ESRS E5 in practice: what EFRAG's 2025 State of Play tells us about circular economy reporting.

The 2024 financial year was the first mandatory reporting period under the Corporate Sustainability Reporting Directive. In July 2025, EFRAG published State of Play 2025, an analysis of 656 Wave 1 sustainability statements issued as of 20 April 2025. The report is not authoritative guidance, but it is the clearest picture available of how the first cohort of large European companies actually applied the European Sustainability Reporting Standards.

For anyone tracking the circular economy, one topical standard matters most: ESRS E5 — Resource use and circular economy. Below is what the EFRAG data says about it, and what we read into it.

1. E5 is now a mainstream topic — but far from universal

Across the 656 Wave 1 preparers, three topical standards were material for almost everyone: E1 Climate Change (98%), S1 Own Workforce (99%) and G1 Business Conduct (93%). E5 sits in the next tier, alongside S4 Consumers and end-users (68%), S2 Workers in the value chain (63%) and — critically — E5 Circular economy at 65%.

That places circular economy fourth on the materiality league table for the first mandatory year of CSRD reporting. Ahead of pollution (E2, 38%), water (E3, 33%) and biodiversity (E4, 39%). Behind only the standards that no serious company can plausibly deem immaterial. It is a strong signal that circularity has crossed from thematic interest into standard board-level disclosure.

2. The FI / non-FI gap is enormous

The headline 65% hides a split that anyone advising both sides needs to see clearly:

• Non-financial institutions: 63% declared E5 material.
• Financial institutions: only 31%.
• Banks specifically: just 16%.

Read carefully, this is a warning. Banks are, in their own view, largely not exposed to circular economy risk. Yet their loan books, mortgages and corporate exposures sit on top of the same manufacturing, real estate and retail assets whose circular transition determines future asset values and default risk. The conclusion is not that banks are wrong to score E5 low today; it is that this number will not hold as the transition data matures. The gap is a leading indicator of where circular disclosure pressure will move next.

3. Sector determines almost everything

Inside the non-financial cohort, E5 materiality tracks physical exposure to materials. This is what preparers told EFRAG:

Source: EFRAG State of Play 2025, Figure 8. Share of preparers declaring E5 Circular economy material, by sector.

Construction, manufacturing, wholesale/retail and real estate all sit above 75%. In practice, if your business handles physical product or built assets, treating E5 as non-material now requires a defence, not a default.

4. Sub-topics are being prioritised more than the standard suggests

EFRAG notes that only ~10% of preparers declared all ten topical standards material, and about a quarter declared four or fewer. The average was six. This is the double materiality assessment working as intended — companies are choosing what to report on, not defaulting to a full sweep. Within E5, this means many preparers are picking specific sub-sub-topics (resource inflows, resource outflows, waste) rather than reporting the whole standard uniformly.

At the extremes, six sub-topics were declared material by fewer than 5% of preparers, including E2 microplastics and E4 animal welfare. Circularity's sub-topics fared better, but the pattern is the same: preparers are learning to be selective. That is healthy — but it means users of these statements cannot assume E5 has been reported in equivalent depth across peers.

5. Data-point disclosure varies wildly

The high-level structure of these first statements is comparable. Almost all follow the ESRS Application Requirement 16 taxonomy at topic and sub-topic level. But at the granular data-point level, the picture fragments. Report length ranges from ~25 to ~440 pages, with the median at 100. Longer reports are not necessarily deeper; EFRAG's manual sub-set check found that extra length often reflected narrative repetition, not additional data.

For E5, this matters because the standard's value depends on quantitative disclosure of material flows, product durability and end-of-life outcomes. Where those data points are missing or buried in prose, comparability breaks. The next few reporting cycles will be defined by whether preparers convert E5 narrative into structured, auditable data.

6. Stakeholder engagement is business-heavy

The double materiality assessments underpinning these E5 calls were built primarily from internal and commercial input: 97% of preparers engaged internal stakeholders, ~70% customers, ~65% suppliers, ~60% investors. NGOs (33%), communities (30%) and academia (~15%) were consulted far less often. For a topical standard whose impacts run through supply chains and end-of-life systems, this is a gap. The E5 disclosures reflect the perspective of those who buy, sell and finance products — not those who inherit the waste.

What this means for leaders

For CFOs and reporting leads

If your company reported E5 as material in 2024, expect investors and assurance providers to focus next on the depth of quantitative data behind that call. If you reported E5 as non-material and you sit in manufacturing, retail, construction, real estate or utilities, the defence needs to be documented against the sector benchmarks in this EFRAG data set.

For financial institutions

Sixteen percent of banks calling E5 non-material is a fragile position. Loan book exposure to circular transition risk is not yet in the disclosure — but the physical assets financed are the same ones that non-financial peers are already reporting on. Building the data infrastructure to translate financed-company E5 disclosures into portfolio-level risk is the next move.

For everyone else

E5 is now the standard-setting benchmark for what "circular economy" means inside a large European company: resource inflows, outflows, waste, and the strategy connecting them. If your circular narrative does not map to those data points, it will not survive the next reporting cycle.