Corporate alignment with the EU taxonomy for sustainable activities: First evidence from financial accounting data

Abstract

What is sustainable? This question is of paramount importance given the trillions of assets invested according to different sustainability criteria. While until now we have had no standard for answering this question, the European Union's (EU) Taxonomy1 for sustainable activities aims to provide a comprehensive classification system for environmentally sustainable economic activities. This research paper examines the disclosures of European companies that are part of the Stoxx 600 index in relation to the EU Taxonomy.

The EU Taxonomy went into effect in July 2020, following the adoption of the Taxonomy Regulation (Regulation (EU) 2020/852). It can be a critical piece of legislation for investors, as it establishes a common language and framework for identifying environmentally sustainable investments. The taxonomy provides clear criteria for determining whether an economic activity can be considered environmentally sustainable based on six environmental objectives, including climate change mitigation, climate change adaptation, sustainable use and protection of water and marine resources, transition to a circular economy, pollution prevention and control, and protection and restoration of biodiversity and ecosystems. Fiscal year 2022 disclosures, the first year that the regulation went into effect, relate only to the first two objectives.

Companies subject to the Non-Financial Reporting Directive (NFRD) are required to report on their alignment with the EU Taxonomy. This includes large, publicly listed companies with more than 500 employees. Activities that contribute to one of the six goals are deemed “eligible.” To be considered “aligned,” a company must not only contribute to one or more of the above objectives, satisfying specific technical criteria, but do so while also meeting a “do no significant harm” standard: while supporting sustainable goals, aligned activities must not otherwise impinge upon any other sustainable goals. These companies must disclose data related to the taxonomy, including information on the proportion of their revenues, capital expenditures, and operating expenditures that are aligned with the taxonomy's criteria. These disclosures could be helpful for investors, policymakers, and other stakeholders to assess the progress of companies toward sustainable activities and specifically, the transition to a low carbon economy.

We undertake a comprehensive analysis of the newly reported corporate disclosures related to the EU Taxonomy, focusing on the alignment of revenues, capital expenditures (Capex), and operating expenditures (Opex) with the taxonomy's criteria. Leveraging this unique dataset for the first year of reporting, we aim to answer several pertinent research questions that can shed light on the current state of corporate sustainable activities in the context of the EU Taxonomy.

We start by documenting several empirical patterns. To be aligned, a sustainable activity first needs to be eligible, by contributing substantively to one of the six sustainability objectives. We find that the largest percentage of eligible revenues are in the Real Estate, Utilities, and Industrials sectors. But for all the implied progress, there is nonetheless significant variation within sectors and across industries. For example, while the Consumer Discretionary sector has a relatively low percentage of eligible revenues overall, the Automobiles industry ranks as one of the highest industries in terms of eligible revenues. This reflects the ongoing transformation due to electric vehicles. Other industries with high eligible revenues include Real Estate Development & Management, Building Products, Construction Materials, Metals & Mining, Transportation Infrastructure, Utilities, and Electrical Equipment.

An activity is aligned if it is eligible and satisfies a host of technical criteria. We find that aligned revenues are much lower than eligible revenues, consistent with the technical criteria representing a bar that most companies are not currently meeting. This trend is evident in the Automobiles industry, which has a high percentage of eligible revenues but very low aligned revenues. In contrast, Transportation Infrastructure exhibits an almost equal percentage of aligned and eligible revenues, suggesting that all eligible revenues are aligned in this industry. Building Products and Electrical Equipment industries also demonstrate higher aligned revenues, while Utilities and Real Estate exhibit the highest alignment at the sector level. Moreover, significantly higher variation exists within sectors and across companies for aligned rather than eligible revenues, suggesting that firm-specific strategies are an important driver of alignment.

Our data is cross-sectional in nature, given the first year of available data required by the regulation has been for the 2022 fiscal year end. The absence of a panel dataset and exogenous shocks to a firm's aligned investments and revenues does not allow us to make any causal claims about the relation between them or with other variables of interest. Nevertheless, we uncover several empirical patterns that are of interest to investors and managers. For example, modeling the relation between aligned capital and operating investments and aligned revenues we find that a 1% increase in aligned operating expenditures is associated with a 0.84% increase in aligned revenues with an additional 0.37% contribution from capital expenditures.

Moreover, given the significant variation in alignment metrics across competing companies within industries, we analyze the relationship between alignment metrics, business fundamentals, and corporate valuation ratios. There are three plausible scenarios for how these might relate. First, some companies may choose not to align because alignment leads to suboptimal financial outcomes, while other firms value the environmental benefits of the alignment. For example, eligible products that meet the technical criteria might face lower demand by customers due to higher prices, or they might require higher production costs leading to lower margins. Second, some firms may choose to align because alignment leads to better financial outcomes by tapping unmet customer demand for new products, while other firms are slower to align as they are reluctant to make the necessary investments, or they do not have the capabilities to adapt. Third, neither strategy may dominate so far. Firms with higher alignment may sell products and services to satisfy emerging customer demands while optimizing their investment profile, while at the same time firms with lower alignment satisfy existing customer needs. Consistent with this third scenario, we find little difference in sales growth, profitability ratios, or corporate valuation ratios across firms with high or low alignment within the same industry. However, we document that companies with higher enabling revenues and capital investments exhibit lower profitability (i.e., ROA) and (1-year sales growth).

Then, we analyze the ability of models that rely on business segment data and a classification of business segments to sustainable activities to predict eligibility of revenues and the correlation between alignment metrics and other environmental data. We find that models significantly overestimate the percentage of revenues that would be eligible as a sustainable activity. This is particularly true in the Information Technology, Communication Services, Industrials, and Utilities sectors. This discrepancy reflects the challenges associated with accurately estimating these figures from corporate business segment disclosures. Moreover, the alignment metrics are only moderately correlated with environmental data and ratings. Strikingly, we find that many firms in industries with high exposure to carbon emissions have close to perfect environmental ratings while having close to zero Taxonomy-related revenues or expenditures. We conclude that the disclosure regulation has provided investors with novel data and a differentiated assessment of firm strategies.

The analyses conducted in this paper are valuable for investors for several reasons. First, by examining the alignment of revenues, capital expenditures, and operating expenditures with the taxonomy across sectors, industries, and firm characteristics, investors can gain a deeper understanding of the extent to which large European companies are embracing sustainable practices. This can help investors identify investment opportunities in firms that are more likely to benefit from the transition to a low-carbon economy and avoid those that may face increasing regulatory risks and stranded assets.

Second, by documenting the intra-industry variation in alignment with the taxonomy, this research can provide insights into the differences in the adoption of sustainable practices within industries. This information can be used by investors to better assess the relative performance of firms within an industry and make more informed investment decisions based on environmental considerations.

Lastly, by investigating the relationship between alignment with the taxonomy and commercial environmental ratings, this research can help investors evaluate the consistency between the EU's classification system and other widely used environmental ratings. This information can support investors in their due diligence process and portfolio construction by identifying potential discrepancies and complementarities between different sustainability frameworks.

Fiscal year 2022 is the first year that companies are obliged to report the amount of revenues, capital expenditures, and operating expenditures aligned with the EU taxonomy for sustainable activities. Moreover, companies need to report the amount of revenues that are eligible for evaluation of alignment with the taxonomy. Finally, companies have the option to also report revenues, capital expenditures, and operating expenditures in enabling or transitional activities.

We collect these data from Bloomberg for all companies that have reported as of the end of December 2023 for the fiscal year 2022. We focus on companies in the Eurostoxx 600, as these companies represent the largest European companies and as a result, we expect that the quality of the disclosures will be robust given that these companies have the accounting resources, internal control systems, and regulatory, as well as market scrutiny, to produce reliable financial data. After removing firms that have no data on Bloomberg as of the end of June as well as all financial sector firms,2 our sample includes 319 firms with reported eligible revenue data and 297 with reported aligned revenue data.

We expect a positive correlation between aligned revenues and investments. Firms that invest to build manufacturing plants and hire people to produce products that are aligned with the EU taxonomy should exhibit higher revenues from aligned activities. For example, an automobile company that invests to build a manufacturing plant that produces electric vehicles and to hire electrical engineers for battery optimization and integration in the vehicles should exhibit higher revenues from the sales of electric vehicles.

However, there are two reasons why we expect the revenue and investment relation to be attenuated in our empirical model. First, not all aligned expenditures are revenue generating. Some expenditures are not going to generate revenues but rather they are directed toward reducing the carbon emissions of a firm. For example, powering the paint shop in an automobile manufacturing plant with electricity from renewable energy instead of natural gas. Second, some expenditures might not produce revenues yet. Following the same example, the automobile manufacturer might not yet be selling any vehicles, but the expenditures are expected to lead to futures sales. Therefore, the observed empirical relation between aligned revenues and expenditures is likely to be smaller than the relation if one was able to identify revenue-targeting investments and measure revenues generated over a multiyear period. Table 9 shows the estimated coefficients on a model that has as a dependent variable aligned revenues and as key independent variables aligned expenditures. We log transform all variables so the coefficients can be interpreted as measures of elasticity. We control for industry fixed effects to allow estimates to be derived only from within industry variation. We find that for every 1% increase in capital and operating expenditures, revenues increase by 0.876%.

However, combining capital and operating expenditures obscures the strength of the relation with revenues across the two types of expenditures. Estimating separately the relation with capital and operating expenditures we find a larger coefficient on the latter. A 1% increase in operating expenditures is associated with a 0.835% increase in revenues while a 1% increase in capital expenditures contributes another 0.367%. The higher association with operating expenditures is sensible given that capital expenditures can be recognized as an asset because they will lead to future economic benefits, consistent with the definition of an asset in financial accounting standards.

Estimating sector-specific models for the three sectors with the highest number of observations, we find that the estimated coefficients on expenditures vary significantly across sectors. For example, in both materials and consumer discretionary, a 1% increase in operating expenditures is associated with a 1.3% increase in revenues. Capital expenditures do not exhibit a significant association with revenues. In contrast, in industrials capital expenditures exhibit a significant association with revenues. A 1% increase in capital expenditures is associated with a 0.688% increase in revenues. A 1% increase in operating expenditures is associated with a 0.454% increase in revenues.

Given the significant variation in firm alignment with the taxonomy, we explore whether companies with higher alignment exhibit different business fundamentals. To keep the analysis tractable, we focus on two fundamental aspects of the business: growth and profit. We use 1-, 3-, and 5-year revenue growth as our growth metrics. We use operating profit margin and return on assets (ROA) as our profit metrics.

If sustainable products experience higher growth or they command premium prices for equal production costs, then we expect companies with higher alignment to exhibit stronger fundamentals. In contrast, if demand for sustainable products is weak, or if their production requires higher costs, then we expect companies with higher alignment to exhibit weaker fundamentals.

In some industries, the percentage of activities classified as sustainable is very small for all companies. In those industries, it would be rather impossible for such a small part of a firm's activities to generate differential sales growth or operating profitability. Therefore, to increase the power of our test, we focus on a subset of industries where there is at least one firm with 10% or greater aligned revenues. This cuts our sample by a little more than half.

Tables 10 and 11 show that the alignment metrics are not significantly associated with past sales growth after controls for GICS industry fixed effects and starting period level of sales. We control for the latter given that it might be easier to grow more from smaller levels of sales. Moreover, the association with profitability margins is insignificant. This suggests that firms that are selling products and services aligned with sustainable activities have not been growing at lower or higher rates nor that their cost and pricing structures have been generating lower profitability ratios.

For enabling and transitional metrics, we find a negative and significant correlation with both past 1-year sales growth and ROA. In unreported results, we separately analyze enabling and transitional metrics and find that the negative association is driven by enabling activities. Firms with more enabling activities, primarily coming from utilities (electric, gas, and multi), machinery, automobiles, semiconductors, chemicals, and electrical equipment industries, are growing at a slower rate and earning a lower rate of return on their assets. While such activities are important in that they enable the implementation of climate solutions that can reduce carbon emissions they do not reduce carbon emissions directly. Companies that engage in more enabling activities do not seem to be adequately compensated so far for the investments they make in terms of profitability or sales growth. In contrast, transitional activities are associated with superior sales growth in some of the unreported models.

Following the fundamental analysis, we analyze if firms with greater alignment with sustainable activities are trading at higher or lower valuation multiples. For example, if investors expect that alignment with sustainable activities might make a firm less risky or that it exposes a firm to future superior business growth, these firms might trade at higher multiples. In contrast, if investors expect that these activities will not experience growth and that these firms are investing resources that will not be monetized later, they might assign lower valuation multiples to those firms.

For this analysis, we focus on two widely used valuation ratios: the price-to-book equity value ratio (PTB) and the price-to-earnings ratio (PE). We control for GICS industry fixed effects, logarithm of market capitalization, past 3-year sales growth, and in the case of PTB also for ROA.

Table 12 shows that firms with higher alignment exhibit no difference in valuation multiples. Across all specifications, none of the estimated coefficients on the Taxonomy variables is significant. This is the case for both PTB and PE.

In this section, we explore whether the new disclosures are predicted by three data types. First, data from models that rely on business segment disclosures and their classifications as sustainable or not. Second, environmental ratings provided by commercial entities. Third, carbon emission metrics reported by companies. If existing data are sufficient to characterize the activities of a company and their positioning relative to competitors, then the disclosure requirements could be obsolete.

We have analyzed data from more than 300 large European companies on their first year of mandatory corporate disclosure of financial accounting data that characterize whether their activities are sustainable or not based on technical criteria. We reach several conclusions based on the analysis of this dataset.

First, our data suggests that because of the technical criteria set forward by the EU Taxonomy, only a small percentage of business activities align with the taxonomy as shown in Tables 3 and 4. This is consistent with the EU taxonomy setting a ‘high bar’ for what constitutes a sustainable activity. However, we expect that over time, the percentage of activities will increase, as companies are transitioning their investments and products toward activities that align with the taxonomy. In other words, we expect that the gap we document in Figure 3 to shrink over time.

Second, we observe significant differences in the percentage of activities that are aligned with the taxonomy across competitors, suggesting that some firms are much faster and more willing than others to align their activities. However, we find little evidence in Tables 10–12 that efforts to attain alignment can observably translate into benefits in operating performance or market valuation multiples at the time of the analysis, but nor do we find evidence suggesting that firms that chose to align more with the taxonomy have put themselves at a competitive disadvantage so far. Perhaps in the future, as product and capital markets reward alignment with sustainable activities, companies with higher alignment will grow their revenues faster, enjoy higher profitability margins, and trade at higher valuation multiples. An indication that this might be the case would be the estimated coefficients on investments in Table 9 increasing over time, suggesting that aligned investments translate at a higher rate to revenues.

Third, we find little evidence that existing data are correlated with alignment metrics. Analyzing correlations with carbon emission metrics and environmental ratings reveals very weak relationships. We infer that carbon emission metrics serve a purpose different from that of alignment metrics. Carbon emission metrics provide a measure of the total carbon emissions produced in the value chain of a company. This in turn provides a measure of the contribution of the operations of a company, its supply chain, and the use of its products to the challenge of climate change, while at the same time gauging a degree of exposure to risks arising from the transition to a low carbon economy as a result of regulatory (i.e., carbon taxes or cap and trade systems) or market changes (i.e., shift to lower carbon products by customers). In contrast, alignment metrics represent the extent to which an organization's products and services can be classified as sustainable and its investments are directed toward sustainable activities. Regarding the low correlation with environmental ratings, we expect that the increase in available data of EU taxonomy-aligned activities will likely increase that correlation as rating agencies will integrate such data into their own rating processes. In other words, we expect over time the dots in the scatter plots in Figure 6 to populate the non-red area.

This material is for informational purposes only. The views expressed in this material are the views of the authors, are provided “as-is” at the time of first publication, are not intended for distribution to any person or entity in any jurisdiction where such distribution or use would be contrary to applicable law, and are not an offer or solicitation to buy or sell securities or any product. The views expressed do not necessarily represent the views of State Street Global Markets® and/or State Street Corporation® and its affiliates.