Fewer, Bigger, Blended: How the Money Behind African Digital Infrastructure Is Being Rewired

By /

A note on where we sit. We are not a lender or a fund. We sit on the independent side of these transactions — commercial, technical and transaction diligence for the investors and operators building this infrastructure. What follows is the pattern we see across mandates. It is an observation about how the capital is behaving, not a lecture to the banks, funds and DFIs who know their own books far better than we ever will.

A question we are asked more and more often, in different forms, comes down to this: has investment in African digital infrastructure slowed down? Tower and fibre companies talk about margins, utilisation and free cash flow rather than build targets. Large greenfield announcements feel less frequent. The instinct is that the money has cooled.

The data tells a more precise story. Capital has not dried up. It has rotated and restructured — moving out of mature passive assets and into compute, backbone and gateways; concentrating into fewer, larger, more structured deals; and drawing in a wider, more patient set of stakeholders even as early-stage venture retreats. Increasingly, the outcome of a deal is decided by how it is structured and executed, not by who spotted the theme first.

The mismatch that forces the discipline

Start with scale, because it frames everything. The African Union and OECD put the continent’s infrastructure need at roughly $155 billion a year through 2040, of which digital — fibre in particular — accounts for about $36 billion a year (a wide range sits behind that headline: the World Bank’s earlier Broadband Commission work costed universal-broadband access alone at closer to $9 billion a year, so the number depends heavily on scope and horizon). On the compute side, McKinsey estimates installed data-centre capacity must roughly quadruple — from around 0.4 GW today to 1.5–2.2 GW by 2030 — implying $10–20 billion of new investment.

Against that need, total African private capital deployed across all sectors in 2025 was $5.1 billion across 530 deals (AVCA). Notably, Africa was the only global region to record deal-volume growth that year — up 8% — while global volumes fell. So the continent is not out of favour. But the pool is a fraction of the need, and that single fact explains most of what follows: capital concentrates in the largest, most bankable, best-structured opportunities and skips the rest. In a market this capital-constrained, structuring is the binding skill, not sourcing.

Trend one: the great rotation, from harvest to greenfield

The instinct that towers and fibre incumbents have gone quiet on building is correct — and it is deliberate, not a blip.

Look at the towercos in their own words. IHS Towers guided full-year 2025 capex down to $240–270 million — a deliberate step-down — while lifting adjusted levered free cash flow 47% to $448 million, generating $983 million of operating cash and cutting net leverage to 3.1x, having exited Peru, Brazil and Colombia to refocus on core African markets. Helios Towers tripled free cash flow, launched its first-ever dividend ($25 million) alongside a $50 million buyback, and brought leverage down from 4.0x to 3.4x. The Africa towers market itself is growing at just 3.35% a year (from $4.03 billion in 2026 to $4.75 billion by 2031), with capex intensity falling from a 21% peak in 2021 toward ~15% by 2027 and tenancy ratios climbing from 1.5x toward 2x in core metros. This is a segment sweating built assets and returning cash — harvest and efficiency mode.

But the greenfield energy did not disappear. It rotated. It moved to the layers where demand is genuinely accelerating:

  • Data centres. Raxio has crossed $380 million of committed capital after contracted power capacity rose sixfold in H1 2026, with existing shareholders Roha and Meridiam raising their stakes. African Infrastructure Investment Managers (AIIM) committed $90 million of growth equity to a joint venture with the Moroccan operator N+ONE Datacenters to build new West African capacity.
  • Backbone and subsea. The DRC’s $1.5 billion Congo River fibre MoU with Genew (1,700 km along the river bed plus 400 km of land links); Google’s new Eastern Cape digital exchange and subsea gateway, anchoring its Umoja cable.

The advisory point for anyone underwriting this: the diligence question has changed. It is no longer “what does it cost to build?” It is “is the demand contracted, and is the power real?” Those two questions now decide more deals than construction cost ever did — which is a very different underwriting exercise from the one that funded the tower rollouts of the last decade.

Trend two: fewer, bigger, blended

The second shift is in the shape of the capital itself. Deals are becoming larger, fewer and more structured. The clearest evidence is in venture, where the pool held but the breadth collapsed: African startups raised $1.44 billion in H1 2026 across just 146 deals, down from 252 a year earlier — the same money, roughly half the companies. Debt made up around 40% of that total, as venture debt fills the gap left by retreating equity.

At the infrastructure end, single-sponsor cheques have given way to layered structures. As one legal advisory firm working these transactions puts it, a current African data-centre financing typically layers contracted-revenue project debt at the core, flexible capital around it, development-finance-institution participation, and a security package built for operational continuity. A single asset now carries equity, mezzanine, senior debt and credit enhancement — each tranche held by a different party with a different risk appetite.

This is the trend that most changes our own work, and we would rather be plain about it than pretend otherwise: more tranches and more counterparties mean more contracts to test and more assumptions to reconcile. What every party in a blended deal needs is not a louder opinion but a single, independent evidence base — on the demand, the technical viability and the security package — that the fund, the bank and the DFI can each rely on. That shared factual footing is what lets a complex structure close.

Trend three: who is at the table

The most useful way to read the market is as an ecosystem of complementary roles, each doing what it is structurally suited to. None is more important than another; they close deals together.

Development finance institutions are the anchor and the de-risker — cornerstone fund commitments and longer-tenor debt that crowd in everyone else. Recent examples: the EIB’s $75 million into Helios Fund V, BII’s $20 million into Africa50’s infrastructure fund, the EBRD’s €270 million to Axian’s Yas, and IFC debt into Atlas Tower and Raxio. Very little of scale gets built without them in the structure.

Commercial banks provide senior debt, and their distinctive contribution is local-currency lending — the one thing that addresses the currency mismatch that dollar financing cannot. The market’s syndication capacity is real and, for the right credit, deep: Standard Bank’s recent $800 million sustainability-linked loan drew over $1 billion of commitments from 30 banks, and Africa Finance Corporation upsized a syndicated facility to $2 billion — both signs of the liquidity available to well-structured African borrowers, even if neither is a digital-infrastructure deal in itself. Within digital infrastructure specifically, that appetite is naturally strongest where cash flows are contracted and securable — the nature of prudent lending, not a limitation we would presume to critique.

Private-equity and infrastructure funds supply the equity, and their behaviour mirrors the wider market: fewer, bigger bets, with data-centre platforms the favoured trade and long-hold vehicles replacing the classic ten-year fund. Actis raised $1.7 billion for its second long-life infrastructure fund and spun out an Africa-focused manager, Grene Capital; AIIM manages $3.8 billion across eight funds.

Strategics and operators are both a source of capital and, increasingly, the exit for everyone else — self-funding network capex (Airtel guided $1.1 billion for FY2027) while consolidating through M&A. MTN’s agreed take-private of IHS Towers, at an all-cash offer valuing IHS at roughly $6.2 billion enterprise value, and Vodacom’s $2.1 billion move to control Safaricom, are as much about internalising cost lines and securing scarce assets as about growth.

Hyperscalers are a newer source of capital — funding regions, cables and exchange points directly — and, just as importantly, their anchor-tenancy contracts are what make third-party data centres bankable in the first place. They often supply the very credit quality the lenders lend against.

Venture capital funds the adjacent layers — fintech rails, connectivity-enabled platforms, AI and data infrastructure — rather than the hard assets, and it is currently retreating from early-stage breadth toward concentration and debt. Egypt led H1 2026 fundraising at $327 million, a reminder that the venture centre of gravity has shifted north.

Family offices and dynastic holding companies are the emerging answer to a gap the rest of the stack cannot fill: patient, long-duration capital, well-suited to assets that compound slowly. Today they mostly participate through co-investment and holding structures rather than direct deal teams. The clearest African examples are the family-controlled holdcos — Remgro, which retains 57% of CIVH and, through it, the Maziv fibre group (Vumatel, at a 34.5% share of the fibre-to-the-home market, and Dark Fibre Africa); and African Rainbow Capital, through the mobile operator rain. The Vodacom–Maziv transaction, which completed on 1 December 2025 with Vodacom acquiring 30% of Maziv, is a textbook illustration of dynastic and strategic capital meeting in a single fibre asset.

Trend four: the market is quietly maturing

Beneath the deal flow, the plumbing is improving. 2025 was one of Africa’s strongest exit cycles on record — venture-backed exits rose 31% to 34, a new high (AVCA) — which matters, because a market that cannot exit cannot recycle capital. Institutional investors now provide around 21% of private-capital commitments, led by sovereign wealth and pension funds beginning to allocate at home. Sustainability-linked pricing has become a standard feature.

Public equity, though, remains thin, which is why exits skew to strategic M&A and take-privates rather than listings — Cell C’s May 2026 JSE listing raised roughly $156 million against a $413 million target, and Airtel is taking its mobile-money unit to the London market rather than a domestic one. The practical implication for sponsors is unavoidable: the exit has to be designed at entry. In a market where the realistic buyer is a strategic or a larger fund, that is a structuring discipline, not an afterthought.

The constraints we actually see kill deals

It would be dishonest to present only the momentum. The recurring reasons good assets fail to get funded are consistent: currency and repatriation (dollar-denominated cost against local-currency revenue); the tenor mismatch between bank debt and asset life; power, now the binding physical constraint on any compute-heavy build; a persistent early-stage equity gap; and thin public markets that narrow the exit. The clearest live illustration is power: Microsoft and G42’s flagship $1 billion Kenyan data-centre project was suspended in mid-2026 once it became clear that a 1 GW facility would draw close to a third of the country’s national grid — held up by power allocation and commercial payment guarantees, not by any shortage of demand. None of these is a demand problem. All of them are structuring and execution problems — which is precisely why they are solvable with the right work up front.

What it means

For investors, the message is that the theme is no longer the edge — everyone can see the demand. The edge is in underwriting the structure: the contracted demand, the power contract, the currency position, the security package and the exit. For operators, bankability increasingly turns on the same questions a lender or a fund will ask on day one, and answering them early is cheaper than answering them in diligence.

Our own role in all of this is deliberately modest and deliberately independent: the commercial, technical and transaction diligence that every party in a blended deal — fund, bank, DFI, operator — relies on to price risk with confidence. Not because we know these markets better than the people committing the capital, but because a complex, multi-party structure needs one shared, evidence-based view of the asset that all sides can trust. As the deals get larger, more structured and more blended, that need only grows.

The money behind African digital infrastructure did not leave. It changed shape. The winners this cycle will be the ones who understand the new shape — and structure for it.

Prepared by Digitalthings. Figures are drawn from primary results and recognised industry sources, verified against their publications; market-sizing forecasts vary by methodology and are cited accordingly. Data is current as of mid-2026.

Sources

Leave a Comment

Your email address will not be published. Required fields are marked *

Scroll to Top