Insight, Climate Finance

The Hurdle Is Not the Risk

A new vehicle aims to channel Nigerian pension savings into climate infrastructure, backed by a US$253m first-loss commitment. First-loss capital addresses credit risk. The evidence suggests credit risk is not what has been keeping Nigerian pension funds out.

Line-art diagram: on the left a small sage-green block labelled Perceived Risk sits under a low gauge; on the right a much taller coral stack of Infrastructure, Regulation, Financing, Distribution, Habits & Trust and Incumbent Advantage forms the Real Hurdle, with a construction barrier beside it.
First-loss capital lowers the perceived-risk block on the left. The stack that actually keeps institutional capital out of infrastructure sits on the right, and no amount of credit enhancement shortens it.

Africa Finance Corporation has launched ICRF Nigeria, a domestic vehicle designed to channel Nigerian pension funds, insurers, and asset managers into climate-resilient African infrastructure. It feeds AFC Capital Partners' Infrastructure Climate-Resilient Fund (target size in the region of US$750m), which already holds a first-loss commitment of around US$253m from the Green Climate Fund and aims to mobilise as much as US$3.7bn across ten to twelve projects. Fund size, first-loss size, and mobilisation targets are as reported by AFC.

The premise is one of the most important in African transition finance, and it is correct. Enormous pools of domestic long-term savings exist. They are overwhelmingly invested in government paper. Redirecting even a modest share into productive infrastructure would do more for the continent's transition than another decade of importing climate capital from Europe.

It is also worth being precise about the diagnosis, because the instrument chosen only works if the diagnosis is right, and the available data suggests the standard diagnosis may be wrong.


What the pension data actually shows

Nigerian pension assets reached about ₦29.5 trillion in March 2026 and a record ₦31.3 trillion by May, roughly US$22bn. Federal Government securities accounted for 58.07% of the total, at ₦17.14 trillion.

That is the number everybody quotes, and on its own it supports the usual story: conservative funds parked in risk-free paper, needing to be coaxed out by de-risking.

Two other numbers complicate that story considerably.

Infrastructure allocation is far below what regulation permits, and it is falling. Pension investment in infrastructure funds stood at ₦224.23bn in March 2026, down 25.26% in a single month, against a total asset base of ₦29.5 trillion. That is under 1%. Allocations to infrastructure and other alternatives remain below 5% despite significantly higher regulatory limits.

Donut chart titled Nigerian pension fund allocation, March 2026, with ₦29.5tn total printed in the centre. Five segments labelled with their share: cobalt Federal government bonds at 43 per cent (largest), butter domestic equities at 25 per cent, sky money market at 15 per cent, plum corporate bonds at 16 per cent, and a small coral segment for infrastructure funds at 1 per cent. A coral callout box points at the tiny coral segment and reads below the regulatory ceiling of 5 per cent. Footer: ₦29.5 trillion total, PenCom data.
Four segments compete for the bulk of the pool. The one segment that would fund the shortfall the country most needs is the sliver.

Meanwhile, the same funds took on substantially more risk elsewhere. Pension fund administrators increased holdings of domestic ordinary shares from ₦3.96 trillion at the end of 2025 to ₦5.46 trillion by March 2026, a 38% rise in a quarter.

Read those together. These are not funds incapable of taking risk, nor funds prevented from taking risk by a regulatory cap. They are funds actively increasing exposure to Nigerian corporate equity, a volatile asset class, at speed, while simultaneously reducing an infrastructure allocation that sits at a small fraction of the permitted ceiling.

Whatever is keeping them out of infrastructure, it is not an unwillingness to bear risk.


What first-loss capital does and does not do

A first-loss tranche absorbs initial losses on a portfolio, improving the risk-adjusted position of investors above it. It addresses one specific thing: the probability and severity of credit loss.

Set that against the four constraints a Nigerian pension fund administrator is actually facing.

Four constraints, one instrument
  1. Relative yield. The competing asset is not a theoretical risk-free rate. It is Nigerian government paper at a high nominal yield in naira, held to maturity, requiring no credit team and no board argument. An infrastructure fund must clear that, plus an illiquidity premium, plus a spread for a credit the fund cannot easily analyse. First-loss capital does not raise the coupon. It reduces the chance of losing money, which is a different variable, and in a high-nominal-rate environment it is often the less binding one.
  2. Liquidity and valuation. Retirement savings account funds are priced regularly and members can switch administrators. An illiquid, infrequently valued holding creates operational and fairness problems entirely separate from whether it is a good investment. There is no established secondary market for African infrastructure fund interests in naira.
  3. Eligible instruments. The recurring explanation given by the industry itself is the absence of eligible, rated, appropriately structured instruments to invest in. This is a plumbing problem, not a risk-appetite problem, and no amount of first-loss capital creates a rated, listed wrapper that a compliance function can approve.
  4. Career and reputational asymmetry. A trustee who buys government bonds and underperforms is doing their job. A trustee who buys an infrastructure fund and loses money on pensioners' savings is answering questions for years. This asymmetry is rational, present in every pension system in the world, and not addressed by improving expected returns at the margin.

First-loss capital solves the first half of one of these four. That may still be worth US$253m. It is unlikely to be sufficient on its own, and mobilisation targets built on the assumption that credit risk was the obstacle will disappoint in a way that gets blamed on the wrong thing.


The part that genuinely does transfer

One element of this deserves unambiguous credit, and it is the currency.

Naira revenue matched against naira liabilities removes the exposure that has destroyed more African infrastructure and energy businesses than weak demand ever has. A domestic institutional investor holding a naira-denominated asset does not care about a devaluation in the way a dollar lender does. We made the same point about the record share of local currency in off-grid solar funding, and it holds here with more force, because pension liabilities are naira liabilities by definition.

If ICRF Nigeria achieves nothing else, moving infrastructure financing into the currency of the revenue is a structural improvement that survives whatever happens to the mobilisation target.


The asymmetry nobody puts in the press release

There is a distributional feature of domestic mobilisation that is worth stating plainly, because the language of the field obscures it.

The money belongs to Nigerian workers. Retirement savings account membership stood at more than 11.3 million in June 2026. These are teachers, civil servants, bank staff, and factory workers whose retirement income is the asset being mobilised.

If this works, it is excellent. Returns stay in the country, domestic savings build domestic infrastructure, and the pattern where African growth is financed abroad and the returns leave begins to break. That is the argument we have made about European capital and African growth, and domestic institutional capital is the strongest available answer to it.

If it does not work, the loss sits with Nigerian pensioners. It does not sit with the development finance institution that recorded the mobilisation, or with the concessional funder whose first-loss tranche was capped at US$253m, or with the fund manager whose fees were earned on committed capital.

That asymmetry becomes dangerous when mobilisation ratios are used as a performance metric. A funder measured on dollars mobilised per dollar of concessional capital has an institutional interest in the pension money moving, and no equivalent interest in whether it should. This is not a claim about anybody's conduct. It is an observation that the incentive is misaligned by construction, and that the party bearing the residual risk is the only one in the chain with no representation in the design.

The fiduciary question sits with the trustees, where it belongs. It would be improved considerably by an independent view of the underlying projects that is not produced by the fund manager, the sponsor, or the concessional funder, all three of whom have a position.


What would actually move the money

If the diagnosis above is right, four things would do more than expanding the first-loss tranche.

A rated, listed, eligible instrument. The single most cited obstacle by the industry itself is the absence of instruments a compliance function can approve. A listed infrastructure bond with a domestic rating and a defined eligibility status is a plumbing exercise, not a financial innovation, and it removes a hard constraint rather than softening a soft one.

Concessional support to the coupon, not only to the downside. If the binding constraint is that infrastructure must beat a high domestic sovereign yield, then subsidising the return is the intervention that matches the problem. This is less fashionable than first-loss structures because it looks like an operating subsidy rather than a catalytic one, and it may nonetheless be the thing that works.

Standardised documentation across projects. A pension fund credit team cannot underwrite twelve bespoke infrastructure projects. It can underwrite a standardised instrument twelve times. Documentation standardisation is the least glamorous item on any blended-finance agenda and consistently one of the highest-leverage.

A secondary market, even a thin one. Any mechanism that allows an administrator to exit a position addresses the liquidity constraint directly, and the liquidity constraint is doing more work here than the credit constraint.

It is also worth noting that PenCom itself is developing an infrastructure vehicle to channel pension assets. Two parallel efforts pointing at the same pool of money, one from a regulator and one from a development finance institution, is either productive competition or duplicated plumbing, and which it turns out to be depends on whether they converge on a common instrument standard.

The Lab works on this at the intersection of financial inclusion and energy access. Where a financing structure depends on outcomes at project level, somebody independent of the sponsor should be establishing what those outcomes actually are, which is ordinary impact measurement applied to a fiduciary question.

If you are a trustee, a sponsor, or a concessional funder in one of these structures and want an independent read on what the projects are actually delivering, tell us what you need to know.


Sources


This is an independent insight piece by Transitions Lab. For the Lab's applied work, see Financial Inclusion & Payment Systems and For Funders. See also Strategic Is Not the Same as Financeable on the same misdiagnosis, when an instrument designed to raise expected returns is used against a constraint that is actually about variance. To discuss a study, see Contact.

Read more Articles & insights See all articles → See it in the field Case studies See all case studies →