By Chidi Nwafor
Two solar projects, in different countries, sought roughly the same amount of capital in the same calendar year: just under $150 million, similar technology, similar construction timelines, similar generation capacity. One reached financial close in eleven months. The other spent two years in structuring, went through three different lender consortia, and closed at a cost of capital nearly four percentage points higher than the first. The difference was not the resource, the technology, the sponsor’s competence, or even, in any simple sense, the risk. Both projects carried comparable currency exposure, comparable offtaker credit quality, and comparable construction risk. The difference was that the first project had been built with a capital stack designed around where its risks actually sat, and the second had been financed, largely, with a single layer of commercial debt asked to absorb every risk in the project at once.
This is the piece of the infrastructure-finance puzzle that receives the least public attention and does the most damage when it is missing. Article 1 in this series established that capital exists in abundance relative to deployable projects. Article 2 established that capital stalls when risk is not properly allocated. Article 3 established that bankability itself has to be built, project by project, through deliberate preparation. This article asks a question that sits underneath all three: even when a project is prepared, and its risks are understood, what determines whether the capital that finances it actually fits the risk it is being asked to carry? The answer is financial structure, and the industry’s habitual response to the infrastructure gap; more financing, more pledges, more capital raised- routinely ignores it.
Why the Capital Stack Exists at All
A capital stack is not a fundraising exercise. It is an engineering decision about how to distribute a project’s risk across capital providers with fundamentally different risk appetites, return requirements, and time horizons, so that no single layer is asked to bear risk it is not built to hold. Commercial debt exists to fund the most predictable, lowest-risk portion of a project’s cash flows, and prices accordingly; it is patient with operational cash flow and impatient with uncertainty. Equity absorbs the residual risk the project carries after debt is serviced, and demands a return that reflects that first-loss position. Subordinated or mezzanine capital sits between the two, absorbing more risk than senior debt for a higher return, often exactly the layer needed to make a project’s risk profile acceptable to senior lenders who would otherwise decline it outright. First-loss capital and guarantees exist specifically to absorb a defined slice of downside risk, protecting the capital above them and allowing it to be priced as though that risk did not exist for its purposes. Concessional finance and grants exist to fund the risk no commercial or even standard development capital will accept at any price, usually early-stage, developmental, or explicitly non-commercial risk.
Each layer, in other words, exists to solve a specific problem that a different layer cannot solve as efficiently. The question that should be asked of every infrastructure transaction is not “how much capital does this need” but “what problem does each layer of this capital stack need to solve, and is the right instrument solving it?” The second solar project above failed to ask this question. A single tranche of commercial debt was asked to price construction risk, currency risk, and off-taker risk simultaneously, because no first-loss or guarantee layer existed to take any of them off its plate. The result was not that the project became unfinanceable. It was that the debt was repriced to compensate for everything at once, and the project became far more expensive to finance than its underlying risk justified.
Capital Stack as Architecture
The useful way to think about a capital stack is architectural rather than financial: as a structure that sequences and distributes risk deliberately, the way a building’s frame distributes load, so that the parts of the structure best suited to bear weight actually bear it. A well-designed stack protects senior capital by ensuring it only absorbs the risk it is priced to hold, which is what allows commercial lenders to offer competitive terms in the first place; a project with no subordinate or first-loss capital ahead of senior debt forces that debt to price as though it were equity, which defeats the purpose of raising debt at all. It sequences risk across the project’s life cycle, matching development-stage capital, which should expect a meaningful probability of loss, to development-stage risk, and construction-stage capital, priced for completion risk, to the construction period, and refinancing the project into cheaper, longer-tenor operational debt once that risk has cleared. It extends tenor by allowing patient capital, DFI debt, pension allocations built for long duration, to take the portion of the project that benefits most from a longer repayment horizon, rather than forcing the entire project onto whatever tenor commercial banks are willing to offer. And it manages currency exposure by matching, where possible, the currency of financing to the currency of revenue, or by allocating the mismatch deliberately to a party equipped to hedge or absorb it, rather than leaving it embedded, unpriced, in a single tranche.
None of this is exotic financial engineering. It is standard project finance discipline, applied inconsistently, and the inconsistency is precisely where the Missing Middle of Financial Structure sits: between the general availability of every layer described above, somewhere in the global financial system, and the specific, deliberate assembly of those layers into a stack matched to a given project’s actual risk profile.
Blended Finance, Examined Rather Than Celebrated
Blended finance has become the default answer to structural gaps in infrastructure capital stacks, and it deserves more scrutiny than the enthusiasm around it usually allows. The honest test of any blended structure is not whether concessional capital was deployed, but what that capital actually did. Did it mobilise additional private capital that would not otherwise have entered, by absorbing a specific risk private capital could not price? That is blended finance working as intended. Did it instead subsidise a return on a project that would have attracted commercial capital anyway, at only modestly worse terms? That is scarce concessional capital spent for no structural gain, crowding out its own future usefulness. Did the structure build a repeatable market, standardised terms, a track record other transactions can reference, or did it produce a single bespoke deal that leaves the next project exactly as unfinanceable as this one was before? And critically, did the structure create a genuine path to graduation, where the sector or geography eventually attracts capital without concessional support, or did it create a dependency, where every subsequent transaction in that market assumes a concessional layer will always be there? Blended finance that cannot answer these questions well is not solving the Missing Middle of Financial Structure. It is a more expensive way of not solving it.
Africa and the Global Comparison
Africa’s experience makes the cost of poor structuring unusually visible, because the margin for error is thinner: a project financed with the wrong stack does not merely become expensive, it frequently becomes unfinanceable outright, since there is often no deep secondary market willing to absorb mispriced risk the way there might be in a mature market. Nigerian and East African renewable energy projects that have successfully layered DFI debt, guarantee instruments and commercial tranches have closed at costs of capital meaningfully below comparable projects financed on commercial debt alone, not because the underlying risk was different, but because the structure matched capital to risk rather than asking one tranche to absorb all of it. The comparison with more developed capital markets, parts of Latin America’s renewable sector, or Southeast Asian toll-road financing, shows the same principle at larger scale: markets that have built standardised, replicable capital-stack templates for a given asset class consistently finance projects faster and more cheaply than markets structuring each transaction from first principles.
What Governments, DFIs and Investors Should Actually Do
Governments can do more here than is generally assumed, not by providing capital directly but by enabling the layers that make private capital comfortable: credible guarantee frameworks, clear rules on currency convertibility and transfer, and regulatory stability that reduces the risk premium every layer above it has to price. Development finance institutions should think of their role less as filling a financing gap and more as designing the specific layer, guarantee, subordinated debt, first-loss tranche, that is missing from a given market’s capital stack, deployed with a clear thesis about what commercial behaviour it is trying to unlock, and a clear measure of whether that behaviour materialises. Commercial and institutional investors should engage with capital-stack design earlier in a transaction’s life, rather than being presented with a single-tranche structure and asked to accept or reject it, since much of what makes a project unfinanceable on commercial terms is not the underlying risk but the absence of a layer built to absorb the part of it commercial capital was never meant to hold.
The lesson across all four articles in this series so far is converging on a single point. Capital exists. Risk can be translated. Bankability can be built. And once all three are true, a project still requires a financial structure engineered specifically to its risk profile, not simply “more financing,” to actually close. Infrastructure’s persistent underfunding is, in a meaningful sense, a myth built on a category error: mistaking the volume of capital in the system for the quality of its architecture.
But even a perfectly structured capital stack solves nothing if the capital it describes cannot physically and institutionally reach the opportunity it was designed for, if the pension fund with the right risk appetite has no vehicle through which to invest in a Nigerian mini-grid, or the guarantee facility with the right instrument has no presence in the market that needs it. That is a different problem, one of access rather than design, and it is where this series turns next.
That is the Missing Middle of Capital Access.
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. chidi.nwafor@de-lazuliconsult.com