In 2025, the largest single foreign direct investment event on the African continent was a subtraction. Anglo American’s $7.2 billion spin-off of its 66.7 per cent stake in Valterra Platinum pushed South Africa to negative inflows of $6 billion, a British miner unwinding a century old position in South African platinum, recorded in the national accounts as capital leaving. In the same year, the French group Canal+ acquired MultiChoice, and inflows ran the other way.
Both facts sit inside the same headline. FDI inflows to Africa reached about $70 billion in 2025, below the exceptional $94 billion recorded in 2024, but still the third highest level since 1990 and roughly one third above the continent’s long term average. Greenfield project values fell by almost one third while the number of announced projects increased, and UNCTAD noted that investors from the Gulf and other Asian economies are becoming important sources of greenfield investment, especially in energy, logistics, real estate and infrastructure.
That observation is accurate and it is being read wrongly almost everywhere. The question being asked is whose money is arriving. The question that matters is what those investors are being given in return, and on that, the change is far larger, far more durable, and almost entirely absent from the coverage.
The controlling argument of this piece: the nationality of capital entering Africa is changing less than the instrument is. Across Cairo, Luanda, Dar es Salaam and Nairobi, African states are converting debt claims and public assets into concessions of twenty or thirty years. That shift, not the passport of the counterparty, redistributes money and power. And the largest pool of capital that could plausibly have bid, African institutional savings now past $2 trillion, is lending to the seller instead.
What the greenfield number can and cannot tell you
Start with the evidence, because the standard framing rests on a figure that does not mean what it is asked to mean. The claim that the United Arab Emirates has become Africa’s leading new investor traces to a single dataset. Between 2019 and 2023, UAE companies announced $110 billion in projects across Africa, including $72 billion targeted at renewable energy, according to FT Locations, a data firm owned by the Financial Times. That is an announcement series, not a deployment series. It represents an announced pipeline of project commitments, not capital that has already entered African economies, and some projects have experienced delays or have not yet moved to active development.
The gap is not theoretical. Mauritania’s $34 billion green hydrogen project, announced in 2023, remains at a preliminary stage, and a $2 billion solar initiative in Zambia has been delayed because of financial struggles at the state owned utility ZESCO. Nor is commitment necessarily persistent. AD Ports announced more than $800 million in planned investments in Egypt, the Republic of Congo, Tanzania and Angola in the three years to April 2025, yet its 2025 annual report, published on 30 March 2026, confirms a focus on upgrading terminals in the UAE, Safaga, Karachi and Latakia.
Scale matters too, and it is rarely supplied. Africa recorded 639 greenfield project announcements in 2025, a 5 per cent rise, helped by Egypt and Côte d’Ivoire. In the same year, drawing on the same underlying fDi Markets data, the UAE alone recorded 1,491 projects, ranking second worldwide by number of announced projects. One country of roughly eleven million people attracted more than twice as many new project announcements as fifty four African countries combined. Whatever is happening, a boom is not it.
Treat the first print of any African FDI figure as provisional. UNCTAD’s January 2026 estimate put Egypt’s 2025 inflows at $11 billion; by July the World Investment Report had revised that to about $15 billion. A revision of that size on the continent’s largest recipient, inside six months, is a fact about the data, not about Egypt.
Finally, flows are not ownership. By stock, meaning the accumulated claim on African productive assets, the picture is almost unchanged. A large share of FDI to Africa comes from or is routed through the Netherlands, the top investor home country by stock, followed by France, the United States, the United Kingdom and China. The Netherlands leading that table should be read for what it is, a statement about tax treaties rather than about Dutch industrialists. We do not in fact know with precision who owns African assets, because the statistics record the last legal jurisdiction rather than the ultimate owner. Set against that, FDI stock from within Africa amounted to $128 billion, larger than any single external source country and almost never discussed.
Three windows closed, and the price of assets fell
The causation here reaches back about a decade, and no further. Three sources of external finance for African states contracted in sequence, and the sequence explains the rest.
Chinese policy lending went first. New Chinese loan commitments to Africa fell from a high of $28.5 billion in 2016 to $995.5 million in 2022. Commitments have stayed below $5 billion a year since 2020: thirteen loans worth roughly $4.61 billion in 2023, followed by six worth $2.1 billion in 2024.
Then the bond market. African sovereign issuance dropped from over $29 billion in 2018 to between $4 billion and $6 billion annually in 2022 and 2023, with only limited recovery through 2024 and 2025.
Then aid. Official development assistance fell from $83.8 billion in 2020 to $73.5 billion in 2023, with further declines expected. The United States accelerated this deliberately. Washington moved to shut down USAID and terminated 83 per cent of existing grants and contracts, and a ninety day foreign assistance review in January 2025 produced a 20 per cent reduction in Africa focused aid even as DFC infrastructure financing increased.
A government that cannot borrow and cannot rely on grants still owns ports, coastline, transmission corridors and mineral rights. A sovereign fund holding a depreciating deposit at that government’s central bank would rather hold the coastline. The concession is the instrument that clears that market.
Ras El Hekma is the transaction that makes the mechanism legible. ADQ committed $35 billion in total, roughly $24 billion in fresh foreign currency cash for the development rights and $11 billion converted from existing UAE deposits at the Central Bank of Egypt into equity, with the Egyptian state retaining a 35 per cent stake. The site spans over 170 square kilometres. Egyptian reserves that sat near $35 billion at the end of 2023 climbed past $47 billion by mid 2024.
Read that structure carefully. A third of the headline figure was not new money at all. It was a creditor converting a deposit into land. That is the clearest single illustration available of what is actually changing in African capital. Not the flag on the money, but the nature of the claim.
Four capital systems, one instrument
Set the four pools of capital side by side and the convergence is what stands out.
Western capital is exiting equity ownership of African operating assets and re-entering as debt and statecraft. Anglo American left platinum. In Nigeria the supermajors left the onshore: Shell’s $2.4 billion sale of SPDC to the Renaissance Africa Energy consortium, ExxonMobil’s $1.28 billion divestment of its shallow water assets to Seplat Energy, Eni’s $783 million sale of its Nigerian onshore assets to Oando, and TotalEnergies’ divestment of its stake in the SPDC joint venture to Chappal Energies. Meanwhile Western public money reappeared as senior debt inside other sponsors’ projects. The IFC provided up to $600 million in financing for AMEA Power’s integrated solar and storage project in Egypt. It also reappeared as strategic corridor finance. The DFC launched its Trade Over Aid initiative at the New York Stock Exchange on 28 April 2026, and has committed $553 million to upgrade the Lobito Atlantic Railway linking Kolwezi to Angola’s Atlantic coast.
Gulf capital is buying operating control of chokepoints under long concessions. AD Ports plans to invest $250 million through 2026 to modernise the Luanda terminal under a concession of twenty years signed with the Luanda Port Authority in April 2024, lifting container capacity from 25,000 twenty foot equivalent units to 350,000. Its Pointe-Noire terminal sits under a concession of thirty years with the Republic of Congo, extendable by another twenty. DP World holds a contract of thirty years to operate and upgrade a section of Dar es Salaam port. Masdar alone has committed $10 billion to deliver 10 gigawatts of renewable capacity in Africa south of the Sahara by 2030, including projects in Angola, Zambia and Ethiopia.
Asian capital has made the most consequential structural move. China converted itself from creditor to owner. Between 2004 and 2024, Chinese companies announced $73.9 billion in greenfield investment and $38.1 billion in mergers and acquisitions in Africa, with greenfield rising substantially since 2013. Sources disagree on the current run rate. Rhodium Group puts completed Chinese deals at $11 billion in 2024, up from less than $5 billion in 2021, while Ministry of Commerce data place China’s invested stock in Africa below 4 per cent of the total as of 2023. That disagreement should be stated rather than resolved. The direction is not in dispute. India arrives as operator rather than financier. Adani Ports entered a concession agreement of thirty years with Tanzania in May 2024 to operate Container Terminal 2 at Dar es Salaam. Japanese and Korean money comes in quietly as minority equity. AMEA Power’s project of over $700 million in Aswan was advanced together with Kyuden International of Japan and the IFC.
African institutional capital has the scale and is not at the table. More on that below.
The common denominator across all four is the long concession over a public asset. That structure is agnostic about nationality and specific about time. A loan matures and the relationship ends. A concession signed in 2026 is a decision about who collects the rent in 2056.
The hybridity of these deals also undercuts the tidy geopolitical reading. AD Ports is developing Pointe-Noire through a joint venture with CMA CGM’s subsidiary CMA Terminals, and awarded a contract worth 184 million dirhams to Shanghai Zhenhua Heavy Industries. Emirati sponsor, French partner, Chinese equipment, Congolese asset, a clock running thirty years. Asking whose capital that is misses the point. Asking who holds the cash flows until 2056 does not.
Who wins, who loses
Sovereign linked investors win, and the trend is global. UNCTAD’s top 100 non financial multinationals now include 26 companies with significant State ownership, up from 15 in 2017, and completed cross border acquisitions by majority State owned multinationals averaged $85 billion a year from 2021 to 2025. Patient, politically insulated capital is exactly what a long African concession requires, and it is precisely what listed Western operators, disciplined by quarterly returns and reputational exposure, were exiting.
Finance ministries win in the near term and pay later. Egypt bought reserve cover and exchange rate stability with coastline. Kenya’s transmission utility was candid about its own predicament. KETRACO said it can mobilise only a fraction of a $250 million annual requirement through traditional sources including development partners, government support and remittances from KPLC. Under that constraint a concession is not a strategic choice. It is the only clearing price available.
The losers are harder to see because they are diffuse and future dated. Structures built on availability payments keep the fiscal liability with the state while transferring the asset. Kenya’s replacement transmission deal, valued at $311 million with a consortium of Africa50 and Power Grid Corporation of India, runs over a concession period of thirty years, with KETRACO making availability payments tied to performance. A future Kenyan finance minister inherits that obligation with no ability to reprice it.
Environmental and decommissioning liability is the other quiet transfer. Nigeria’s regulator has been unusually alert here. NUPRC approved the Seplat and ExxonMobil transfer only after confirming operational readiness and securing over $400 million in pre sale decommissioning and abandonment liabilities. That is good regulation, and it is the exception rather than the norm.
And there is the flow nobody counts as investment at all. Swissaid found that the United Arab Emirates imported 2,569 tonnes of gold from Africa between 2012 and 2022 that was not declared for export by the source countries, worth $115.3 billion. In 2024 the UAE imported 748 tonnes of African gold, an increase of 18 per cent on the previous year. Over a comparable decade, the undeclared gold leaving the continent for one destination is of the same order of magnitude as that destination’s entire announced investment pipeline into it. Any honest account of the Gulf’s position in African capital has to hold both numbers at once.
One group has done unambiguously well. The Nigerian divestments added approximately 200,000 barrels per day to national output while unlocking over $5.5 billion in final investment decisions. Nigerian independents now own onshore assets that Shell held for most of a century. Whether they bought value or liability will be settled over fifteen years, not three.
The bidder that isn’t bidding
Here is the structural fact that the framing built on Gulf and Asian capital obscures entirely. The Africa Finance Corporation’s State of Africa’s Infrastructure Report 2026 found that non bank domestic capital pools exceed $2 trillion against cumulative external flows of approximately $1.7 trillion between 2014 and 2024, and that domestic pension and insurance assets crossed $1 trillion for the first time. Including banks, the domestic capital base reaches $4.4 trillion. Africa’s own savings now exceed a decade of everything the world sent it.
That capital is lending to governments. Nigeria’s pension funds held 77 per cent of their assets in government securities in April 2026. In some countries nearly 70 to 80 per cent of institutional portfolios sit in government debt. Nigeria’s regulator permits funds to hold up to 15 per cent of assets in private equity, with industry assets now above 31 trillion naira and actual take up far below the ceiling, while Ghana’s regulator allows up to 25 per cent in private funds against actual allocations to alternatives of just 0.58 per cent.
Note what that implies. The ceilings are not binding. This is not a regulatory failure but a behavioural and institutional one, rooted in the absence of bankable pipeline, in fee and fiduciary structures that punish illiquidity, and in the scar tissue left by domestic debt restructuring. AFC’s chief executive Samaila Zubairu put the diagnosis precisely. “The constraint is no longer capital, it is intermediation,” he said at the report’s launch in Nairobi.
He is right about the mechanism, and the framing deserves one correction. As Brookings has argued, these are not idle trillions. The money is already at work financing governments, and the stock, honestly measured, would cover under three years of annual shortfalls before running dry. The $2 trillion is a reason to build intermediation machinery, not a reservoir to drain. Anyone pitching it as a quick substitute for external capital is selling a stock as though it were a flow.
There is one working counter example, and it is worth watching closely. Africa50, an infrastructure investment platform backed by 33 African governments, two African central banks, the African Development Bank Group and the Public Investment Corporation of South Africa, now holds the Kenyan transmission concession that Adani lost. AFC joined the Lobito Corridor as lead developer of the greenfield rail extension into Zambia, backed by a financing commitment of $500 million from the African Development Bank. African institutional money is entering the concession economy. It is entering through multilateral vehicles rather than through the pension funds that hold the savings.
What to watch, and what would prove this wrong
Intermediation catches up: Pipeline, guarantees and project preparation improve, and African pension money co-invests alongside Gulf and Asian sponsors rather than watching. Ghana’s 5 per cent Pension and Insurance Compact, launched in April 2025 alongside a locally managed fund of funds of $70 million and subsequently backed by a government mandate, is the cleanest natural experiment on the continent, because Ghana’s problem is demonstrably not its ceiling.
Concession lock in: The best assets are contracted out for two or three decades before domestic capital mobilises. The ownership question is then settled by default, and the debate about it becomes retrospective.
Sponsor retreat: Announcements do not convert. Mauritania, Zambia and AD Ports’ own capital plan already supply evidence for this scenario.
Concrete watch points over the next twenty four months: whether any African pension fund takes a direct equity position in a port, grid or rail concession; whether Masdar’s Zambian and Angolan projects reach financial close rather than announcement; construction milestones at Ras El Hekma against the stated schedule; PenCom’s reported take up against its ceiling of 15 per cent for private equity; decommissioning provisioning on the transferred Nigerian assets; and the size of UNCTAD’s 2027 revisions to its own 2026 prints.
What would prove this analysis wrong is evidence that announced Gulf and Asian capital expenditure is converting to deployment at rates comparable to historical Western investment, which would make the nationality story real rather than a re-papering story. That evidence is not currently available at project level, and its absence is itself the finding.
For the people who have to act. A finance minister signing a concession of thirty years without a computed reservation price is a price taker, and there are perhaps a handful of African sovereigns with the internal analytical capacity to compute one. A pension trustee holding 77 per cent of retirees’ savings in a single sovereign issuer is running a concentration risk that no asset allocation committee in any developed market would sign off on, and calling it prudence does not change the exposure. An investor pricing sponsor nationality rather than sponsor delivery record is pricing the wrong variable.
Africa is not attracting more foreign capital. It attracted about $70 billion in 2025 out of roughly $1.6 trillion moving globally, in a year when flows to developed economies rose 43 per cent. What is changing is what the continent is selling. For two decades it sold debt, which matures, and equity in commodity extraction, which the buyers have now largely sold back. It is now selling time: twenty and thirty years of the cash flows from the ports, grids, corridors and coastline its own economies will depend on. That is a durable transfer, and it is being made at a moment when the continent’s own institutions hold more capital than they ever have and are deploying almost none of it into the assets being sold.
The useful question is no longer whose money it is. It is what kind of claim it buys, and how long the clock runs.
