The proposed merger between Vodafone and Three, two of the UK's four mobile network operators, drags on like the 2010 Wimbledon tennis match between John Isner and Nicolas Mahut. That finished with a final-set score of 70-68 on day three and left John Isner, the eventual winner, so exhausted that he could barely stand in round two. Each telco already insists it is too weak to survive on its own. As they continue to return the stinging shots of regulatory opposition, they are unlikely to grow any stronger.

The operators, which announced their deal in June last year, are battling competition authorities unconvinced a merger will be good for the country. It would not be the first tie-up of so-called Tier 1 telcos. In 2016, UK fixed-line incumbent BT bought mobile operator EE in a £12.5 billion (US$16.5 billion, at today's exchange rate) deal. Five years later, cable operator Virgin Media came together with O2, another mobile operator. The difference this time round is that both Vodafone and Three are in the mobile networks game. A merger would reduce the supply of mobile networks to UK businesses and consumers, turning the market into an oligopoly, say critics.

The Competition and Markets Authority (CMA) published its latest edict this month on Friday the thirteenth, and executives at Vodafone and Three may have read it in horror. While the CMA has not served up an ace that kills the deal, it has raised serious objections after its more detailed review and is demanding remedies that sound onerous.

Its overarching concerns are twofold. The first is that a merger would drive up prices for UK consumers, partly by limiting the options available to mobile virtual network operators, which rent network capacity. The finding is hardly a surprise. Regardless of the sector, higher prices are a natural economic consequence of a reduction in the supply of something when demand for it is unchanged. The CMA's other concern, in its own words, is that "the merged firm would not necessarily have the incentive to follow through on its proposed investment" after the merger. In simple terms, it might not do what it promised.

The risk of broken promises

The pricing concerns are short of merit. Besides promising higher investment, Vodafone and Three have insisted there will be "no annual price increases" if they are allowed to combine. This does not mean they can be trusted. Analysis carried out by the CMA shows average prices would rise by 7% for Three and 3.8% for Vodafone after a merger. It measures the "harm" to consumers of this and higher prices charged by rivals at between £328 million ($433 million) and £1.1 billion ($1.5 billion). But the regulatory obsession with price looks misguided when UK tariffs are already among the lowest in the developed world.

Today, a consumer can sign up for a SIM-only deal that provides unlimited calls, minutes and a generous allocation of monthly gigabytes for less than £15 a month. This is well below the cost of other essentials such as energy and water, the prices of which have risen steeply in recent years. Average monthly revenue per user for a Vodafone postpaid customer is now less than £18. In the US, a postpaid phone customer of T-Mobile, widely seen as the value leader, generates almost $49. Even adjusting for higher income levels and other factors in the US, this is considerably more expensive. UK phone users, from a telco perspective, have long enjoyed a bargain.

The CMA's concerns about investment promises by Vodafone and Three, by contrast, look far more reasonable. The case previously made by the companies is that their weighted average cost of capital has exceeded their return on capital employed, inhibiting their ability to invest in 5G. There is some basic logic in this. Serving 30.6 million customers with a single nationwide network should be more economical than dividing them between two operators and their parallel systems. The healthier and more profitable Vodafone-Three would pump £11 billion ($14.5 billion) into nationwide 5G over the subsequent decade, say the operators. As on prices, they cannot be trusted, according to the CMA. But a broken promise here would be much harder to forgive.

The actual numbers are a concern. For a start, the direct linkage of £11 billion to 5G "standalone" (SA), a newer version of the technology, is misleading in the original statement on the deal. The small print shows the operators are really talking about "combined capex spend," which inevitably covers more than just 5G. What's more, over a ten-year period £11 billion works out at just £1.1 billion annually, which is about £400 million ($528 million) less than Vodafone and Three collectively spent in the fiscal year before the deal was announced.

Supporters would probably argue that channeling £1.1 billion into a single network is better than splitting £1.5 billion ($2 billion) across two. But it sounds bad for suppliers, despite the supportive remarks Ericsson and Nokia tend to make about any planned European mergers. This is, of course, not a direct concern of the CMA. Yet Ericsson and Nokia are arguably the UK's only viable 5G options since it banned doing business with China's Huawei. A collapse in 5G investment globally has already weakened them and there is known to be some concern at senior telco levels about their wellbeing. A merger that reduces the overall level of spending could add to their pain.

The CMA's justifiable worry, however, is that Vodafone and Three would not deliver the full joint business plan (JBP) in a less competitive three-player market. Separately, each telco maintains about 18,000 sites. The stated goal of the JBP is to operate a network of 26,000 after the merger, implying that around 10,000 will be decommissioned. But Ofcom, the UK telecom regulator, fears a bigger number will be taken out of service, according to the CMA.

Despite what they say, the operators might also have little commercial incentive to splurge on 5G rollout. In the absence of new applications that demand a 5G (or 5G SA) as opposed to a 4G service, 5G networks hold little value for the average customer. During a survey of consumers carried out by the CMA, 76% of respondents said they would be unwilling to pay for faster speed. "Ofcom also noted the current limited evidence of customer willingness to pay a premium for services that rely on 5G SA capabilities," said the CMA.

Big Brother is watching you

Merger remedies in the mobile market are typically "structural," meaning a deal is blocked outright or that parties must divest assets to address competition concerns. Both remain on the table, according to the CMA's paper on possible remedies. But the alternative it seems to prefer is one or more "behavioral" remedies whereby Vodafone-Three would essentially be supervised and held to account on its pricing and investment pledges. A combination of structural and behavioral remedies also remains an option.

Vodafone and Three have already offered to sell spectrum to Virgin Media O2, a rival, if the deal is allowed. But a divestiture that introduces a new entrant to the market would undermine the case they have made for combining. They are, at least, now unlikely to face any regulatory pressure to withdraw from or somehow unwind a network-sharing agreement. BT and Three partner on MBNL, while VMO2 and Vodafone share infrastructure through Beacon. Vodafone-Three, then, would be active in both. There is fear it would have access via MBNL to commercially sensitive information about BT's plans, which it could use to shape its network strategy. The CMA, however, is now dismissive, insisting there are already enough safeguards to limit the risks.

Given all this, the likely next step in the process is the offer of behavioral remedies by Vodafone and Three. But this would be fraught with danger for both the CMA and the companies (and, by implication, the wider market). The CMA itself recognizes the risk of market distortion associated with any strict monitoring of wholesale rates or consumer tariffs. As for supervising Vodafone and Three on investment alone, this would probably not offset "anti-competitive effects" the CMA has observed in retail and wholesale markets, it said.

It would also be extremely difficult to measure spending on 5G, which telcos do not ordinarily break out. What's more, requiring Vodafone-Three to prove it has met 5G investment promises each year could hurt the operator if the market deteriorates or the execution of the merger runs into problems.

A confusing jumble of vendors

Integration of the two networks is unlikely to be easy because each company relies on a different mix of vendors. In the core, the control center, Vodafone is shifting from a 4G platform provided by Cisco to a "converged" core, supporting multiple generations, from Ericsson. But Three uses a Nokia core it says is overdue an upgrade. The migration of customers and systems between cores can be time-consuming, tricky and potentially expensive.

But on the radio access network (RAN) side, there is even more confusion. Vodafone now relies on Ericsson at about 15,500 of its 18,000 or so sites. It is stripping Huawei equipment out of the other 2,500 and replacing it with a mix of vendors including Samsung, NEC and various IT companies (Dell, Intel and Wind River among them). As the controlling player in a joint venture with Three, holding a 51% stake, it seems unlikely to abandon this effort. It is among Europe's biggest supporters of the "open RAN" move to create more interoperable network components, allowing an operator to use products from multiple vendors at the same site. But the Huawei rip-and-replace job, while not behind government targets, seems to be taking forever.

Meanwhile, Three now has five RAN vendors in its network, perhaps more than any other major European telco, including Nokia (for 3G), Samsung (4G), Huawei (4G and 5G), Ericsson (4G and 5G) and Mavenir (now being used for small cell coverage in parts of the UK). This mishmash is partly the consequence of picking a different vendor for each new generation when rivals were taking "single RAN" products, which lumped multiple Gs onto a radio platform sold by one vendor. It may also be partly responsible for Three's return-on-investment plight and looks hard to unravel. In Glasgow (and presumably other places), radios from Nokia, Samsung, Huawei and Ericsson are intermingled, Three revealed at the start of the year.

The CMA does not buy the sob stories advanced by Vodafone and Three, which depict the UK as a dysfunctional duopoly of BT and VMO2. It is VMO2 that has the worst mobile network, according to the CMA. Business plans it reviewed show Vodafone and Three would continue to invest in networks as separate companies, it said. It does not believe their claims they cannot compete effectively unless they are joined. The merger's prohibition, in the CMA's words, remains possible.