The Principal Adverse Impacts: A key feature in the EU’s efforts to eliminate greenwashing

The European Union’s Sustainable Action Plan, which was released in early 2018, has set out to reform the broader sustainable finance marketplace by creating a financial market which truly integrates sustainable finance. The Action Plan is expected to be the largest regulatory shift that the sustainable finance industry has ever experienced - and will have an exceedingly large ripple effect as other countries seek to adopt similar regulatory frameworks. The regulation consists of three primary pillars: Sustainable Finance Disclosure Regulation (SFDR), taxonomy regulation and changes to MiFID, UCITS, and AIFMD.

In this blog, we explore the SFDR pillar, specifically, the data intensive nature of one of its key features: The Principal Adverse Impacts. The Principal Adverse Impacts (PAI) refer to a specific list of indicators that, at least in principle, allow investors to monitor the adverse impacts of the securities they invest in. This means that there is a new wave of ESG data that investors must learn to interpret and analyse. We hope that by sharing our insights we can help investors better understand the inherent challenges and capitalise on potential opportunities.

For a more in-depth review of the Action Plan’s objectives and structure, you can refer to our previous blog post - A journey to eliminate greenwashing.

The Principal Adverse Impacts defined

The Principal Adverse Impacts, or PAIs, are measured by an assortment of 14 mandatory corporate indicators, with two additional indicators for sovereigns and two real-estate specific indicators. The lineup is expanded by a list of 46 additional voluntary indicators, of which firms must select at least two additional indicators to report on. These indicators cover a broad scope of environmental and social metrics, including scope one, two, and three GHG emissions, biodiversity impacts and gender pay gap ratios.

Uncovering the data landscape

To help improve our understanding of the data that is currently available to meet the upcoming regulations, we completed a multi-faceted review of major data providers1. The goal was to better understand the current landscape, challenges and opportunities, and to determine potential use cases for the data, other than SFDR disclosure. We want to go beyond a tick-the-box exercise and uncover other ways that this data could be integrated into the investment process.

For our review, we’ve focused on four data providers. Our analysis focuses on corporate and sovereign issuers (covering 16 of the 18 mandatory indicators). The reason why this is important to highlight is that the regulations do not exclude private assets such as real estate, private equity and private debt from the scope of reporting requirements. Thus, alternative solutions may have to be considered for multi-asset managers, including engagement efforts with underlying managers to push for increased data disclosure.

The challenges ahead

The first and obvious challenge related to measuring the PAIs is lack of data availability. Data availability is typically measured in terms of coverage, or percent of the portfolio or investment universe where data exists. In the case of PAIs, two types of coverage are relevant: percent of companies or entities where the data is available at all, independent of provider, which we’ll call the “real” coverage and the second is coverage by a given data provider.

Why is coverage from a data provider different from the “real” coverage? The PAIs represent a significant shift for providers away from score and qualitative methodologies. They provide a very prescriptive requirement for the raw environmental and social data. A quick example of this is how, historically, analysts at the data providers could conduct qualitative assessments of firms and their respective gender and diversity policies, environmental performance and forward-looking initiatives. This resulted in a score or assessment being assigned, like an “A” rating on diversity policies, rather than collecting a raw number or metric. Now, to help firms comply with the PAIs, the providers will need to collect metrics such as the unadjusted gender pay gap, which is expressed as the actual pay gap percentage. As data providers increase their coverage of these indicators, meaning that they start looking for and filling in the data precisely in line with the PAIs, provider coverage will increase to 100%, meaning they are covering the whole universe, but the data available to the end investor will still indicate a lot of “NA”, or data not available.

This is because companies have yet to start consistently reporting on these metrics, and therefore, the data is simply not available. Fortunately, we do expect that corporate data disclosure will continue to meaningfully improve as more disclosure legislation is passed at the corporate level. One such example is the recently drafted Corporate Sustainability Reporting Directive (CSRD) which seeks to improve upon the current Non-Financial Reporting Directive (NFRD) by expanding its scope to include a broader range of companies as well as additional disclosure metrics, including the principal adverse impacts.

Figure 1: Average Coverage of PAI Indicators across four Data Providers, Sample Global Multi-Asset Portfolio
Scope 1 GHG emissions
Scope 2 GHG emissions
From 1 January 2023, Scope 3 GHG emissions
Total GHG emissions
Carbon footprint
GHG intensity of investee companies
Companies active in the fossil fuel sector
Share of non-renewable energy consumption and production
Energy consumption intensity per high impact climate sector
Activities negatively affecting biodiversity-sensitive areas
Emissions to water
Hazardous waste ratio
Violations of the UNGC principles and OECD Guidelines for Multinational Enterprises
Lack of processes and compliance mechanisms to monitor compliance with UNGC principles and OECD Guidelines for Multinational Enterprises
Unadjusted gender pay gap
Board gender diversity
Exposure to controversial weapons

After reviewing our results, we found that four of the eighteen mandatory indicators, on average, experienced disproportionally lower real coverage due to this very challenge. These included:

  • Share of non-renewable energy consumption and production
  • Emissions to water
  • Hazardous waste ratio
  • Unadjusted gender pay gap

Real estate investors (excluding listed REITs) are exposed to further challenges. The provider coverage of the two real estate specific indicators is likely to be rare; therefore, investors and advisors may have to look elsewhere for data coverage specific to this asset class.

Where are the data opportunities?

Lack of transparency and standard definitions are rampant in the environmental, social and governance (ESG) investing space. PAI legislation, in its ideal form, makes small steps toward increased transparency and comparability. Consider ESG assessments, or scores, as a counterexample. ESG assessments are designed to address different things, with no standard definition. Scores range from those identifying highest (financial) risk from ESG-related issues, to companies actively contributing to sustainability outcomes.

While differences may seem subtle, they lead to significant divergence in end scores. Berg et al. (2019) found that correlations among providers averaged 0.54 and ranged from 0.38 to 0.71, meaning the information to form ESG-based decisions upon is noisy2. This makes it difficult for an investor to draw conclusions on the basis of the average ESG score of their portfolio. What is the score really telling the investor? PAIs, on the other hand, address very specific issues and do so precisely. In a hypothetical world where all the data exists, PAI allow the investor to answer: what are the adverse impacts of the securities I’m invested in? With PAIs we can see that impact against a very specific range of metrics. And this measurement is independent of the asset manager or data provider, and therefore more easily facilitates comparison across the industry.

While company coverage is certainly a known challenge for embracing the PAIs, it is not completely discouraging. Metrics for GHG emissions, fossil fuel involvement, board gender diversity and controversial weapon exposure all flaunt high real coverage percentages. Furthermore, we found that the two country specific indicators were highly covered by most providers, with data on approximately 160 to 190 different countries.

The push for more granular ESG data reporting, as prioritised in the PAIs, may very well usher in a new era for the way investors evaluate the ESG quality of firms. That being said, we imagine a hybrid approach that considers indicators such as the PAIs alongside qualitative assessments will likely prevail as the preferred methodology.

Use-case examples

One of the overarching themes of the Action Plan is to enable investors to make informed decisions. The PAIs facilitate this goal by consolidating metrics from a diverse range of ESG issues. In addition to meeting the PAI disclosure requirement, we see the following as other potential uses for the provider datasets.

Non-PAI Disclosure Frameworks

PAI indicators can be part of an investor’s strategy for identifying funds that meet their requirements for either Article 8 (light green) or Article 9 (dark green) status. The indicators provide a diverse set of metrics that can be used to highlight a strategy’s ESG characteristics and can be compared to relevant benchmarks or absolute thresholds (i.e., no violations of the United Nations Global Compact). Looking beyond the EU, PAIs can be integrated into other disclosure frameworks such as the Task Force on Climate-Related Financial Disclosures (TCFD). By incorporating and understanding these datasets now, investors can better prepare themselves for future regulatory requirements, such as those likely to materialise in the U.S. and around the globe.


Using relevant PAI metrics in client reports, fund factsheets, and firm marketing material provides further transparency into the impacts of an investor’s portfolio. In recent years, asset managers have expanded their reporting capabilities to include a myriad of ESG metrics, including carbon emissions, ESG risk scores and product involvement percentages. PAIs offer an opportunity to complement these metrics by providing insight into a portfolio’s gender pay gap or water emissions, ultimately allowing investors to develop a better understanding of the adverse impacts of their investments.

Engagement Opportunities

Active ownership and engagement are a critical pillar of any responsible investment strategy. The PAIs can be an additional tool in the shareholder’s toolbox by helping them track a firm’s ESG progress year over year. Furthermore, these indicators can provide a common language and a set of industry-accepted metrics with little room for misinterpretation or vague policy guidance, ultimately providing opportunities to confirm that firms are indeed taking the actions that they have previously indicated.

Strategy Development/Implementation

A common refrain in the sustainability space is “what gets measured, gets managed”. A natural progression after measuring is that investors may have questions around mitigating the adverse impacts of their investments. Raw numeric data lends itself well to being incorporated as primary or secondary input within a quantitative investment strategy - for example, our decarbonisation 2.0 methodology tilts on carbon emissions and renewable energy production metrics, alongside material ESG scores. In instances where data quality is sufficiently high, the PAI metrics may eventually provide the basis for portfolio tilts to broad environmental or social themes.

Bottom line

Sustainable investing suffers from no shortage of data-related issues. The EU’s sustainable finance regulations seek to make meaningful strides by providing a common language and access to diverse quantitative data. Providers are working diligently to improve their capabilities to meet the constantly evolving regulations and ultimately, the needs of their clients. While there remain significant challenges, mostly with respect to the reporting done by companies, we expect these gaps to be resolved in due time and for the benefits to far outweigh the growing pains.


1 To assess coverage, we asked providers to run their data solutions on a representative, globally diversified multi-asset portfolio that included listed equities, public REITs, and corporate and sovereign bonds. The portfolio had positions from both developed and emerging markets and included roughly 1,150 equity securities, 775 corporate bonds, and 325 sovereign bonds with the goal being to accurately represent the portfolios we build for our clients. Our findings have been reported on a generalised basis and it is important to note that meaningful variations do exist among the providers, with each offering unique strengths and weaknesses.

2 Berg, Florian and Kölbel, Julian and Rigobon, Roberto, Aggregate Confusion: The Divergence of ESG Ratings (May 17, 2020). Available at SSRN: or