Sustainability in finance: ESG reporting for banks

Banks, asset managers, insurers and investment firms influence environmental outcomes through the activities they finance, invest in and support. Their own offices, business travel and purchased services matter, but a meaningful assessment also needs to examine the companies, assets and projects connected to their portfolios.

Sustainability in the finance sector is therefore a data and decision-making challenge, not simply an annual reporting exercise. Financial institutions need reliable information to measure emissions, understand climate-related exposure, substantiate product disclosures and assess how counterparties are managing their transition.

The carbon footprint of the financial sector extends beyond an institution’s direct operations. Portfolio-related emissions can dominate its inventory, although their share varies by business model, portfolio composition and measurement boundary. A single percentage should not be assumed to apply to every bank, fund or insurer.

The regulatory landscape adds another layer of complexity. Corporate sustainability disclosures, product-level transparency, taxonomy indicators and prudential risk requirements draw on overlapping information, but they use different definitions and serve different purposes.

The practical objective is to collect relevant data once, govern it consistently and adapt it to reporting, risk management and operational decisions. This guide explains the main challenges, the frameworks financial institutions need to distinguish and the processes that make environmental data useful across the business.

Key sustainability challenges in finance

Financial institutions connect capital with economic activity. Lending decisions, investment allocations and client engagement can support projects with very different environmental profiles. Understanding those profiles helps institutions assess exposures and discuss realistic transition needs with their counterparties.

Environmental information can also reveal financially relevant dependencies. A property’s flood exposure, a manufacturer’s energy demand or a borrower’s reliance on water-intensive processes may affect costs, asset values or business continuity. These factors need to be evaluated alongside established financial indicators, not treated as substitutes for them.

Reliable data supports accountability as well. Investors, supervisors and customers need to understand what an institution has measured, which parts of its portfolio are covered and how much of the analysis relies on estimates. Clear boundaries and transparent limitations are more credible than broad claims unsupported by evidence.

Financed emissions measurement

Financed emissions relate to emissions attributed to a financial institution’s lending and investment activities. The relevant investment activities sit within Scope 3 Category 15, but calculating the inventory requires more than collecting each counterparty’s total emissions and adding them together.

The institution must define the portfolio boundary, identify the relevant asset classes and apply appropriate attribution methods. The PCAF standard for financial institutions distinguishes financed emissions, facilitated emissions and insurance-associated emissions in separate parts. Capital-market facilitation and insurance underwriting should not simply be folded into one undifferentiated financed-emissions figure.

Portfolio records need to be matched with the correct counterparty or asset. Institutions also need to manage the relationship between financial reference dates and emissions reporting periods. A borrower may provide its previous year’s inventory while the institution holds more recent exposure data, making the treatment of timing differences important.

The broader Scope 3 emissions measurement process illustrates why boundaries, methods and evidence must remain visible. Improvements in portfolio data should be documented so that changes in the reported result can be separated from genuine changes in the underlying activities.

Portfolio ESG data quality

Financial institutions depend on information from borrowers, investees, property owners and other counterparties. Some provide independently reviewed inventories and detailed activity data. Others have limited measurement capabilities or report information that does not match the institution’s needs.

The first challenge is comparability. Two companies may publish emissions for different periods, use different consolidation boundaries or include different Scope 3 categories. Treating their figures as equivalent without checking these differences can create misleading portfolio totals.

The second challenge is missing information. Estimates, sector averages and proxies can help build an initial assessment, but their use needs a defined hierarchy and documented assumptions. A missing value is not zero, and an estimate should remain distinguishable from a counterparty-reported figure.

Data requests should be proportionate to the counterparty and the intended use. A small borrower may need a focused request for energy, location and activity information, while a large corporate borrower may be able to provide more complete disclosures. The objective is to improve usable coverage rather than distribute the same lengthy questionnaire to everyone.

Regulatory complexity

Corporate disclosures, financial-product disclosures and prudential reporting often require information about the same counterparties. However, shared source data does not make the resulting indicators interchangeable. The relevant scope, denominator, methodology and reporting date may differ.

This creates operational risk when each team maintains its own files. Sustainability, investment, risk and compliance teams may receive different versions of the same counterparty information and publish results that cannot be reconciled easily.

A controlled data environment helps separate the source record from its different uses. The institution can maintain one validated emissions figure, for example, while documenting the specific calculations applied in each portfolio or disclosure process.

The principles behind governance, risk and compliance management are relevant here. Clear ownership, version control and approval routes make it easier to implement regulatory changes without losing the history needed to explain previous results.

Climate risk integration

An emissions inventory and a climate-risk assessment answer different questions. Emissions accounting examines an attributed environmental impact. Risk analysis examines how environmental changes, policies, technologies or market developments could affect financial outcomes.

Physical risk may require information about asset location, vulnerability and dependence on infrastructure. Transition risk may require information about energy sources, production technology, sector exposure and the credibility of a counterparty’s investment plans. Neither can be assessed reliably from an emissions total alone.

This distinction matters in credit and investment decisions. A low-emission asset may still face material flood risk, while a high-emission company may have a well-supported transition plan that requires a more detailed assessment than a simple exclusion rule.

Financial institutions therefore need to connect environmental information with established risk processes. The connection should include documented assumptions, model validation and a clear explanation of how the analysis influences decisions, rather than merely adding another dashboard.

Regulatory landscape for finance

The applicable framework depends on the institution, its activities, the products it offers and the jurisdictions in which it operates. A bank, an insurer and an asset manager should not assume that they have identical disclosure obligations.

The starting point is an applicability assessment. It should identify the legal entity responsible, the reporting period, the relevant templates and the teams accountable for producing and approving each output.

CSRD for financial institutions

For institutions within scope, the CSRD establishes corporate sustainability reporting requirements using the applicable European Sustainability Reporting Standards. The European Commission’s corporate reporting guidance also records the evolving scope, timing and standards. Institutions should verify current requirements rather than rely on an old implementation calendar.

The practical work includes defining the reporting boundary, assessing material impacts, risks and opportunities, and establishing reliable information for relevant disclosures. Climate may be particularly significant for a financial institution, but workforce, business conduct and other topics may also require attention.

Relevant topics can include climate under ESRS E1, the institution’s own workforce under S1, value-chain workers under S2 and business conduct under G1, subject to the standards and materiality requirements applicable to the reporting period. Outsourced operations, responsible lending and financial-crime controls may also raise issues to consider within the appropriate governance and risk processes.

Material topics should be determined through the relevant assessment process, not assumed to be identical across all financial institutions. Portfolio activities, outsourced services, customer relationships and the institution’s own operations can raise different issues.

The CSRD guide for finance teams is useful for connecting environmental information with financial controls. Sustainability teams should not be expected to reconcile portfolio exposures, consolidation boundaries and accounting information without participation from finance and risk.

SFDR and product classification

SFDR addresses sustainability-related transparency for relevant financial market participants, financial advisers and products. Its product disclosures need to be distinguished from the institution’s corporate sustainability statement.

The familiar Article 6, Article 8 and Article 9 references describe different disclosure situations under the existing framework. They should not be presented as a simple quality ranking. Article 8 concerns products promoting environmental or social characteristics, while Article 9 concerns products with a sustainable investment objective; neither reference alone explains every feature of a product.

The Commission presented an SFDR revision proposal in November 2025. A proposal should not be treated as an applicable requirement without checking its legislative status and implementation arrangements.

Operationally, institutions need to connect product commitments with the relevant underlying holdings, indicators and evidence. When assessing software for SFDR compliance, teams should test how portfolio changes affect calculations and whether information remains consistent across pre-contractual, website and periodic disclosures.

EU Taxonomy for green finance

The EU Taxonomy distinguishes activities covered by its classification from activities that satisfy the applicable alignment conditions. Eligibility alone does not establish environmental sustainability under the taxonomy.

For credit institutions subject to the relevant disclosure requirements, the Green Asset Ratio concerns taxonomy-aligned assets within a defined regulatory calculation. It is not simply the percentage of the entire balance sheet labelled green, and eligibility should not be confused with the aligned numerator.

The calculation requires careful treatment of exposure types, counterparties, exclusions and the evidence supporting alignment. Revenue and CapEx data from non-financial counterparties may contribute to the relevant calculations, but financial institutions have their own templates and indicators.

Understanding the relationship between the EU Taxonomy, CSRD and ESRS helps institutions reuse source information without collapsing different requirements into one metric. Teams should verify the rules and simplifications applicable to their reporting period before designing the calculation process.

National regulations

Disclosure is only part of the supervisory picture. Institutions also need to consider how environmental, social and governance factors affect their existing risk categories and how those factors are identified, monitored and managed.

The EBA guidelines on ESG risk management address these processes. Their application arrangements distinguish institutions generally from small and non-complex institutions, so the relevant timetable should be checked for the institution concerned.

For banks, applicable Pillar 3 disclosures create another use for risk and exposure information. For insurers, the prudential framework and supervisory context differ, including relevant EIOPA guidance and national supervisory expectations. These requirements should be mapped to the responsible business and risk functions rather than treated as a universal ESG reporting template.

The important operational question is whether the institution can explain how environmental information enters credit assessment, portfolio monitoring and planning. Supervisory processes need a defensible connection between data, analysis and decisions, not just a published emissions number.

European requirements interact with national legislation and supervisory arrangements. Institutions should assess these obligations at entity level, particularly when operating through subsidiaries or branches in several countries.

In Spain, the EINF reporting framework remains relevant to the assessment of applicable non-financial reporting obligations. The roles of Banco de España and CNMV should be considered according to the institution’s activities and supervisory status.

In Germany, institutions need to consider the relevant national corporate-reporting framework, including the legacy CSR-RUG context, alongside BaFin supervision and subsequent legislative developments. A historical reference should not be used as a substitute for checking the law applicable to the reporting year.

A maintained regulatory register helps manage this complexity. It should record the source requirement, its effective period, the responsible entity, the required information and the internal owner, with changes reviewed by the appropriate legal or compliance team.

Practical strategies for ESG management

The strongest approach starts with shared sources and controlled transformations. Rather than building a new collection process for every disclosure, institutions should establish a common data foundation and make each output’s specific rules visible.

This requires cooperation between finance, risk, compliance, investment teams and the people managing counterparty relationships. Technology supports the process, but responsibilities and methods need to be agreed before large-scale automation begins.

Centralize counterparty ESG data

Build a repository that links counterparty identities, exposure records, environmental information and supporting documents. Consistent identifiers are essential when the same company appears in lending, investment and product-management systems.

Each environmental record should include its source, reporting period, organisational boundary, unit and review status. The original evidence should remain accessible so that users can understand what the number represents and whether it is appropriate for the intended calculation.

Environmental and ESG data collection software can help manage imports, requests and validation workflows. However, public-disclosure extraction and third-party data feeds still need checks for entity matching, methodology and the rights governing how the data can be used.

Implement PCAF methodology

Segment the portfolio by the relevant asset classes before selecting methods. Corporate lending, project finance, commercial property and mortgages do not necessarily use identical attribution approaches or input requirements.

Begin with a portfolio segment that is significant and sufficiently understood to test the process. Record the selected method, necessary financial inputs and treatment of missing information, then review the calculation before extending it to additional segments.

Retain the information needed to explain changes between periods. A lower result may reflect reduced exposure, a different portfolio mix or improved estimation, rather than lower emissions from the financed activities. These effects should be analysed separately wherever practical.

Align CSRD and SFDR data flows

Map the information shared by CSRD, SFDR, taxonomy calculations and internal portfolio monitoring. Then identify the differences in boundaries, reporting dates and calculation rules that must remain separate.

For example, a counterparty emissions record may support several outputs, but each output may use different holdings, exposure values or aggregation methods. The source can be shared without requiring the resulting totals to match.

This distinction prevents two common mistakes. The first is recollecting the same information for every team. The second is forcing one result into every framework even when the definitions differ. A common data model should avoid both.

Build climate scenario analysis capability

Forward-looking analysis should start with a defined question. The institution may want to examine how transition costs affect a borrower, how physical hazards affect collateral or how a portfolio behaves under different policy and economic assumptions.

Scenario resources from bodies such as NGFS can provide inputs, while ECB and Bank of England exercises offer relevant supervisory context. However, a scenario is not a forecast, and a supervisory exercise should not automatically be treated as a mandatory internal model for every institution.

Connect the analysis with suitable location, activity and financial information, then document uncertainty and model limitations. Environmental data platforms can provide inputs, but dedicated financial-risk modelling, expert review and validation may still be required.

Strengthen ownership and evidence

Data ownership should follow expertise. Finance should validate exposure and accounting information, risk teams should own risk interpretation, compliance should assess disclosure requirements, and relationship managers can support counterparty engagement.

A documented materiality assessment helps determine which impacts, risks and opportunities need attention in corporate reporting. It should remain distinguishable from credit-risk materiality and from the selection of emissions-intensive portfolio segments.

Escalation routes are equally important. Teams need a process for resolving inconsistent boundaries, disputed counterparty information and missing evidence before these issues reach the final approval stage.

Create an evidence record as information enters the system, rather than assembling documents shortly before assurance begins. Preserve the source, calculation method, assumptions, reviewer and changes affecting the result.

A controlled reporting baseline helps prevent moving targets during review. Subsequent adjustments should be logged separately, with an explanation of why they were made and which outputs they affect.

The preparation of a CSRD data room for assurance provides a useful model for organising evidence. The same discipline can support other portfolio and disclosure processes, while the reviewer retains responsibility for evaluating whether the evidence is sufficient.

How Dcycle supports financial institutions

Dcycle is a data platform for companies that need to collect environmental information, preserve its supporting evidence and reuse it across reporting, operational decisions and improvement processes. For financial institutions, this approach addresses the data-governance layer that sits beneath multiple disclosure and management needs.

The platform is not an auditor, a credit-rating provider or a substitute for a financial-risk modelling system. Its role is to help structure the information and workflows that those processes require, with traceability between sources and results.

The principles for choosing an ESG data platform are particularly relevant in financial services. Institutions should test source connectivity, evidence handling, consolidation and methodology requirements with their own data before assuming that a platform covers every financial-sector calculation.

Dcycle supports collection from business systems, spreadsheets and documents, with information organised in a shared environment. This can reduce repeated requests and help teams maintain the connection between a metric and its supporting records.

For portfolio-related use cases, the institution should validate how borrower, investee and exposure records will be imported and linked. Counterparty identity, reporting periods and boundaries need to remain clear throughout the process.

Data collection needs controls for completeness, consistency and unusual values. Dcycle’s approach combines automated checks with review workflows so that teams can investigate exceptions and preserve the reasoning behind approved information.

Financial institutions should define their own acceptance rules and escalation paths. A software flag identifies a point for review; it does not establish that a counterparty disclosure is complete, accurate or suitable for every regulatory purpose.

A common dataset can support several reporting and management processes without being recollected each time. Dcycle’s broader approach is to keep the information organised and adapt it to the frameworks and business uses the organisation requires.

For financial services, the precise coverage of CSRD, SFDR, taxonomy indicators and prudential outputs must be confirmed during scoping. Shared data infrastructure should not be confused with a guarantee that every specialist template, calculation or filing is available out of the box.

Linking data with documents, calculation assumptions and ownership helps teams explain how a result was produced. This is useful for internal controls, external review and recurring requests from stakeholders.

Dcycle prepares the information environment; it does not issue an independent assurance opinion. Reviewers still need to assess the evidence, while the institution remains responsible for its methods, classifications and published statements.

Before committing to a financial-sector implementation, test the relevant asset classes, attribution methods and counterparty data flows. Confirm the availability of any PCAF-specific calculations, taxonomy ratios, product-level outputs and integrations that the project requires.

The pilot should also clarify user permissions, data-provider usage rights, implementation responsibilities and export formats. A platform is valuable when it supports a repeatable, controlled process, not when specialist capabilities are assumed without verification.

Dcycle is therefore most relevant to institutions looking for a stronger environmental data foundation that can work alongside their financial, portfolio and risk systems. The intended result is less repeated collection work, more accessible evidence and information that remains useful beyond a reporting deadline.

Measure portfolio data quality and coverage

A portfolio emissions total is important, but it does not reveal whether the underlying process is improving. Institutions also need indicators that show coverage, quality, responsiveness and control maturity.

These management metrics should have defined denominators and consistent periods. Their purpose is to identify operational improvements, not to create another set of figures that cannot be reconciled.

Track the share of the defined portfolio included in each analysis, measured against the relevant exposure or holdings denominator. Explain exclusions and distinguish assessed positions from positions that lack sufficient information.

Coverage should be reviewed by asset class and important portfolio segment. A strong aggregate percentage can conceal significant gaps in a smaller segment with high environmental or financial relevance.

Monitor how much of the assessment uses counterparty-reported information and how much relies on estimates. Distinguish independently reviewed information from unreviewed disclosures where this affects the methodology or intended use.

The aim is not to remove estimates at any cost. It is to make their use transparent and prioritise improvements where better data would most meaningfully affect the analysis.

Track the reporting periods represented in the dataset, overdue requests and the time needed to resolve missing or inconsistent information. Old data may remain usable for a particular purpose, but its age should not be hidden.

Review whether requests are proportionate and whether counterparties understand what is needed. Improving request design can be more effective than repeatedly sending reminders for a questionnaire that is too complex.

Review portfolio emissions alongside exposure movements, portfolio composition and methodological changes. This helps explain whether a lower total reflects changes in financed activities or changes in the institution’s holdings and calculations.

Emissions tracking software can support recurring comparisons, provided the underlying boundaries and methods remain visible. A trend should be accompanied by an explanation, not treated as proof of real-world decarbonisation by itself.

Monitor the proportion of significant metrics with approved sources, documented methods and completed reviews. Track unresolved exceptions and the time required to answer review questions.

These measures reveal whether reporting readiness is maintained continuously. A process that produces consistent evidence every cycle is more resilient than one that depends on last-minute reconstruction by a few individuals.

Build a practical implementation plan

The following sequence is an illustrative planning approach, not a guarantee that every institution can complete implementation within three months. Portfolio size, integration complexity and data availability will influence the pace.

The aim is to establish a controlled pilot and a credible expansion plan. A smaller process that works end to end is more useful than a broad deployment whose results cannot be explained.

Identify the intended outputs, select a representative portfolio segment and map the available financial and environmental information. Confirm who owns the sources, who approves methods and who will use the results.

Review the relevant sustainable finance frameworks and create a requirement-to-data map. Record where the same source can support multiple outputs and where separate rules must be maintained.

Import a controlled dataset, check counterparty matching and test the selected methods. Use a mix of complete and incomplete records so the pilot demonstrates how exceptions and estimates are handled.

Ask finance, risk and reporting users to trace a sample result back to its source. The test should expose missing metadata, unclear ownership or calculation assumptions before the process is extended.

Prepare an evidence package and run an internal review of representative results. Resolve significant issues and document what remains outside the pilot’s coverage.

Use the findings to prioritise additional asset classes, data integrations and counterparty engagement. Agree the resources, dependencies and acceptance criteria needed for the next phase rather than assuming the pilot automatically proves full-portfolio readiness.

From reporting to better decisions

Well-governed data can improve more than disclosure preparation. It gives relationship managers a clearer basis for discussing counterparties’ energy use, investment plans and information gaps, while helping risk and investment teams understand the limitations of their analysis.

Verified ESG data for green finance can also support more credible conversations about financing and investment decisions. Reliable evidence does not automatically create favourable financing terms, but it reduces ambiguity about the information available for assessment.

The business value comes from using information consistently. Reporting, portfolio monitoring and client engagement become more useful when they draw on shared sources, documented methods and visible responsibilities instead of disconnected annual exercises.

Conclusion

Sustainability in the finance sector depends on understanding both the institution’s own activities and the portfolios it finances or manages. Meaningful progress requires clear measurement boundaries, reliable counterparty information and a distinction between environmental impact and financial risk.

The most effective operating model collects relevant information once, retains its evidence and applies the specific rules required by each output. CSRD, SFDR, taxonomy and prudential processes can share sources without using identical metrics or serving identical purposes.

Financial institutions should prioritise controlled portfolio coverage, transparent estimates and methods that remain explainable as the data improves. A practical pilot can reveal integration gaps and strengthen responsibilities before the process is expanded.

Dcycle can support this approach by organising environmental data, validation workflows and traceability. Combined with appropriate financial methodologies and specialist risk systems, that foundation helps institutions move from fragmented disclosures to information that supports reporting, client engagement and operational decisions throughout the year.

Frequently asked questions (FAQs)

Related articles

• CSRD for finance: the complete guide for CFOs
• CSRD resource hub
• Carbon footprint collection