The financial model ran to forty tabs. It described a refurbished gas-fired plant, a network upgrade and a meter factory, and it listed six separate revenue streams, each with its own margin. Then a credit analyst at a development finance institution asked a question so simple that the room needed a moment to realise it had no answer. “Which account does the first naira land in, and who put it there?”
The model had a tariff, but no contract obliged anyone to pay it. It had customers, but no licensed seller entitled to bill them. Only a fraction of the headline price was the price of generating electricity; the rest was network, metering and vending charges and a tax pass-through, stacked onto one bill and counted as the project’s income. The fuel the plant would burn appeared nowhere on the cost side. Everyone had assumed that somewhere, someone would pay.
This is the gap beneath most conversations about financing Africa’s infrastructure. We argue about capital, risk, preparation and structure. Rarely do we ask the analyst’s question. Infrastructure finance is a claim on future cash: a lender advances money today in exchange for payments over twenty years. Guarantees, tranches and covenants exist to make that stream more certain. A project with immaculate preparation and elegant structure but an uncertain stream of payments is a well-designed building on sand.
Four things must hold. The price must cover the full cost of supply, including fuel, maintenance and the cost of capital. A specific party must have a legal obligation to pay. Cash must actually arrive, in full and on time, through a mechanism the lender can see. And the arrangement must survive a change of government. Fail any one, and the project has a revenue gap, whatever its model says. Most project models test the first and simply assume the other three.
The evidence suggests price is the first to fail. A World Bank study of utilities in Sub-Saharan Africa found that fewer than half covered even their operating costs, with several losing more than 25 US cents on every kilowatt-hour sold. A later World Bank review of 15 jurisdictions concluded that tariffs are rarely high enough to cover the full cost of service, even where that cost is low. And a 2025 World Bank working paper covering 67 utilities in 47 countries found that cost recovery from revenue worsened between 2012 and 2018, despite a 16 percent rise in tariffs.
Read also: The power of poetry: CCPC’s capturing times with words
Nigeria shows what this looks like in practice. The World Bank reported that distribution companies’ aggregate technical, commercial and collection losses averaged about 50 percent, against 15 percent regarded as international good practice and 26 percent allowed in the regulator’s tariff order. Between 2015 and 2019, the federal government paid a cumulative N1.68 trillion to cover the gap between the tariff and the cost of supply. The 2019 shortfall alone, N524 billion, exceeded the federal health budget of N428 billion that year. In 2024 the Minister of Power put the wholesale tariff shortfall at N1.6 trillion. The Bank estimates that unreliable power costs Nigeria some $25 billion a year in lost output.
The gap did not vanish; someone carried it. Generators went unpaid, gas suppliers withheld supply, and the minister himself attributed recent grid disruptions to unsettled debts across the industry. The 22 million gasoline generators that, according to the Bank, power about a quarter of Nigerian households and nearly a third of small businesses are what a failed revenue system looks like from the customer’s side. Nigeria has since begun raising tariffs for higher-consuming customers, which is a step in the right direction, and the data above should be read as a diagnosis of the problem rather than a verdict on the reform.
This is why the endless debate over sovereign guarantees often misses the point. A guarantee is a proxy. It stands in for a creditworthy payer with an enforceable obligation to pay a price that covers cost. Political risk insurance can protect an investor against expropriation, but it cannot turn an unpaid bill into cash. Nor can a guarantee manufacture a customer. A lender who accepts a guarantee in place of a payer has not solved the revenue problem; he has deferred it to the day the guarantee is called, which is the worst moment to discover that the guarantor’s own budget is stretched.
A tariff that is politically comfortable but below cost does not remove the cost; it relocates it to the party least able to refuse, usually the private supplier or the lender. The real question is not only whether customers can afford a cost-reflective tariff, which is a serious question deserving honest data and targeted support for poor households. It is who bears the difference if the tariff stays below cost, and under what contract. If nobody can answer, the project has an unbooked subsidy, and the lender is holding it.
Why does this persist? Because no single institution owns the chain. Regulators set prices, ministries sign obligations, distribution companies collect and banks lend, and each can say its own part is complete. Tariff reform is among the hardest political decisions any government makes, so leaders prefer to announce capital rather than adjust prices. And advisers, sponsors and lenders are rewarded at financial close. A fragile revenue assumption does not block it; it simply moves the failure to year three, when everyone responsible has moved on.
Honesty requires a limit, and naming it makes the case stronger, not weaker. Revenue architecture does not explain every failure. A plant whose fuel costs more than any customer can pay needs a different business, not a better collection account, and a community with almost no income may need grants rather than project finance. The argument applies where a viable economic activity exists and the system for turning it into dependable cash is missing.
Where it does apply, the responsibilities are clear. Regulators should treat tariff-setting as part of the financing system, with cost-reflective prices, transparent adjustment rules and any subsidy shown as a funded budget line rather than a hidden receivable. Governments should ensure that whoever signs a payment obligation has the legal authority and fiscal capacity to honour it; a memorandum promising to “facilitate” a power purchase agreement is not one. Development finance institutions should test the revenue chain before the capital structure, and use their instruments to build metering, escrow accounts and distribution-level credit enhancement, rather than to compensate for their absence. Investors should read a model that counts pass-through charges as profit, or omits fuel, as information about the project. Sponsors should resist presenting a stacked tariff as a business model.
The analyst’s question deserves to become routine: where does the first naira land, who put it there, and what happens when they stop? Capital exists and risk can be translated, but none of it holds without a payer who can be named, a price that covers cost, a payment that can be seen and an arrangement that outlasts an election cycle. Even then, one problem remains. The money collected arrives in naira, cedis or shillings, while the money lent must be repaid in dollars. That is a question for another day.
Nwafor is founder and lead strategist at De-Lazuli Consult, an advisory practice focused on energy transition, project finance, DFI engagement and infrastructure investment facilitation. He writes from Lagos and Abuja. [email protected]
Related News
