European simplification: what a group keeps when the threshold no longer targets it
The February 2026 Omnibus directive sharply raises sustainability reporting thresholds and cuts the number of companies in scope by roughly 80%. Why a structure that was never in scope should not conclude it has nothing to maintain.
Directive (EU) 2026/470, known as "Omnibus I", was adopted on 24 February 2026 and entered into force on 18 March 2026. It raises the thresholds of the sustainability reporting directive (CSRD): a company now falls in scope only if it simultaneously exceeds 1,000 employees and 450 million euros in net turnover. According to published analyses, the perimeter moves from roughly 42,500 companies across the Union to roughly 10,000 — a reduction of about 80%. Transposition into national law is expected by 19 March 2027, and companies already in scope continue their obligations in the meantime (Norton Rose Fulbright, Portail RSE).
UGM-Group is far below those thresholds, and never came close. The question that interests us is therefore not "are we in scope" — we are not — but a more useful one: what does a small structure take from a text that does not target it?
A moving threshold is not a disappearing question
The temptation, with every simplification, is to read it as an exemption. That is a misreading, for two reasons. First because a raised threshold can be lowered again: the CSRD itself has gone through three calendars in four years. Second, and above all, because the duty to publish is only the wrapper. What the text asks at bottom — knowing where your suppliers come from, what risks the business carries, and on what data you base your claims — is the same question a ten-person company already asks itself, in simpler form.
The value-chain effect does not run through the threshold
This is what simplification does not remove. A company in scope must document its value chain, and therefore questions its suppliers — including those who, taken on their own, fall under no obligation. The mechanism is exactly the one we described for cybersecurity in our article on director liability: the constraint comes down through the contract, not through the threshold.
The Omnibus did provide safeguards on this point, framing what a large company may demand of its smaller partners. That limits questionnaire overload; it does not remove the demand.
What we keep, and why
Three disciplines strike us as worth keeping regardless of any obligation, because they cost little to maintain and a great deal to rebuild:
- Knowing who our critical suppliers are and what would happen if one of them stopped. It is the same mapping as for concentration risk, applied beyond the cloud.
- Dating and qualifying what we assert. A figure published without source or date goes stale in silence. The architecture framework we are rolling out across our subsidiaries requires marking every finding — declared, observed, inferred, validated — and we hold to it even when the result is uncomfortable.
- Keeping decisions written down at the level where they are taken. A holding company that cannot say why it decided something two years ago reruns the same debate every year.
The real cost is not the report
The companies that suffered under the CSRD were not, for the most part, complaining about writing a document. They were complaining about having to produce, in a few months, data that nobody had collected — because the question had never been asked upstream. That is the transferable lesson at our scale: what costs dearly is never the deliverable, it is the absence of a system that makes it possible.
Regulatory simplification is good news for the administrative burden. It teaches nothing new about governance, and we do not ask it to.