Business
Know the Business
MTN is a pan-African oligopoly play wrapped around a low-cost prepaid SIM, where ~85% of group EBITDA comes from four markets (Nigeria, Ghana, South Africa, Uganda), a fintech rail (MoMo) sits on top of the same SIM at near-zero marginal cost, and reported earnings whip around with the naira and rand rather than with subscribers. The right way to underwrite it is a sum of three things: an FX-noisy connectivity engine, a fintech option that is real but unfinished, and a self-funded tower/data-centre buildout (the IHS deal). The market mostly judges MTN on the FX-noisy headline number, which is exactly why the constant-currency picture and the per-OpCo cash upstream are where edge sits.
1. How This Business Actually Works
MTN sells three products on one SIM card, in 16 mostly oligopolistic African markets where the regulator only licenses two-to-four operators. The economics start with a heavy fixed-cost radio network: once the tower, spectrum and core are paid for, every extra subscriber and every extra gigabyte drops a high incremental margin straight to EBITDA. That is why the 17% capex-intensity line and the 44% EBITDA-margin line both look so stable in constant currency even as the headline P&L lurches around with the local currency.
Where the incremental rand of profit actually comes from. Voice and data revenue cover the network's fixed cost; everything above ~30% subscriber-utilisation is operating leverage. Fintech is the second leverage point — once a customer is on the SIM, signing them up to MoMo costs almost nothing and the agent network is already paid for by airtime distribution. The third is the tower-and-fibre layer: by buying IHS Holdings' African business for $6.2B (signed Feb-26, ~75% MTN didn't already own), MTN reverses 15 years of tower-margin leakage to landlords like IHS and recaptures roughly $0.5-0.7B of annual EBITDA that used to leave the group as lease payments.
Bargaining power is split three ways and MTN is on the right side of two of them. Against customers: in markets like Ghana, Nigeria and Uganda there is no real alternative for a rural prepaid user — MTN is either #1 dominant or co-leader. Against suppliers: scale on procurement and the IHS reinternalisation pull pricing power onto MTN's side of the tower lease. Against regulators: this is where MTN is weakest — Nigeria delayed a tariff hike for 12 years, Ghana's regulator put MTN under Significant Market Power remedies in 2020, and South Africa's ICASA continues to police prepaid pricing. That regulatory exposure is what keeps EBITDA margins in MTN's mature markets capped at ~37% rather than the 50%+ that uncontested operators like Maroc Telecom and Safaricom achieve.
The single thing newcomers misread: MTN's FY2024 loss of -R10.9B looked like a broken business — it wasn't. Service revenue in local currency grew double-digits in Nigeria and Ghana; the loss was almost entirely an R18.9B FX translation hit from the naira devaluation. In FY2025, the same engine produced R27.4B of net income because the naira stopped falling. Always read constant-currency service revenue and per-OpCo EBITDA first; the reported group line is the output of the engine plus the FX dial.
2. The Playing Field
MTN's listed peer set is short — there are only five other public operators whose economics genuinely substitute for MTN's: two South African JSE rivals (Vodacom, Telkom), one pan-African pure-play in USD (Airtel Africa), and two European-listed groups with material African exposure (Orange, Maroc Telecom). Compared with those five, MTN is the scale leader on subscribers, mid-pack on capex intensity, top-quartile on EBITDA margin among the multi-market operators, and currently trading at the cheapest EV/EBITDA in the group — which is the market's way of pricing in Nigerian FX risk and the IHS payment.
Latest reported full-year for each: MTN, Orange and Maroc Telecom on Dec-25 (Maroc Tel on Dec-24); Vodacom, Airtel Africa and Telkom on Mar-25. Multiples computed on closing prices 19-May-2026 with net debt at the most recent reporting date. MTN EV includes net interest-bearing debt but excludes lease liabilities for like-for-like comparability with peers.
What the peer set reveals. Maroc Telecom anchors the top-right "high quality" corner — a single-country dominant operator with 50%+ EBITDA margins, where the market pays a 5.5x EBITDA multiple for predictability. Airtel Africa, MTN's closest economic substitute, trades at a 30% premium (6.3x vs MTN's 4.9x) despite a similar EBITDA margin and a worse capex intensity, mostly because Airtel reports in USD and the market hates translating MTN's naira earnings into rand. Vodacom is the cleanest like-for-like since both are JSE-listed and have heavy SA exposure; the fact that Vodacom trades roughly in line with MTN despite a 8pp lower EBITDA margin tells you MTN's discount is about perceived FX risk, not operational quality. Telkom SA is the cautionary outlier — what an African telco looks like when it does not have scale: 22% EBITDA margin and a 3.9x multiple, because the SA market punishes sub-scale players brutally.
The best peer does two things MTN does not yet do. Maroc Telecom prints 51.6% EBITDA margin in a regulated single-player market (Morocco), and Airtel Africa reports in USD. MTN's Ghana operation (60% EBITDA margin) shows the Maroc-style economics are possible when MTN is dominant. The Airtel-style reporting fix is structural — MTN's investor case would re-rate materially if the parent ever moved primary reporting to USD or if the Nigeria float (MTN Nigeria is separately listed on NGX) was used to ring-fence FX optics.
3. Is This Business Cyclical?
Wireless telecom is not classically cyclical, but MTN runs three overlapping rhythms — and the one that drives reported earnings is FX, not subscribers.
Read the chart this way. The blue EBITDA-margin line stays in a narrow 31-45% band even through the worst FX shock in MTN's history — that is what underlying operating cyclicality looks like for this business: barely any. The darker net-margin line whipsaws from +9.5% in FY2020 to -5.0% in FY2024 back to +8.9% in FY2025 — that swing is almost entirely the FX-and-hyperinflation cycle flowing through finance costs and the IAS 29 net-monetary-gain line. MTN's FY2024 reported revenue actually fell 14.9% in ZAR (R221B → R188B) while constant-currency service revenue grew about 14%; the gap was pure translation.
Where the cycle hits next, in order of speed: (i) FX line (immediate, the moment a peg breaks or a currency revalues), (ii) ARPU and pricing (one to two quarters after a regulator approves a tariff change — Nigeria's Jan-25 hike added ~32% to ARPU within 2Q), (iii) capex intensity (a year or two when a spectrum auction or 5G push hits), (iv) subscriber growth (rarely declines — even Covid did not dent it). What MTN's history does not show is a demand cycle — there has been no year since 2015 when group subscribers actually fell.
4. The Metrics That Actually Matter
For African telco analysis, the standard ratios (P/E, EV/EBITDA, ROE) are noisy on a one-year basis because FX rewrites them. The five metrics below cut through that noise and predict reported earnings 1-3 quarters earlier than the income statement.
Why these five and not P/E or net income. A P/E on reported earnings would have said MTN was uninvestable in FY2024 (loss) and a generational bargain at 5x in FY2025 — neither was right. The five operating metrics moved much less: CC service revenue growth stayed in a 13-23% band, EBITDA margin (CC) in a 39-45% band, capex intensity in a 13-17% band. Those bands are the actual business; the rest is FX optics. The forward signal in the metric set is MoMo: MAU has stayed flat at 63-72m for four years even as MTN added 70m+ connectivity customers, which says either fintech penetration has hit a near-term ceiling or competitive pricing pressure is biting — both are decision-useful before they appear in reported revenue.
5. What Is This Business Worth?
Value here is mostly determined by cash that actually reaches the HoldCo from a small number of African OpCos, capitalised at a multiple that compensates for naira/cedi FX risk and political risk. Reported earnings are too FX-noisy to anchor on, so the right lens is constant-currency EBITDA times a peer multiple, sanity-checked by an OpCo-by-OpCo sum-of-the-parts because the four big OpCos have materially different growth, margin and risk profiles.
Why a sum-of-the-parts is genuinely useful here, not vanity. Three things make the consolidated EBITDA multiple misleading: (i) the four big OpCos have very different risk profiles — Ghana deserves a Maroc Telecom multiple (~5.5x), SA deserves a Vodacom multiple (~5x), Nigeria deserves a discount for naira risk but a premium for tariff-hike runway, and the long tail of WECA/SEA markets a deeper discount; (ii) MTN Nigeria is separately listed on the NGX, which gives a real market reference for the largest piece; (iii) MoMo has private-round valuations (~$85/active user from past raises, per Afridigest) and an in-progress structural separation in three markets, which means a chunk of value is currently bundled into a connectivity EBITDA multiple that does not reflect fintech economics.
This is a directional SOTP, not a price target. The point is to show that the current market cap (~$22.8B) sits in the middle of a reasonable parts-based range — meaning the market is paying for a normal HoldCo discount but not penalising MTN for FX or political risk beyond peer-set norms. The stock becomes obviously cheap only if (a) the IHS deal closes cleanly, (b) Nigeria tariff durability is proven over 4+ quarters, or (c) MoMo separation crystallises a visible fintech multiple. It becomes obviously expensive on naira re-devaluation or a Ghana SMP price-floor cut.
The single sentence to internalise: MTN is worth roughly what the next four to eight quarters of constant-currency EBITDA say it is worth, capitalised at a 4-6x multiple that reflects African oligopoly economics minus naira-translation risk, with a real but discounted option on fintech separation and tower-margin recapture. Everything else is noise around that.
6. What I'd Tell a Young Analyst
Don't trust the reported income statement on a one-year view. ZAR translation moves it more than the operating business does. Build your model in constant currency at the OpCo level — Nigeria in naira, Ghana in cedi, SA in rand — and translate at the end. The four OpCos that matter (Nigeria, SA, Ghana, Uganda) drive 85%+ of group EBITDA; the rest is geographic diversification, not earnings.
Watch four signals in this order: (i) Nigeria CBN/naira spot rate — it explains more of MTN's reported volatility than any operating decision; (ii) MTN Nigeria SENS filings on the NGX — they post quarterly service revenue and ARPU in naira a few weeks ahead of group, and they are the cleanest read on whether the Jan-25 tariff hike is sticking; (iii) MoMo MAU and advanced-services share — the fintech option only crystallises if MAU growth re-accelerates and advanced-services share keeps creeping toward 50%; (iv) IHS deal closing milestones at FCCPC and SEC — the deal economics are the largest discrete value lever on the next 18 months.
What the market is most likely getting wrong, both ways: it under-prices MTN's MoMo and tower optionality because they are buried in a consolidated 5x EBITDA multiple; it over-prices the Nigeria tariff hike because tariff approvals in Africa often get politically reversed before they fully flow through ARPU. The thesis genuinely changes if (a) MTN moves primary reporting to USD or executes a clean MoMo separation, (b) the IHS deal closes on terms close to the announced $6.2B, or (c) Nigeria does a hard FX repeg or rolls back the 2025 tariff. Watch those, not the consensus EPS line.
One unforced error to avoid. Do not anchor on the FY2024 reported loss as a base for valuation or growth — it was a translation artefact, not an operating year. Anchoring there will either make MTN look perpetually broken (it is not) or make FY2025's 1,113c EPS look unrepeatable (the operating engine that produced it has not changed; only the FX dial moved). Use a 3-year average of constant-currency EBITDA, not a single reported year, for any normalised analysis.