OECD Review on Aligning Finance with Climate Goals 2026

This chapter provides an overview of developments in approaches and metrics to track climate performance, opportunities, and risks in the financial sector, based on the analytical dimensions defined in (OECD, 2024[1]) and summarised in Box 4.1. It addresses selected dimensions identified as particularly pertinent to address assessment gaps in ongoing tracking efforts (Chapter 3) and to inform the further development of policy frameworks and disclosure requirements (Chapter 2). This chapter first dives into climate alignment and transition metrics for financial institutions (Section 4.1) and corporates, both listed (Section 4.2) and private (Section 4.3). It then discusses how to integrate dimensions relating to resilience to climate change in corporate climate assessments (Section 4.4). The chapter also zooms on approaches to assess sovereign bonds beyond those labelled as green (Section 4.5). Finally, the chapter reflects on using innovative solutions to fill data gaps across layers of the financial sector (Section 4.6).

Existing metrics have provided the foundation for climate target setting and engagement. Evidence highlights that banks still channel more financing to fossil fuels than low-carbon energy (see Chapter 3 on tracking). Moreover, over 80% of financial institutions globally have set some type of climate target and over 70% have engaged with clients and stakeholders on climate as of 2024 (Figure 4.2). These shares increased from 57% and 34% respectively in 2019. Such targets and engagements build on the increasing availability of emissions data. As of 2024, 23% of financial institutions disclose Scope 1 and 2 emissions and 19% Scope 3 emissions (OECD, 2026[2]).

Recent developments highlight a shift to rebalance tracking efforts beyond GHG-metrics to better support environmental and economic credibility. Aside from emissions-related information, voluntary metric frameworks and mandatory disclosure requirements also request information on portfolio composition, engagement, and strategy and governance (OECD, 2024[1]; OECD, 2023[5]). Financed emissions and other emissions-based metrics are necessary but insufficient as they can fail to accurately reflect real-economy change. Robust tracking efforts require complementary core metrics, transparent assumptions, reliable comparable data, and credible benchmarks. The incentive effects of broad classifications and binary alignment can push capital to divest from large emitters rather than supporting transition plans and capital expenditure shifts. Such decisions result in portfolio decarbonisation but emissions impacts in the real economy are uncertain. While metrics on GHG emissions capture outcomes or outputs, information points on portfolio composition, engagement and on strategy capture decisions, inputs, and actions (Figure 4.3).

Information on changes in input metrics relating to portfolio composition and engagement strengthens the credibility of GHG emissions reduction targets. While climate awareness and target setting in the financial sector increased, some research finds that financial institutions with net-zero emissions targets do not scale up green lending more than financial institutions without targets (Sastry, Verner and Ibanez, 2024[11]). Some evidence shows that banks that signed Principles of Responsible Investment (PRI) temporarily shortened the maturity of loans extended to borrowers in emissions-intensive sectors, but banks with voluntary climate commitments did not contribute to syndicated loan reallocation away from those sectors (University of California, Santa Cruz/Columbia University/Federal Reserve Bank of San Francisco, 2024[12]). While PRI signatories maintain lower portfolio carbon footprints in their holdings, they are not characterised by significant reductions in carbon footprint over time. However, the impact varies based on institutions’ investment approaches (Allahdadi, 2025[13]), suggesting practical frameworks with measurable decision inputs could strengthen information signals.

The emergence of transition planning contributed to the demand for and the development of action-oriented metrics. Transition plans set out an entity’s strategic response to risks and opportunities that emerge from the impacts of climate change and the transition to a low-emission economy (NGFS, 2024[14]). Frameworks such as the Transition Plan Taskforce Disclosure Framework, which cover both financial and non-financial institutions, apply three guiding principles of ambition, action, and accountability (Transition Plan Taskforce, 2023[15]). Compared to initial climate-alignment frameworks primarily focused on measuring progress towards targets, this more holistic approach places greater emphasis on actions, for example in relation to business operations, products and services, and financial planning. Such actions are captured by portfolio composition metrics for financial institutions (Figure 4.3), and composition of activities and related investments for listed corporates (Figure 4.5).

Assessing how financial institutions are capturing climate transition opportunities can be based on a selection of granular metrics relating to their portfolios. Assessing portfolio composition helps stakeholders track changes in a financial institution’s investment or lending approach and can reveal whether capital is being redirected toward green businesses and climate solutions (OECD, 2023[5]). Greenwashing risks arise when granular data are over-aggregated, input assumptions (e.g. scenario choice) are opaque, or assessments are oversimplified to a single binary “aligned/not aligned” metric (Noels and Jachnik, 2022[16]). Aggregate transition metrics should be presented in dashboards that preserve some information on sectoral and activity break downs and on rates of change in metrics. Any headline figures should be paired with appropriate portfolio coverage and data quality indicators. Climate transition dashboards need to combine backward- and forward-looking metrics and qualitative evidence. However, a key challenge for aggregate metrics remains that much of the data available at asset level remains primarily qualitative.

Financial institutions could benefit from more clearly defined metrics tailored to different actors, activities, asset classes, and sectors. For example, the SBTi Financial Institutions Net-Zero Standard proposes five activity segments: lending, asset owner investing, asset management investing, insurance underwriting, and capital markets (SBTi, 2025[17]). Underlying decision contexts differ across financial actors, calling for tailored climate transition metrics. For instance, transition metrics for banking tend to be more anchored in sectoral approaches (UNEP FI, 2025[18]), while insurers place greater weight on climate risk concentrations (IAIS, 2025[19]).

Metrics that capture climate transition opportunities, such as investments and lending in climate solutions, have emerged across financial sector activities but need to be further developed. Existing frameworks demand that entities disclose information on climate-related risks and opportunities (OECD, 2023[5]). For example, investors can disclose amounts invested in green bonds for fixed income, and banks on volumes of lending and underwriting for climate solutions (Dai et al., 2023[20]; TPI, 2025[21]). However, some climate transition opportunity metrics are difficult to compare and others need to mature.

Climate transition opportunity metrics for lending and capital markets can reflect shifting business models and how financing is being directed toward transition-enabling uses. In lending activities, metrics on the amount or share of lending, commitments, project finance, trade finance, or loan-book exposure classified as transition finance could be considered. Other metrics include low-carbon-to-fossil financing ratios (BloombergNEF, 2025[22]; TPI, 2025[21]). For capital-markets activities, analogous metrics could track the amount or share of arranged or underwritten debt, equity, or syndicated-loan activity that qualifies as transition finance, the share directed to climate-solution issuers or use-of-proceeds instruments, and sustainable issuance volumes such as green, transition, or sustainability-linked instruments (Net-Zero Banking Alliance, 2024[23]). Metrics can be linked to regional taxonomies (as discussed in Chapter 2 on policies). Green revenue metrics track how business models are changing. For example, lenders generated roughly USD 3.7 billion of revenue from climate-related loans and bond underwriting in 2025, compared with about USD 2.9 billion from oil, gas and coal (Bloomberg, 2026[24]).

Climate transition metrics for asset owner and manager investing can capture green capital or revenue shares of portfolio companies. For asset owners, metrics can capture the amount or share of assets under management (AUM) invested in climate solutions or transition sectors, of assets aligned or aligning with net-zero emissions, and of residual carbon-intensive exposures (IIGCC, 2024[25]; TPT, 2024[26]). For asset managers, the same logic extends to their mandate. For example, disclosing the AUM committed to be managed in line with net zero and setting a quantitative objective for scaling up climate-solutions investment, which supports metrics such as the share of managed assets to which climate metrics and targets apply and the share of AUM invested in transition-enabling products and funds (TPT, 2024[27]). A Weighted Average Green Revenue (WAGR) metric can be used to integrate climate solutions measurements into portfolio construction (Dai et al., 2023[20]).

Gaps in metric interoperability of disclosure initiatives across jurisdictions remain a costly challenge. Fragmentation of disclosure standards and taxonomies (as analysed in Chapter 2 on policies) raises costs, reduces comparability, and risks resulting in box-ticking exercises. For example, criteria in taxonomies as to what constitutes a climate solution, a climate enabling or a climate transition activity are not consistent across jurisdictions. Priorities for improving interoperability include greater mutual recognition between disclosure regimes, convergence on a core set of metrics, clear crosswalks between global guidance and regional taxonomies, and open-source methodologies. Further, interoperability and avoiding that disclosures are not repurposed inappropriately also requires clarifying the objective(s) that metrics are serving, including supporting an orderly transition, managing prudential risk, and/or climate policy alignment.

Compliance with disclosure requirements for financial institutions hinges on decision-useful data for high-impact sectors and filling data gaps for more opaque financial asset classes. Chapter 2 on policies indicates that financial institutions often face higher data demands than non-financial corporates. Improving data disclosure in the real economy is, however, critical, notably in relation to issuers’ transition plans with interim targets and capital expenditure details (see Section 4.2). More efforts are also needed to develop proxies where primary data are unavailable, such as for SMEs, private assets, parts of real estate and infrastructure (see Section 4.6). However, to avoid being a sole compliance exercise, tracking and disclosure needs to focus on fit-for-purpose data and credible metrics to help set incentives to transition.

Existing frameworks agree on categories of climate transition metrics for corporates but do not systematically refer to specific and quantitative metrics. Climate assessments of corporate-related asset classes beyond those identified as already green based on labels for corporate bonds and sectoral analysis for equity (see Chapter 3 on tracking) require a range of metrics to characterise a given entity. Similar to broad categories for financial institutions identified in OECD work (OECD, 2023[5]), metrics for non-financial companies are proposed along four types: (1) GHG emissions, (2) composition of activities and related financing and investments, (3) engagement, and (4) strategy and governance (Figure 4.5). The greatest convergence among frameworks is on GHG emissions metrics. Only few quantitative engagement, strategy and governance metrics are proposed by frameworks. For composition-of-activities and related financing and investment metrics, there is some convergence on the need for quantitative expenditure-related and sector-specific metrics, but other metrics are still maturing.

The decision usefulness of climate performance metrics depends on their use cases, characteristics, and ability to measure shifts and effects in the real economy. No climate metric is decision-useful in the abstract. Its usefulness depends on whether its characteristics fit the use case and whether it captures real shifts and effects in the economy rather than only stated ambition (OECD, 2024[1]; UNEP FI, 2022[29]). Different metrics capture different dimensions of change in the real economy. They are decision-useful only insofar as they help users understand effects on business models, strategy, financial positions, and performance (IFRS, 2023[30]).

Metrics on companies’ activities and related investments indicate how they are capturing climate transition opportunities and adapting their business models. As for financial institutions, climate transition metrics for non-financial corporates are increasingly focussing on “inputs” related to transition investments and corporate activities, in addition to GHG emissions outcomes. This trend is reinforced by action-oriented transition planning metrics in relation to business operations, products and services, and financial planning (Transition Plan Taskforce, 2023[15]). Information on changes in “input” metrics relating to corporate activities, investments and supply chain engagement can strengthen the credibility of GHG emissions reduction targets. Such complementary information supports robust tracking efforts (Box 3.1) and highlight how companies are capturing growth opportunities from the transition.

Activity-composition metrics can include research and development, expenditures, real assets, energy use, products and services, revenues, sectoral or other financial indicators. Research and development metrics include counts of green patents and green R&D investments over revenue (Figure 4.5). Expenditure metrics cover “green”, taxonomy-aligned, or carbon-intensive capital and operational expenditures. Real-asset metrics can capture the share of physical assets, production capacity, or reserves that are low-carbon, transition-enabling, or carbon-intensive. Energy-use metrics track electricity consumption renewable energy sourcing, or energy intensity, helping to reveal operational progress in decarbonisation. Products-and-services metrics can measure the share of output, sales, or business lines linked to climate solutions, transition-enabling activities, or high-emitting activities. Revenue-based and other sector-specific indicators can complement these measures to show how current activity and future investments are distributed across climate-relevant activities.

Different activity-composition metrics capture various dimensions of corporate climate transition and transformations (Figure 4.5). For example, metrics on green research and development reflect corporate innovation efforts and transition potential. However, companies can transition without in-house innovation. Expenditure metrics can signal ongoing changes in business models and operational priorities. However, they can be volatile with technology costs and implementation uncertainties. Further, planned expenditure does not always translate into realised transition outcomes. Low-carbon and carbon‑intensive real assets metrics connect emissions to core business activities and can reveal structural exposure in firms’ production systems. Low-carbon and carbon‑intensive revenue shares also connect emissions to core business activities. However, revenue may be volatile with macroeconomic conditions.

Available data to compute corporate climate transition metrics beyond GHG emissions remains sparse. Corporate emissions data have become relatively widely available. Over 40% of listed non-financial companies globally disclosed emissions data in 2024 (Figure 4.6). This is the case for both Scope 1 and 2 emissions. Over 30% of companies also disclosed Scope 3 emissions data. Similar shares can be observed for emissions intensity data. On the other hand, corporate data on action-oriented metrics remains limited. For most activity-composition and investment metrics less than 10% of companies disclose quantitative information. One exception is green capital expenditure data, which is available for around 50% of listed companies. Information disclosed on engagement is difficult to summarise in comparable quantitative metrics.

Major data gaps hinder the comparability of climate-related financial metrics and limit the ability of financial investors and policymakers, to evaluate transitioning business models. Globally, around half of listed companies provide some information on transition-related capital expenditures (Figure 4.6). Sectoral and graphical trends differ. For example, over 60% of Asian-Pacific listed companies do not provide information on climate-relevant capital expenditures. Data gaps are often bigger in high-impact sectors. Even in sectors where low-emissions investment opportunities are expanding (e.g. utilities, transport, manufacturing), green capital expenditures disclosure is far from universal. This indicates that, despite growing investor demand for forward-looking, investment-based alignment indicators, the availability, comparability, and granularity of capital expenditures data remain insufficient to assess whether corporate investment plans are consistent with climate-mitigation pathways.

More quantitative metrics are a priority for disclosure policies, especially to better track transition opportunities in emissions-intensive sectors and climate solutions. While policymakers need to ensure disclosure requirements are not overly complex and resource-intensive, more quantitative information is needed (OECD, 2026[2]). Much disclosure initiatives remain too qualitative. Backtracking and simplification of existing requirements will result in a further reduction of quantitative information. Empirical work to identify critical quantitative metrics can support an effective simplification of disclosure requirements. Moreover, simplification of disclosure requirements needs to acknowledge that the transition is asymmetrical. More detailed metrics are needed for companies in a limited number of high-impact sectors, while fewer metrics can be enough for other sectors. Further, requirements must be cost-effective and proportional, notably for smaller institutions, especially in emerging markets and developing economies (OECD, 2024[31]).

Gaps in metric interoperability of disclosure initiatives across jurisdictions remains a costly challenge. Fragmentation across standards and taxonomies raises costs, reduces comparability, and risks resulting in box-ticking exercises. For example, criteria in taxonomies as to what constitutes a climate solution, green or transition activity on are not consistent across jurisdictions. Priorities for improving interoperability include greater mutual recognition between disclosure regimes, convergence on a core set of metrics, clear crosswalks between global guidance and regional taxonomies, and open-source methodologies and public data repositories. Further, interoperability and avoiding that disclosures are not repurposed inappropriately also requires clarifying the objective(s) that metrics are serving, including supporting an orderly transition, managing prudential risk, and/or climate policy alignment.

A lack of transparency and disclosure requirements in private equity makes it difficult to understand the climate impact of private equity investment portfolios. Private equity firms are well positioned to capture climate transition opportunities. As discussed in Chapter 3 on tracking, private equity investment in low-carbon activities is catching up with fossil fuels. However, the existing lack of transparency a critical barrier in more comprehensively and credibly assess the role private equity plays in the low-carbon solutions and transition opportunities.

Initiatives have emerged that partly address the lack of climate transparency in private markets by assessing the energy portfolios of private equity firms. Some metrics applicable to listed corporates can apply to private equity (Private Equity Task Force, 2024[32]). Emissions metrics and targets provide incentives to reduce climate impacts in the real economy. Most large private equity managers are disclosing Scope 1, 2, and 3 emissions as of 2025 (Unwritten, 2025[33]). Analyses based on this metric revealed that the energy portfolios of 21 leading private equity firms invested over a trillion dollars in energy investments between 2010 and 2024. Fossil fuel companies typically represent two-thirds of energy companies in their portfolios (Duong et al., 2024[34]), adding up to an estimated 1.17 gigatons of annual GHG emissions (Duong et al., 2024[34]). Such analyses, however, remain a resource-intensive process.

A low carbon to fossil fuel energy supply investment ratio (ESIR) metric has emerged but remains data constrained. Calculating this ratio for private equity comes with data and methodological challenges, including having complete and consistent data on (energy) holdings of private funds, attributing energy supply capital expenditure by companies to funds, and classifying activities as climate (mis)aligned. Initial analysis testing this metric for over 1,000 private markets funds and almost 70,000 securities-focused ETFs and mutual funds finds that they tend to invest more in low-carbon than fossil fuel energy supply (BloombergNEF, 2025[35]). Capital-intensive renewable energy assets are particularly suited to infrastructure funds because of contracted revenues, high barriers to entry, and large physical assets. This demonstrates the role private equity could play in financing the transition.

Disclosure and data to compute other metrics for assessing the climate performance of private funds remain elusive. Consistent with metrics put forward by voluntary disclosure frameworks, these could include the disclosure of portfolio-wide emissions, science-based climate targets, transition plans metrics, and climate-relevant capital expenditure ratios. Considering a sample of large private funds, disclosure across these indicators is generally lacking (Duong et al., 2024[34]).

Efforts to increase climate resilience are gaining further momentum, yet existing climate-related frameworks remain less comprehensive and consistent in addressing adaptation than for climate mitigation (OECD, 2024[36]). The lack of a common language and standardised categories for adaptation and resilience metrics hinders comparability and coherent tracking (OECD-NGFS, 2024[37]). Although some jurisdictions have incorporated adaptation and resilience considerations into sustainable finance taxonomies and disclosure requirements (see Chapter 2 on policies), there remain gaps in quantitative metrics for outcome-based progress measurement (Noels et al., 2024[38]). As companies and investors make decisions based on a financial business case, these gaps limit their effectiveness in driving adaptation finance (NGFS, 2025[39]). Pilot analyses of investment portfolios, however, illustrate opportunities for making the business case for resilience investments based on applying emerging physical climate risk assessments methods (IIGCC, 2025[40]; IIGCC, 2025[41]).

Integrating adaptation to climate change in transition plans enables a holistic approach to climate resilience and transition-related investment planning. Transition planning is a strategic process through which an entity plans actions towards both a low-carbon and climate-resilient economy (IFRS, 2025[42]). Unlike mitigation, which can be anchored to a single quantified goal such as net-zero emissions, adaptation has no equivalent global target. The NGFS framework therefore structures the integration of adaptation around two objectives: (1) managing exposure and vulnerability to physical risks and, (2) where relevant, seizing adaptation-related opportunities (NGFS, 2025[39]). Company disclosure on adaptation and resilience through transition plans can improve wider enabling environments for adaptation and lower corporate risk profiles and improve lending terms (TPT, 2024[43]). However, uncertainties around incorporating adaptation and resilience considerations into transition plans impedes the scaling up of adaptation finance (G20 SFWG, 2025[44]).

Adaptation metrics can be progressively developed and integrated in transition plans as their design and robustness matures. As adaptation-related data and metrics are still developing, approaches for identifying key metrics can take a step-by-step approach. The NGFS proposes a maturity model (Figure 4.7) that starts with a stocktake of data and coverage status to facilitate a baseline of adaptation metrics and targets. It then progresses step by step towards a meaningful set of metrics: from (1) baseline exposure and vulnerability to (2) inputs applied towards adaptation activities, to (3) output-led metrics that quantify the impact of adaptation activities and set these against a target (NGFS, 2025[39]). Similarly, OECD work identifies five main interrelated analytical dimensions for tracking adaptation and resilience alignment (Figure 4.8): (1) Measuring physical climate risks to physical assets, (2) Aggregating physical climate risks to real economy entities, financial flows and financial assets, (3) Analysing adaptation and resilience actions and strategies by real economy entities and financial institutions, (4) Identifying resilience related public policy goals and targets that can be used as reference points to assess alignment, (5) Evaluating progress towards aligning finance with these reference points (Noels et al., 2024[38]).

Tracking corporate adaptation starts with physical risk baseline metrics that measure exposure and vulnerability to physical risks linked to estimated financial impacts. Metrics for non-financial corporates include the share of assets exposed to 1-in-100 or 1-in-200 hazards and expected asset value at risk or lost revenue due to climate-related hazards (Figure 4.8). Metrics for financial institutions include the share of portfolio exposed to key physical risks by geography/sector and the number and value of mortgages in flood zones and expected losses under climate scenarios. With location- and portfolio-relevant data, institutions can assess (1) whether exposure to physical hazards exists, and if so, (2) whether this exposure could lead to financial impact (NGFS, 2025[39]). Understanding the inherent vulnerability of assets to physical risk is the first step towards then managing physical risks.

Adaptation input metrics track actions and strategies to build resilience. Metrics for non-financial corporates include amount of capital expenditure deployed towards climate adaptation, board oversight over adaptation measures, and calculations of the cost of measures required to build resilience to physical climate risks (Munday, Parker and Panichi, 2026[45]). Metrics for financial institutions include the amount of adaptation finance mobilised and people trained in resilience measures (Figure 4.8). Adaptation input metrics can capture both risk management and opportunity (NGFS, 2025[39]).

Resilience outcome metrics analyse the impact and effectiveness of adaptation actions. Simpler output measures can include hectares restored, assets strengthened, reduced downtime, or reduced repair costs (NGFS, 2025[39]). More advanced approaches could consider risk-based metrics, such as avoided losses, reduced value-at-risk from physical climate impacts, a lower share of asset value exposed to acute and chronic risks, or ensuring a given share of risk-exposed assets is protected to a specified hazard standard. For the most advanced institutions, output metrics would reflect risk appetite and be benchmarked against common risk thresholds as well as against targeted levels of resilience to climate hazards and impacts. Such levels can be institution specific as well be linked to national or regional adaptation and resilience policy goals and targets where these exist.

Resilience-related metrics, whether relating to baselines, inputs, or outputs, are most decision-useful when paired with outcome-based and time-bound targets. Effective targets reveal both the gap to a desired outcome and the timeframe to close it, thus anchoring adaptation planning and reinforcing the business case for investment (NGFS, 2025[39]). Practice still lags this principle: a survey of OECD member countries found that only around 30% attach a timeframe to their adaptation objectives, and none systematically include a baseline (Noels et al., 2024[38]).

Credible assessments of the climate alignment of bonds are faced with different challenges for corporate and sovereign issuers. As highlighted in Chapter 3 on tracking, green-labelled sovereign bonds remain a very small and stagnating share of total sovereign bond issuance, while general purpose and unearmarked bonds issued by sovereigns cannot be easily linked to specific economic and climate-relevant sectors. In contrast, climate assessments of the corporate bond asset class can rely on sectoral analyses to identify companies dedicated to fossil fuel or high-GHG intensive sectors, in addition to the tracking of green-labelled corporate bonds.

Green-labelled bonds are debt securities whose proceeds are allocated to environmental projects, with a commitment to transparency and impact measurement (Flammer, 2025[46]). They are a financial mechanism designed to enable the mobilisation of capital for activities and projects with a positive environmental impact, while providing investors with visibility on the use of funds (Flottmann et al., 2025[47]). Bonds can be labelled as green by complying with taxonomies and frameworks developed in specific countries and jurisdictions or referring to voluntary frameworks. Among the latter includes the Green Bond Principles of the International Capital Market Association (ICMA, 2025[48]).

Green labelling of corporate bonds contributes to market transparency, but empirical evidence highlights a need for careful evaluation of impact and additionality. On the one hand, there is evidence that increased issuance of green bonds has followed stricter emissions policies aimed at reducing country-level emissions, and that green bond can be a good indicator of reduced corporate emissions for firms in carbon-intensive sectors reductions (Demski et al., 2025[49]; Fatica and Panzica, 2021[50]; Lorente et al., 2025[51]). In this context, third party verification is found to be an important factor to strengthen the likelihood of positive impacts (Flammer, 2023[52]). On the other hand, the additionality of corporate green bonds has been questioned as they are often used to refinance existing corporate debt rather than to fund new green activities (Lam and Wurgler, 2024[53]).

Robust climate transition bond frameworks, criteria and metrics can contribute to broadening the ability to assess and drive bond markets to finance decarbonisation across economic sectors. Transition bonds are a relatively new financial instrument, thus not reflected yet in trends of issued and outstanding bonds analysed in Chapter 3 on tracking. Transition bonds typically fund activities that help high-emission sectors, such as energy, industries, transportation, transition to greener practices, rather than financing already low-carbon activities. Unlike green bonds, which require proceeds to be allocated to already green activities, climate transition bonds focus on enabling improvements to current practices and long-term decarbonisation. Their integrity, however, depends on being anchored in robust frameworks and credible metrics that ensure transparency, accountability, and measurable environmental impact (OECD, 2022[54]), including to avoid locking in GHG-intensive activities and emissions at levels and over timeframes inconsistent with long-term climate policy goals (OECD, 2023[55]).

Assessment of the environmental integrity of corporate bonds can be strengthened and broadened to transition and general-purpose bonds based on metrics relating to corporate transition plans and activities. Metrics building on quantitative and qualitative information included in corporate transition plans can help expand climate-related assessments beyond green-labelled bonds. As discussed above in Section 4.2, this is notably the case for data points relating to corporate capital expenditures and investment plans, which can allow to identify not only green and climate-aligned spending, but also assess the climate consistency of more, if not all, recent and forthcoming corporate investments. Such approach can be made sector-specific by linking to analyses of corporate bonds based on climate-relevant sectors as presented in Sections 3.2.1 and 3.2.2 of Chapter 3 on tracking.

For sovereign issuers, the requirements for proceeds from sovereign debt and fiscal revenues to be fungible can prevent accurate evaluations of additionality. Green-labelled sovereign bond frameworks require governments to earmark and provide transparency on the use of proceeds. However, sovereign green bonds may be integrated into existing debt management frameworks and thus finance activities that were budgeted in any case or even already implemented (Chesini, 2024[56]). However, sovereign green bond issuance is found to have quantitative and qualitative benefits for the development of private sustainable bond markets (Cheng, 2024[57]).

Climate-related assessments of general-purpose sovereign bonds can, similar to corporate bonds, consider a range of metrics addressing ambition, progress, plans, and risks. The categories of metrics considered by providers assessing the climate performance and alignment of countries has remained broadly consistent since the first edition of the OECD Review, typically spanning historic and forward-looking GHG emissions and energy-related metrics, as well as indicators relating to existing policies and plans (OECD, 2024[28]). The choice and relative priority placed on individual metrics, however, depend on the target audience and use case with, to date, a notable difference between assessments of the alignment of sovereign bond issuance and portfolios with climate policy goals on the one hand, and, on the other hand the integration of climate-related risks in sovereign credit rating assessments.

Credible assessments of sovereigns’ climate plans depend on robust data on the implementation and characteristics of climate-related real economy and financial sector policies. Comprehensive and comparable information about all public policies implemented by governments is critical to evaluating the likelihood of sovereign issuers to both decarbonise their economies and to reduce their exposure and vulnerability to climate physical risks. As analysed and illustrated in Chapter 2 on policies, dimensions relevant to assessments include the scope, ambition, stringency, bindingness, predictability, and changes in policies. Similarly, other initiatives highlight the relevance of tracking the adoption and stringency of real economy climate policies towards assessing the climate performance of countries (OECD, 2025[58]). Efforts relating to climate budget tagging (World Bank, 2021[59]) and green budgeting (OECD, 2024[60]) have the potential to enhance the credibility of such assessments, but they have to date not resulted in increased data availability at scale.

The increasing focus on making national climate plans investable places metrics relating to transition management, risks, and opportunities at the forefront. Most climate change mitigation-related assessments of sovereigns have to date been primarily anchored in evaluations of whether each country’s current and targeted contribution to global efforts can be considered as sufficient and equitable (OECD, 2024[1]; Noels and Jachnik, 2022[16]). However, considering the widening ambition and implementation gap to reach the Paris Agreement goals, assessment frameworks increasingly emphasise forward-looking elements. These elements can help increase the credibility of nationally determined contributions (NDCs) and national adaptation plans (NAPs) towards unlocking economy-wide investments (CBI, 2025[61]) and opportunities (TPI, 2025[62]). Examples notably include sector-specific decarbonisation pathways and policy action plans (OECD/UNDP, 2025[63]) as well as quantified investment needs and associated financing strategies to sign post investment opportunities (IIGCC, 2025[64]).

The integration of physical climate risk in sovereign assessments remains primarily focused on exposure metrics that do not reflect adaptation efforts and opportunities. Current and future climate impacts can challenge sovereign debt sustainability, especially in countries already experiencing debt distress, constrained fiscal space, and high cost of debt servicing. Geospatial data discussed in the following section are improving capacities to assess the exposure of territories to a range of climate hazards. Sovereign climate exposure varies widely across geographies but a significant share of countries is expected to face future climate-related credit rating downgrades (Fitch Ratings, 2025[65]). However, a range of proposed frameworks and illustrative analyses highlight the need to better incorporate into credit risk assessments variables and metrics that make it possible to better assess vulnerability and resilience (ECB, 2023[66]), including data points on existing or planned policies, measures, and investments to adapt and increase resilience (Climate Financial Risk Forum, 2025[67]).

Corporate disclosure has significantly improved data availability, yet large data gaps remain across asset classes. In 2025, over 22 100 corporates, representing nearly two thirds of global market capitalisation, reported on climate change to CDP, an independent environmental disclosure system (CDP, 2026[68]). Considering sustainability disclosure more broadly, almost 12 900 companies representing 91% of listed companies by global market capitalisation disclosed sustainability‑related information in 2024, up from 9 600 companies representing 86% of market capitalisation in 2022 (OECD, 2025[69]). Yet, Sections 4.2, 4.4 and 4.5 exemplify important gaps in measuring transition planning and adaptation actions for corporate and sovereign asset classes. Further, data for private markets and different types of financial institutions is lacking particularly, as show in Chapter 3 on tracking.

Innovative data approaches, such as natural language processing and geospatial data techniques are being used to fill climate data gaps. Technical innovation offers promising solutions to address some climate-related data gaps (Nefzi et al., 2025[70]; OECD-NGFA, 2026[71]). Examples of such solutions include (i) text analysis and natural language processing of corporate reporting, (ii) large-scale extractions from sustainability reports and tables, (iii) satellite and other remote-sensing data, (iv) geospatial asset-level mapping, (v) firm-level emissions imputation with machine learning and accounting data, and (vi) supply-chain/network reconstruction (OECD, 2026[72]).

Text analysis and natural language processing (NLP) are increasingly used for extracting corporate climate data at scale, which can improve consistency and comparability across companies. Advances in machine learning have enabled the extraction of structured information from text in climate disclosures, sustainability reports, regulatory filings, and news reporting. Overall, such extraction relies on three main approaches:

• Keyword searching is a rule-based text analysis approach that can be effective for identifying relevant passages when terminology is consistent and standardised. For example. keyword searches have been used to identify climate risk-related language in corporate filings and consistency of climate reporting with TCFD recommendations (Doran and Quinn, 2009[73]; Moreno and Caminero, 2021[74]). This approach is transparent and reproducible, as any result can be traced directly back to the initial keyword glossary. However, developing a keyword glossary can be subjective (Loughran and McDonald, 2011[75]). Moreover, a predefined keyword list may be less effective in capturing sector-specific topics, linguistic nuance or context, and areas that use highly technical or heterogenous language.

• Basic word vectorisation methods convert text into numerical vectors so it can be analysed computationally. Common approaches include Bag-of-Words (BoW), Term Frequency–Inverse Document Frequency (TF-IDF), and Word2Vec. BoW represents a document by counting how often each word appears, while TF-IDF adjusts those counts by giving more weight to words that are frequent in one document but relatively rare across a broader corpus. Instead of counting words directly, Word2Vec learns vector representations from patterns of word co-occurrence in large corpora (Mikolov et al., 2013[76]). Such methods have been used to identify SDG-related language in corporate disclosures, construct sustainability vocabularies, and classify climate-related reporting content (Amel-Zadeh et al., 2021[77]; Acosta-Smith et al., 2023[78]). Although vectorisation itself can be unsupervised, many practical applications built on these representations still need labelled training data, which is often costly to create. BoW and TF-IDF also ignore much of the surrounding context, while Word2Vec gives each word only one fixed meaning, making it hard to handle ambiguity. As a result, these methods may be less effective for nuanced or longer disclosures.

• More sophisticated natural language processing (NLP) approaches rely on deep learning techniques, notably transformer-based architectures such as BERT and GPT, which are designed to capture the meaning of words in context rather than treating them as isolated terms (Vaswani et al., 2017[79]; Brown et al., 2020[80]). Large language models (LLMs) have been used to extract metrics from corporate reports, assess firms’ alignment with climate reporting frameworks, and extract data on lobbying behaviour across corporate reports (Bingler et al., 2022[81]; Colesanti Senni et al., 2024[82]; Dave et al., 2024[83]; IFRS, 2024[84]; Kolli et al., 2025[85]). These models offer greater flexibility and scalability than earlier text-analysis methods due to their ability to interpret nuanced or domain-specific language with little or no extensive training (Vaswani et al., 2017[86]). However, these models can lack transparency as their internal decision processes are difficult to interpret. Moreover, prompt-based interactions with generative models may generate different results depending on how instructions are framed, creating challenges for robustness, consistency and reproducibility. Additionally, as LLMs can hallucinate by generating plausible but incorrect information, recent work has developed approaches to address this risk, such as through retrieval-augmented generation (RAG), which grounds model outputs in retrieved external documents and can improve accuracy, transparency and timeliness (Pisaneschi, 2025[87]).

NLP approaches are being used not only to fill climate data gaps, but also to construct new metrics on climate disclosure quality, greenwashing, and selective reporting. For example, the Cheap Talk Index developed by Bingler et al. (2022[88]) uses a BERT-based NLP model to rate the specificity of climate claims in annual reports, on the premise that vague, non-committal language may signal low credibility or a lack of concrete action. Related work examines the extent to which firms disclose favourable ESG and climate-related indicators while omitting negative climate information, thus producing metrics of “selective disclosure magnitude” (Lublóy, Keresztúri and Berlinger, 2025[89]; Marquis, Toffel and Bird, 2015[90]). Other approaches may track the disappearance of previously stated targets (GreenWatch, n.d.[91]). While these approaches are relatively scalable and data-efficient, they may measure disclosure behaviour or public-relations management rather than actual climate performance (OECD, 2025[92]).

NLP approaches are also leveraging alternative data sources to construct climate-related metrics beyond those relating to corporate reporting. Alternative data that are being used to develop climate metrics using NLPs include news and media sentiment data, real estate data, transcription data, supply chain and logistics data, and patent data among others (Pisaneschi, 2024[93]). For example, NLP techniques have been used to process news and media sources for possible corporate climate misconduct, by classifying and scoring controversy events related to issues such as GHG emissions, fossil fuel activities, or environmental misrepresentation (RepRisk, n.d.[94]; Hafez et al., 2022[95]).

Geospatial data approaches are also relied upon to fill corporate disclosure data gaps, address more opaque financial asset classes, and cross-validate reported climate data. Geospatial data can improve transparency, attribution and accountability by making it possible to connect localised environmental conditions to specific corporate entities and financial decision-making Geospatial approaches shift the unit of analysis from the legal entity to physical assets and operations, enabling a bottom-up analysis of a company, sector, or portfolio (Christiaen, 2023[96]). This is particularly relevant for asset classes and firms, for which reporting is sparse or opaque, notably private markets. Beyond filling disclosure gaps, geospatial data can also support due diligence, risk management, assessment of historic liabilities, monitoring of sustainability outcomes between reporting cycles, and the verification of company disclosures or statements by external stakeholders.

Earth Observation satellite data is already being used to estimate companies’ exposure to physical climate risk, GHG emissions, energy production, and land use impacts (Alonso-Robisco et al., 2026[97]; Rapach et al., 2024[98]; Climate TRACE, 2025[99]). Earth observation satellite data can provide granular insights to assess climate-relevant developments and address shortcomings and inconsistencies in traditional and disclosed data. It can reduce information asymmetries across scattered data sources and limit the potential for greenwashing practices (Nefzi et al., 2025[70]; ESA, 2023[100]). While Earth observation systems can enhance data availability, accessibility remains a challenge, with barriers such as proprietary databases and high costs for newcomers needing to process raw data (Alonso-Robisco et al., 2026[97]). These methods do not eliminate the need for corporate disclosure, but they can provide an important independent layer of evidence, and a means of cross-checking reported information.

Geospatial analysis for sustainable finance requires geolocating corporate facilities and linking them to ownership structures. Geospatial data facilitates linking physical assets to parent companies, and in some cases to financial instruments. A growing body of data sources that combine asset location with ownership linkages, including public emissions-trading platforms, sector-specific geospatial asset databases, and open-source trackers such as Global Energy Monitor (Noels et al., 2024[38]). The location-ownership connection is essential as it allows geospatial insights to be attributed to firms and then to loans, bonds and equity holdings (Christiaen, 2023[96]).

[78] Acosta-Smith, J. et al. (2023), Understanding climate-related disclosures of UK financial institutions, Bank of England Staff Working Paper No. 1,017, https://www.bankofengland.co.uk/-/media/boe/files/working-paper/2023/understanding-climate-related-disclosures-of-uk-financial-institutions.pdf.

[13] Allahdadi, M. (2025), “Beyond Green Signaling: Are Institutional Investors Decarbonizing Their Portfolios?”, International Review of Economics & Finance, Vol. 104, p. 104767, https://doi.org/10.1016/j.iref.2025.104767.

[97] Alonso-Robisco, A. et al. (2026), “Houston, we have a problem: Can satellite information bridge the climate-related data gap?”, Latin American Journal of Central Banking, Vol. 7/1, p. 100173, https://doi.org/10.1016/j.latcb.2025.100173.

[77] Amel-Zadeh, A. et al. (2021), “NLP for SDGs: Measuring Corporate Alignment with the Sustainable Development Goals”, SSRN Electronic Journal, https://doi.org/10.2139/SSRN.3874442.

[126] Battiston, S. et al. (2017), “A climate stress-test of the financial system”, Nature Climate Change, Vol. 7/4, pp. 283-288, https://doi.org/10.1038/nclimate3255.

[107] Battiston, S. et al. (2022), The NACE – CPRS – IAM mapping: A tool to support climate risk analysis of financial portfolio using NGFS scenarios, https://doi.org/: https://ssrn.com/abstract=4223606 or http://dx.doi.org/10.2139/ssrn.4223606.

[81] Bingler, J. et al. (2022), “Cheap talk and cherry-picking: What ClimateBert has to say on corporate climate risk disclosures”, Finance Research Letters, Vol. 47, https://doi.org/10.1016/j.frl.2022.102776.

[88] Bingler, J. et al. (2022), “How Cheap Talk in Climate Disclosures relates to Climate Initiatives, Corporate Emissions, and Reputation Risk”, SSRN Electronic Journal, https://doi.org/10.2139/ssrn.4000708.

[24] Bloomberg (2026), Banks Notch Higher Fees From Green Bonds Than Fossil Fuel Debt, https://www.bloomberg.com/news/articles/2026-01-02/banks-notch-higher-fees-from-green-bonds-than-fossil-fuel-debt.

[35] BloombergNEF (2025), Energy Supply Fund-Enabled Capex Ratio, https://assets.bbhub.io/professional/sites/44/ESFR_report_20250603_final_summary.pdf.

[22] BloombergNEF (2025), Fourth Energy Supply Ratios: Little Progress, https://www.bnef.com/.

[80] Brown, T. et al. (2020), “Language Models are Few-Shot Learners”, Advances in Neural Information Processing Systems, Vol. 2020-December, https://arxiv.org/pdf/2005.14165 (accessed on 4 July 2025).

[61] CBI (2025), Unlocking Investment for Sovereign Transition, https://www.climatebonds.net/data-insights/publications/unlocking-investment-sovereign-transition.

[68] CDP (2026), Keeping Pace: Disclosure Data Factsheet 2025, https://cdp.net/en/insights/keeping-pace-disclosure-data-factsheet-2025.

[57] Cheng, G. (2024), “Sovereign Green Bonds: A Catalyst for Sustainable Debt Market Development?”, IMF Working Papers, Vol. 2024/120, p. 1, https://doi.org/10.5089/9798400277030.001.

[56] Chesini, G. (2024), “Can Sovereign Green Bonds Accelerate the Transition to Net-Zero Greenhouse Gas Emissions?”, International Advances in Economic Research, Vol. 30/2, pp. 177-197, https://doi.org/10.1007/s11294-024-09900-6.

[96] Christiaen, C. (2023), State and trends of spatial finance 2023: Fundamental Building Blocks for Transparency and Accountability in Green Finance, https://www.cgfi.ac.uk/wp-content/uploads/2023/03/State-and-Trends-of-Spatial-Finance-2023.pdf.

[67] Climate Financial Risk Forum (2025), From risk to resilience: Integrating adaptation into finance, https://www.fca.org.uk/publication/corporate/from-risk-resilience-integrating-adaptation-finance.pdf.

[99] Climate TRACE (2025), When, Where, and How: Source-level solutions for turning climate ambition into on-the-ground action, https://media.climatetrace.org/Climate_TRACE_whitepaper_When_Where_And_How_Source_Level_Emissions_Reduction_Solutions_202511_v_Final_a13f795b93.pdf.

[82] Colesanti Senni, C. et al. (2024), “Combining AI and Domain Expertise to Assess Corporate Climate Transition Disclosures”, SSRN Electronic Journal, https://doi.org/10.2139/SSRN.4826207.

[114] Cortina Lorente, J. et al. (2025), Green versus Conventional Corporate Debt: From Issuances to Emissions (English), Policy Research Working Paper;Prosperity;Paper is funded by the Knowledge for Change Program (KCP) Washington, D.C. : World Bank Group, http://documents.worldbank.org/curated/en/099204010022526543.

[3] CPI (2025), Net Zero Finance Tracker Methodology 2025, https://www.climatepolicyinitiative.org/wp-content/uploads/2025/10/NZFT-Methodology.pdf.

[123] CPI (2025), State of OECD Pension Funds’ Climate Transition: Insights and recommendations from the Net Zero Finance Tracker, https://www.climatepolicyinitiative.org/publication/state-of-oecd-pension-funds-climate-transition-insights-and-recommendations-from-the-net-zero-finance-tracker/.

[4] CPI (n.d.), Net-Zero Finance Tracker, https://netzerofinancetracker.climatepolicyinitiative.org/aggregated-data?tab=target_score (accessed on 21 March 2026).

[20] Dai, L. et al. (2023), Weighted Average Green Revenue (WAGR): Integrating climate solutions into portfolio construction, https://www.gmo.com/globalassets/articles/white-paper/2023/weighted_average_green_revenue_-_integrating_climate_solutions.pdf.

[83] Dave, A. et al. (2024), “Climate AI for Corporate Decarbonization Metrics Extraction”, Vol. 1, https://arxiv.org/pdf/2411.03402 (accessed on 6 June 2025).

[49] Demski, J. et al. (2025), “Growth of the green bond market and greenhouse gas emissions”, BIS Quarterly Review, https://www.bis.org/publ/qtrpdf/r_qt2503d.htm.

[73] Doran, K. and E. Quinn (2009), Climate Change Risk Disclosure: A Sector by Sector Analysis of SEC 10-K Filings from 1995-2008, https://papers.ssrn.com/abstract=1416279 (accessed on 11 June 2025).

[34] Duong, D. et al. (2024), Private equity climate risks scorecard 2024, https://www.peclimaterisks.org/wp-content/uploads/2024/10/PECR_Report_Climate-Risks-Scorecard_2024.pdf.

[66] ECB (2023), Climate change and sovereign risk, https://www.ecb.europa.eu/press/financial-stability-publications/fsr/special/html/ecb.fsrart202305_03~f51dd11fd7.en.html.

[100] ESA (2023), Earth Observation for Environmental, Social, and Governance Schemes, https://gda.esa.int/wp-content/uploads/2023/12/D17-EO-for-ESG-Schemes.pdf.

[50] Fatica, S. and R. Panzica (2021), “Green bonds as a tool against climate change?”, Business Strategy and the Environment, Vol. 30/5, pp. 2688-2701, https://doi.org/10.1002/bse.2771.

[110] Financial Times (2024), fDi Markets, Database, https://www.fdimarkets.com/.

[65] Fitch Ratings (2025), Sovereigns Climate Vulnerability Signals: Discussion Paper, https://www.fitchratings.com/research/sovereigns/sovereigns-climate-vulnerability-signals-discussion-paper-27-01-2025.

[46] Flammer, C. (2025), “Sustainable Finance: Tools, Effectiveness, and Challenges”, Asian Economic Policy Review, Vol. 21/1, pp. 14-19, https://doi.org/10.1111/aepr.70007.

[52] Flammer, C. (2023), “Green bonds and carbon emissions”, Oxford Review of Economic Policy, Vol. 39/4, pp. 752-764, https://doi.org/10.1093/oxrep/grad040.

[47] Flottmann, C. et al. (2025), “Green debt: a systematic literature review and future research agenda”, Management Review Quarterly, https://doi.org/10.1007/s11301-025-00511-x.

[121] FSB (2025), Global Monitoring Report on Nonbank Financial Intermediation 2025, https://www.fsb.org/2025/12/global-monitoring-report-on-nonbank-financial-intermediation-2025/.

[129] G20 DG13 (2025), Recommendation 4: self commitments overview december 2025, https://www.imf.org/-/media/files/data/statistics/wgsd/rec-4-self-commitments-overview-december-2025.xlsx (accessed on 17 January 2026).

[44] G20 SFWG (2025), Sustainable Finance Working Group: 2025 Presidency and Co-chairs Note on Agenda Priorities, https://g20sfwg.org/wp-content/uploads/2025/02/2025-G20-SFWG-Note-on-Agenda-Priorities-rev.pdf.

[6] GFANZ (2022), Recommendations and Guidance on Financial Institution Net-zero Transition Plans, https://assets.bbhub.io/company/sites/63/2022/06/GFANZ_Recommendations-and-Guidance-on-Net-zero-Transition-Plans-for-the-Financial-Sector_June2022.pdf.

[91] GreenWatch (n.d.), GreenWatch Methodology, https://greenwatch.ai/greenwatch-methodology/ (accessed on 29 March 2026).

[95] Hafez, P. et al. (2022), RavenPack ESG Controversy Scoring Framework, RavenPack Quantitative Research.

[19] IAIS (2025), Application Paper on the supervision of climate-related risks in the insurance sector, https://www.iais.org/uploads/2025/04/Application-Paper-on-the-supervision-of-climate-related-risks-in-the-insurance-sector.pdf.

[48] ICMA (2025), Green Bond Principles Voluntary Process Guidlines for Issuing Green Bonds, International Capital Market Association, https://www.icmagroup.org/assets/documents/Sustainable-finance/2025-updates/Green-Bond-Principles-GBP-June-2025.pdf.

[106] IEA (2025), World Energy Investment, https://iea.blob.core.windows.net/assets/904392c0-caa0-45c1-a53a-90a6eacd2bfe/WorldEnergyInvestment2025.pdf.

[104] IEA (2025), World Energy Investment 2025 Datafile, IEA, Paris, https://www.iea.org/data-and-statistics/data-product/world-energy-investment-2025-datafile, Licence: Terms of Use for Non-CC Materia.

[42] IFRS (2025), Disclosing information about an entity’s climate-related transition, including information about transition plans, in accordance with IFRS S2, https://www.ifrs.org/content/dam/ifrs/supporting-implementation/ifrs-s2/transition-plan-disclosure-s2.pdf.

[84] IFRS (2024), Progress on Corporate Climate-related Disclosures—2024 Report, IFRS Foundation, Basel, https://www.ifrs.org/content/dam/ifrs/supporting-implementation/issb-standards/progress-climate-related-disclosures-2024.pdf (accessed on 11 July 2025).

[30] IFRS (2023), IFRS S2 Climate-related Disclosures, https://www.ifrs.org/issued-standards/ifrs-sustainability-standards-navigator/ifrs-s2-climate-related-disclosures/.

[7] IFRS ISSB (2023), Sustainability Disclosure Standard: Climate-related Disclosures, https://www.ifrs.org/content/dam/ifrs/publications/pdf-standards-issb/english/2023/issued/part-a/issb-2023-a-ifrs-s2-climate-related-disclosures.pdf?bypass=on.

[64] IIGCC (2025), Making NDCs investable - the investor perspective, https://www.iigcc.org/resources/making-ndcs-investable-the-investor-perspective.

[41] IIGCC (2025), Making the case for resilience investment: PCRAM 2.0 case study, ferry and port (PIDG), https://www.iigcc.org/resources/making-case-resilience-pcram-2.0-case-study-ferry-port-pidg.

[40] IIGCC (2025), Making the case for resilience investment: PCRAM 2.0 case study, solar plant analysis (Octopus Energy Generation), https://www.iigcc.org/resources/resilience-investment-pcram-2.0-case-study-solar-plant-analysis-octopus-energy-generation.

[25] IIGCC (2024), Net Zero Investment Framework updated: NZIF 2.0, https://www.iigcc.org/resources/updated-net-zero-investment-framework-nzif-2.0.

[8] IIGCC (2021), Net Zero Investment Framework, Institutional Investors Group on Climate Change, https://www.iigcc.org/resource/net-zero-investment-framework-implementation-guide/.

[116] IMF (2020), What is debt sustainability?, IMF Finance and Development Magazine, https://www.imf.org/en/Publications/fandd/issues/2020/09/what-is-debt-sustainability-basics#:~:text=A%20country's%20public%20debt%20is,assistance%20or%20going%20into%20default.

[102] Jachnik, R., M. Mirabile and A. Dobrinevski (2019), “Tracking finance flows towards assessing their consistency with climate objectives”, OECD Environment Working Papers, No. 146, OECD Publishing, Paris, https://doi.org/10.1787/82cc3a4c-en.

[85] Kolli, I. et al. (2025), Automated Evidence Extraction and Scoring for Corporate Climate Policy Engagement: A Multilingual RAG Approach, https://aclanthology.org/2025.emnlp-demos.9.pdf.

[53] Lam, P. and J. Wurgler (2024), Green Bonds: New Label, Same Projects, National Bureau of Economic Research, Cambridge, MA, https://doi.org/10.3386/w32960.

[51] Lorente, C. et al. (2025), Green versus Conventional Corporate Debt: From Issuances to Emissions, World Bank Group Policy Research Working Paper, http://documents.worldbank.org/curated/en/099204010022526543.

[75] Loughran, T. and B. McDonald (2011), “When Is a Liability Not a Liability? Textual Analysis, Dictionaries, and 10-Ks”, Journal of Finance, Vol. 66/1, pp. 35-65, https://doi.org/10.1111/J.1540-6261.2010.01625.X;PAGE:STRING:ARTICLE/CHAPTER.

[89] Lublóy, Á., J. Keresztúri and E. Berlinger (2025), “Quantifying firm-level greenwashing: A systematic literature review”, Journal of Environmental Management, Vol. 373, p. 123399, https://doi.org/10.1016/J.JENVMAN.2024.123399.

[90] Marquis, C., M. Toffel and Y. Bird (2015), “Scrutiny, Norms, and Selective Disclosure: A Global Study of Greenwashing”, SSRN Electronic Journal, https://doi.org/10.2139/SSRN.1836472.

[119] McKinsey & Company (2025), Global Private Markets Report 2025: Braced for shifting weather, https://www.mckinsey.com/industries/private-capital/our-insights/global-private-markets-report.

[120] McKinsey & Company (2024), Global Private Markets Report 2024: Private markets in a slower era, https://www.mckinsey.com/industries/private-capital/our-insights/global-private-markets-report-2024#/.

[76] Mikolov, T. et al. (2013), Efficient Estimation of Word Representations in Vector Space, https://arxiv.org/pdf/1301.3781.

[74] Moreno, Á. and T. Caminero (2021), Application of text mining to the analysis of climate-related disclosures, https://www.bis.org/ifc/publ/ifcb56_25.pdf (accessed on 10 July 2025).

[117] MSCI (2024), Sizing Up the Global-Market Portfolio, https://www.msci.com/research-and-insights/blog-post/sizing-up-the-global-market-portfolio (accessed on 15 January 2026).

[45] Munday, P., Z. Parker and R. Panichi (2026), Mercury Rising: European Entities Show Some Adaptation Gains As Physical Risks Mount, https://www.spglobal.com/ratings/en/regulatory/delegate/getPDF?articleId=3536929&type=COMMENTS&defaultFormat=PDF.

[70] Nefzi, D. et al. (2025), Addressing climate change data needs: the global debate and central banks’ contribution, https://www.bis.org/ifc/publ/ifcb63_01_rh.pdf.

[23] Net-Zero Banking Alliance (2024), Target Setting for Capital Markets Activities: Supporting note, https://www.unepfi.org/wordpress/wp-content/uploads/2024/10/Target-Setting-for-Capital-Markets-Activities-An-NZBA-Supporting-Note.pdf.

[39] NGFS (2025), Integrating adaptation and resilience into transition plans, https://g20sfwg.org/wp-content/uploads/2025/07/SFWG-P2a_Adaptation-and-Transition-Plans.pdf.

[14] NGFS (2024), NGFS: Transition Plan Package, https://www.ngfs.net/en/publications-and-statistics/publications/ngfs-transition-plan-package.

[38] Noels, J. et al. (2024), “Towards assessing the alignment of finance with climate resilience goals: Exploring options, methodologies, data and metrics”, OECD Environment Working Papers, No. 251, OECD Publishing, Paris, https://doi.org/10.1787/9446d65e-en.

[118] Noels, J. and R. Jachnik (2025), Climate transition metrics for the financial sector: Translating targets into action, https://www.oecd.org/content/dam/oecd/en/events/2025/06/transition-metrics-for-the-financial-sector/workshop-climate-transition-metrics-policy-summary-2025.pdf (accessed on 7 February 2026).

[16] Noels, J. and R. Jachnik (2022), “Assessing the climate consistency of finance: Taking stock of methodologies and their links to climate mitigation policy objectives”, OECD Environment Working Papers, No. 200, OECD Publishing, Paris, https://doi.org/10.1787/d12005e7-en.

[9] NZAOA (2021), Inaugural 2025 Target Setting Protocol: U.N.-convened net-zero asset owner alliance monitoring reporting and verification track, https://www.unepfi.org/wordpress/wp-content/uploads/2021/01/Alliance-Target-Setting-Protocol-2021.pdf.

[72] OECD (2026), Innovative data for sustainable finance: Informing climate performance assessments and financial risk management, https://www.oecd.org/en/events/2026/04/Innovative-data-for-sustainable-finance---Informing-climate-performance-assessments-and-financial-risk-management.html.

[2] OECD (2026), “Mapping climate-related metrics in the financial sector”, OECD Business and Finance Policy Papers, No. 96, OECD Publishing, Paris, https://doi.org/10.1787/e91736e8-en.

[127] OECD (2026), OECD Environmental Performance Reviews: Colombia 2026, OECD Environmental Performance Reviews, OECD Publishing, Paris, https://doi.org/10.1787/968398f7-en.

[92] OECD (2025), Behind ESG ratings: Unpacking sustainability metrics, OECD Publishing, Paris, https://doi.org/10.1787/3f055f0c-en.

[105] OECD (2025), Environment at a Glance Indicators, OECD Publishing, Paris, https://doi.org/10.1787/ac4b8b89-en.

[69] OECD (2025), Global Corporate Sustainability Report 2025, OECD Publishing, Paris, https://doi.org/10.1787/bc25ce1e-en.

[115] OECD (2025), Global Debt Report 2025: Financing Growth in a Challenging Debt Market Environment, OECD Publishing, Paris, https://doi.org/10.1787/8ee42b13-en.

[130] OECD (2025), Sustainable Bonds: Trends and Policy Recommendations, OECD Publishing, Paris, https://doi.org/10.1787/26726c68-en.

[58] OECD (2025), The Climate Action Monitor 2025, OECD Publishing, Paris, https://doi.org/10.1787/1819c631-en.

[28] OECD (2024), “Aligning finance with climate goals”, OECD Net Zero+ Policy Papers, No. 4, OECD Publishing, Paris, https://doi.org/10.1787/aa7c23b2-en.

[36] OECD (2024), Climate Adaptation Investment Framework, Green Finance and Investment, OECD Publishing, Paris, https://doi.org/10.1787/8686fc27-en.

[109] OECD (2024), FDI Qualities Indicators Visualisation Platform, https://www.oecd.org/en/data/dashboards/fdi-qualities-indicators-visualisation-platform.html.

[60] OECD (2024), Green Budgeting in OECD Countries 2024, OECD Publishing, Paris, https://doi.org/10.1787/9aea61f0-en.

[31] OECD (2024), Implementing Sustainability Reporting Requirements that Work for SMEs, https://g20sfwg.org/wp-content/uploads/2024/06/P3-G20-SFWG-OECD-Implementing-sustainability-reporting-that-works-for-SMEs.pdf.

[1] OECD (2024), OECD Review on Aligning Finance with Climate Goals: Assessing Progress to Net Zero and Preventing Greenwashing, Green Finance and Investment, OECD Publishing, Paris, https://doi.org/10.1787/b9b7ce49-en.

[5] OECD (2023), “Assessing net-zero metrics for financial institutions: Supporting the monitoring of financial institutions’ commitments”, OECD Business and Finance Policy Papers, No. 37, OECD Publishing, Paris, https://doi.org/10.1787/dedcfe56-en.

[55] OECD (2023), Mechanisms to Prevent Carbon Lock-in in Transition Finance, Green Finance and Investment, OECD Publishing, Paris, https://doi.org/10.1787/d5c49358-en.

[122] OECD (2023), “Towards Orderly Green Transntion: Investment Requirements and Managing Risks to Capital Flows”, OECD Business and Finance Policy Papers, No. 57, OECD Publishing, Paris, https://doi.org/10.1787/24520e7e-en.

[108] OECD (2022), FDI Qualities Policy Toolkit, OECD Publishing, Paris, https://doi.org/10.1787/7ba74100-en.

[124] OECD (2022), Long-term investing of large pension funds and public pension reserve funds 2022, OECD Publishing, Paris, https://doi.org/10.1787/809eff56-en.

[54] OECD (2022), OECD Guidance on Transition Finance: Ensuring Credibility of Corporate Climate Transition Plans, Green Finance and Investment, OECD Publishing, Paris, https://doi.org/10.1787/7c68a1ee-en.

[112] OECD (2021), Financial markets and climate transition: Opportunities, challenges and policy implications, OECD Publishing, Paris, https://doi.org/10.1787/ab4655d6-en.

[113] OECD (n.d.), Financial corporations debt to equity ratio, https://www.oecd.org/en/data/indicators/financial-corporations-debt-to-equity-ratio.html (accessed on 8 March 2025).

[63] OECD/UNDP (2025), Investing in Climate for Growth and Development: The Case for Enhanced NDCs, OECD Publishing, Paris, https://doi.org/10.1787/16b7cbc7-en.

[71] OECD-NGFA (2026), OECD-NGFS Workshop: Innovative data for sustainable finance: Informing climate performance assessments and financial risk management, https://www.oecd.org/en/events/2026/04/Innovative-data-for-sustainable-finance---Informing-climate-performance-assessments-and-financial-risk-management.html.

[37] OECD-NGFS (2024), Workshop summary: Assessing the Climate Resilience of Finance - From physical risk to resilience alignment, https://www.oecd.org/content/dam/oecd/en/events/2024/6/workshop-on-assessing-the-climate-resilience-of-finance/OECD-NGFS-workshop-policy-summary.pdf.

[128] OeNB (2020), Financial Stability Report 40, https://www.oenb.at/en/Publications/Financial-Market/Financial-Stability-Report/2020/financial-stability-report-40.html.

[87] Pisaneschi, B. (2025), RAG for Finance: Automating Document Analysis with LLMs, CFA, https://rpc.cfainstitute.org/research/the-automation-ahead-content-series/retrieval-augmented-generation (accessed on 2 April 2026).

[93] Pisaneschi, B. (2024), Unstructured Data and AI: Fine-Tuning LLMs to Enhance the Investment Process, https://rpc.cfainstitute.org/research/reports/2024/unstructured-data-and-ai.

[32] Private Equity Task Force (2024), Private Markets Decarbonisation Roadmap 2.0, https://www.bain.com/how-we-help/private-markets-decarbonization-roadmap/.

[98] Rapach, S. et al. (2024), “A taxonomy of earth observation data for sustainable finance”, Journal of Climate Finance, Vol. 6, p. 100029, https://doi.org/10.1016/j.jclimf.2023.100029.

[94] RepRisk (n.d.), RepRisk | RepRisk methodology overview, https://www.reprisk.com/insights/resources/methodology (accessed on 3 July 2025).

[11] Sastry, P., E. Verner and D. Ibanez (2024), Business as Usual: Bank Net Zero Commitments, Lending, and Engagement, National Bureau of Economic Research, Cambridge, MA, https://doi.org/10.3386/w32402.

[17] SBTi (2025), Financial Institutions Net-Zero Standard Version 1.0, https://files.sciencebasedtargets.org/production/files/Financial-Institutions-Net-Zero-Standard.pdf.

[125] ShareAction (2018), Pensions in a changing climate, https://cdn2.assets-servd.host/shareaction-api/production/resources/reports/AODP-PensionsChangingClimate.pdf.

[10] TCFD (2021), Implementing the Recommendations of the Task Force on Climate-related Financial Disclosures, https://assets.bbhub.io/company/sites/60/2021/07/2021-TCFD-Implementing_Guidance.pdf.

[62] TPI (2025), ASCOR in Practice: Use Cases and Insights, https://www.transitionpathwayinitiative.org/publications/2025-ascor-in-practice-use-cases-and-insights.

[21] TPI (2025), State of the Banking Transition 2025, https://www.transitionpathwayinitiative.org/publications/uploads/2025-state-of-the-banking-transition-2025.pdf.

[27] TPT (2024), Asset Managers Sector Guidance, https://www.ifrs.org/content/dam/ifrs/knowledge-hub/resources/tpt/asset-managers-sector-guidance-apr-2024.pdf.

[26] TPT (2024), Asset Owners Sector Guidance, https://www.ifrs.org/content/dam/ifrs/knowledge-hub/resources/tpt/asset-owners-sector-guidance-apr-2024.pdf.

[43] TPT (2024), Building Climate-ready Transition Plans: Including adaptation and resilience for comprehensive transition planning approaches., https://itpn.global/wp-content/uploads/2024/11/Adaptation-1.pdf.

[15] Transition Plan Taskforce (2023), TPT Disclosure Framework, https://www.ifrs.org/content/dam/ifrs/knowledge-hub/resources/tpt/disclosure-framework-oct-2023.pdf.

[18] UNEP FI (2025), Guidance for Climate Target Setting for Banks - Version 4, https://www.unepfi.org/industries/banking/guidance-for-climate-target-setting-for-banks-version-4/.

[29] UNEP FI (2022), Theory of Change for Climate Mitigation, https://www.unepfi.org/wordpress/wp-content/uploads/2022/09/14-Theory-of-Change-for-Climate-Mitigation-D1_JD.pdf.

[12] University of California, Santa Cruz/Columbia University/Federal Reserve Bank of San Francisco (2024), “Industrial Composition of Syndicated Loans and Banks’ Climate Commitments”, Federal Reserve Bank of San Francisco, Working Paper Series, Vol. 2024/23, pp. 01-37, https://doi.org/10.24148/wp2024-23.

[33] Unwritten (2025), Benchmarking private market investor climate disclosures, https://www.unwritten.co/resources.

[86] Vaswani, A. et al. (2017), “Attention Is All You Need”, p. 1, https://arxiv.org/pdf/1706.03762 (accessed on 4 July 2025).

[79] Vaswani, A. et al. (2017), Attention Is All You Need, https://proceedings.neurips.cc/paper_files/paper/2017/file/3f5ee243547dee91fbd053c1c4a845aa-Paper.pdf (accessed on 4 July 2025).

[111] Wilson, C. and B. Caldecott (2023), “Investigating the role of passive funds in carbon-intensive capital markets: Evidence from U.S. bonds”, Ecological Economics, Vol. 209, p. 107792, https://doi.org/10.1016/j.ecolecon.2023.107792.

[59] World Bank (2021), Climate Change Budget Tagging, World Bank, Washington, DC, https://doi.org/10.1596/35174.

[103] World Bank (n.d.), Gross fixed capital formation (current US$), https://data.worldbank.org/indicator/NE.GDI.FTOT.CD (accessed on 29 June 2024).

[101] World Bank Group, IMF and OECD (2023), Activating Alignment: Applying the G20 Principles for Sustainable Finance Alignment with a Focus on Climate Change Mitigation, World Bank, https://www.imf.org/external/np/g20/091323.htm.