Reformers keep asking governments to raise the ceilings on where domestic institutional capital can go. The ceilings are nowhere near being reached and one December in Ghana explains why. This article is published on African Arguments ahead of the African Development Summit which takes place on 18 September 2026.

Nigeria’s pension regulator permits funds to hold up to 15 per cent of assets in private equity. Industry assets now stand above ₦31 trillion. Actual take-up sits far below the ceiling. Ghana’s regulator allows up to 25 per cent in private funds; actual allocations to alternatives stand at just 0.58 per cent.

In Nigeria, PenCom itself points to a supply-side constraint, a shortage of qualifying, investable funds yet the broader pattern across the continent suggests the deeper friction is not the availability of paper, but the appetite for it.

That is the fact that should reorganise the argument about domestic capital in Africa. The headroom exists. It goes unused. Whatever is stopping African savings from financing African production, it is not the rulebook. That means the reform most often demanded, a higher permitted allocation, would change almost nothing.

The ACD Summit on 18 September 2026 will debate major questions of domestic investment, development financing, and capital flight.

The question matters more this year than last. When the OECD released its preliminary figures in April, aid from the world’s wealthiest donors had fallen 23.1 per cent in 2025 to $174.3 billion. That was the steepest single-year contraction on record and the second consecutive annual decline. Bilateral aid to sub-Saharan Africa fell 26.3 per cent. In a subsequent June projections report, the OECD forecasts a further 11.6 per cent drop in bilateral aid to sub-Saharan Africa in 2026, with health financing potentially falling by as much as 63 per cent below its 2022 peak.

The reflex has been to point at remittances. Roughly $124 billion was sent home by Africans abroad in the same year, about twice what the continent received in official development assistance. It is the wrong comfort, and not only because the two flows are different sizes of the same thing. They do different work. Remittances are transfers between households. They pay school fees, clinic bills, rent and funeral costs, and they are extraordinarily good at that. They do not build a district hospital, capitalise a fertiliser plant, or underwrite a twenty-year power purchase agreement.

The pool that could do those things is institutional and already domestic. African pension funds and insurers hold around $775 billion in assets, according to the Africa Finance Corporation’s State of Africa’s Infrastructure Report 2025, of which $455 billion is in pensions and $320 billion in insurance. Add sovereign wealth funds and the African Development Bank’s Solomon Quaynor puts the figure at $2.1 trillion. The AFC’s 2025 report notes that in some countries, nearly 70 to 80 per cent of institutional portfolios are held in government debt. Africa’s largest pool of patient capital is, in other words, mostly lending to African governments to cover recurrent spending and service existing debt.

Capability, not permission

Ask why, and the honest answer starts with staffing. As the OECD notes, most African funds lack in-house teams able to price infrastructure risk, run diligence on an unlisted company, or sit on the other side of a negotiation from an experienced fund manager. A trustee board that cannot evaluate a toll-road concession is not being cautious when it declines one; it is being accurate about its own capacity.

This is a solvable problem, but it is solved by unglamorous instruments. Pooled diligence vehicles let several funds share the cost of expertise none can afford alone. First-loss tranches make an unfamiliar asset class survivable on a first attempt. Trustee training that regulators actually fund rather than merely recommend. None of it is a policy announcement. All of it is a line item.

The second answer is incentive, and it is harder. Treasury bills pay well, settle predictably, and have never cost a trustee their job. An underperforming government bond is a market condition. An underperforming private placement is a decision with a name attached to it. Until that asymmetry is addressed directly through how trustees are appointed, indemnified and assessed, raising the ceiling higher simply widens a door nobody is walking through.

The leak on the way out

When African institutional money does reach private markets, a further loss occurs that rarely appears in the allocation statistics. The vehicle receiving it is often domiciled in Mauritius, Luxembourg or Delaware. The fees, the legal work, the fund administration and the professional expertise accumulate offshore, and the domestic industry that would eventually supply the missing capability never quite takes root. The sharper cost is denomination. A Ghanaian manufacturer earning cedis but funded in dollars is running an unhedged currency bet alongside its actual business. When the currency moves, an otherwise sound company becomes a distressed one. Ghana’s venture capital association launched a 5 per cent Pension and Insurance Compact in April 2025, a voluntary industry commitment, alongside a $70 million locally managed fund of funds. A subsequent government mandate directed pension and insurance funds to meet that 5 per cent allocation, precisely to test whether local-currency, locally domiciled capital can close that gap. It is small. It is also one of the few live experiments worth watching, because it treats domicile and denomination as design choices rather than accidents.

The thing nobody wants to say out loud

All of this runs into an obstacle that no technical fix can resolve. Pension assets are not development capital. They are workers’ deferred wages, held under fiduciary duty, and African trustees have recent and specific reason for caution.

When Ghana launched its domestic debt exchange in December 2022, pension funds fell inside the perimeter. They were exempted only after labour unions threatened to strike. The exemption held through the February 2023 exchange, though in August 2023 the finance ministry completed a separate swap of 95 per cent of local debt held by pension funds. When the government reopened the programme in September 2023, pension funds were again explicitly excluded. Bondholders who did participate faced coupon terms that some analysts valued as losses of 60 to 70 per cent.

That episode explains the unused headroom better than any regulatory audit could. It is also why the conventional sequencing is backwards. The usual programme runs: liberalise allocation rules, build capability, mobilise domestic savings, finance development. The order has to be reversed. Governments that want access to domestic long-term savings must first accept binding limits on their own claims to it. They need statutory protection of pension assets in any future restructuring, and an end to using captive institutional demand as a substitute for fiscal discipline.

Domestic capital mobilisation is not a technocratic programme. It is a bargain, and the state has to move first.

What progress would look like

The virtue of framing the problem this way is that it is measurable on a timescale short enough to embarrass anyone who claims progress without it. Two numbers tracked to 2030 would settle the question: the share of African institutional assets held in productive non-government assets, and the share of Africa-focused funds domiciled and denominated on the continent. A third would test whether the bargain is real: how many jurisdictions have placed pension assets beyond the reach of a domestic debt exchange by statute rather than by negotiation.

Neither of the first two requires anyone’s permission. The third requires a government to accept a constraint on itself in advance of a crisis, which is the hardest thing any government does, and the only thing that would make the other two move.

The author is Secretary General of the African Coalition for Development the global summit of which takes place on 18 September 2026.