Data
ESEF filing quality in the EU/EEA: what 14,804 filings show
Annual financial reports of companies listed on EU regulated markets have been filed in the European Single Electronic Format (ESEF) since the 2020 financial year. These files — XHTML with iXBRL tags — are among the first content the European Single Access Point receives. We analysed the public index of those filings to answer one question: how clean is the data that is about to become visible EU-wide?
Headline numbers, financial years 2024–2025
| Financial year | Filings | Issuers | With errors | With warnings | Fully clean | Issuers with ≥1 flagged filing |
|---|---|---|---|---|---|---|
| 2024 | 2,668 | 2,123 | 5.4% | 53.9% | 44.3% | 57.1% |
| 2025 | 2,218 | 1,854 | 3.5% | 54.3% | 44.2% | 55.9% |
| 2024 + 2025 | 4,886 | 2,278 | 4.5% | 54.1% | 44.3% | 65.3% |
EU/EEA only; filings tagged as United Kingdom excluded. "Issuers" are distinct legal entity identifiers (LEIs). An issuer can have more than one filing per year (language versions, amendments), and each is counted. The 2025 financial year is still being filed at the time of analysis.
What a "warning" is, and is not. The index operator validates each file against the XBRL specifications (errors) and against the formula rules defined in the ESEF taxonomy (warnings), and records inconsistencies. A warning does not block a filing and is not a regulatory finding; some are immaterial. But every one of them is a machine-detectable inconsistency in a file that, from July 2027, any user of ESAP can see. The index does not check the rules of ESMA's ESEF Reporting Manual or national rules, so the real number of issues is, if anything, higher.
By country, all filing years in the index
Twenty-six countries, 14,804 ESEF filings, 3,271 issuers. Germany and Ireland are not covered by the index; the United Kingdom is excluded here. Sorted by number of filings.
| Country | Filings | Issuers | With errors | With warnings | With inconsistencies | Fully clean |
|---|---|---|---|---|---|---|
| Denmark | 2,218 | 371 | 16.8% | 78.9% | 17.8% | 16.7% |
| Sweden | 1,838 | 441 | 2.1% | 71.8% | 24.8% | 27.3% |
| Norway | 1,199 | 317 | 2.8% | 75.2% | 35.0% | 22.1% |
| France | 1,185 | 292 | 9.2% | 73.1% | 32.7% | 21.2% |
| Finland | 1,169 | 170 | 5.6% | 62.8% | 46.4% | 28.8% |
| Poland | 876 | 326 | 9.5% | 40.1% | 2.3% | 51.4% |
| Italy | 874 | 225 | 1.3% | 77.7% | 18.0% | 21.3% |
| Spain | 732 | 139 | 0.4% | 49.3% | 2.9% | 50.5% |
| Belgium | 709 | 112 | 1.0% | 73.2% | 36.0% | 23.8% |
| Greece | 709 | 121 | 1.1% | 65.0% | 7.5% | 32.2% |
| Netherlands | 657 | 162 | 1.4% | 71.5% | 13.7% | 27.2% |
| Austria | 601 | 89 | 3.8% | 81.5% | 34.9% | 17.1% |
| Croatia | 280 | 95 | 3.2% | 66.1% | 15.0% | 31.8% |
| Portugal | 274 | 49 | 4.4% | 54.0% | 23.7% | 41.2% |
| Luxembourg | 266 | 88 | 4.5% | 68.8% | 18.4% | 28.9% |
| Malta | 265 | 64 | 16.6% | 52.5% | 3.8% | 31.7% |
| Lithuania | 191 | 27 | 0.5% | 50.8% | 1.0% | 48.7% |
| Hungary | 157 | 40 | 6.4% | 58.0% | 14.6% | 36.9% |
| Slovenia | 145 | 26 | 3.4% | 69.7% | 32.4% | 24.1% |
| Iceland | 119 | 29 | 0.8% | 69.7% | 21.0% | 30.3% |
| Romania | 115 | 39 | 12.2% | 67.0% | 33.9% | 26.1% |
| Slovakia | 80 | 18 | 2.5% | 63.8% | 0.0% | 33.8% |
| Estonia | 46 | 18 | 8.7% | 65.2% | 13.0% | 26.1% |
| Latvia | 36 | 13 | 0.0% | 63.9% | 16.7% | 36.1% |
| Cyprus | 34 | 33 | 0.0% | 79.4% | 20.6% | 14.7% |
| Czechia | 29 | 26 | 17.2% | 58.6% | 13.8% | 31.0% |
Country = country recorded in the index for the filing. Figures cover all financial years present (2020–2026) and therefore mix early-adoption years with recent ones; use the 2024–2025 table above for the current picture. "Inconsistency" flags fell sharply across the index from the 2023 financial year onward, which most likely reflects a change in the validation rules applied rather than a sudden improvement in filings.
Reading the numbers
- Technical errors are rare but not negligible. One filing in twenty-two in 2024–2025 does not conform to the XBRL specifications. Such a file may not open reliably in the software investors and data vendors use.
- Warnings are the norm, not the exception. More than half of recent filings carry at least one formula-rule warning — typically totals that do not add up to their components, or sign and period inconsistencies.
- Country differences are large. Among the larger markets, Poland and Spain show the highest share of fully clean filings; several Nordic and Alpine markets show warning rates above 70%. Part of this reflects national validation at submission, part the tools and service providers used locally.
- Most issuers are affected at some point. Two in three issuers had at least one flagged filing in 2024–2025. For a company, the relevant question is not "is the market clean?" but "what does my own history look like?"
Method
- On 29 October 2026 we downloaded the full public index of filings.xbrl.org through its documented API (27,626 records across the ESEF, UKSEF and Ukrainian filing systems). The index operator states that there are currently no restrictions on the use of the data.
- We kept records from the ESEF filing system and excluded filings tagged as United Kingdom, leaving 14,804 filings from 3,271 LEIs in 26 EU/EEA countries.
- For each filing we used the index's own counts of validation errors, warnings and inconsistencies. "Fully clean" means all three counts are zero. "Flagged" means at least one of them is non-zero.
- Financial year is the period-end year recorded in the index. Issuers are counted once per LEI.
Limitations
- The index is not complete: Germany and Ireland are not covered, and the operator notes that some filings cannot be processed or identified and are left out.
- Validation reflects the index operator's checks (XBRL specifications and ESEF taxonomy formula rules), not ESMA's Reporting Manual or national rules.
- We publish aggregate statistics only. We do not publish rankings of named companies.
The aggregate tables on this page may be reused under the Creative Commons Attribution 4.0 licence with a link to this page. If you need the per-country, per-year table as a file, or a check of a specific company's filing history, write to us.