Comment: Baby don’t harm me

By Patricia Volhard (partner) and John Young (international counsel) at Debevoise & Plimpton.

As a matter of best practice, impact investors have generally to date considered, alongside potentially positive outcomes, the possible negative effects of their investments. This practice is enshrined in the EU Sustainable Finance Disclosure Regulation (SFDR), which requires investors to ensure, within the definition of sustainable investment, that the investment, in addition to contributing to an environmental or social objective, does no significant harm (DNSH) to any other objective.

Investors must conduct the DNSH test at the outset and on an ongoing basis, with investors in practice concentrating on the test at the due diligence stage, with a degree of monitoring during the holding period. Although the DNSH test remains a subjective exercise, requiring investors to use considerable judgement, the process required under the DNSH principle is becoming clearer.

Use of PAI indicators
The SFDR does not prescribe specific criteria for the DNSH test, but it is clear from clarifications made by the supervisory authorities and the relevant disclosure and reporting templates that firms should take into account the principal adverse impact (PAI) indicators in Table 1 of Annex 1 of the SFDR Regulatory Technical Standards (RTS) and, where relevant, the additional factors in Tables 2 and 3.

It seems reasonable in this context that funds will only apply the relevant indicators to a given investment, and justify omission of particular indicators.

Use of thresholds

Alongside use of the PAI indicators, firms are expected to set thresholds to determine where an investment causes harm, at least for those indicators that generate quantitative data. For example, where an investment emits greenhouse gases, a firm will need to define significant harm by reference to acceptable and unacceptable levels of emissions in the investment’s sector. Following comments in the European Supervisory Authorities’ (ESAs) recent proposals to update the RTS, these thresholds should not be sector agnostic, so will generally need to be considered and tailored to a given investment, although sponsors will need to ensure consistency of approach across the portfolio.

The ESAs also gave some clarifications in June 2022 on the DNSH test in its Q&A, stating: “Best practice could be to disclose DNSH for sustainable investment by extracting the indicators from Table 1 of Annex I, and any additional relevant indicators from Table 2 and 3 of Annex I, and show the impact of the sustainable investments against those indicators, proving through appropriate values (e.g. where feasible in compliance with the Climate Delegated Act and the Complementary Climate Delegated Act) that the sustainable investments do not significantly harm any environmental or social objectives.”

Here, the ESAs suggest, when setting a threshold for DNSH, that firms could look to the DNSH criteria in the Taxonomy Technical Screening Criteria for guidance. The DNSH criteria in the Taxonomy are partly qualitative (such as acceptable level of GHG emissions for a given activity) and partly quantitative (for example, with a low energy-consuming cement manufacturing plant, compliance with the DNSH test in relation to impact on marine resources and biodiversity require compliances with EU standards for environmental degradation to water and an EU standard environmental impact assessment).

As the ESAs state, it is only guidance for firms to use the strict criteria in the Taxonomy Regulation for the purpose of the DNSH tests, and it is open for firms to use equivalent non-EU standards to those standards specified in the taxonomy.

Use of transition plans to address DNSH

In some cases, firms may identify a harm, or a possible harm, and commit to a transition plan to address it. There is debate currently as to how transition plans fit into the DNSH test. The Commission’s recent answer to questions posed by the ESAs on interpretation of SFDR was that “investments considered as ‘sustainable investment’ shall not significantly harm any of the objectives referred to in Article 2(17)”, adding: “Therefore, referring to a transition plan aiming to achieve that the whole investment does not significantly harm any environmental and social objectives in the future could for instance not be considered as sufficient.”

Based on this guidance, there are two possible approaches. The first is that, if a firm identifies significant harm at the outset (through the use of the thresholds referred to above), it may take the view that it cannot qualify the investment as a sustainable investment until the transition plan has addressed the harm. The second is that firms can qualify the investment at the outset as a sustainable investment on the basis of a transition plan, other than where the transition plan relates to the whole investment.

There is certainly support for the second view in that it reflects existing practices by impact investors, and the fact that the significant harm test is fairly subjective – it is open for firms to conclude that, in relation to their transition plan, the harm is not significant because it is regarded as a risk of harm rather than actual harm. For instance, where a firm sees that the portfolio company has not diligenced its supply chains appropriately, the firm may conclude that there is risk of human rights breaches in the supply chain, and qualify the investment as a sustainable investment on the basis that it will address that risk in the ownership period. Admittedly, this leaves open the risk of a subsequent human rights breach in circumstances where the firm fails to address the risk, although arguably that risk arises with any investment, regardless of the due diligence conducted.

Monitoring harm during the life of the investment

It is clear that firms must monitor harm or the risk of harm during the life of the investment. For a fund investing in public securities, the expectation is that it will divest its position if and when it determines the investment is causing significant harm. The challenge in the private funds context is that, where a firm sees a harm arise, it cannot easily divest. It seems reasonable to treat such a scenario in the same manner as passive investment breaches and allow firms a leniency period to address the harm, although there is no clear guidance on this point.

Data sources and use of estimates

The ESAs gave guidance in their November 2022 Q&A on the DNSH text in the context of the Taxonomy Regulation. This guidance is helpful to the DNSH test in the broader context of the SFDR concept of sustainable investments. The ESAs stated that “it should be possible to use estimates for the DNSH test”, but not make conclusions based on a company’s track record, such as “media-based environmental controversies”.

So it appears that a sponsor can use estimated data for the purpose of the DNSH test, most likely in practice where the sponsor judges there not to be a severe risk of harm. To address data gaps, private fund sponsors can of course commit portfolio companies in the transaction documents to provide specific data going forward. Similarly, it seems possible for funds to use third-party data in the DNSH test, such as reports from international organisations and established market data providers.

It may also be possible for firms to make some assumptions in relation to DNSH criteria. There is support for this practice from guidance given by the EU Technical Expert Group (TEG) in the context of the EU Taxonomy, which stated in its March 2020 report that “for DNSH criteria that reflect legal requirements under EU regulations, it would be reasonable for Taxonomy users to assume these criteria have been met in the normal, lawful conduct of business, unless evidence to the contrary is demonstrated”.

What level of due diligence is required?

Again, in considering the level of due diligence required, the Taxonomy TEG provides some helpful statements. In particular, the TEG’s guidance is that due diligence should be risk-based and proportionate to the severity and likelihood of the adverse impact, with firms needing to prioritise investigation work on harms that are more likely to arise, in consideration of factors such as the industry sector, operations, business model and its position in supply chains.