ESEF financial data: how to read an iXBRL annual report

ESEF financial data is the set of machine-readable figures tagged inside the annual financial report that every company listed on an EU regulated market must publish in the European Single Electronic Format: an XHTML report with inline XBRL (iXBRL) tags that link each number to an IFRS concept, a period and a unit. Reading it correctly means reading the tag, not just the printed number.

We use ESEF filings as a primary source for the annual figures of the Nasdaq Helsinki and Nasdaq Stockholm companies we cover. This guide explains how an ESEF report is built and where it trips people up, using the FY2025 report of KONE Oyj (KNEBV, Nasdaq Helsinki).

What ESEF is and which companies file in it

ESEF is set by the EU's ESEF Regulation (Commission Delegated Regulation 2019/815), developed by the European Securities and Markets Authority. It applies to issuers whose securities trade on an EU regulated market, which covers the main lists of Nasdaq Helsinki and Nasdaq Stockholm. The annual financial report is published as an XHTML file that opens in a browser like any web page, and companies that prepare IFRS consolidated statements must tag them with XBRL.

ESEF covers the annual report only. Interim reports, half-year reports and stock exchange releases are not tagged, so quarterly Nordic figures still have to be read from the published report pages.

How iXBRL tags turn a printed table into data

Inline XBRL hides the tags inside the human-readable document. Each tagged number, called a fact, carries:

• Concept: what the number is, for example ifrs-full:Revenue or ifrs-full:ProfitLoss.
• Context: the reporting entity and the period, either a duration (1 Jan to 31 Dec 2025) or an instant (31 Dec 2025 for balance sheet items).
• Unit: the currency, such as EUR, or EUR per share for earnings per share.
• Scale and decimals: how the displayed number converts to the full value. A figure shown in millions is tagged with scale 6, and the decimals attribute records how precisely it was rounded.
• Sign: whether the stored value is negative, which is separate from how the number is printed.

Because the tagged value is stored in full units, a revenue line printed as 11,245.2 in a table headed MEUR is stored as 11,245,200,000 EUR.

The ESEF taxonomy, company extensions and block tagging of notes

The ESEF taxonomy is based on the IFRS Accounting Taxonomy published by the IFRS Foundation. When a company reports a line that has no matching IFRS element, it creates an extension element and anchors it to the nearest broader IFRS element, so software can still place it in the hierarchy. Extensions are common for alternative performance measures and industry-specific lines.

The level of detail differs between the statements and the notes. The primary statements (income statement, statement of comprehensive income, balance sheet, cash flow statement and statement of changes in equity) are tagged figure by figure. Since financial years beginning on or after 1 January 2022, the notes are block tagged: each note, or each accounting policy, is tagged as one text block. A segment table inside a note is therefore readable as text, but its individual figures are usually not tagged as separate facts.

Worked example: ESEF financial data from KONE's FY2025 report

The table below shows the income statement facts we read from KONE's ESEF filings for FY2025 and FY2024, with the IFRS concept each figure is tagged with.

Source: KONE Oyj ESEF annual financial reports for the years ended 31 Dec 2025 and 31 Dec 2024, consolidated income statement.¹²

Three things stand out. Income tax is stored as a positive EUR 334.8 million, because the concept is defined as an expense: the minus sign belongs to the presentation, not the value. Profit for the period (EUR 991.9 million) includes non-controlling interests, so the figure behind earnings per share is the EUR 980.1 million attributable to owners of the parent. And earnings per share carries a different unit (EUR per share) from the other lines.

Common pitfalls when you read ESEF financial data

• Sign conventions. Expense concepts hold positive values. If you sum tagged facts to rebuild a subtotal, subtract expenses rather than adding signed values as printed.
• Scale. Use the stored value, not the displayed one. Reading 11,245.2 as euros instead of millions is a factor-of-a-million error.
• Parent company versus group. Many annual reports include the parent company's own statements under local accounting rules after the consolidated statements. Check the entity and the statement before taking a figure.
• Extensions. A company-specific element may look like a standard line. Check its anchor before comparing it across companies.
• Notes as text blocks. Segment, lease or goodwill figures in the notes generally need to be read from the text, not queried as facts.
• Comparatives. Each report also tags the prior year. If the prior year was restated, the newer report's comparative differs from the original filing.

How to get ESEF figures without parsing iXBRL yourself

You can open any ESEF report in a browser and inspect tags with an iXBRL viewer, which works for one company. For many companies and years, you need the facts extracted, mapped to standard fields and kept with their source. Our Data API returns income statement, balance sheet and cash flow in one consistent shape across US and Nordic companies, and every figure links back to the filing it came from. For KONE, the FY2025 figures above link to the ESEF viewer of the filing itself. Endpoints and fields are in our API docs.

Our MCP server makes the same data available in Claude and ChatGPT. A question you can ask there: