In results published at the start of September, Shoprite, Africa’s largest food retailer, recorded the end of its pan-African project. In the year to 28 June 2026 it completed the sale of seven stores and a warehouse in Ghana for R70 million, sold its five Malawian stores for R39 million, and agreed to dispose of its remaining property developments in Nigeria. Chief executive Pieter Engelbrecht described what is left as seven countries, “all situated relatively close to our South African home base.”
The numbers explain the retreat. South African supermarkets produced R228.7 billion of the group’s R270.8 billion in merchandise sales, or 84.5%. Supermarkets outside South Africa produced R22.8 billion, about 8%. A group that, at its peak, led food retail in roughly 15 African countries has left Nigeria, Kenya, Uganda, Madagascar, the Democratic Republic of Congo, Ghana and Malawi since 2020.
Shoprite is the clearest case of a pattern that runs through African corporate results. Companies cross borders, accumulate a long tail of small subsidiaries, then fall back to a core of one to three markets that generate most of the profit. This analysis argues that the barrier between a multi-country company and a continental institution is rarely market access. It is whether each foreign operation can fund, price and grow itself in its own currency, and most business models built in one African market cannot.
Footprints that outgrow their earnings
Dangote Cement shows the shape of the problem most precisely, because it reports Nigeria and the rest of Africa as separate segments. In 2025 it operated plants or terminals in ten countries outside Nigeria. Those ten countries carried 38% of the group’s volume but produced only 14% of its segment EBITDA.
| Dangote Cement, FY2025 | Nigeria | Ten other African markets | Other markets’ share |
|---|---|---|---|
| Volume sold (Mt) | 17.7 | 11.0 | 38% |
| Revenue (₦ billion) | 2,956.5 | 1,456.0 | 33% |
| Segment EBITDA (₦ billion) | 1,763.5 | 294.1 | 14% |
| EBITDA margin | 59.6% | 20.2% | — |
| Capital expenditure (₦ billion) | 729.8 | 131.3 | 15% |
Source: Dangote Cement FY2025 audited results, 28 February 2026. Shares exclude inter-company sales and central costs.
Non-Nigerian EBITDA fell 14.8% in 2025, which the company attributed to elections in Cameroon, Senegal and South Africa and to liquidity constraints in Ethiopia. In the first half of 2026, BusinessDay reported, the non-Nigerian segment’s profit fell to ₦2.8 billion from ₦139.9 billion a year earlier as finance costs absorbed an operating recovery.
Rather than replicate integrated plants market by market, it is shipping clinker from Nigeria: 34 vessels to Ghana and Cameroon in 2025, with a target of 10 million tonnes of exports by 2030. Management now describes Nigeria as a low-cost regional hub. That is a hub-and-spoke model, not a continental network of self-contained subsidiaries.
When Jumia closed its South African and Tunisian businesses in late 2024, the two markets accounted for 3.5% of orders and 4.5% of gross merchandise value. When MTN sold its units in Guinea-Bissau and Guinea-Conakry the same year, together they contributed less than 1.6% of group revenue; the Guinea-Bissau unit changed hands for a nominal $1, with the buyer taking on its net liabilities.
In each case the market was small relative to the group, yet it consumed a full set of fixed costs: a licence, a local board, a tax relationship, a treasury function, a regulator to manage. A company with five such subsidiaries does not have five growth options. It has five claims on scarce management attention and hard currency.
The currency line
The six frictions usually listed (capital, regulation, management, logistics, currency, fragmentation) are real, but they are not equal. The evidence points to one that governs the others: the mismatch between where a subsidiary earns and what it must pay for in hard currency.
Jumia’s fourth quarter of 2024 shows the effect cleanly. Reported revenue fell 23% year on year; in constant currency it fell 2%. Gross merchandise value fell 12% in dollars but rose 13% in constant currency. The business grew in the markets where it operated and shrank in the currency in which its investors measure it.
Shoprite’s results statement makes the same point in plainer language: economic factors outside the group’s control drive devaluations in its markets, whatever hedging it does. Coverage of the Ghana and Malawi decision listed the specific costs: currency volatility, double-digit inflation, import duties and dollar-denominated rents. A food retailer imports much of what it sells and often pays rent in dollars, while it collects cedis, kwacha or naira.
MTN has turned this into an explicit rule. Its risk framework, as the company described it when announcing the Guinea exits, judges a market by whether it can fund its own growth. A subsidiary that needs the parent to keep converting rand or dollars into a weakening local currency fails that test regardless of subscriber growth. Secondary reporting at the time said MTN kept Liberia off the exit list partly because the economy is dollarised.
Convertibility, not only the exchange rate
A stable peg is not the same as access to hard currency. The International Air Transport Association counted $774 million of airline revenue blocked from repatriation in African countries at the end of March 2026. Algeria held the most, at $258 million, but the Central African CFA franc zone, whose currency is pegged to the euro, held $105 million. For any company, trapped cash means the parent cannot recycle profit from a strong market into a weak one, which is the basic operation of a multi-country group.
Capital that is short, expensive and local
The cost of money varies so widely that a single group operates under several financial regimes at once. Dangote Cement’s own filing notes a benchmark rate of 21.5% in Ghana and 15% in Ethiopia, against about 5.25% in the West African CFA zone. Subsidiaries in high-rate markets struggle to borrow locally, so they borrow from the parent in dollars, and the currency mismatch reappears on the balance sheet.
Equity is scarce too. Africa’s exit markets run mostly through trade sales, not listings; the Africa Venture Capital Association found trade buyers took 38% of private equity exits in 2025 and 88% of venture exits in the first quarter of 2026. Capital that expects a sale within a fund’s life is poorly suited to the decade-long build-out a continental network needs.
Regulation, and where AfCFTA does and does not reach
The African Continental Free Trade Area now has 50 ratifications, the latest from Somalia on 4 September 2026. Yet only about two dozen countries had gazetted the tariff schedules needed to trade under its preferences, and rules of origin were agreed on 92.4% of tariff lines, with textiles and automotives outstanding. The agreement lowers tariffs; tariffs are one cost among several in the exits documented here, and rarely the decisive one.
The regulation that bites is national and operational. Shoprite’s sale in Malawi required approval from the Reserve Bank of Malawi as well as the competition authority. In one of MTN’s two Guinea markets, the regulator reportedly closed its local headquarters over unpaid taxes and fees months before the sale. Uber left Tanzania after years of disputes over fare controls and commission caps.
Logistics and management as multipliers
Poor logistics raises the working capital each subsidiary must hold and the hard currency it must spend on fuel and imported inputs. Dangote’s Congo plant reported that logistics constraints cut its export volumes in 2025, and Afreximbank’s 2026 trade report names infrastructure gaps among the main obstacles to regional trade.
Management capacity is harder to measure, and here our reading is an inference. Units worth 1% to 5% of group revenue rarely command a group executive committee’s time, yet each carries the full regulatory and currency risk of a large market. The exits above suggest that boards eventually price that attention as a cost.
Not an African management problem
A hostile reader might argue that African firms simply manage expansion badly. Foreign firms hit the same wall. Uber, the world’s largest ride-hailing company by revenue, now operates in four African countries (Egypt, Ghana, Kenya and South Africa) after leaving Côte d’Ivoire, Tanzania, Nigeria and Uganda. The emerging-markets economist Charlie Robertson told Semafor that the markets Uber kept all have income per head above about $2,000, pointing to middle-class size as the filter.
What crossing the threshold looks like
Some African companies have crossed. Standard Bank’s operations outside South Africa produced R10.4 billion of headline earnings in the first half of 2026, 40% of the group total, up from about 25% a decade earlier. The bank told analysts that those earnings have grown by an average of 14% a year since 2015.
The structural reason, in our reading, is that a bank subsidiary lends in the currency in which it takes deposits. Its revenue, its funding and much of its cost base sit in the same currency, and local regulators require it to hold local capital, which forces each unit to become self-funding. The parent’s currency exposure falls mainly on dividends sent home rather than on day-to-day operations.
Even so, the currency line shows up. Standard Bank’s Africa Regions earnings rose 7% in rand and 11% in constant currency in the half. Crossing the threshold does not remove currency risk; it moves the risk from the operating model to the translation of profits.
Three models recur among the companies that hold multi-country positions:
- The local balance sheet. Banks, and telecoms operators whose subsidiaries fund network investment from local cash flow, earn and spend in the same currency. MTN’s self-funding test is an attempt to impose this discipline.
- The export hub. Dangote Cement’s clinker shipments concentrate capital in the lowest-cost plant and sell across borders, rather than replicating plants in every market.
- The near-home cluster. Four of Shoprite’s seven remaining markets (Namibia, Botswana, Lesotho and Eswatini) sit in the Southern African Customs Union with South Africa, and three of them peg their currencies to the rand through the Common Monetary Area. The retreat has been towards the rand’s orbit.
The models that struggle are those in which the parent must keep supplying hard currency: importers paying for stock in dollars, businesses paying dollar rents, and platforms whose unit economics depend on imported capital and a middle class that devaluation shrinks.
Who gains from the retreat
Retrenchment redistributes assets, and the winners are identifiable. Local and regional buyers acquire operating businesses at low prices: Shoprite’s Malawian stores went to Karson Investment Trust for R39 million, its Nigerian retail business earlier went to Ketron Investment, and MTN’s two Guinea units passed to Telecel and to the Guinean state. For buyers who already earn in the local currency, the mismatch that defeated the seller largely disappears.
Parent shareholders gain in the short run. Shoprite’s cash generated from operations rose 26% to R27.6 billion in the year, and cash and equivalents rose 60% to R13.0 billion, as capital moved to its home market. Dangote Cement’s capital spending went 85% to Nigeria, where its margin is three times that of its other markets.
The costs fall elsewhere. Smaller economies such as Malawi, Guinea-Bissau and Tunisia lose the investors most able to bring supply chains, formal jobs and competition, and become markets served by one fewer credible operator. Workers carry transition risk; the Malawi sale carried conditions that staff who wished to stay be retained and those who left be compensated.
The larger loss is to the integration project itself. A single continental market needs companies that operate across it. If the firms best placed to build continental networks conclude that the rational footprint is one home market plus its currency neighbours, AfCFTA will lower tariffs for trade that few companies are organised to carry.
What would move the line
Three paths are plausible over the next five years, and they are not mutually exclusive.
- Currency clusters (our base case). Groups organise around a home market and its monetary neighbours: the rand’s Common Monetary Area, the two CFA franc zones, or dollarised economies. Footprints shrink in country count while profits concentrate.
- Hub exporters. Firms with a low-cost plant in one market sell across borders under AfCFTA preferences. Dangote’s target of 10 million tonnes of exports by 2030, against 1.4 million tonnes in 2025, is the test case.
- Settlement reform. If local-currency settlement systems such as the Pan-African Payment and Settlement System gain real volume, and governments honour repatriation commitments, the cost of running subsidiaries in weak-currency markets falls and the case for replication returns.
Five indicators will show which path is winning: IATA’s quarterly count of blocked funds in Africa ($774 million at end-March 2026); the number of AfCFTA states trading under gazetted tariff schedules; Dangote Cement’s export tonnage; the gap between reported and constant-currency growth in Standard Bank’s Africa Regions; and whether Shoprite’s non-South African supermarkets, about 8% of sales, grow or shrink as a share.
Each group of decision-makers faces a sharper question than the usual one about market size:
- Founders: before entering a third market, can that subsidiary fund its own growth in local currency within a defined period, and which of its costs are priced in dollars?
- Investors: is the business underwritten on constant-currency growth, on reported dollar growth, or on cash that can actually be repatriated? These are three different companies.
- Policymakers: for the firms that would build continental networks, a reliable right to move profits home is likely worth more than the next round of tariff cuts.
The threshold is financial
A continental institution is less a company present in many countries than one in which no subsidiary needs to ask head office for dollars. Standard Bank has built that; Dangote Cement is trying to route around it through exports; Shoprite has decided the answer lies within reach of the rand. AfCFTA is steadily taking tariffs off the table. Whether Africa produces its next generation of continental companies depends on whether governments can make the movement of money as routine as the movement of goods is becoming.
