A Japanese leviathan retreats from the box-moving trade

NEC, a Japanese technology group that once built telephone exchanges for half the developed world, has agreed to sell its overseas network equipment business. The buyer is a consortium led by Assist Group S.r.l., an Italian firm with a name that invites the question of what it is assisting. The deal, announced on August 5th, covers NEC's resale of partners' telecom gear across Latin America, the Asia-Pacific and Europe, as well as parts of its IT services operations for foreign governments and municipalities. It is expected to close within fiscal year 2026.

The headline suggests a tidy portfolio edit. The reality is a wider strategic retreat. NEC is not merely trimming a distant limb; it is abandoning the hardware battlefield on which it spent years, billions of yen, and not a little pride.

Until recently NEC was the world's sixth-biggest vendor of radio access network (RAN) equipment - the towers and antennas that connect mobile phones to the internet. Sixth, to be clear, is the equivalent of finishing behind the world's best athletes in every major relay. According to Omdia, an analyst firm, NEC's global RAN market share in 2024 was 0.9%. Huawei, Ericsson and Nokia, the big three, accounted for 77.4% of all RAN sales, and their share had been growing. Samsung, NEC's nearest rival outside the leading trio, held 4.8%.

The incentive structure of telecom hardware is unforgiving. RAN equipment requires vast, sustained investment in research and development - an investment whose returns depend on scale that scale begets. Ericsson alone spent 53.5bn Swedish kronor (about $5.8bn) on R&D in 2024, roughly 22% of its sales. Not even Nokia could match that. NEC, by contrast, saw its 5G revenue guidance for fiscal 2026 slashed from ¥85.4bn ($540m) to ¥31bn ($200m) in November 2023, as open-RAN deployments - a telco-led attempt to unbundle proprietary network components - failed to materialise at expected speed. Global RAN product sales themselves dropped from about $45bn in 2022 to $35bn in 2024, as the initial 5G buildout stalled and brownfield operators clung to their existing vendors.

In February this year NEC terminated its conventional physical base-station business outright, recording a restructuring charge of ¥18bn ($120m). The remaining network infrastructure work will be folded into its aerospace and national-security unit. IT-related telecom services will be transferred to NEC's IT-services segment. The overseas resale operation now going to Assist Group is the last piece of the puzzle to be handed away.

The broader lesson is that telecom hardware has become a brutal game of industrial arithmetic. Three firms, backed by the R&D budgets of sovereign states, national champions or decades of compounding, command three-quarters of the market. A fourth (Samsung) is a conglomerate for which mobile networking is a side project rather than a core identity. NEC, for all its engineering pedigree, found itself in the unprofitable middle: too small to compete on R&D spend, too specialised to retreat without damage.

But NEC is not retreating into thin air. The divestiture is the flip side of an aggressive pivot to software, where margins are higher and scale is less binary than in hardware. In June NEC completed its acquisition of CSG Systems, an American provider of software for telecom operators' billing, customer care and payments. The deal is being folded into Netcracker Technology, NEC's own software subsidiary, to create what management describes as one of the industry's most complete digital platforms for telecom service providers. The logic is clear: rather than building boxes that operators buy once every few years, NEC wants to sell software that they pay for every month.

The numbers so far suggest the pivot is working. In fiscal 2026, which ended in March, NEC's revenue rose 9% to ¥3.58tn. Non-GAAP operating profit grew 27.6% to a record high. The stock now carries a market capitalisation of about ¥6.14tn. These figures reflect not only the telecom restructuring but also strength in digital government, biometric identification, enterprise IT in Japan, and financial-software operations inherited from earlier acquisitions.

To be sure, the new owners of NEC's overseas equipment business have not inherited a prize. Reselling partners' hardware is a low-margin activity in a contracting market. The countries where NEC still operated - much of Latin America, parts of southern and eastern Europe, and scattered markets in Asia - are precisely those where the big three have been consolidating their hold. Assist Group is a relatively obscure Italian entity, and the consortium around it is described only as being led by "business leaders who have extensive experience across a wide range of industries". That vagueness is not unusual for employee- or management-led buy-outs, but it raises the question of how much strategic direction these buyers can bring to a business whose fundamental challenge is structural, not managerial.

For NEC, the move is the right one. NEC's recent strategy materials frame the group as a trusted operator of social and digital infrastructure rather than a general electronics conglomerate. Biometrics, submarine cables, digital government platforms, and telecom software are businesses where domain expertise and trust matter more than manufacturing scale. They are also businesses where recurring revenue and gross margins of 33% or more - NEC's latest reported figure - provide a cushion that box-moving never could.

The danger is not that NEC will lose its way, but that it will repeat the mistake of believing software scale is easy once hardware scale is abandoned. The telecom-software market is itself competitive, and Netcracker, even with CSG added, faces larger rivals in Oracle, SAP and specialist players backed by deep pockets. NEC's ability to extract value from the combined platform depends on integration discipline, not aspiration.

The story of NEC's retreat is not a tale of Japanese decline. It is an example of a company recognising that the economics of an industry have changed and adapting accordingly. Hardware in telecom has become a scale monopoly. Software, for now, still rewards those who understand the customers who use it. Whether NEC's new owners of the old equipment business see the same structural headwinds, or merely a discount purchase, will be a test of their own sophistication. For NEC itself, the arithmetic is unambiguous: the boxes were never going to be enough.