Cell C Claims the Widest 5G With Zero Towers. That's Not a Telecom Story.

Cell C says it now has the widest 5G coverage in South Africa. It owns zero towers. Both halves of that sentence are true, and neither is the story.

The story is that a telecom just did the thing most of crypto's infrastructure thesis assumes but rarely has to prove at scale: rent someone else's base layer instead of building your own. Telecom is the hardest possible venue for that bet. It is the industry that turned "you must own the physical plant" into a business-school axiom, where the two biggest operators each pour hundreds of millions of dollars a year into their networks. If the rent-don't-build model holds up here, every rollup renting security from a base chain and every startup reselling a hyperscaler's compute gets its strongest validation yet. If it cracks, they should take that failure personally.

Here is what actually changed. In July 2023, Cell C switched off the towers and radio access network it owned — the antennas and masts that physically carry phone signals — and moved its prepaid and virtual-operator traffic onto competitor MTN's network through a virtualized arrangement. It has kept its spectrum licenses and leases MTN's physical layer to run them on. It went from roughly 5,500 sites it owned to roaming across more than 28,000 rented sites on the networks of MTN and Vodacom. A former CTO described the model in exactly those terms: Cell C is no longer a network owner, but a "buyer of network services," one that avoids the annual capex burden its larger rivals carry. The 5G coverage claim announced this week rides entirely on that arrangement — what is now officially a highly leveraged roaming play.

That is the same structural bet as a Layer 2: keep the customer-facing logic, rent the security and settlement underneath, and let someone else's sunk capital be the floor. The difference is that an L2's landlord is a protocol with no pricing power over its tenants. Cell C's landlord is a competitor that sets wholesale prices.

The reason this matters goes beyond the coverage map. It is not about who has the widest network. It is about who captures the money flow once the network is a commodity. Cell C's real business is no longer selling connectivity; it is selling access to connectivity and renting out its customer-facing layer to other brands. It calls itself the "home of MVNOs" — mobile virtual network operators, brands like a bank or a supermarket that sell phone service without owning a network. Capitec Connect, FNB Connect, and Shoprite's K'nect Mobile all run on Cell C's system. The numbers make the shift legible: in its latest full year Cell C counted 8.88 million direct subscribers and another 5.7 million MVNO subscribers living on its platform; in its first reported half as a listed company, wholesale revenue rose 22.5 percent to R840 million and it posted a 16.1 percent normalized EBITDA margin, a rough proxy for operating cash profit.

This is the platform-versus-application pattern playing out in reverse of the usual plot. Amazon built AWS for its own e-commerce first, then externalized. Cell C dropped the AWS part and kept the application and the billing layer — and it is finding that the customer record is the moat. Every phone number on Capitec Connect lives in Cell C's database, giving it the durable relationship while Capitec does the marketing and pays for access. Capitec Connect alone has signed up 1.5 million subscribers, one of the fastest mobile launches South Africa has seen, launched without a single new tower. A retail bank built a phone business on rented rails, and the renter ended up owning the flow.

Now the counter-frame, because this is where the thesis usually overreaches. Rented capacity is a lease, not a moat. MTN controls the physical network, the wholesale price, and the upgrade roadmap, and it can reprice at every renewal. Cell C's own quality story is really a story about renegotiations — the quality gap closed only after it reworked its roaming terms. The landlord can also move into its tenant's business: MTN has been adding hosted MVNOs of its own, and Vodacom entered the space too. Cell C's "widest coverage" claim is also exactly the kind of thing this writer distrusts: a claim about someone else's asset that the claimant alone cannot verify. A South African advertising regulator already sided with MTN's complaint that an earlier "best network" campaign was misleading. Independent measurement is blunter: OpenSignal's most recent South Africa report hands MTN, not Cell C, the 5G experience and 5G availability awards, which is a reminder that "widest coverage" (a lease of footprint) and "best experience" (a measured outcome) are not the same sentence.

The balance sheet adds the final hedge. This model was adopted as survival architecture before it became strategy. As recently as December 2022 Cell C was technically insolvent, with negative equity of R9.294 billion — its liabilities exceeded its assets after years of losses and a court-supervised restructuring. The headline numbers it reported this week, earnings up 57.4 percent with prepaid revenue strong, are a genuine turnaround, but a turnaround proves follow-through, not permanence.

Where does that leave the investor verdict? We would place this not on the speculation curve but on the adoption one, with an asterisk. The structural part — the MVNO hosting engine compounding on rented rails — is real and durable, and it is showing up in earnings, not just in press releases. The froth is the stock market wrapper: Blue Label Telecoms, which still holds a majority of Cell C through a subsidiary, saw its shares roughly quadruple into the JSE listing people spent a year anticipating. Crypto markets show the same split right now: fear and greed run hot at 72 and total market cap stands near $2.6 trillion, yet Bitcoin dominance sits around 60 percent — capital still pays its biggest premium to the base-layer incumbent, not to the layers renting its rails.

That parallel is the takeaway. Owning the base layer is no longer automatically a moat: note that Telkom, an owner-operator still running its own network, trails badly in independent grading, finishing a distant fourth on real-user tests while the zero-tower renter ties with the best landlord in the country. But renting is only a strategy when you own the margin-bearing layer on top — the customer record, the payment flow, the retail relationship — because the landlord can reprice at any time. Cell C wins if the MVNO host position keeps compounding. The 5G crown is a marketing artifact of a lease; the earnings under the lease are the proof. That is the whole game.