The CMA Says No to nexfibre/Netomnia: The Overlap Map Decides It

On Friday 2 October 2026 the CMA provisionally found that nexfibre's acquisition of Substantial Group (Netomnia, Brsk and YouFibre) may be expected to cause a substantial lessening of competition (SLC) in the wholesale supply of fixed broadband. In plain terms, the deal has been provisionally blocked.

In July we argued in The Statistic at the Heart of the CMA's nexfibre / Netomnia Probe that concentration would be the crux of the inquiry, and that the most probable outcome was conditional clearance.

Network overlap figures hold up

The CMA's case rests on network overlap, and its headline figures line up closely with the postcode-level overlap data we published in May in The Geography of Fibre Competition.

Table 1: Point Topic overlap data vs CMA interim report findings. Source: Point Topic, CMA.

Two arguments from July also survive. The CMA did treat Openreach as context rather than the deciding frame: it included Openreach in the market but found that, as a regulated operator, it cannot freely compete on price or service, and the CMA rejected the parties' scale argument, much as we expected it to look past the "16% of the real market" defence.

An update to the July analysis and CMA reasoning

The national HHI pointed in the right direction, but the CMA's reasoning shows it was the wrong lens. Two things drove that.

The counterfactual.  Our July HHI compared the merger against Netomnia staying independent. The CMA instead concluded that, absent this deal, CityFibre would most likely have bought Substantial and sold off YouFibre. Re-running the HHI on our latest footprint data shows how much that changes the picture: the CMA's preferred outcome is the more concentrated one nationally.

Table 2: Altnet wholesale FTTP concentration under each scenario. Shares of FTTP premises passed by CityFibre, Netomnia, nexfibre, Hyperoptic and Gigaclear; premises covered by both merging networks are counted once. Source: Point Topic UPC.

Geography beats national share.  On national concentration alone, the CMA's own counterfactual looks worse. It still prefers it because CityFibre and Netomnia barely overlap (c.31,000 premises on our data, against c.214,000 between nexfibre FTTP and Netomnia), so combining them removes almost no local choice. nexfibre and Netomnia do overlap. In those streets, the deal cuts wholesale options from three (Openreach, VMO2/nexfibre, CityFibre-owned Netomnia) to two. That shrinks the share of VMO2/nexfibre's network facing three wholesale networks from c.32% to c.18%.

The lesson for altnet M&A is clear: a national HHI can flag a deal, but the overlap map decides it.

Why conditional clearance now looks unlikely

In July we expected conditional clearance, built on wholesale non-discrimination commitments, pricing safeguards or a targeted divestment. The interim findings make each of these hard to land.

• Wholesale commitments add little. The CMA already assumes VMO2/nexfibre would wholesale its network without the deal, so promising to do so does not restore the lost third network.

• Divesting the overlap would hollow out the deal. If VMO2 completes its FTTP upgrade, c.82% of Substantial's network sits inside the VMO2 footprint.

• Efficiencies were rejected as not merger-specific. The CMA found the extra scale small once VMO2's own cable upgrade is counted.

  
  

The parties must submit any remedies by 5pm on 16 October 2026.

What comes next

Responses to the provisional findings are due by 5pm on 23 October 2026, and the statutory deadline for a final decision is 15 December 2026. The parties could withdraw before then if no workable remedy emerges.

The wider signal matters more than this deal. The CMA stressed that its findings reflect this deal's specific overlap, not a view on altnet consolidation generally. That points to a two-track market: deals between networks with little overlap, such as CityFibre buying urban altnets or VMO2/nexfibre buying rural specialists, look clearable; deals that remove a third network from the same streets do not.

For anyone weighing the next altnet deal, the first question is no longer "how big is the combined footprint?" It is "how many premises lose a wholesale network, and where?" That is a question only postcode-level overlap data can answer.

Sources: CMA case page (Summary of Interim Report, 2 October 2026); Point Topic UK footprint data, Q1 and July 2026.