio disclosure identifies concrete problems of emissions coverage and comparability across instruments, including value-chain measurement. Pension analysts should report the market-value share of assets covered, the observation date, model-estimated versus reported records, instrument and sector omissions, and uncertainty induced by incomplete holdings.
For illustration, weighted average carbon intensity can be expressed WACI=Σ_i w_i(E_i/R_i), with investment weights w_i, specified issuer emissions E_i, and revenue R_i. Reporting boundaries, currency, Scope definitions, consolidation and denominator availability are essential. WACI is a characterization of carbon intensity under an accounting convention, not a default frequency or a pension deficit probability. Missingness may correlate with geography and issuer size, making naive interpolation or complete-case averages biased.
Pension capacity and information costs
The IOPS 2025 proportionality guidance recognizes that risk assessment must reflect materiality and resources. Large institutions may sustain specialized research and direct engagement, whereas smaller plans can face high per-member data expenses. Reliable assessment can therefore use material exposure thresholds, shared datasets, independent reviews of assumptions, and escalation procedures. The aim is auditable decision quality rather than maximum disclosure volume.
Long-Duration Funding Sensitivity and Comparative Evidence
A transparent asset-liability derivation
For a defined-benefit scheme with assets A>0 and valuation liabilities L>0, define the simple funding ratio F=A/L. For small parallel yield shifts Δy, first-order modified-duration approximations yield ΔA/A≈−D_AΔy and ΔL/L≈−D_LΔy. Thus Δln F≈(D_L−D_A)Δy. If liabilities have greater duration than assets, rising discount yields can improve the reported ratio despite reducing asset market value. This is a local approximation: it neglects convexity, changing expected cash flows, credit spreads, inflation, derivative collateral and the precise legal valuation basis.
A numerical example is deliberately hypothetical. With A=L=100, D_A=12, D_L=18 and a +1 percentage-point yield movement, first-order assets become 88 and liabilities 82, leaving F≈1.073. A higher reported ratio alone demonstrates neither increased liquidity nor improved benefit purchasing power. Any conclusion about actual members requires plan-specific valuations and enforceable obligations.
The 2022 British LDI episode
The Bank of England's December 2022 financial stability analysis describes pressure on leveraged liability-driven investment strategies after abrupt UK gilt yield rises. Hedge and derivative positions generated collateral demands; asset sales into strained markets contributed to financial stress. Liability valuations could fall with higher rates while the investment vehicles faced immediate requirements for cash or high-quality collateral.
Subsequent UK oversight of pension LDI addresses liquidity preparedness and risk management. The episode supports a general mechanism: matching interest-rate sensitivities need not guarantee settlement liquidity. Its empirical magnitudes and institutional features remain specific to UK gilt markets, pension derivatives, and the conditions of 2022.
Accordingly, a meaningful stress has at least three outputs: asset and liability valuations, realizable cash after payments and collateral, and sponsor contribution capacity under the same state. Liquidity buffers, time to mobilize collateral, derivative documentation, and forced-sale costs can dominate short-term outcomes even for a pension scheme with a favorable accounting funding movement.
EIOPA climate transition stress
The EIOPA 2022 occupational-pension climate test models a sudden disorderly transition. The official results report a 12.9% asset value decrease, approximately €255 billion, and a 2.9-percentage-point deterioration in aggregate funding ratio after changes in liabilities and other effects. These are conditional results of an imposed stress and a specified institutional sample, not realized losses or a statistical probability of future transition.
An independent illustration with initial A=L=100, a hypothetical 12.9% asset loss and an assumed 10% reduction in measured liabilities gives F=87.1/90≈0.968. The 10% liability reduction is chosen solely to explain the algebra; it is not EIOPA's reported liability figure and is not a reproduction of its aggregate finding. Both denominator changes and the ability to pay benefits must be reported before an asset markdown can be converted into a funding conclusion.
Defined-contribution losses affect member account wealth according to investment allocation and any guarantee or collective mechanism. There is no automatic DB discount-liability offset to restore a DC saver. The time of loss relative to retirement, subsequent contributions, withdrawal design, and insurance annuitization shape the final economic result.
Inflation, longevity, and income adequacy
A nominal fixed pension and an indexed pension respond differently to inflation. Higher inflation can increase indexed obligations and reduce purchasing power of fixed benefits, while assets react according to their real cash flows, nominal coupons, refinancing, and currency. Longevity improvements extend payment duration and may increase benefit costs. Real return and cash-flow sequence are therefore crucial to sustainability.
A simple DC terminal real wealth expression is W_T=Σ_t c_t Π_{u=t+1}^T [(1+r_u)/(1+π_u)], where contributions c_t are expressed consistently, r_u are nominal net investment returns and π_u inflation; fees, withdrawals and mortality-contingent conversions require appropriate treatment. Two return sequences with the same simple average may deliver different retirement outcomes when contributions or withdrawals occur along the path.
The OECD Pensions Outlook 2026 distinguishes contributions, investment income, and benefit payments. Its earlier Pension Markets in Focus 2025 uses an earlier reporting period; different asset denominators or currency translations must be checked before constructing growth comparisons.
Comparative governance across institutions
EU occupational pension investment under IORP II, UK trust governance, DC member investments, public reserve operations, and insurer annuity portfolios differ in investment powers and liabilities. A member's protection depends on contribution laws, pension contract terms, supervisor powers, insolvency procedures, pension guarantee systems, and any insurer transaction. The legal distinctions determine whether a stress falls on the employer, members, guarantor, taxpayers, or insurer. International comparison should identify plan type, effective regulatory date, reporting basis, funding rules, demographic maturity, and local financial-market depth.
Robustness, Independent Validation, and Source Criticism
Reproducible analysis and falsification
Each assessment should begin with an identified valuation date, complete or sufficiently described holdings, pension benefit rules, discounting basis, inflation assumptions, participant cohort, currency, sponsor covenant, collateral contracts, and scenario vintage. Analyses of alternative allocations must preserve baseline risk budget and fees. Otherwise, an observed difference may reflect sector selection, factor returns, maturity, or transaction costs instead of the sustainability metric under study.
Sensitivity checks should vary climate hazard intensity and adaptation, transition timing, issuer repricing, funding discounts, sponsor credit, longevity improvement, inflation, collateral requirements, liquidation costs, and data missingness. A conditional stress is not a probability forecast. Empirical backtests may support nearer-term factors, but future climate states decades away remain beyond complete historical validation.
The following table describes diagnostic experiments, not measured portfolio losses:
| Exposure | Illustrative test | Output | Main confounding risk |
|---|---|---|---|
| Yield curve | Parallel +100 bps | Assets, liabilities, funding, collateral | Convexity and liquidity |
| Transition | Issuer spread/earnings shock | Portfolio value and sponsor covenant | Existing priced exposure |
| Physical climate | Alternate site hazards/adaptation | Issuer cash flows | Exposure mapping errors |
| Inflation | +200 bps expected inflation | Indexed obligations and real benefits | Nominal/real inconsistency |
| Longevity | Cohort mortality shift | Liability cost and payment duration | Mismatched population |
| Liquidity | Stressed sales and collateral | Days of cash coverage | Book value treated as cash |
| Data gaps | Reweight missing issuers | Exposure interval | Nonrandom missingness |
| Mandate | Comparable alternative portfolio | Net returns, fees, tracking | Sector and factor confounding |
A well-supported conclusion states whether the ranking of alternative investments remains stable under plausible perturbations. Where results are unstable, that uncertainty is a substantive finding. Data documentation should permit independent replication of any calculation that is publicly testable.
NGFS physical-risk correction
A material source-integrity restriction attaches to selected outputs of the NGFS long-term Phase V scenarios. In its December 2025 methodological statement, NGFS reports the retraction of Kotz et al. (2024), whose damage function underpinned some chronic physical-loss estimates. The NGFS distinguishes affected output subsets from unaffected variables and states that short-term scenario results are not affected by that study.
An analyst using Phase V must record which variables depend on that retracted source and compare alternative damage functions, scenario vintages, or unaffected components. The retraction does not negate observed climate hazards or all transition modeling. It lowers confidence in particular derived estimates and prevents their use as unqualified forecasts. The NGFS short-term scenarios remain an independent scenario family whose assumptions and limits also require disclosure.
Evidence hierarchy and inference
Different records support different conclusions. Primary legislation establishes legal obligations in scope. A supervisory stress test demonstrates a modeled conditional result. A financial stability report describes an observed market episode. A scientific climate synthesis establishes hazard evidence. A quantitative valuation study illustrates an assumption-dependent model. OECD and investor interviews describe reporting, behaviors and institution-level perspectives. Combining them requires explicit causal links; no one type can automatically stand in for another.
The OECD regulatory inventory and interviews establish heterogeneous current practice, not that implementation caused higher pension returns. The EIOPA 2022 test establishes vulnerability under one assumed shock, not the frequency of that shock. The Bank of England's LDI evidence establishes that collateral mechanics can create forced-sale dynamics, not that every duration hedge is destabilizing.
Confidential holdings, private market asset valuations, and plan-level joint observations of portfolio climate risk and beneficiary outcomes remain major data limitations. Statistical causal claims would require well-characterized counterfactual portfolios, comparable sponsor obligations, and robust controls for asset allocation, fund size, jurisdiction, manager incentives, and market pricing. A model's complexity is not a substitute for externally observable validation.
Integrated Analytical Synthesis and Scientific Agenda
The central synthesis is entitlement-dependent financial analysis. Pension portfolio sustainability is determined through the interaction of investment assets, projected benefit payments, sponsor and insurer commitments, realizable liquidity, and governing contracts. A single loss event may improve one reported funding metric through liability repricing while worsening asset values and near-term settlement. A parallel shock in a DC plan may appear primarily as lower individual retirement wealth. The material differences are generated by institutional risk allocation.
The second synthesis concerns temporal consistency. Financial sustainability assessments must join future climate and demographic scenarios to actual cash settlement dates. Long-term hazards matter through security and sponsor fundamentals; short-term market dysfunction can crystallize losses even for portfolios whose expected long-run earnings remain substantial. Interest-rate duration, benefit horizon, and collateral horizon require distinct measurement.
The third synthesis concerns evidence quality. Decision-useful integration of sustainability risk requires verifiable mechanisms, provenance of source observations, model sensitivity and a feasible governing response. Aggregate carbon intensity, taxonomy labels, and reporting-framework adoption can help organize information, but cannot independently establish financial resilience or pension adequacy.
Future research should obtain linked security-holdings, employer-covenant, liability and benefit-outcome data for comparable funds; test matched portfolios net of fees and sector exposures; estimate liquidity and funding jointly under transition and physical shocks; and compare cohort-level real-income outcomes in different DB/DC designs. Corrected and version-controlled scenario data are a prerequisite for reproducible climate-financial stress research. Comparative investigations should trace effective laws, prudent-person discretion, pension-protection systems, and sector development rather than treating surveyed jurisdictions as interchangeable.
These conclusions provide an original analytical synthesis of the assembled evidence and explicit mathematical sensitivities. They do not constitute a new named model, theory, principle, or doctrine.
Conclusion
Pension portfolio sustainability risk is a financial and institutional condition measured by the ability to sustain retirement entitlements across long time horizons and adverse states. It is shaped by climate hazards, transition changes, governance quality, demographics, inflation, market prices, legal claims, and liquidity requirements. An adequate assessment links each risk to the specific issuer, pension promise, fund, and risk-bearing institution that can transmit its consequence.
Official international standards support proportionate analysis of financially material risks within pension mandates. Supervisory and market evidence show that asset losses, liability repricing and collateral liquidity can pull financial outcomes in different directions. Scenario calculations are conditional, and retraction-sensitive climate estimates demand explicit qualification. The durable research conclusion is methodological: source-verified exposure-to-benefit transmission, independent sensitivity testing, and accurate pension entitlement classification provide a stronger foundation for long-horizon investment decisions than investment labels or unqualified aggregate figures. Further causal and cohort-based empirical testing remains warranted.
References
Bank of England. (2022). Financial Stability Report: December 2022. The Resilience of Liability-Driven Investment Funds. London: Bank of England. https://www.bankofengland.co.uk/financial-stability-report/2022/december-2022
Dietz, S., Bowen, A., Dixon, C., and Gradwell, P. (2016). Climate Value at Risk of Global Financial Assets. Nature Climate Change 6: 676–679. https://doi.org/10.1038/nclimate2972
European Central Bank. (2025). ECB Adds Indicator of Nature Loss in Climate-Related Financial Disclosures as Portfolio Emissions Continue to Decline. June 12. https://www.ecb.europa.eu/press/pr/date/2025/html/ecb.pr250612~484d529368.en.html
European Insurance and Occupational Pensions Authority. (2022). Climate Stress Test for the Occupational Pensions Sector 2022. December 13. https://www.eiopa.europa.eu/browse/financial-stability/occupational-pensions-stress-test/climate-stress-test-occupational-pensions-sector-2022_en
European Parliament and Council. (2016). Directive (EU) 2016/2341 on the Activities and Supervision of Institutions for Occupational Retirement Provision (IORPs) (Recast). Official Journal of the European Union L 354. https://eur-lex.europa.eu/eli/dir/2016/2341/oj/eng
International Accounting Standards Board. IAS 19 Employee Benefits. IFRS Foundation, official standard overview, accessed November 15, 2026. https://www.ifrs.org/issued-standards/list-of-standards/ias-19-employee-benefits/
International Energy Agency. (2025). World Energy Outlook 2025: Executive Summary. Paris: IEA. https://www.iea.org/reports/world-energy-outlook-2025/executive-summary
International Organisation of Pension Supervisors. (2019). Supervisory Guidelines on the Integration of ESG Factors in the Investment and Risk Management of Pension Funds. Financial Stability Board listing, October 22. https://www.fsb.org/2019/10/supervisory-guidelines-on-the-integration-of-esg-factors-in-the-investment-and-risk-management-of-pension-funds/
International Organisation of Pension Supervisors. (2025). Explanatory Notes for IOPS ESG Guidelines: Implementation Note. May. https://www.oecd.org/content/dam/iops/en/iops-principles-and-guidelines/Implementation-note-IOPS-ESG-guidelines-May-2025.pdf
Intergovernmental Panel on Climate Change. (2023). Summary for Policymakers. In Climate Change 2023: Synthesis Report, edited by H. Lee and J. Romero. Geneva: IPCC. https://www.ipcc.ch/report/ar6/syr/summary-for-policymakers/
International Sustainability Standards Board. (2023). IFRS S1: General Requirements for Disclosure of Sustainability-related Financial Information. IFRS Foundation. https://www.ifrs.org/issued-standards/ifrs-sustainability-standards-navigator/ifrs-s1-general-requirements/
International Sustainability Standards Board. (2023). IFRS S2: Climate-related Disclosures. IFRS Foundation. https://www.ifrs.org/supporting-implementation/supporting-materials-for-ifrs-sustainability-disclosure-standards/ifrs-s2/
Network for Greening the Financial System. (2024). NGFS Climate Scenarios for Central Banks and Supervisors: Phase V. Official report and subsequent qualification. https://www.ngfs.net/en/publications-and-statistics/publications/ngfs-climate-scenarios-central-banks-and-supervisors-phase-v
Network for Greening the Financial System. (2025). NGFS Publishes the First Vintage of Short-Term Climate Scenarios. May 7. https://www.ngfs.net/en/press-release/ngfs-publishes-first-vintage-short-term-climate-scenarios
Network for Greening the Financial System. (2025). Statement Regarding Physical Risk Estimates in Phase V of NGFS Long-Term Scenarios. December 3. https://www.ngfs.net/en/press-release/statement-regarding-physical-risk-estimates-phase-v-ngfs-long-term-scenarios
OECD. (2016). OECD Core Principles of Private Pension Regulation. Paris: OECD Publishing. https://www.oecd.org/en/publications/oecd-core-principles-of-private-pension-regulation_c950bb36-en.html
OECD. (2025). Pension Markets in Focus 2025. Paris: OECD Publishing. https://www.oecd.org/en/publications/pension-markets-in-focus-2025_b095d0a0-en.html
OECD. (2026). Pension Markets in Focus. In OECD Pensions Outlook 2026: Are We Saving Enough for Retirement? Paris: OECD Publishing. https://www.oecd.org/en/publications/oecd-pensions-outlook-2026_a51aa775-en/full-report/pension-markets-in-focus_f12b3e30.html
OECD. (2026). Pension and Insurance Investors' Integration of Sustainability Risks into Investment Processes. In OECD Pensions Outlook 2026: Are We Saving Enough for Retirement? Paris: OECD Publishing. https://www.oecd.org/en/publications/oecd-pensions-outlook-2026_a51aa775-en/full-report/pension-and-insurance-investors-integration-of-sustainability-risks-into-investment-processes_7ac8f0b8.html
OECD Council. (2022). Recommendation of the Council for the Good Design of Defined Contribution Pension Plans. OECD/LEGAL/0467. https://legalinstruments.oecd.org/en/instruments/OECD-LEGAL-0467
The Pensions Regulator (United Kingdom). (2025). Market Oversight: How Well Pension Schemes Are Prepared for LDI Risk and What Trustees Need to Know. https://www.thepensionsregulator.gov.uk/en/document-library/research-and-analysis/market-oversight-how-well-pension-schemes-are-prepared-for-ldi-risk
UK Department for Work and Pensions. (2021). Governance and Reporting of Climate Change Risk: Guidance for Trustees of Occupational Schemes. https://www.gov.uk/government/publications/governance-and-reporting-of-climate-change-risk-guidance-for-trustees-of-occupational-schemes/governance-and-reporting-of-climate-change-risk-guidance-for-trustees-of-occupational-schemes