Testing the taxonomy: insights from the PRI Taxonomy Practitioners Group

Further reading

This report assumes a baseline level of understanding of the taxonomy. For those new to the topic, here are suggestions for further reading.

• TEG Final Report on EU Taxonomy
• Technical Annex to the TEG Final Report on the EU Taxonomy
• PRI Article on the EU Sustainable Finance Taxonomy
• PRI EU Taxonomy Investor Briefing
• Technical Expert Group on Sustainable Finance (TEG) – Frequently Asked Questions

The conclusions and recommendations of this report, set out in the following sections, are based on the case studies below. To look at individual approaches to taxonomy implementation, for different asset classes and geographies, please follow the links under each organisation name.

_{The case studies are being published periodically on the PRI website, so at any given moment fewer than 37 case studies may be available. All analysis in this document, including any stated statistics, are based on analysis of the full suite of 37 case studies.}

Scope of case studies

The investors in the group considered a range of asset classes, regions and investment styles in their case studies.

Some 89% of investors were investment managers and the remainder were asset owners. Almost one quarter of the investors came from the US, with the others based in Europe. The regions represented in their case studies were Europe or countries within Europe (32%), North America (16%), South America (5%) and Asia (3%). Some 57% of case studies considered activities globally.

Most case studies considered listed equity and fixed income, with a smaller number analysing real assets (infrastructure, real estate, forestry). None of the case studies assessed sovereign bonds, reflecting the absence of a clear methodology to do so.

Assessing alignment

To assess whether an issuer was performing taxonomy-aligned activities, the majority of participants used third party data providers, although many supplemented this with in-house research, direct assessment of corporate reports and data available through not-for-profit organisations, such as the CDP or Climate Bonds Initiative (CBI). External validation consultants were used to a greater extent when assessing real assets, in sectors such as real estate, infrastructure and forestry.

Many participants started by assessing the extent to which their existing data services might be repurposed for taxonomy assessment, initial screening or as a point of reference for their analysis. Participants referenced a range of services including Bloomberg’s WATC EUTAX tool, MSCI, ISS, FactSet, Trucost S&P, Carbon Delta, GS Sustain Taxonomy mapping tool and RepRisk. Several participants developed proprietary frameworks for mapping taxonomy activities to existing classifications systems.

Participants typically found it easier to access climate change mitigation data. Some participants were able to draw on firm-level assessments of elements of the minimum safeguards, such as the OECD Guidelines for Multinational Enterprises and ILO Core Labour Conventions. However, for a fuller assessment of Do No Significant Harm (DNSH) criteria and minimum safeguards, investors used a range of proxies, such as UN Global Compact compliance, IFC performance standards or NGO assessments, while noting that this did not always perfectly match the DNSH requirements of the taxonomy.

Financial metrics

The most common method of calculating alignment for listed equity was quantifying the proportion of turnover aligned with the taxonomy. Capex (and to a lesser extent Opex) were considered in sectors such as utilities, where these metrics were recognised as more relevant to the asset base. The analyses were limited by the fact that expenditure data was typically not available at the required level of granularity, and some investors sought to engage corporates on this point. In the case of Green Bonds (GBs), investors relied on use-of-proceeds to calculate the assessment. Calculating taxonomy alignment for property, infrastructure and other single assets that do not have a revenue base was more challenging, with some participants relying on assessments of the financing of projects or assets, while others sought to have all assets within a fund validated externally to show compliance with the taxonomy. In forestry, only taxonomy-aligned revenues derived from certified assets that could be linked to sustainable activities were considered. These taxonomy-aligned revenues for each underlying investment were weighted relative to the investment’s share of the total capital deployed. In real estate, once a suitable building or buildings had been identified, the financial metrics considered for evaluation of the renovation process included fees generated, total investment costs and energy cost savings resulting from renovation measures.

Establish a framework

• Ensure adequate resources are set aside and that management is aware of this regulatory requirement. Decide whether there are sufficient internal resources to collect data from companies to implement the taxonomy independently, or whether external liaison with data providers is required.
• Integrate the taxonomy into the investment strategy.
• Manage expectations. The current taxonomy criteria are advisory and may change. Participants typically found relatively low percentages of taxonomy-alignment, reflecting differences between the taxonomy framework and some existing methods of assessing sustainability. Learn from the process but recognise current limitations.

Develop a process

• Start early. Allocate time and expertise for detailed analysis – especially for identifying DNSH criteria and screening against them at company level. Consult with specialists.
• Quantify findings as far as possible. Initially, it is better to integrate ESG criteria deeper into investment activities, to avoid higher implementation costs in future.
• Start small. Test one sector/product/region - even if initially case-by-case, using a small sample group on a share of a portfolio or one sector/product/region.
• Apply a step-by-step approach. Work through DNSH and minimum safeguards screening before the more complex substantial contribution (SC) screening.
• Take a bottom-up approach to fairly assess company alignment with the taxonomy. There is insufficient reported company data to make a top-down assessment.

Identify the challenges

• Strictly adhere to thresholds wherever possible. Do not pick and choose areas to assess or liberally interpret guidelines. Participants highlighted enabling activities as a particular risk. These should only be considered where the activity is explicitly defined as enabling in the taxonomy, rather than including any company providing a service to another taxonomy-aligned company.
• Carefully consider reliability levels for different sources of data, noting that in the absence of high quality corporate disclosure, all in-house or third party assessments will involve estimation, modelled data or use of proxies.
• Verify with companies directly when in doubt or if there is a lack of data.
• Provide context for results, which may help explain a lack of alignment.

Find solutions

• Engage on data. Engage early with investee companies, particularly smaller companies or non-EU companies, being clear as soon as possible about requirements to publish relevant technical data for more aligned reporting. Doing so will allow investors to make allocations and fund managers to include companies in a portfolio or opportunity set.

• Share with partners. Cross-check findings with other investors investing in the same companies and collaborate on publicly available information. Be clear: agree on best practice, clarify the data set and when it is likely to be available, be transparent about methods (data, models and sourcing of data).
• Work with data providers. Use data provider tools to ease the burden of manual input. Some participants mentioned using several data providers to compare results, build confidence and fill gaps in data provision.
• Support innovation and improvement from data providers to deliver the data required to carry out the assessments (more granular Opex/Capex data disclosure as opposed to mainly revenue disclosure), assess data gaps and encourage holistic, off-the-shelf solutions and further product development. Data providers should be encouraged to cover all steps of the taxonomy assessment.
• Investigate validation and external assurance of both underlying taxonomy data and processes followed by external managers and data providers.

“While detailed in-house analysis and application of the EU Taxonomy is very time-consuming, we believe there is value in understanding all the pitfalls. This should help to assess the robustness of the approaches taken by different data providers, even if applied on a limited sample of issuers or GBs.”

“There is value in nurturing an understanding of a company’s culture, regulatory environment and the industry in which it operates. This will avoid overburdening companies to report information that may not add value to the investment process.”

“Industry groups and trade associations have a role to play in translating the taxonomy into an industry reporting standard, supported by best practice processes and assurance requirements.”

Group members identified the key challenges they faced in implementing the taxonomy, and the solutions they found. The challenges can be broadly defined as problems around sourcing data, matching data to the taxonomy, adapting to the granular taxonomy approach and creating new processes to address the taxonomy. Various solutions were proposed – but not all will be suitable for every investor. There were also specific challenges for those dealing with Green Bonds.

Table summarising main challenges and solutions

Focus on green bonds

The taxonomy is an innovative disclosure regulation. While the high-level framework is clear, successful deployment of the taxonomy will benefit from greater support from policymakers and regulators. This section summarises the recommendations made by the group to policymakers. It is divided into three key themes:

1. Data
2. Guidance and supervisory expectations
3. Taxonomy development.

Data

• Future corporate disclosure obligations should go beyond the scope of the existing corporate reporting requirement (and implicitly, the scope of the current Non-Financial Reporting Directive) and consider private companies, as well as potentially strengthening expectations on issuers of green bonds.
• Much of the available data is not yet at sufficient granularity. More activity-level reporting, and particularly reporting of revenues per business or product, would help investors who are undertaking detailed company analysis to avoid the use of estimates, or company-wide measures as proxies. However, this may not be necessary if corporates provide reliable firm-level disclosure of taxonomy-alignment.
• Standardised reporting expectations for some of the more subjective or qualitative elements of the taxonomy (e.g. minimum safeguards, DNSH criteria) would help to improve provision of data.
• Companies should be encouraged to standardise their reporting methodologies, potentially using tools, templates and the possible future development of an EU database of taxonomy reports.
• The EU should encourage audit or assurance of corporate taxonomy-alignment corporate data.
• Some investors suggested a role for policymakers to engage data providers to limit discrepancies, agree on common definitions on key terms and standardise the approach to subjective elements, while others suggested that investors should collaborate to achieve the same goals.

Guidance and supervisory expectations

Investors piloting the taxonomy identified several substantive issues on which further guidance should be provided. These are in addition to recommendations for further development of the taxonomy framework, discussed below. Key guidance requests included:

• Development of a more structured approach for demonstrating compliance with minimum safeguards and qualitative DNSH criteria (such as a checklist). This is particularly challenging outside of the EU as some DNSH criteria are aligned to EU regulations.
• Tools to support the application of the taxonomy in non-EU markets. In particular, the development of correspondence tables for alternative, non-EU standards as proxies (building from the TEG’s own recommendations, for example, in green buildings).
• Further development of the adaptation criteria, which are challenging to interpret due to their qualitative nature.
• Clarity on the time period for a taxonomy assessment. For example, how frequently should investors review companies for potential violations of DNSH or minimum safeguards?
• Guidance on reporting investments that may substantially contribute to more than one environmental objective.
• Further guidance on treatment of financial metrics. For example, whether to use an income statement or cash flow approach when calculating turnover.
• Further specificity on expenditure metrics – for example, identifying situations when Opex would be an appropriate metric.
• Further guidance on accounting for GBs. For example, in cases where GBs proceeds directly finance loans or equity investments rather than expenditures.
• How to address potential double counting in multi-asset strategies.
• How to interpret taxonomy requirements for funds beyond the typical listed equity funds. For example, for funds with a defined fundraising and deployment period, advice is needed for different stages of investment (fundraising capital fully committed, deployed or in the process of being invested). We noted that none of those undertaking case studies attempted to assess the alignment of sovereign bonds, which indicates that investors do not yet consider this assessment feasible. Significant methodology development is likely to be necessary here.
• Though the proposal to allow a “potentially aligned” category was welcomed, investors also identified it as a significant possible source of divergence, dependent on strictness of interpretation, and that greater clarity would help preclude this risk.

Alongside this further guidance, greater clarity around supervisory expectations was requested. In particular:

• Due to the time periods within the taxonomy regulation, the data required to properly perform a taxonomy assessment may not be available before the first mandatory reporting period. Capacity gaps mean it is likely to be imperfect even after it becomes mandatory. Supervisors should ensure that investors are able to clearly communicate the limitations of their analysis and the underlying data.
• Ensure greater alignment with other EU disclosure requirements, in particular the forthcoming Regulation on Sustainability-Related Disclosures in the Financial Services Sector.
• Finally, supervisors should give consideration to how taxonomy regulation interacts with other regulatory requirements, including marketing and product communication.

Taxonomy development

When considering the specific indicators and presentation of the taxonomy, group members suggested that policymakers should:

• Where possible, simplify presentation on the criteria and remove subjectivity, or provide greater clarity through guidance (for example, what does it mean to be “near” a sensitive site for biodiversity in the context of DNSH criteria?)
• Develop some of the existing criteria to provide greater certainty and clarity (for example, the IT criteria were felt to be under-developed).
• Widen the sectors covered. Investors indicated a desire to see more enabling activities/solutions reflected in the taxonomy.
• Provide greater clarity on which sectors have been intentionally excluded from the taxonomy and therefore are always considered to cause significant harm.
• Clarify the role of “neutral” activities that neither substantially harm nor significantly contribute to environmental objectives – this may help to contextualise the relatively low proportion of taxonomy alignment seen in some case studies.

Beyond the specific technical screening criteria, policymakers should:

• Consider and reflect on the use of expenditure metrics. Capex and Opex, while useful for assessing how a company’s spending is in line with environmental commitments, may lead to unintended consequences such as bias against capital-intensive industries.
• Work internationally to encourage harmonisation of taxonomies (while also supporting investors required to analyse non-EU securities for immediate taxonomy implementation, for example through the development of correspondence tables).
• Provide greater clarity on the future development of taxonomy criteria.
• Support the development of taxonomy-aligned activities through sectoral regulation (e.g. the forthcoming renovation wave policy programme).
• Carefully monitor implementation to avoid unintended consequences (for example, bias towards sectors with simpler compliance requirements). Passive products in particular should be carefully examined, as they do not benefit from human judgement to counterbalance potential biases.

“Ensuring that the taxonomy is aligned with company non-financial reporting requirements is integral to the success of the regulation.”

“Compliance, at least at this stage, will be on best-efforts basis. This will hopefully change as better data becomes available.”

“We would suggest the following: Providing an exhaustive mapping tool with all NACE sub-sectors and systematic matching with the following information: Activity, definition, and if such activity (1) has a positive contribution and technical criteria; (2) has no positive contribution; (3) has positive contribution but no technical criteria yet; (4) has not been covered yet.”

“Push to ensure that the taxonomy becomes a global standard that is widely used by leveraging market participants and institutional investors.”