Contents
The IMF’s latest reading of the region’s economy, published on 15 October, leads with growth of 4.5% in 2026 and again in 2027: steady, slightly better than April’s forecast, and of almost no use to anyone running a business.
Seven pages in, there is something far more useful — a table of how much of this year’s rise in international fuel prices actually reached the pump in 21 African markets. The range runs from −36% to 133%: one shock, 21 different cost curves. If you move goods, run a generator or buy freight in these markets, that table maps your cost exposure, and it says something uncomfortable about the markets where fuel still looks cheap.
What is happening?
On 15 October 2026 the IMF’s African Department published its Regional Economic Outlook Update: Sub-Saharan Africa, subtitled Resilient Growth, Clouded Outlook. Regional output is projected to grow 4.5% in both 2026 and 2027, 0.2 percentage points better than April’s forecast for this year, while median inflation rises from 3.4% in 2025 to 4.3% in 2026 and 5.0% in 2027 (IMF, 15 October 2026).
The pressure comes from the Middle East war. The October 2026 World Economic Outlook, which the Update builds on, assumes a gradual reopening of the Strait of Hormuz, with oil averaging $96 a barrel in 2026 and easing to $87 in 2027. Diesel markets stay tight after losses of refining capacity, and fertiliser prices are projected to rise 26% in 2026 on 2025.
Then the part specific to each market. Using retail prices, national sources and US Gulf Coast spot benchmarks, IMF staff estimated pass-through for 21 countries between 23 February and 14 September 2026 — the retail price change as a percentage of the change in the matching benchmark. A selection:
| Market | Petrol | Diesel |
|---|---|---|
| Rwanda | 125% | 111% |
| South Africa | 77% | 106% |
| Kenya | 52% | 62% |
| Nigeria | 46% | 60% |
| Mozambique | 32% | 92% |
| Côte d’Ivoire | 23% | 10% |
| Cameroon | −6% | −5% |
| Zambia | −36% | 12% |
Across the full 21, diesel pass-through runs from −5% in Cameroon to 133% in Sierra Leone, and the two products diverge inside a single market: Lesotho passed through 82% on petrol and 116% on diesel.
Three limits. These are preliminary estimates over a single window. Pass-through is not a subsidy rate — the IMF says so explicitly; a low figure may reflect taxes cut, a currency that appreciated, or a price simply held. And the window closed on 14 September: Kenya’s regulator has since raised maximum retail prices for the cycle from 15 October, diesel in Nairobi to KSh 227.76 a litre, up 4.55% (EPRA, 14 October 2026).
What stays uncertain is larger than what is settled: whether Gulf energy flows normalise as assumed, whether the governments holding prices down can keep doing so, and how hard El Niño lands. Risks, the IMF says, are “decidedly tilted to the downside”.
Why it matters economically
Diesel is not a line item. It sits inside almost every delivered price: freight rates, fleet running costs, generator hours, cold chain, irrigation pumps. A business rarely sees “fuel” move; it sees a haulier’s rate rise, or a margin thin out.
The first implication is that a low pass-through figure is not a cost advantage but a timing difference. The IMF is blunt about the mechanics: where relief is not fully funded, losses at fuel importers and state-owned utilities turn into arrears, shortages or rationing. In Comoros, which suspended a planned increase after protests, holding prices through 2026 could cost about 2.5% of GDP in quasi-fiscal losses at the state oil company. The money is spent either way; what varies is whether it reaches your cost base through the pump, a shortage, or later through tax.
Not all low pass-through is suppression. Zambia’s negative petrol figure comes with higher copper receipts, a stronger maize harvest, fuel-tax cuts and an appreciating kwacha — a better position, not a deferred bill.
The second implication is the cost of carrying any of it. Interest payments absorb a median of about 12% of government revenue excluding grants across the 2025–27 estimates, banks in the region hold more sovereign debt as a share of assets than in any other region measured, crowding out private credit, and the IMF argues most countries now warrant a bias towards monetary tightening. Working capital is scarcer and dearer exactly as input costs rise.
What it means for African businesses
The exposure is not continental, nor even uniform by country. It is per market and per product.
Where pass-through has been high or above full — Rwanda, Sierra Leone, Cabo Verde, South Africa on diesel — the cost has landed, and the question is whether your prices moved with it. Where it has been low or negative — Cameroon, Benin, Madagascar, Côte d’Ivoire, Zambia on petrol — the risk is a step change you cannot time, plus supply interruption if importer losses accumulate. Plan there for a discontinuity, not for cheap fuel.
Weather compounds it in the south. The Update expects El Niño to persist through February 2027, with below-normal rainfall favoured from October to March across much of the region, including parts of Malawi, Mozambique, Zambia and Zimbabwe, and lower reservoir levels constraining hydropower. In the 2023/24 drought Zambia’s maize harvest fell by about half, and falling reservoir levels contributed to prolonged power cuts. Drought and fuel inflation then arrive as one bill: fewer grid hours, more generator hours, at a higher price per litre.
Retail and distribution, logistics and trade, agro-processing, hospitality, and clinics and schools on generators hold this exposure whether or not they measure it. Businesses whose costs are people and software, on reliable or solar power, largely do not. Which market you are in changes the shape of it, which is why we treat market conditions as design constraints rather than background.
Technology and AI implications
Start with what technology cannot do. No system changes the price of diesel, the rainfall forecast, or whether foreign currency is available this week. Hedging a fuel bill is a commercial instrument — a forward, a fuel-indexed clause — not a software feature. Forecasting fuel prices with AI is not a serious proposition either: the IMF’s own baseline is conditional on an assumption about a strait reopening, and no model trained on past prices anticipates that. We would not build it. The IMF also notes that most economies here, unlike technology exporters in Asia, receive little direct support from the AI investment boom: AI is a cost-side tool in this market, not a demand-side windfall.
What technology does change is the lag between a cost moving and a price moving — a data problem before it is a software problem. It needs three things: fuel, freight and energy costs recorded against whatever consumed them — a route, a vehicle, a site, an order — and not only against a bank statement; price lists and quotations in which an input cost is a variable to be refreshed rather than a number typed once; and generator hours logged somewhere other than a notebook. Reading fields off supplier invoices and fuel receipts is a fair use of AI: it gets cost data into a system that can be queried, and the constraint is the process, not the model. That difference between adopting a tool and integrating it into decisions is the gap the World Bank measured in African firms this month. Across several markets, keep the data separated by market and currency; an average hides the divergence that matters.
Practical business implications
Four questions worth answering this month.
- Per market: has the pump price already absorbed the import cost move? Read the IMF’s figures, then your own fuel invoices over six months.
- What is fixed, and for how long? Quote validity, supply contracts, price lists, flat delivery fees. Where a regulated diesel price moved 4.55% in one cycle, a 30-day quote validity is a pricing decision, not an administrative default — and the contracts you issue rarely carry the input-cost clause your hauliers put in theirs.
- What does a 10% diesel increase do to gross margin, by product and route? If answering takes a week, that is the gap to close.
- What do public-sector customers owe you, and for how long? The IMF names supplier arrears as a channel carrying fiscal stress into corporate cash flow.
An illustrative example, on stated assumptions and not a measurement. Take a distributor running twelve routes on 40,000 litres of diesel a month. At about KSh 227.76 a litre that is roughly KSh 9.1m of fuel, so a 10% move adds about KSh 911,000 a month. On revenue of KSh 60m at a 6% net margin, that is about a quarter of net profit. Fuel is usually a larger share of profit than of revenue, which is why it is under-managed.
None of this is a reason to replace your systems. It is a reason to answer question 3 inside a day, which usually means joining two sets of records you already keep — and if they sit in spreadsheets that no longer reconcile, that is a more familiar problem.
RevenueStack perspective
The least useful number in the Update is 4.5%. The useful content is the dispersion around it.
Our reading is that the report describes two balance sheets absorbing one shock. Where the state passed the price through, the business carries the cost now and can price for it. Where the state held it down, the state borrowed the cost — and it returns as a step increase, a shortage, or higher taxes. Either way it lands in the business’s numbers; only the timing and the visibility differ. Treating a low pass-through market as a cheap one is reading a deferral as a discount.
What follows has little to do with forecasting. When input costs move sharply and often, the capability that pays for itself is the time between a cost change and a price change. Most businesses cannot measure that lag: cost data sits in bank statements, fuel cards and invoices, price data in quotes and spreadsheets, and nothing joins them. Closing it from a month to a week is worth more than any prediction, and the work is ordinary — decide which costs attach to which unit of work, record them there, and make the price a calculation rather than a memory. That is how we approach any operating constraint: the economics first, then only what the decision requires.
Conclusion
The Update’s headline is stability; its content is divergence. Regional growth of 4.5% tells an individual business nothing. The pass-through table tells it a great deal: the cost of moving goods has risen by very different amounts in markets that look similar from outside, and where it has not risen yet, the adjustment is still owed. The response is not prediction but visibility and speed — knowing what fuel and energy cost per route, per site and per order in each market, and changing a price in days when that number moves.
Sources
- Regional Economic Outlook Update: Sub-Saharan Africa, October 2026 — Resilient Growth, Clouded Outlook — International Monetary Fund, African Department, 2026-10-15
- Resilient Growth, Clouded Outlook — Regional Economic Outlook Update, Sub-Saharan Africa (full text, figures and endnotes) — International Monetary Fund, African Department, 2026-10-15
- Maximum Retail Petroleum Prices in Kenya for the Period 15th October 2026 To 14th November 2026 — Energy and Petroleum Regulatory Authority (Kenya), 2026-10-14