Telcos bristle whenever governments hand operating licenses to new entrants, often to preserve the four-networks status of a mobile market after a merger. Throughout Europe, the telcos argue, networks are crammed into a confined space, like bodies in a lift, leaving no one with room to move. But they appear less worried about their similarly squeezed suppliers. In the market for radio access network (RAN) products, where mobile operators spend a lot of their money, they have held the doors open and invited others to enter. It has made life uncomfortable for Ericsson and Nokia.

"Tricky situation" was how one top industry executive recently described it for the Nordic suppliers, although without linking their perceived problems to the telco push for supply-chain diversity. The market, today, is less a confined space than a shrinking one, a temple-of-doom chamber with a ceiling that slowly descends. Revenues generated by all suppliers fell from nearly $45 billion in 2022 to $40 billion last year, according to Omdia, a Light Reading sister company. This year, it thinks sales will drop to below $37 billion. After partial buildouts of 5G, telcos everywhere have cut spending.

The industry assumed a backlash against Huawei and ZTE, Chinese manufacturers now unpopular with some governments, would have left space for others. But few European countries have done what the European Commission urged in its "5G toolbox" recommendations and moved to ban so-called "high-risk vendors." Only ten had done so by June last year, complained Thierry Breton, the former EU commissioner for the internal market and services, in a speech. "This is too slow, and it poses a major security risk," he said.

With Breton's resignation this week, and the temporary vacuum this creates, the issue of Huawei seems unlikely to receive much immediate political attention. "There is nothing happening in the EU before you have a new commission," said John Strand, CEO of Danish advisory firm Strand Consult. "Breton has been very tough on these things and he's out now." Since his June 2023 speech, moreover, there has been little of the progress he would have liked to see, as well as continued resistance to bans.

Sound and fury about Huawei

A focal point is Germany, Europe's biggest economy, where telcos rely heavily on Huawei. In July, its government tried to mollify EU officials by ordering the operators to delete Huawei's "core" network software by 2026 (something they had already done) and its management software for RAN and transport networks by 2029. Under the new rules, however, they are allowed to retain Huawei's RAN products, including baseband software that could theoretically – according to numerous telecom experts – contain malicious code. Deutsche Telekom and Vodafone are reportedly lobbying the Czech government to adopt the same approach.

Anti-Huawei efforts, meanwhile, have triggered a response in China, where Ericsson and Nokia have been effectively crushed by Huawei and ZTE. Ericsson's China sales plummeted from about $1.8 billion in 2020 to less than $990 million the following year, as 5G contracts went primarily to local vendors. Nokia's revenues from "Greater China" fell from more than $2 billion to about $1.5 billion.

Unfortunately, this has left the Nordic vendors relatively inert in what remains the world's biggest 5G market. A sign of that can be found in the latest investor presentation from China Mobile, the world's biggest mobile operator, which said it has put another 351,000 5G basestations into use this year, bringing its total to around 2.29 million. Its commercial deployment of 5G-Advanced, an evolution of 5G that is merely a talking point elsewhere, is underway in more than 280 cities, claims the telco.

Even if this figure of 2.29 million includes small cells and does not directly map to mobile sites, it seems to dwarf numbers published for other parts of the world. The US, for example, now has around 430,000 cell sites altogether, according to the latest figures from the Cellular Telecommunications Industry Association.

Outside China, Europe's unwillingness to abandon Chinese vendors means Huawei and ZTE continue to exert pricing pressure on Ericsson and Nokia. The same appears true in parts of the Americas. "We are seeing sharply increased competition from Chinese vendors in Europe and Latin America," said Börje Ekholm, Ericsson's CEO, on his company's most recent results call, warning analysts he expects to lose some deals.

Open RAN planning, Machiavelli-style

Despite offering words of encouragement to alternative RAN suppliers, telcos with brownfield networks have bought almost none of their products. Huawei, Ericsson and Nokia, the big three kit vendors, still account for roughly three-quarters of the global market, according to Omdia's data, and their respective market shares have barely changed since the birth of "open RAN," a bid to cultivate new suppliers by ensuring one vendor's products can be joined easily to another's (not possible in a traditional RAN).

But while they have failed to grab market share, those suppliers have certainly provided competition against Ericsson and Nokia in tenders that would previously have been less contested. Mavenir, a US software company that has branched into making radios, is known to have responded to a major Vodafone tender for supposedly 170,000 basestations across Europe and Africa. Rakuten Symphony, an offshoot of the Japanese e-commerce company, has competed vigorously for business in North America. Other Asian suppliers have done likewise.

All this reinforces industry suspicion that telcos never intended to do major deals with smaller vendors. Their somewhat Machiavellian open RAN scheme, say critics, was to maintain pressure on their traditional vendors and drive down prices. Ericsson, Nokia and Samsung, active in this market since the 4G era, now show up as the main open RAN winners. Without apparent shame or any sense of irony, AT&T has even marketed its decision to rip out Nokia and become almost solely reliant on Ericsson under the banner of open RAN.

AT&T's Ericsson deal is the most egregious example of what the industry now calls "single vendor open RAN." In this phenomenon, telcos demand compliance with open RAN specifications but take all products from the same vendor. Commercial arrangements are largely identical to those of a traditional RAN deal.

Ericsson, then, is now AT&T's sole provider of RAN software, service management and orchestration, a RAN intelligent controller (RIC) and a containers-as-a-service platform for that RAN. It is also down to supply most of the radios. Dell and Intel will contribute some IT hardware, but the only remaining opportunities for third parties are as developers of RIC applications or suppliers of radios in the few areas where AT&T has decided not to use Ericsson. Fujitsu, a longstanding Ericsson partner, is the only radio alternative AT&T has identified since announcing the deal nine months ago.

In what should come as no surprise, the economics of a multivendor RAN comprising numerous specialists do not make obvious sense. "Saying that you can make it cheaper and better if you take components from a lot of small players and put them together is basically saying that you can produce a PC more cheaply by buying different components from different suppliers and putting them together, and then compete against Dell, HP, Lenovo and so on, which have high volumes," said Strand.

While former specialists would undoubtedly oppose this view, they have themselves shape-shifted to accommodate the industry's single vendor needs. The most obvious example is Mavenir's expansion from RAN software into radio hardware. Yet besides defeating the original purpose of open RAN, this development makes it difficult for companies with limited resources to compete against the Tier 1 players.

"Let's say you spent half a billion dollars on R&D to develop all these radios, but you don't sell anything, and the radios have been there for a year," said Earl Lum, the president of EJL Wireless Research. "Come next year, those designs are old, because nothing stays still. So now you've spent all this money for a product no one wants. It's a bleeding thing that Ericsson and Nokia understand. The new guys who are software players thought they could build a radio. But you need to build 20 radios or 50 radios, all different flavors."

An unnecessary technological distraction

Maintaining the competitive pressure on suppliers is instinctive behavior for any business. But after years of mergers between kit makers, the recent initiatives linked to supply chain diversity and open RAN have arguably backfired, leaving telcos in a precarious position, with less solid choice than before. Not only has open RAN, judging by Omdia's data on market shares, failed to produce new viable alternatives to Huawei, Ericsson and Nokia, but it has also weakened the Nordic vendors as an additional source of pricing pressure and technological distraction that has nevertheless gobbled resources.

At a basic level, executives from Ericsson and Nokia have had to field questions from reporters, analysts and investors about open RAN's impact. In an industry that was already drowning in associations, they have had to commit resources to groups such as the O-RAN Alliance, the body in charge of specs, or face the charge they are unsupportive.

They have also had to invest in O-RAN compatibility that will have all the usefulness of the male nipple in a largely "single vendor open RAN" market. Incorporating new interfaces is not as straightforward as simply unlocking a door, as endless disputes about functional splits show. Ericsson, for instance, would rather build its radios with equalizers, which address interference, and its baseband products without them. But to be O-RAN compliant, it must put equalizers on both sides, even when a customer buys everything from Ericsson and the baseband equalizer is redundant.

It is a similar story for Nokia. Mismatched algorithms could affect the performance of massive MIMO, an advanced 5G technology, in multivendor networks, said Tommi Uitto, the head of its mobile networks business group, at the start of the year. "It can be done, but the performance is lesser, and some work is needed to make it feasible, make it practical, make it something that actually can be used in the field," he said. Yet this would all seem like wasted effort if the market dynamics do not change.

Complaints about high equipment prices look unsubstantiated. Telcos' capital intensity (expenditure as a percentage of revenues) usually tracks at between 15% and 20%, depending on the stage of the investment cycle, but is expected by Analysys Mason, a market research company, to fall several percentage points in the next few years. Meanwhile, Ericsson and Nokia look in worse health than their customers.

Stripping out the effect of a one-off impairment charge, Ericsson recorded an operating margin of 6.7% for the first half of 2024, down from the full-year margin of 13.7% it reported for 2021. On a like-for-like basis, its first-half sales dropped a tenth, compared with the same period of 2023. Nokia's operating margin of 11.5% for the first six months of the year, compared with the 12.5% full-year margin for 2021, looks better. But first-half revenues at its mobile unit dropped 32% year-over-year and its operating margin there was a troubling 3.6%. These are not the performance metrics of suppliers that overcharge. Each company has already shed thousands of employees and further cuts are expected.

Speculation that Nokia is trying to sell its mobile networks business group, and in talks with Samsung about a deal, have been repudiated by most experts. But without an improvement in the market, once-unthinkable remedies may be up for discussion. Further consolidation at this level, though, would leave Huawei-banning countries with only two realistic RAN choices: Ericsson or some Finnish and South Korean crossbreed, struggling to assimilate parts. For telcos that need a dependable supply of equipment, that would be a tricky situation indeed.