The year in UK mergers: long-term trends and highlights from 2024/25
18 June 2025The CMA recently released its merger outcomes data up to and including March 2025. In this article, we look at how that data compares to that from the first ten years of CMA merger enforcement, and pick out some key trends and themes from the year’s decisions.
2024/25 will likely be viewed as a pivotal year for UK merger control. In the face of criticisms of its approach by the government and business community, the CMA introduced proposals aimed – amongst other things – at making merger enforcement more proportionate. The year also saw some notable outcomes, with transactions being cleared in ways that might not previously have seemed achievable to merging parties.
Against that context, there is some merit in seeing what can be discerned from shifts in CMA merger outcomes over time – both to assess whether the accusations of over-interventionist enforcement were justified, and to see whether the CMA’s shift in focus is already bearing fruit.
2024/25 at a glance
| Phase 1 | Phase 2 |
41 inquiries concluded
| 7 inquiries concluded1
|
Digging into the data – Phase 1
Number of Phase 1 cases continues downward trend
Looking at the long-term trend since the CMA’s establishment in 2014, the number of mergers being investigated by the CMA at Phase 1 appears to be declining.
Whilst 2023/24 saw an uptick in the number of decisions, it should be remembered that the CMA’s two inquiries in the vets sector accounted for 25 mergers between them. Some adjustment should also be made for the years 2020-2022, as many transactions will have been postponed or delayed due to the Covid-19 pandemic.
In all, however, it appears that the evolution of the briefing paper process - (with exponential growth in the numbers of briefing papers submitted since 2017) - has (more than) succeeded in offsetting the large increase in Phase 1 case numbers expected post-Brexit.
Phase 1 outcomes: unconditional clearances have become significantly less common, but recent years show signs of recovery
Phase 1 outcomes have – as one would expect – fluctuated year-on-year. That said, unconditional clearances at Phase 1 have generally become less common than they were in the earlier years of the CMA’s existence. At the same time, there has been an upward trend in the proportion of cases resulting in referral. The last two years since 2022/23 have, however, bucked those trends, with unconditional clearances creeping back above 50% of Phase 1 cases, and the proportion of cases referred remaining relatively low.
Perhaps the simplest metric by which one can measure Phase 1 outcomes is the proportion of relevant merger situations in which the prospect of an SLC was found (the purple dotted line above). This reinforces the conclusion that Phase 1 outcomes have generally become less favourable to merging parties over time, whilst indicating that 2022/23 and 2023/24 may represent the high point of CMA intervention.
Digging into the data – Phase 2
Phase 2 clearances were more common in the CMA’s early years – but are beginning to recover
With a much smaller sample size to work from, one would expect a greater degree of year-on-year variance in Phase 2 outcomes compared to Phase 1. However, 2018-2023 appears to have been something of a consistent fallow period for Phase 2 clearances, with the last two years reverting close to levels seen in the CMA’s early years.
In cases that are not cleared, outcomes remain broadly stable
Despite business concerns about the predictability of CMA merger enforcement, the proportion of cases in which an SLC was identified that resulted in prohibition has remained relatively low, and broadly flat, as has the proportion of such cases abandoned (bar a spike in 2020/21).
2024/25 highlights – Phase 1
CMA intervention this year largely focused on traditional horizontal theories of ham. Whilst a substantial proportion of the cases examined by the CMA included vertical elements, none of the 12 cases in which an SLC was found were purely vertical in nature, and nearly all vertical theories of harm considered by the CMA were ruled out (AlphaTheta/Serato was a notable exception – here the CMA was concerned the acquirer might degrade rival DJ hardware manufacturers’ interoperability with Serato’s market-leading DJ software, or use their commercially sensitive information to its advantage).
The CMA’s Mergers Intelligence function (MIF) continues to be very active: out of the 41 concluded cases, 26 were identified by the MIF as warranting investigation (17 anticipated and nine completed) – although many of these will have stemmed from briefing paper processes rather than being ex officio investigations.
A resurgence in de minimis clearances
In 2023/24 the CMA did not clear any cases on the basis of the de minimis exemption, and across the four years before that, there were only seven de minimis cases in total.
However, in April last year the CMA amended its de minimis policy, increasing the market size threshold to £30m2. This saw three mergers cleared through application of the discretionary de minimis exemption: Arla/Volac, XSYS/MGS and Keysight/Spirent (in which the CMA found a prospect of an SLC in five different markets), although the CMA rejected de minimis arguments in Theramex/European Rights to Viatris’ Femoston and Duphaston products, including on the basis that the transaction could be expected to result in detriment to patients that rely on the hormone replacement therapy drugs under consideration.
The CMA’s AI partnership inquiries explored the boundaries of the relevant merger situation
In this high-profile series of investigations, the CMA examined five transactions between established Big Tech players and innovative generative AI and foundation model developers. These cases accounted for all four found-not-to-qualify (FNQ) decisions published by the CMA in 2024/25.
The published FNQ decisions (Microsoft/OpenAI, Microsoft/Mistral, Amazon/Anthropic and Alphabet/Anthropic) were unusually detailed. They are helpful in shedding additional light on the boundaries of the concepts of material influence (which was found in Microsoft/OpenAI but not the other cases, despite them all involving combinations of minority debt/equity investments and compute and distribution agreements), de facto control (which was examined but not established in Microsoft/OpenAI) and enterprises (with key employees and an IP licence being sufficient to comprise an enterprise in Microsoft/Inflection).
For more information on these cases, see our recent articles on Microsoft/OpenAI and Microsoft/Inflection.
In some (but not all) cases, the CMA was able to see past high shares of supply
For example, in TGS/PGS UK, the CMA found that the parties had combined shares of up to 50% of the supply of marine seismic data and were two of a limited number of global multi-client seismic data suppliers. However, the investigation also showed they had differentiated offerings, and would continue to be constrained by a number of large competitors post-merger. Similar reasoning led to the CMA clearing Thermo Fisher Scientific/Olink and Acerinox/Haynes.
In Altrad/Stork UK, the CMA considered that the merging parties’ combined 50% share of the supply of fabric maintenance services to the offshore oil and gas sector was not fully informative of closeness of competition between the parties and the constraints imposed by alternative suppliers, in a market that was characterised by bidding processes.
On the other hand, GXO/Wincanton was referred to Phase 2 despite the parties’ combined share of the supply of mainstream contract logistics services being no higher than 30%. The CMA focused on the fact the parties competed particularly closely for certain customers, with only two major alternative suppliers for that demand, and the tail of suppliers imposing limited competitive constraint.
The CMA was also unwilling to depart from the decision rule it applied in Barratt/Redrow (namely, a share in excess of 40% and two or fewer competitors in the supply of new homes) – refusing to entertain additional evidence submitted by the parties on the one local area that failed that rule, and requiring the divestment of one of the merging parties’ developments in that area.
“Proportionality” and “pace” in action?
Earlier this year, the CMA formally launched its “4Ps” programme – promising to deliver rapid change to the UK’s merger control system through a package of measures aimed at improving its pace, predictability, proportionality and process.
Certain decisions could be viewed as early evidence of the CMA’s new approach in action. For example, it has recently begun publishing much shorter decisions (only 15-20 pages in length) where possible, and has cleared some cases extremely quickly (XSYS/MGS was cleared on a de minimis basis in only 13 working days; William Grant & Sons/The Famous Grouse was cleared almost three weeks early). The CMA also adopted a proportionate approach in Nationwide/Virgin Money – choosing in its decision to focus on a few areas of overlap and ignoring numerous others, and clearing the deal a week early with a 25-page decision, despite the scale of the transaction.
On the other hand, the average length of pre-notification was a (joint) record high: 71 working days. The CMA faces a challenge in meeting its new target of 40 working days for pre-notification (at least, for parties that do not opt out of being subject to that target).
2024/2025 highlights – Phase 2
This was a relatively quiet year for Phase 2 mergers, given the low number of references in 2023/24.
Substantively, as with Phase 1, the focus was on traditional horizontal theories of harm. No SLCs were found as a result of vertical effects (although they may have been substantiated in AlphaTheta/Serato, if the deal had not been abandoned), nor as a result of conglomerate/ecosystem effects or a loss of dynamic competition between the merger parties.
Behavioural remedies make a comeback
In the most high-profile decision of the year, the CMA approved the merger of Vodafone and Three, subject to a package of behavioural remedies aimed at locking in rivalry-enhancing efficiencies, incorporating commitments to: invest £11bn in network improvements over the next eight years; retain certain low cost retail tariffs for a period of three years; and make available to virtual network operators a predetermined wholesale reference offer, again for a three-year period.
This is the fourth example of the CMA agreeing to behavioural remedies at Phase 2, but the first since Bauer Media Group in 2020. Overall, it is the tenth time the CMA has agreed to a purely behavioural remedies package. You can read more about the decision in our previous article. The CMA’s review of its approach to remedies, announced in March, signalled that the CMA may have greater appetite for such remedies in future (read more on the review here).
A willingness to revisit prior conclusions
Global Business Travel Group/CWT Holdings was the first merger investigated by the CMA under its revised Phase 2 process, which amongst other things replaced the Provisional Findings stage with an earlier Interim Report. In this case, the Interim Report provisionally found an SLC as a result of horizontal unilateral effects. The CMA then considered additional evidence and representations from the parties, before reversing its SLC finding in a Supplementary Interim Report published three months later, on the basis that the target was becoming a less important competitive force than its market share suggested.
This outcome – which you can read more about in our previous article – highlights the benefits for merging parties of being able to engage earlier with the CMA Phase 2 panel on the competitive assessment, as well as a welcome willingness on the panel’s part to revisit its initial views3.
Similarly – although not quite as dramatic a reversal – in Lindab/HAS-Vent, the CMA reversed its Phase 1 conclusion that competition for the supply of ducts and fittings for ventilation systems took place principally at the national level. By looking at local overlaps – rather than the parties’ >50% share of supply at the national level – the CMA was able to clear the transaction at Phase 2, subject to the divestment of two local sites.
A cautionary tale
Whilst Vodafone/Three stole the headlines, Spreadex/Sporting Index threw up some of the year’s most interesting procedural points. First and foremost, the CMA declined to defend its prohibition decision on appeal, conceding that it had not properly summarised confidential third-party information before sharing it with the merger parties. The Competition Appeal Tribunal therefore remitted the merger to the CMA, with a further decision due in August.
The stakes are high for Spreadex. The case involved a completed transaction: Spreadex effectively shut down Sporting Index’s betting operations after acquiring most (but not all) of its assets, turning the Sporting Index website into a white-labelled version of Spreadex’s own spread-betting platform.
The CMA rejected an offer by Spreadex to divest the assets it had acquired as a remedy to the SLC, as they did not constitute a viable standalone business, and their sale would not restore the pre-merger competitive position. Ultimately, Spreadex was required to develop a bespoke technology platform enabling Sporting Index to operate on a standalone basis again, and enter into a transitional services agreement with the purchaser of the business whilst it built up its own capabilities.
On 5 June the CMA published its Provisional Findings on the remittal, indicating that it had not changed its position on either the SLC or how to mitigate or reverse it. It therefore appears likely that Spreadex will have to go ahead with this commercially unpalatable remedy.
Cross-cutting theme: the failing firm defence
Alongside the de minimis clearances and the Phase 2 behavioural remedy in Vodafone/Three, this year also witnessed a resurgence of the failing firm defence, with exiting firm arguments being accepted in three separate cases, at both Phase 1 and Phase 2. Prior to this, there were only two cases since 2019 in which such arguments had been accepted.
Two of the successful cases came in short succession, as detailed in our previous article: Eurofins/Cellmark, which the CMA cleared at Phase 1 in only 11 working days; and TLS/Tereos, in which the CMA rejected the target’s exiting firm arguments at Phase 1, only to accept them in Phase 2 – notwithstanding that the target was not loss-making and a definitive decision to exit the market had not yet been taken by the seller.
These were followed in March 2025 by Bidvest/Citron Hygiene, in which the CMA agreed at Phase 1 that the most likely counterfactual was Citron’s exit from the national market for the supply of washroom services, as there was evidence of it abandoning its national account strategy. At the local level, Citron would have continued to compete, as the relevant customer contracts were not loss-making, but with sufficient local competition in the overlap areas, the transaction was cleared unconditionally.
For completeness, however, it should be noted that exiting firm arguments were rejected in Spreadex/Sporting Index (both at Phase 1 and Phase 2) and in Roche/LumiraDX (at Phase 1, although it was ultimately cleared). Merging parties still need to be able to provide solid evidence to persuade the CMA that both Limb 1 (will the target likely exit the market?) and Limb 2 (is there no viable, less anti-competitive alternative purchaser?) of the relevant test are met.
Conclusion
The CMA generally warns against seeking to infer changes in its internal policy or approach from individual decisions or trends in outcomes, insisting that all mergers are assessed on a case-by-case basis, on their own merits.
Nevertheless, there are signs here that the CMA is trying to find more ways to clear transactions that might previously have been referred or prohibited – an inference which accords with the CMA’s stated goal of prohibiting only “truly problematic merger[s], where the harm to businesses and consumers cannot be effectively addressed through remedies”. It is also clear that the CMA is trying to conclude its processes more quickly where possible, including through more targeted investigations.
Deal-makers should find both developments encouraging.
1 One decision (Spreadex/Sporting Index) was remitted to the CMA after the CMA declined to defend it on appeal. To retain as complete a picture as possible we have nevertheless included it in this year’s data.
2 For mergers falling below this threshold, the CMA will consider: the market size; whether the transaction is replicable (i.e. whether it could form part of a “roll-up” strategy); what detriment might result from the merger; and the areas of priority in its Annual Plan, before deciding whether to make a reference. The Digital Markets, Competition and Consumers Act 2024 also introduced a statutory exemption applying automatically where each party’s annual UK turnover falls below £10m.
3 It was not, however, the first time the CMA has reversed a provisional SLC finding in Phase 2 – it did so in Copart/Hills Motors, under the old Provisional Findings process.
Get in touch
