Future Tense: The Draft for Germany's 12th Competition Act Amendment

Many reforms have a governing tension, and this one has it in plain view. The coalition agreement asked for two things that sit awkwardly together: relieve business of unnecessary filings, and make sure genuinely harmful deals do not escape. In merger control those goals compete, and they address the twin fear of over- and under-enforcement, i.e., that the net may catch too many ordinary deals and too few harmful ones.

The under-enforcement worry has a clear European pedigree. In Illumina/Grail (C-611/22 P, 3 September 2024), the Court of Justice shut down the Commission's use of Article 22 EUMR as a catch-all for sub-threshold so-called killer acquisitions, holding that a Member State may refer a deal to merger review in Brussels only where it would itself have had jurisdiction, and pointing national legislatures to their own thresholds (and the prospect of revising them) if they wish to catch more (meaningful) deals. As we have already set out elsewhere, the alternative is a patchwork of national call-in powers that trades one gap for a maze of timing rules. Germany's draft laudably picks the cleaner route, a transparent value-based test rather than discretionary call-in (see 2. below).

At the same time, there is always a latent risk of over-enforcement in merger control, as indicated by the total number of deals caught by the framework (bearing in mind that the vast majority of deals is cleared without even an in-depth review). The number of German merger filings had its peak in 2019 (1,425 filings), fell to 832 in 2022, and has hovered around 870 since (876 in 2025). That drop of roughly 40 percent followed the last threshold increase under the 10th amendment of the ARC. The new draft seeks to build on that by introducing yet another turnover threshold increase (see 3. below).

Germany’s transaction value threshold, currently in Section 35(1a) ARC, has always been a much-debated part of German merger control. Under that threshold, a deal is caught where combined worldwide turnover clears the first hurdle (combined global turnover of more than EUR 500 million) and one party clears the first domestic hurdle (individual turnover of more than EUR 50 million), but the target's German turnover stays low, i.e., below the second domestic turnover threshold of EUR 17.5 million. If against that backdrop the consideration for the transaction exceeds EUR 400 million and the target has significant domestic activities (erhebliche Inlandstätigkeit), a merger filing is required.

The draft folds this test from its current standalone provision into the general catalogue of Section 35(1) ARC, technically ranking it co-equal rather than subsidiary to the general turnover thresholds, and extends it to targets that are not yet active in Germany butwill likely become so. When this provision becomes law, it will mark the end of a winding road with many twists and turns.

For a long time, the prevailing view in academia and at the Higher Regional Court of Düsseldorf had read the rule exactly the other way around. Based on that view, what counts is the target's current market activity, not the acquirer's plans for the period after closing – which is precisely the point we flagged here: the Düsseldorf Court had curtailed the FCO's attempts at a more expansive interpretation. However, the Federal Court of Justice later confirmed the FCO's expansive reading in Meta/Kustomer (KVR 77/22, 17 June 2025), finding significant domestic activities even in a SaaS provider's access to the data of German end-customers. That said, it expressly left open whether future activity could suffice and instead ruled that the threshold’s purpose, to capture deals involving targets with a competitive potential, is only served if all current activities that have a link to Germany and are indicative of that potential are assessed. While this already had a certain forward-looking element under the guise of the “potential” analysis, the draft now goes one step further and enshrines a fully forward-looking concept as an alternative to current domestic activities. Specifically, it is sufficient if “the company to be acquired is active, or is expected to become active, to a significant extent in the domestic market.”

Interestingly, Austria adopted its transaction value threshold at the same time as Germany, yet in March 2025 its Supreme Court (Oberster Gerichtshof) held that a target's potential future activities have no place in the domestic-activity test at all (OGH, 26 March 2025, 16 Ok 2/25t). Berlin has now written into its draft what Austria's highest court wouldn’t have. But in a way, this is a homecoming: 2017’s draft for the 9^th ARC amendment already proposed to cover targets that would only become active in Germany later. What the legislator could not push through then, is now lined up for a revival.

The draft points out that this will, for instance, clearly cover all preparatory activity for market entry (even if it stems from foreign countries), such as research and development aimed at a product to be marketed in Germany, or the pursuit of a German drug approval. Even where the target has no present German footprint of any kind, targets in situations like these will become candidates for merger review, unless any “go to market” with nexus to Germany is completely remote – or too far out in time. The explanatory language that comes with the draft (but not the actual statute!) states that “in most cases”, the forecast will look out around two years, and three to five years maximum. It also states a purely hypothetical entry will not do – an appreciable, not merely marginal, domestic effect is required.

Even so, ruling significant domestic activity out with any confidence will, for many targets, be a challenging exercise. In practice that means that the formal, entry-level threshold test as to whether or not a deal qualifies for review now more than ever requires a comprehensive and deep substantive analysis as to the target’s activities – effectively preceding much of the analysis that would normally be undertaken in the actual merger review (after jurisdiction has been established).

It is worth pointing out that the draft's “crown witnesses” for the changes to the transaction value threshold are two cases which the FCO in the end could not pick up: Microsoft/OpenAI and Microsoft/Inflection. However, the former turned on the absence of a notifiable concentration, with the lack of domestic effects a mere secondary problem. It was only the Microsoft/Inflection case that (regardless of the questionable notion in the press release’s headline that “Taking over employees may be subject to merger control”) escaped review due to a lack of significant domestic activities. This shows that the reform will have its limits in terms of catching “interesting” deals – even though it seems questionable based on the draft’s reasoning that the legislator is aware of that.

While the above suggests more front-loaded complexity for the analysis of international merger control filings, there is also significant relief for merging parties. All three turnover thresholds rise at once for the first time: the worldwide threshold and the first domestic threshold each climb by 50%, from EUR 500 to 750 million, and from EUR 50 to 75 million, respectively. The second domestic threshold climbs more modestly, from EUR 17.5 to 20 million, on the deliberate logic that the 10th amendment had already lifted it to 350 percent of its original level, so that a steeper rise would let regionally significant deals slip away. Still, the increase now proposed sits around 14% – which, incidentally, matches the decrease projected by the Ministry in terms of the total number of German merger control filings per year (around 120 cases fewer than currently).

And for the transaction value threshold specifically, there is another change that is supposed to bring regulatory relief: the introduction of a new Phase 0 notice procedure, essentially the “shock absorber” for the forecasting problem flagged above. For deals caught only by the transaction value threshold, the parties will first file a “simplified notice”, and the FCO then has two weeks to say whether it will waive a full notification. Silence clears the deal once the deadline passes, which would effectively offer a review sped up by around two weeks (compared to the current situation where all filings, including those based on the transaction value threshold, must undergo a full Phase I review of up to one month).

And while the Ministry assumes that the effectively broadened transaction value threshold will increase the number of such filings based on it to 32 a year, 30 (!) of those will resolve in the new Phase 0 already – meaning there might effectively be less red tape for deals caught by the transaction value threshold than currently. Whether that assumption will hold and whether the “simplified notice” actually decreases the regulatory burden on substance (compared to a regular filing) will only be revealed in practice. Notably, the new provision in Section 39(7) ARC would require parties to in particular set out their strategic and economic rationale, evidenced by internal documents – and the FCO will grant its waiver only if that disclosure is so substantive that it can confidently rule out substantive concerns and forgo a full Phase I review. If it doesn’t do that, the deal timetable will effectively be prolonged by the two weeks for Phase 0 – and the FCO's "no waiver" message needs no reasons and cannot be appealed.

If there is relief on the merger control side, there is at the same time a whole new investigative reach in another area, and arguably this is the centrepiece of the whole draft: a comprehensive procurement screening mechanism. A new Section 32h ARC lets the FCO analyse public-tender data, the bids of unsuccessful bidders included, systematically and without prior suspicion, for signs of bid-rigging, with contracting authorities obliged to feed in the data and a five-year retention period. The constitutional weight is plain, given the suspicionless, systematic processing of personal bidding data that the draft itself describes as "not sample-based". Yet the draft’s reasoning carries that weight lightly, asserting EU-law compatibility and leaning on a 15 to 20 percent overcharge figure associated with bid-rigging. While that figure traces back to research work emanating from the OECD rather than any specific FCO finding, it is used as a benchmark to legitimize the FCO’s far-reaching new power: the economic harm potentially caused by bid-rigging in a market with a public procurement volume of around EUR 120 billion is, in the Ministry’s view, reason enough to introduce Section 32h.

And while it is true that EU Member States such as Spain and Denmark have similar instruments in place already, the FCO’s enforcement history in recent years does not exactly suggest a lack of bid-rigging scrutiny. Even without a screening mandate, a road-repair cartel drew fines of roughly EUR 10.5 million in 2025; one company paid EUR 2.79 million over the Cologne zoo-bridge tender; and further fines have landed across the construction and steel-supply sectors. But if the current, reactive toolkit already produces eight-figure fines, the case for suspicionless mass screening may, constitutionally, need more than a borrowed statistic: the legislator would have to show that the FCO’s expanded power adds something to its cartel enforcement power which the current system cannot provide. A similar concern surfaced with the recent fuel-sector amendment, where the mix of a reversed burden of proof and lowered thresholds shifts the balance between authority and business, and where the constitutional test will turn on how strictly the courts read the high requirements the legislature so confidently invokes (see our article here). Procurement screening invites a similar question – and the draft’s reasoning on that point seems fairly superficial.

Another element of the draft seeking to foster enforcement by the FCO is the energy-sector abuse control under Section 29 ARC. The Section 29 regime, currently set to lapse at the end of 2027, would under the draft run to the end of 2032 instead, carried by a live enforcement record in district heating and household base supply. The legislator assumes there will be an ongoing need for strict review of such sectors.

And there is another part granting more reach to the FCO, although it is one offering more guidance to companies: the right to an FCO no-action decision under Section 32c(4) ARC, until now limited to horizontal cooperation, opens to vertical agreements through the simple deletion of the words "with competitors". Today vertical cooperation can rely only on a so-called chairman's letter pursuant to Section 32c(1) ARC, but without being entitled to such a decision. Now such a right is offered to all companies, provided they can show a significant legal and economic interest in such a decision. The draft explains this by pointing out that inter alia data pools have a keen interest in legal certainty, irrespective of whether they are operated by competitors or non-competitors.

Finally, the draft also does a lot in terms of institutional, procedural and administrative housekeeping. Much of it will matter little to most businesses: third-party file access for damages claims stays much as it was, the long-unused formal oral hearing gives way to a more flexible public format, the FCO is put on the same footing as the public prosecutor in fine appeals, the press-publisher cooperation privilege becomes permanent, and the obsolete competition-rules register (Sections 24 to 27 ARC) is abolished. The essence:

• From 2028, all merger filings will be electronic only.
• For appeal proceedings, the draft restores third-party standing against a ministerial authorisation of any merger (Ministererlaubnis) by repealing the 9th ARC amendment's restriction, under which only those able to assert a violation of their own rights could appeal; the older rule – under which it sufficed to have been an (intervening) party to the underlying merger control proceedings – returns. Restoring broad standing fits the exceptional character of an instrument that overrides a prohibition the FCO has already reasoned out.
• For appeal proceedings to the Federal Court of Justice in general, the requirement of leave on points of law falls away. This is to be welcomed wholeheartedly as it frees appellants seeking to gain legal clarity from a highest-instance ruling from either hoping to be granted leave to file their appeal by the lower court (i.e., the Higher Regional Court of Düsseldorf) or else taking up a separate legal fight against the rejection of such leave; the latter happened more and more often in practice, resulting in protracted proceedings and fewer opportunities for the Federal Court of Justice to rule on the substance of high-impact cases and corresponding development of the law.
• Closing during an appeal gets safer: Where a clearance is struck down on the appeal of an intervening party – typically a competitor – or, in a hostile bid, of the target, the draft routes the deal into a fresh Phase II rather than into divestiture, and it confirms that the standstill obligation does not revive for steps already taken in reliance on the clearance or a derogation. For parties that close while such an appeal is pending, that removes a gun-jumping tail-risk (Sections 40(6), 41(1) ARC).
• The FCO president's term, in turn, becomes a fixed eight years without reappointment. This is a genuine novelty, since the Act currently sets no term limit. And while the European comparison supports such a limit (all other Member States have it), it is hard to escape the impression that the cap now proposed is linked to the public criticism lately directed at the authority and its president, much of it tied to the trajectory of fuel prices. That criticism has little real foundation, since the FCO holds no lever over the price at the pump, but it appears to be leaving a mark, and a reform billed as a matter of principle reads, in this light, partly as a response to mood.
• The FCO’s publication obligations are widened. The FCO already publishes many of its decisions as a matter of course. The draft now formally obliges it to publish various types of decisions such as Phase II merger decisions, Section 19a ARC and abuse decisions, commitments and infringement decisions – online in full text, subject to redactions for business secrets; this applies even before any appeal is exhausted (new Sections 43(4), 61(3) ARC). For the companies concerned that cuts both ways: a non-final decision in the public domain carries reputational exposure, but it also gives them an earlier and more exhaustive read on how the FCO is deciding.
• And the fee schedule, frozen since 1989, is altered substantially, most strikingly with a thirtyfold jump for Section 19a proceedings, from EUR 25,000 to 750,000. This means a new financial burden for the large digital firms in the FCO's Section 19a sights. For merger control, fees are doubled, with the maximum now sitting at EUR 100,000 (although that the average fee in merger control will only rise from EUR 7,500 to 10,000 in Phase I and from EUR 30,000 to 50,000 in Phase II).