Response to discussion on Technical Advice on selected KPIs under the Taxonomy Disclosures Delegated Act under Article 8 of the Taxonomy Regulation

Q1. Please identify your organisation (if applicable) and indicate the capacity in which you are responding to this consultation:

Q2. Do you agree with the analysis presented?

As a general remark, Climate Strategy & Partners — an advisory firm working on climate finance, EU budget
instruments and industrial decarbonisation with public and private financial institutions, is responding to this
consultation with its own anaylsis that sums up to the detailed technical thinking of WWF European Policy
Office and Better Europe, whose analysis we have drawn on in developing our positions.

While not a preparer as such, Climate Strategy do analyse and evaluate sustainability data, specially for SMEs, trying to see how to close the current climate data gap that we face without putting the burden on SMEs. 

Following WWF suggestion, we agree with the four-service mapping. But being flow-based is the KPI's function, not its flaw: the GAR captures the stock of lending and structurally misses banks whose sustainability impact runs through deal facilitation. Deleting this KPI would make the most capital markets-intensive banks the least transparent under Article 8. Volatility is handled by multi-year presentation, as for any flow metric; the confidentiality risk can be addressed: it is confined to thin deal volumes, is managed by the 10% threshold and aggregation, and shrinks as EuGBS issuance grows. Finally, paragraph 35 is circular: the EBA proposes to delete the FINREP datapoints and then cites prospective data unavailability against the KPI — supervisory reporting should follow the disclosure decision, not pre-empt it. 

In your response, where applicable, please assess whether voluntary disclosure of this KPI would be meaningful and feasible, and hence should be considered.

Option (b) — mandatory, narrowed, with the 10% threshold. 

This is real simplification (only ~16% of institutions would exceed the threshold, i.e. very few) while keeping disclosure exactly where it matters: the large capital-markets banks. Following WWF suggestion, we oppose option (a): qualitative-only 'composition, trends, objectives and policy' text is not comparable, likely not auditable, and invites narrative greenwashing without a quantitative anchor. Voluntary disclosure is no substitute — leaders disclose, laggards abstain, comparability collapses. We also ask EBA to retain the four relevant FINREP 22.1 datapoints in the parallel ITS revision, so the narrowed KPI costs preparers nothing extra. 

Q4. Should additional items from FINREP Template 22.1 also be included in the revised KPI? Please consider in particular that (i) have been assessed as unrelated to capital market activities, e.g. ‘custody and other related services’, and (ii) those that have not been mapped to the Fees and Commissions KPI, such as ‘structured finance’, ‘loans granted’, ‘commodities’.

Assess structured finance; keep the transactional services out. 

Green securitisation and green ABS have a direct use-of-proceeds link to aligned assets and will grow under the securitisation review and the EuGBS — structured finance therefore merits assessment for inclusion. Custody, payments, clearing and FX are rightly excluded as transactional; 'loans granted' belongs analytically to the GAR. 

Q5. Do you have evidence and/or arguments on potential costs and benefits associated with disclosing the Fees and Commissions KPI?

Costs are front-loaded and shared; benefits grow with green issuance. 

The FINREP mapping exists and the counterparty alignment data is the same data that banks must obtain for the GAR — the marginal cost of the narrowed KPI is modest. The benefit scales up with green issuance volumes, so a cost-benefit test anchored to 2026 volumes structurally understates the steady-state value; a KPI deleted now would likely have to be rebuilt later at higher cost. It is also the only instrument allowing users to compare investment banking franchises on their contribution to sustainable capital flows. 

The benefit side also grows for reasons outside private issuance volumes. From 2028 the ECF InvestEU Instrument becomes, under the exclusivity clause in Article 23 of the ECF Regulation, the single access point for budgetary guarantees and financial instruments across the whole EU long-term budget, and credit institutions are among the intermediaries through which those instruments reach end-beneficiaries. Where a bank's capital-markets services support activities backed by that architecture, the narrowed KPI is one of the few standardised measures of that contribution. A cost-benefit test anchored to today's private green-issuance volumes therefore understates the steady-state benefit twice over.

Q6.Do you agree with the analysis presented?

Partially — fair description, wrong conclusion. 

The descriptive points are fair, but the test applied is again 'is this the GAR?'. The KPI's job is different: to give visibility on whether a EUR 5 trillion asset class supports the market for sustainable instruments. Leaving it with zero quantitative Article 8 coverage is a blind spot, not proportionality. And paragraph 66 inverts its own logic: that market making is only one liquidity channel argues for precise labelling of a narrower KPI, not for abandoning quantitative disclosure altogether. 

In your response, where applicable, please assess whether voluntary disclosure of this KPI would be meaningful and feasible, and hence should be considered.

Option (b), strengthened — a mandatory Market-Making KPI for Taxonomy-aligned securities. 

The positive case, in four steps: 

• Liquidity comes first. Investors buy green bonds and green issuers' shares at issuance only if they can trade them afterwards. Dealers' commitment to quote aligned instruments directly conditions the cost of green capital and the scale-up of the EU Green Bond Standard. This is the single most policy-relevant slice of the trading book — exactly the enabling function the Commission's Call for Advice identifies.
• The scope is documented and auditable. Market making runs on written agreements with trading venues (RTS (EU) 2017/578), so the covered instruments are known in advance and stable. This answers the EBA's operational objection (para 61): no trade-by-trade assessment across the whole book — only a finite, contract-defined instrument list, assessed with the same issuer data the bank already needs for the GAR.
• A ratio design removes the noise. Construct the KPI as aligned instruments under market-making agreements over all instruments under such agreements. Numerator and denominator move with the same market conditions, so the 'external market factors' the EBA worries about (para 60) largely cancel out — what remains is the bank's own choice of what it commits to make markets in.
• Do not shrink it to bonds only. Equity liquidity matters most for smaller pure-play green issuers. Scope should be all aligned debt and equity under market-making agreements, with a mandatory sub-line for European Green Bonds (trivially identifiable at ISIN level). Qualitative disclosures (Annex XI) should complement, not replace the KPI. 

Q8. Do you have evidence and/or arguments on potential costs and benefits associated with disclosing the Trading Book KPI?

The market-making redesign is cheap; the information exists nowhere else. 

The instrument universe under market-making agreements is small and stable relative to the full trading book; alignment data is reused from the GAR; volume infrastructure already exists via the Basel G-SIB reporting the current KPI is modelled on. The benefit: the only standardised indicator of banks' contribution to secondary-market liquidity for sustainable instruments — increasingly valuable as the stock of aligned and EuGBS instruments grows.  

Q9. Do you agree with the analysis presented and the conclusion to narrow down the investment firms’ KPI?

Yes. 

The four retained services are exactly those where the firm exercises discretion or influence over where capital goes; the excluded services are client-driven and administrative. No decision-useful information is lost. 

Please explain if you think that other services, e.g. execution of orders on behalf of clients, should be included in the revised KPI.

Please explain if you think that other services, e.g. execution of orders on behalf of clients, should be included in the revised KPI.

Yes — and keep execution of orders out. 

Execution is client-driven; adding it back would dilute the KPI with flows the firm does not allocate. Discretionary situations resembling order handling are already captured under portfolio management. 

Q11. Do you consider it more appropriate for the KPI to reflect the value of assets covered by the investment services rather than the monetary benefits generated?

Yes — with three conditions. 

An asset-based measure is more stable, insensitive to pricing structures, and comparable across the three fiduciary KPIs — the same 'one method for managed assets' point our response makes to ESMA (Q23 in ESMA public consultation). Conditions: (i) discretionary and advisory assets reported as separate lines, since the firm's influence differs; (ii) the weighting methodology fully aligned with the asset managers' template; (iii) for underwriting/placing — flow activities with no asset stock — capture the value of instruments placed during the period, so the primary-market facilitation role is not lost in the redesign. 

Q12. Do you agree with the analysis presented and the conclusion to align the grandfathering rules with the approach set out in EU Green Bond Regulation?

Qualified agreement — consistency yes, loosening no. 

Cross-framework consistency is valuable, but the EBA should say plainly that seven years instead of five extends the time exposures count as aligned against outdated criteria. We support alignment only as a package: (i) the EuGBS Article 8(3) transparency duties — a published alignment plan with external review — become mandatory in the DDA, not discretionary; (ii) the EuGBS's unlimited grandfathering of already-allocated proceeds is not imported — convergence should be on the stricter current DDA reading (para 88), which caps both; (iii) a dedicated, digitally tagged disclosure line shows the share of each KPI numerator relying on grandfathered criteria, so users can discount stale alignment and track its run-off. That one line converts grandfathering from an invisible integrity risk into usable information at negligible cost. 

The same line has value beyond private users: public support committed against a technical-screening-criteria vintage carries the same stale-alignment risk, and a tagged run-off figure is what lets any tracking system — private or public — discount it consistently rather than each building its own adjustment

Q13. Do you agree with the analysis presented and proposed conclusions?

Yes: users need retrievability, not a blend. 

Averaging a bank's GAR with other businesses' KPIs produces a number with no economic meaning, and the para 114 double-counting point is technically correct. What users need is (i) the group KPI under the parent's rulebook, (ii) separately identifiable, digitally tagged KPIs for the materially different businesses, and (iii) machine-readable access — then any user builds whatever aggregation suits their purpose. Simplified subsidiary templates are acceptable if standardised (one Commission-specified format) and tagged, so 'lighter' does not become 'non-comparable'. 

i. Should the KPI disclosed by the parent undertaking, reflecting the group’s main or prevalent activity, be used? If not, what alternative methodology should be applied?

Yes, on one condition.

Agree, as in our ESMA response (Q10 there): downstream rules that need one figure should take the parent-rulebook KPI — a synthetic aggregate would recreate the weighted-average problem. Condition: the KPIs of materially different subsidiaries remain separately disclosed and tagged, so an institution financing a specific subsidiary can use the KPI that describes what it finances.  

ii. In the case of a credit institution-led group, such as a financial conglomerate, should the consolidated KPI disclosed by the parent undertaking also incorporate the assets of the insurance (and non-financial) undertaking, rather than accounting for the insurance (and non-financial) undertaking solely through the equity method? Please justify your answer in terms of feasibility, usability and transparency of sustainability information.

No — keep the equity method. 

Folding insurance assets into the GAR would break the link with prudential consolidation on which the GAR's verifiability rests and would blend incomparable methodologies. Transparency on the insurance business is better served by the subsidiary's own insurance-template disclosures, cross-referenced from the group report

Q15. Which KPIs you do think are necessary and should be required for groups with mixed activities where the credit institution is the parent company such as credit institution-led financial conglomerates, credit institution-led financial holding companies and relevant mixed financial holding companies?

Parent GAR + off-balance-sheet AuM KPI, plus tagged subsidiary KPIs above a cumulative 10% threshold. 

At parent level: the GAR and the off-balance-sheet AuM KPI. From material subsidiaries: the investment-firm KPI, the insurance underwriting and investment KPIs, and the asset-manager KPI (ideally on the merged template, Q17), each at subsidiary level and cross-referenced. The 10% trigger should carry the same safeguards sets out to ESMA (Q12 there): expressed as cumulative across all 'other' businesses with an anti-fragmentation clause; below-threshold groups still disclose the nature and turnover share of those businesses, tagged; and for financial subsidiaries a supplementary total-assets test, since their economic weight can far exceed their turnover share

One fact pattern deserves explicit treatment in this design: groups led by national or regional promotional banks and other publicly-owned financial institutions. These are directly in scope of the credit-institution rules, and their role is set to expand — amendments under negotiation to the NRPP Regulation would require that national promotional banks and publicly-owned banks or financial institutions be involved from the outset in preparing National and Regional Partnership Plans wherever financial instruments are envisaged. Their group structures routinely combine lending with guarantee, fund-management and advisory subsidiaries deploying EU-backed instruments, which is exactly the mixed-activity pattern the threshold and sub-group rules must handle. The cumulative reading and anti-fragmentation clause matter here in particular, since deployment functions are often spread across several small entities, none individually material. WWF's supplementary total-assets test for financial subsidiaries applies with equal force, as guarantee capacity is poorly captured by turnover share.

Q16. Do you agree with the analysis presented and the conclusion that OpEx financing should not be explicitly incorporated in the methodology for calculating financial undertakings’ KPIs? Please provide your rationale for supporting the explicit inclusion of provisions allowing financial undertakings, on a voluntary basis, to use the OpEx KPI disclosed by their counterparties when calculating their own KPIs.

Disagree with the blanket 'no' — support the voluntary option with every transparency disclosure mandatory (same answer as to ESMA). 

For companies whose green performance lives in their operating costs — the rail/grid/water/forestry/fleet family — a ratio built on CapEx alone structurally understates how green the financing is; the EBA's own Box 5 (18% vs 32%) shows the size of the gap. Allowing the voluntary blend closes a real measurement gap, including for working-capital lending to already-aligned operations. The EBA's comparability and greenwashing concerns are answered by conditions, not prohibition: 

• the three disclosures are a mandatory condition of use — whether the blend is used, the CapEx/OpEx split inside the blended figure, and confirmation of consistent application across the whole book — each digitally tagged, so a blended GAR can never quietly inflate reported alignment;
• only company-reported OpEx KPIs may be used — never estimates produced by the bank or a data vendor;
• the option, its conditions and its disclosures must be identical across the EBA, EIOPA and ESMA frameworks, so one company's OpEx is treated the same way in a bank's GAR, an insurer's investment KPI and an asset manager's portfolio KPI;
• usefulness is contingent on ESMA's redesign of the OpEx KPI itself — Following WWF's proposal there (a two-part mandatory KPI: R&D for all, plus operation/maintenance of aligned assets for a fixed high-relevance activity list) is what makes the input meaningful; the voluntary-use provision should enter into application together with that reform. 

Where a bank uses the blended figure, the digital tag should link the portfolio-level number back to the underlying company KPI so any user can trace and recompute it — the KPI-to-KPI connection the central digital tool proposed under Q19 would make automatic

Q17. Should the credit institution’s AuM KPI (Annex V) and the KPI for asset managers (Annex II) be merged? Please explain why and how, including the content of the information that should be retained in a potential merged KPI.

Yes — identical to our suggestion to ESMA (Q23.1 there). 

The activity is identical, so the method — ideally the template too, with an entity-type identifier — should be one and the same. The merged methodology should also be the reference for the AuM-based redesign of the investment-firm KPI (Q11), so all three fiduciary KPIs converge on one method. 

Q18. What are your views on the design of off-balance sheet item reflecting financial guarantees to financial and non-financial undertakings in a potential merging of the asset management activities as presented?

Keep guarantees separate — split the composite KPI in two. 

Financial guarantees are credit substitutes, functionally closer to lending than to asset management; folding them into an AuM figure would obscure a distinct impact channel. If the merger proceeds, split the composite off-balance-sheet KPI into a FinGuar KPI and an AuM KPI, each with its own numerator and denominator. 

This distinction is about to become materially more important. Financial guarantees are the principal channel through which EU budget support reaches the real economy: the ECF InvestEU budget guarantee achieves leverage ratios of around 14x, against roughly 4.8x for many locally-designed guarantee instruments, and the derisking case for the ECF rests on exactly that multiple — a previous InvestEU generation achieved a 12x multiple on private capital mobilised. From 2028 a single guarantee architecture becomes the access point for budgetary guarantees across the whole EU budget, with credit institutions among the implementing and intermediating channels.

Folding guarantees into an assets-under-management figure would therefore obscure, at precisely the wrong moment, the one KPI capable of showing how much guarantee capacity is supporting Taxonomy-aligned activity. AuM and guarantees also answer different questions: AuM measures assets held on behalf of clients, while a guarantee is a contingent credit substitute that changes what a third party is able to finance. Following WWF's proposal to split the composite into a separate FinGuar KPI and AuM KPI, each with its own numerator and denominator, should be retained on that basis. Where guarantees are counter-guaranteed by a public budget instrument, that share should be a separately tagged line, so users can distinguish a bank's own risk-taking from publicly-backed capacity.

Q19. Do you have any additional comments or suggestions to improve, simplify or clarify the disclosure requirements within the scope of this Call for Advice, including through redrafting of instructions, and clarification of definitions, or the streamlining of terminology?

Four points — the central one identical to our Q24 proposal to ESMA. 

• Evaluate each KPI against its own job. The review should test each Article 8 KPI against its own information function (stock, facilitation, liquidity, fiduciary) rather than against the GAR; several conclusions in this paper follow from an implicit 'is it the GAR?' test that is not relevant, as no complementary indicator can pass it.
• Simplification through infrastructure, not deletion. The largest untapped burden reduction is a Commission-hosted interactive reporting and comparison online tool: guided questionnaires generated from machine-readable criteria, automatic template generation with validation, direct ESAP filing — and a public interface where every disclosure is screenable across firms and where a bank's GAR links back to the company KPIs it is computed from. Every 'relevance' problem this paper diagnoses is partly an access problem, solvable centrally, once.  
• Data quality transparency. Templates should separately identify, as tagged datapoints, the share of numerators relying on grandfathered criteria (Q12) and the share relying on estimates or proxies, so headline KPIs can be quality-adjusted by users.

• Fifth: design for the users who will deploy the next EU budget.

• The relevance assessments throughout this paper are framed around investors and preparers. A third user class is absent: the public financial institutions that will deploy EU budget resources over 2028–2034 and the credit institutions intermediating for them. From 2028 the ECF InvestEU Instrument becomes the single access point for budgetary guarantees and financial instruments across the whole EU long-term budget, and National and Regional Partnership Plans will deploy some €783 billion required to align with Union climate objectives.

• These institutions already screen counterparties and assets for climate performance — and the fragmentation costs of not having a common reference are visible today. Under InvestEU sustainability proofing, implementing partners may use the Taxonomy's DNSH criteria on climate mitigation "to the extent possible", but may equally use the EIB's Paris alignment low-carbon criteria or another internationally recognised methodology, so the same investment can be screened three different ways. Environmental proofing runs on a largely qualitative natural-capital checklist rather than an objective-by-objective assessment across the six Taxonomy objectives, with the result that circular economy and pollution prevention are not systematically assessed as discrete categories.

• Two consequences for this review. First, each KPI or datapoint deleted here widens the space in which parallel methodologies get built over the same undertakings, and that cost falls on the same preparers — it belongs in the cost-benefit assessment, not outside it. Second, the infrastructure proposed in the second bullet above should be specified as interoperable with EU budget climate tracking, with ESAP as the common access point, so that one filing serves both. Timing makes this actionable: amendments to the Disclosures Delegated Act are due to enter into force in Q3 2027, immediately before the 2028–2034 budget period opens.

• An SME layer inside the same infrastructure. Post-Omnibus, counterparties below the CSRD scope are excluded from financial undertakings' KPI denominators — which means banks cannot evidence aligned SME lending. This is a banking-specific gap with a banking-specific cost. The VSME standard offers a proportionate basis for closing it, and working practice already exists: positive lists of eligible clean assets and technologies attached to standardised financial products — such as the EBRD Green Eligibility Checker now used under InvestEU, and Spain's ICO green-line positive-list tool — permit self-declarations of compliance for small-scale retail transactions. CS suggests the EBA recommend that standardised, VSME-based counterparty data, digitally tagged and voluntary for the SME. National promotional banks and local retail channels reach client networks of thousands, and in some cases millions, of SMEs; without a standardised data layer, that entire population stays outside.

Terminology and drafting. Post-Omnibus, service labels should be harmonised between the DDA annexes and FINREP, and stable interpretations moved from Commission FAQ notices into the legal text, so the reporting rules are complete in one single instrument.

Name of the organization