Summary

  • Group net income reached KSh58.2bn, up 35.9% from KSh42.8bn, while M-PESA revenue grew 13.6% to KSh100.1bn: the profit jump did not come from faster mobile money.
  • Safaricom Ethiopia's EBITDA loss closed to KSh434.4m from KSh12.56bn on KSh10.6bn of service revenue, but its EBIT loss stayed at KSh11.4bn — the operating bleed has stopped; depreciation and financing on spent capital have not.

Two growth rates that do not match

Safaricom's headline for the half-year is a 35.9% increase in group net income, to KSh58.2bn from KSh42.8bn (TechTrendsKE, 5 November 2026). The companion number is less flattering. M-PESA, the product that carries the group's economics, grew revenue 13.6% to KSh100.1bn. A company whose profit rises nearly three times as fast as the revenue of its most profitable service is not suddenly growing faster. Something below the revenue line moved, and it moved in Ethiopia.

The KSh6.7bn that did the work

Begin with the line Safaricom's shareholders actually own: the group's share of the Ethiopian net loss. It fell to KSh8.8bn from KSh15.5bn, an improvement of KSh6.7bn (Tech-ish, 6 November 2026). Group net income rose KSh15.4bn over the same comparison. One subsidiary's shrinking loss therefore supplies roughly 43.5% of the increase in group profit.

Strip Ethiopia out and the rest of the group looks ordinary in the best sense. Add the Ethiopian share back to group net income in both periods and the remaining business earned about KSh67.0bn against KSh58.3bn, growth of roughly 14.9% — close to M-PESA's 13.6%. That is my arithmetic on two reported figures, not a company reconciliation; neither retained report publishes an ex-Ethiopia income statement. But the shape of the result is clear enough. A Kenyan business compounding at a mobile-money rate, plus a much smaller Ethiopian loss, produced a headline that reads like a transformation.

What closed, and what did not

Safaricom Ethiopia grew service revenue 71.6% in Kenyan shillings to KSh10.6bn, and narrowed its EBITDA loss 96.5% to KSh434.4m from KSh12.56bn. Read that against revenue and the Ethiopian operation now covers its operating costs almost exactly, with an EBITDA margin of roughly minus 4%. That is a genuine milestone, and it is the part of this story the company is entitled to call progress.

The EBIT line refuses to follow. Safaricom Ethiopia still recorded an EBIT loss of KSh11.4bn. Roughly KSh11.0bn therefore sits between the EBITDA result and the EBIT result in Ethiopia. The retained reports do not divide that gap into depreciation, amortisation, impairment or other adjustments, so the honest statement is narrow and consequential: Ethiopia is no longer an operating-cash shortfall on the reported base. It is a fixed-cost and financing burden sitting on capital that has already been spent.

The consolidation agrees. The Kenyan business generated EBIT of KSh101.8bn while group EBIT was KSh91.1bn, leaving about KSh10.7bn absorbed outside Kenya — consistent with the KSh11.4bn Ethiopian EBIT loss once rounding and the edges of the non-Kenyan perimeter are allowed for.

Three names need separating before any of this is read as one company's performance. Safaricom Limited is the Kenyan operating company and the listed group's home-market business. Safaricom Telecommunications Ethiopia PLC is the Ethiopian operating subsidiary whose EBITDA, EBIT and net result are reported separately. The consolidated group combines them, and every figure here belongs to one of those three layers, not to a single undifferentiated actor.

The evidentiary limit belongs beside the numbers it qualifies: both retained accounts are secondary reports of the half-year results. No primary results filing, presentation, cash-flow statement, subscriber disclosure, spectrum record or return-on-capital metric was read for this article, and neither report discloses Ethiopian subscriber numbers, Ethiopian local-currency revenue, or how the Ethiopian loss splits between depreciation, amortisation and interest.

Capital still has to earn its keep

First-half capital expenditure ran at roughly KSh35bn in Kenya and about KSh5bn in Ethiopia, against full-year guidance of KSh64bn to KSh70bn — KSh58bn to KSh61bn for Kenya and KSh6bn to KSh9bn for Ethiopia. Ethiopian capex alone equals roughly 47% of Ethiopian service revenue; subtract it from the EBITDA result and the Ethiopian business consumed about KSh5.4bn before working capital and tax in six months. A network can run that deficit while it is being built. It cannot call the asset finished while it does.

Guidance also implies a second-half group spend of roughly KSh24bn to KSh30bn, so the heavier half is still ahead. The question that decides whether this is a turning point or a plateau is whether the KSh11.0bn gap between Ethiopian EBITDA and Ethiopian EBIT starts to compress as capex decelerates and the asset base stops growing faster than revenue. That is a utilisation and depreciation question. Nothing in the retained record answers it.

So keep three judgements apart. Operating performance: resolved favourably, and materially so. Asset economics: open, because nobody has yet shown a return on the accumulated Ethiopian asset base. Financing and equity: open, because the reported net loss is still KSh8.8bn and the reports do not say what portion of it is interest rather than depreciation. Safaricom has bought itself the right to be judged on the second and third questions. It has not yet answered them. The company's own record in the directory is a membership listing, nothing more (Safaricom Limited in the BTW directory).