The CMA says a Virgin Media O2 broadband deal cuts three-provider competition from 32% of its network to 18%. That finding rests on an upgrade its own report calls uncertain, never recalculated without it.
Estimated reading time: 6 minutes
In short. The Competition and Markets Authority's Interim Report into nexfibre's takeover of the broadband altnet group Substantial finds that, relative to its chosen counterfactual, the deal "reduces the proportion of VMO2/nexfibre's total network in which VMO2/nexfibre, CityFibre and Openreach are all present from c.32% to c.18%" [1]. The same report says those percentages "are calculated on the basis that VMO2 upgrades its entire network to FTTP, but the percentages are similar if we were to instead use VMO2/nexfibre's current FTTP network only" [1], a claim stated but never shown with its own figures, five paragraphs after the report itself admits "some uncertainty as to the exact timing and extent of future network upgrades, including as a result of recent financial conditions that may have impacted VMO2's plans" [1]. Two documented mistakes: four stars.
nexfibre is the wholesale fibre broadband joint venture of Liberty Global and Telefonica, who also jointly own Virgin Media O2, together with the private equity investor InfraVia [1]. It wants to buy Substantial, the holding company behind the altnet fibre builders Netomnia and Brsk and the retail ISP YouFibre [1]. The CMA referred the deal to an in-depth Phase 2 investigation on 1 July 2026, with a statutory deadline of 15 December 2026 [2]. On 2 October 2026 the Inquiry Group running that investigation published its Interim Report: a provisional finding that the deal gives rise to a realistic prospect of a substantial lessening of competition (SLC) in the wholesale supply of fixed broadband services [1]. This piece tests the working behind that one finding, the only quantified figure in the whole report, against what the report itself says about the assumption it rests on.
The percentage that disappears
The CMA's method is to measure network overlap. VMO2/nexfibre's fibre network is "fully overlapped by Openreach, c.14% overlapped by Substantial, and c.18% overlapped by CityFibre" [1]. In the CMA's chosen counterfactual, in which Substantial is bought by the rival network builder CityFibre instead of nexfibre, three providers would compete across all of that overlap, "ie c.32% of the total VMO2/nexfibre network" [1]. If nexfibre buys Substantial instead, CityFibre never gets a foothold in the areas that used to overlap with Substantial, and the report finds this "reduces the proportion of VMO2/nexfibre's total network in which VMO2/nexfibre, CityFibre and Openreach are all present from c.32% to c.18%" [1]. Run the comparison the other way and the same pattern holds: Substantial's own network is "c.26% overlapped by VMO2/nexfibre's FTTP network, and would be c.82% overlapped if VMO2 were to upgrade its cable network to FTTP" [1]. That 82% figure, not the 26%, is the one the report treats as the live competitive picture, because VMO2's cable-to-fibre conversion is not yet finished.
An upgrade the same report calls uncertain
VMO2 currently runs most of its network on cable, not fibre, and the CMA's own text says plainly that this matters for the sums: "The percentages presented below are calculated on the basis that VMO2 upgrades its entire network to FTTP, but the percentages are similar if we were to instead use VMO2/nexfibre's current FTTP network only" [1]. No alternative set of percentages, current-network-only, appears anywhere in the published Interim Report or its Summary. The claim that the two bases produce similar numbers is asserted once and never demonstrated.
Five paragraphs earlier, in the same document, the Inquiry Group explains why that upgrade cannot simply be assumed to happen on schedule: "we recognise that there is some uncertainty as to the exact timing and extent of future network upgrades, including as a result of recent financial conditions that may have impacted VMO2's plans" [1]. The report resolves that uncertainty in its own favour, finding "a clear long-term incentive on VMO2 to upgrade its cable network" and concluding that, absent the deal, "VMO2 would continue to upgrade its network to FTTP" [1]. That may well be the right call. But the headline competition figures, the 32% that falls to 18%, the 26% that the report treats as really an 82%, are the post-upgrade numbers, not the current ones, and the one sentence that says the choice does not matter much either way is the one sentence in the report that shows no working at all.
A probability that hardens into certainty
The report's case for why this matters to ordinary broadband customers, not just to the three companies, runs on a similar pattern. Its own evidence is stated as a tendency: "where ISPs receive better wholesale terms, this is often reflected in better deals for end-consumers and businesses, including lower prices and improved service quality" [1]. "Often" is not "always", and the report does not say what distinguishes the cases where better wholesale terms pass through to customers from the cases where they do not, or which kind this one is. The very next sentence drops the qualifier: "A reduction in competition at the wholesale level would therefore directly impact end-consumers and businesses (including, disproportionately, those in the midlands and north of England given the location of Substantial's network)" [1]. A pass-through pattern the report's own evidence describes as occurring "often" becomes, one sentence later, something that "would therefore directly impact" named regions, with no finding in between that this transaction sits in the "often" category rather than an exception to it.
Credit where due
The Inquiry Group is not hiding its own uncertainty; it states it outright, both on the network-upgrade timing above and on the harder question of who would have bought Substantial if nexfibre had not. Working out whether Substantial would have stayed independent or been bought by CityFibre instead, the report calls "an inherently uncertain exercise" in its own words, and explains that it is applying the CMA's published Merger Assessment Guidelines rather than simply assuming nothing would have changed [1]. The Inquiry Group also declines to overreach: asked by third parties to treat this case as evidence for or against consolidation across the wider altnet sector, it says its findings "are based on our assessment of this specific acquisition" and "do not reflect any view on consolidation in the wider altnet sector" [1]. And the timetable is genuinely open: the CMA is inviting representations on these provisional findings until 5pm on 23 October 2026, before any final decision is made [1].
The claims, tested
| The document's own words | What the record shows | Verdict |
|---|---|---|
| "The percentages presented below are calculated on the basis that VMO2 upgrades its entire network to FTTP, but the percentages are similar if we were to instead use VMO2/nexfibre's current FTTP network only" [1] | No current-network-only percentages are published anywhere in the Interim Report or its Summary; five paragraphs earlier the same report says there is "some uncertainty as to the exact timing and extent" of the VMO2 upgrade the headline figures assume [1] | The claim that the choice of baseline barely matters is asserted, not shown |
| "where ISPs receive better wholesale terms, this is often reflected in better deals for end-consumers and businesses" [1] | The next sentence drops "often" and says a reduction in wholesale competition "would therefore directly impact" consumers and businesses, disproportionately in the midlands and north [1] | A stated tendency hardens into a direct consequence one sentence later, with no finding that this case is the typical one |
Verdict
Four stars, two documented mistakes. The CMA's Inquiry Group is working through genuinely difficult counterfactual questions, which bank on deals that were never signed and network upgrades that have not happened, and it says so plainly rather than pretending otherwise. But the single number this provisional finding turns on, competition across 32% of a fibre network falling to 18%, is built on an assumption the report itself calls uncertain, and the reassurance that the number would not change much either way is never tested against the numbers it claims would prove it. Representations on these provisional findings are open until 23 October 2026; anyone making one has grounds to ask the CMA to publish the current-network figures it says it has already calculated.
The star score counts two documented mistakes: first, the report's central overlap percentages (32% falling to 18%; 26% treated as 82%) are calculated assuming VMO2 completes a cable-to-fibre upgrade that the same report, five paragraphs earlier, says carries "some uncertainty as to the exact timing and extent" because of VMO2's own recent financial conditions, and the claim that the percentages would be "similar" using the current network instead is never shown with its own figures; second, the report's basis for consumer harm shifts from a qualified "often reflected" pass-through pattern to an unqualified "would therefore directly impact" conclusion within the same paragraph, without establishing that this transaction falls into the "often" case. This piece makes no finding against any named individual; the CMA's Inquiry Group is referred to only in its institutional, decision-making capacity, and its findings are explicitly provisional, open for challenge until 23 October 2026. All sources below were fetched directly on 14 October 2026 and returned HTTP 200.
Sources
- Summary of Interim Report: Anticipated acquisition by Liberty Global, Telefonica and InfraVia (through their joint venture, nexfibre) of Substantial (PDF), Competition and Markets Authority, published 2 October 2026
- nexfibre / Substantial merger inquiry, Competition and Markets Authority case page, referral 1 July 2026, statutory deadline 15 December 2026