Publication
Publication
The CMA’s phase 2 docket might look like mere crumbs in comparison to years gone by, but that doesn’t mean there are no lessons to be learned. Take the two most recent Phase 2 cases. One: a bread merger that brought together two out of three leading brands for sliced bread, one of which had decades of losses, in the context of a market in long-term structural decline (ABF/Hovis). The other: a £3 billion-plus digital content giant agreeing to acquire its arguably closest rival in a global stock image market being disrupted by generative AI (Getty/Shutterstock). Both went to Phase 2. Neither ended in prohibition. Together, they offer what might be the clearest snapshot yet of a CMA that is doing what it said it would – i.e. clearing all deals capable of clearance (which the prohibition of Aramark / Entier shows is still not all deals).
The CMA’s Phase 2 deal mortality rate — which for years hovered stubbornly around 60-70% — has fallen to 28% in 2025-26: odds of clearance that were worse than 1:3 on the numbers alone have flipped. So, what happened, and what does it mean for parties thinking about their next transaction? Grab a slice and read on.
Associated British Foods (ABF), owner of Allied Bakeries (AB) and their flagship branded bread product Kingsmill, agreed to acquire Hovis from Endless in August 2025. In headline terms the deal involved combination of two of the three largest national brands with shares in excess of 80% in some market segments. Two brands, stacked in virtually every supermarket aisle, covering a product that the CMA itself acknowledged is an important staple product, particularly for consumers on lower incomes. On its face, this looked like precisely the kind of deal that even a CMA stepping back on global mergers might have appetite to intervene in.
An early test case of the DMCC fast-track procedure in a full Phase 2, the CMA made the reference on day 10 of Phase 1, skipping straight to the in-depth inquiry at the parties’ request. In place of the usual Phase 1 decision, the CMA published an “Areas of Focus” document to flag the theories of harm it intended to examine. Fast-track is palpably not a backdoor to a more cursory review: the Phase 2 inquiry was rigorous, evidence-heavy, and ran its course (albeit without any extension); but it does appear to have given the parties meaningful process benefits and allowed them to engage constructively on the substantive issues from the outset.
The centerpiece of the CMA’s analysis was the counterfactual and failing firm, historically a high hurdle to overcome but an area where the CMA has recently shown more willingness to engage as we explained here. The Parties submitted clear evidence supporting a failing firm defence, with AB being heavily loss-making (since 2011/12), not of strategic importance to its parent company and a weak brand (in 2023 Tesco shelved the vast majority of Kingsmill branded products). Moreover the CMA agreed that the plant bread market was — and is — in structural decline: consumer tastes are shifting (pre-sliced packaged bread in particular declining), and AB’s national daily-delivery distribution model is expensive. In short: AB was, in the CMA’s own assessment, likely to exit the market absent the merger.
That’s the first limb of the exiting firm framework. The second: would there have been an alternative purchaser who would have kept AB alive as a competitor? Here, the CMA went digging — and found very little grain of comfort for an alternative buyer. The pool of potential strategic purchasers was thin. No one who was approached indicated genuine interest in buying AB, or AB’s GB assets, and running them as a going concern at a price above liquidation value.
As is often a feature in retail cases, the CMA assessed the impact of the transaction in Northern Ireland separately from Great Britain. The CMA also settled on a narrow approach to product market with a detailed qualitative assessment of competitive dynamics for pancakes, soda farls and potato farls. In a world of increasingly complex economic analysis and data, it was refreshing to see some back-to-basics evidence gathering, with the CMA citing customer views that soda farls and potato farls are not substitutable because a traditional Ulster Fry includes both of these products. Not in the yeast bit controversial for the strong contingent of Northern Irish platypus.
However, the CMA’s March 2026 Interim Report provisionally found an SLC in Northern Ireland, concluding that an alternative purchaser for AB NI might have kept the business running, particularly given an ongoing parallel sale process of parts of AB NI (the AB NI Sales Process). In a world where the CMA’s actions on the global stage have diminished, this decision would have been a strong endorsement of the CMA’s impact in the smallest of the four nations in the UK…
ABF disagreed with the conclusions in NI, and pushed back with new evidence, including from the live sales process. Of 17 potential purchasers, only four made any form of offer — and three of those four implied a net cost to ABF higher than outright liquidation. In other words, it would have cost ABF more to sell AB NI than to shut it down. The remaining offer was predicated on and further “red flag due diligence” and was therefore considered by the CMA to involve material uncertainty. In short, blocking the deal over NI seemed likely to lead not to an alternative purchaser, but to liquidation which would take thousands of jobs with it.
The CMA issued a Supplementary Interim Report in May 2026 — a rare procedural step — reversing its provisional NI finding. The Final Report, published on 16 June 2026, confirmed clearance in both GB and NI. No SLC. No remedies. Unconditional clearance.
Getty agreed in January 2025 to acquire Shutterstock for approximately £245 million in cash plus shares in a transaction valuing the combined entity at over £3 billion. As two of the world’s largest visual content platforms, the UK news media sector was quick to raise concerns. The Phase 1 process resulted in two SLC findings in relation to the supply of editorial content in the UK, and the supply of stock content globally. A remedy package for the sale of Shutterstock’s global editorial business was offered in Phase 1 but the CMA considered that the “complex package of remedies” offered at a “late stage in the Phase 1 process … did not fully address it concerns” and the parties were sent packing for Phase 2.
Consistent with its focus on deals with a distinct and direct UK impact the CMA issued a split ruling in its May 2026 Final Report. In the supply of editorial content in the UK, it found an SLC: the merged entity would combine Getty (already the clear UK market leader, with close to or above 50% combined share of supply) with one of the few meaningful alternatives. In the supply of stock content globally, by contrast, it found no SLC, even though the parties competed closely.
The stock content finding is where the CMA’s forward-looking analysis came into focus. The CMA did not just look at today’s market; it asked where competition was heading. The answer was: towards generative AI. Adobe Firefly has been adopted far more widely than the parties’ own GenAI offerings. Canva is growing rapidly, recently targeting large enterprise business customers directly. A quarter of customers surveyed expected their use of AI-generated content to increase considerably within two years. One of the clearer recent instances of dynamic competition from tech-based challengers being used by the CMA to support a “no SLC” conclusion.
Having initially tabled a broader editorial divestiture, the parties narrowed their formal Phase 2 remedy proposal to divestiture of just Splash and Backgrid, leaving the core Shutterstock Editorial business (the former Rex Features, with its news, sport, and entertainment content) intact within the merged group.
The market’s reaction? Unanimously negative. The CMA required instead the Wider Shutterstock Editorial Divestiture: the full editorial business. Helpfully, the CMA accepted that Splash and Backgrid need not be sold to the same purchaser as Shutterstock Editorial.
For the antitrust nerds who've made it this far — and the deal teams who loafed straight to this bit first —our wide-angle take is that both cases clearly demonstrate the “5th P” of CMA pragmatism in action. The CMA digging its heels in and asserting concerns in stock content in Getty/Shutterstock and in Northern Ireland in ABF/Hovis looks likely, in both cases, to have led the CMA to ultimately prohibiting the transactions. In each case that risked being an unpopular outcome for different reasons (intervening in a global deal which had been unconditionally cleared by the DoJ for the former, and significant job losses in an already struggling industry for the latter).
Instead, the CMA treated live sales process data as real-time counterfactual evidence leading to the reversal of an earlier provisional SLC finding; and gave full competitive weight to generative AI as a present and future constraint in a traditional content market. This meaningful substantive engagement with evolving and forward-looking evidence is the 4Ps(+) in action, and demonstrates an authority that is not reflexively interventionist but attuned to commercial and economic reality (one could even say that when the facts change, the CMA changes its mind).
For parties considering CMA engagement, the message is clear: the new CMA rewards careful, evidence-driven analysis and early, credible engagement. The best pictures, as any photographer knows, come from getting the exposure right. So does the best merger strategy. The CMA of 2026 is a much more welcoming environment for deals - but get it wrong, and your deal could still be toast.
Authors: Aoife Monaghan, Jonny Ford, Verity Egerton-Doyle, Helen Crossley
