Blended Finance: From Labels to Impact
The structure exists. The public money is in the wrong layer. In Part III, we argued that the label does not reliably reflect the impact, nor does the price reflect the risk and that the fix is therefore not a better score but a better structure. Blended finance is that structure.
SUSTAINABLE INVESTING INSIGHT SERIES | PART IV | OCTOBER 2026
Blended finance has mobilised some $213 billion in total deal volume since it was first codified. Across developing countries, each dollar of public money has attracted $0.37 of private finance in low-income markets and $0.75 overall. The leverage ratio has not moved since 2018.
The usual explanation is that the system lacks enough concessional capital.
New tranche-level analysis from the OECD suggests otherwise. Across structured blended funds in emerging and frontier markets, junior tranches show an average leverage of 4.35 and the highest mobilisation of any tranche type despite being used far less often than any other.
Africa offers the sharpest illustration. OECD donors made 598 commitments to blended finance funds on the continent. Only 93 of those went to junior, first-loss tranches; nevertheless, they mobilised over $6.8 billion, more than the 455 common-share commitments, which produced $5.7 billion between them.
What separates those 93 commitments from the 455 is not how much money was committed. It is where in the capital structure it was placed: at the bottom, where the first losses fall, rather than alongside everyone else.
This fourth instalment of our Green Finance Insight Series is about that observation, and it should be read as an argument for blended finance, not against it.
The instrument answers a real problem, and the evidence on what works is unusually clear. Properly positioned first-loss capital mobilises private investment at multiples of its size. Guarantees are among the most efficient instruments in development finance. Portfolio structures are beginning to solve the ticket-size constraint that has kept emerging and frontier market assets out of institutional portfolios for two decades. None of that is in doubt.
What is in doubt is the aggregate result and the data points at practice rather than at donor budgets.
Start with the evidence nobody uses
Before the structures, the credit data is the part of this debate that gets the least attention and should get the most. Twenty-nine development finance institutions pool their lending histories into a single database, GEMs. It covers more than 10,000 private entities across 169 countries, from 1994 to 2024. It is the largest emerging-market credit dataset, and lenders publish it themselves.
GEMs, 1994–2024 | Default rate | Recovery rate |
Private lending, all markets | 3.54% | 72.9% |
Sub-Saharan Africa, private | 6.05% | 78%+ |
Renewable energy projects | 3.27% | 64.5% (median) |
Sovereign and sovereign-guaranteed | 0.77% | 95.1% |
A 3.54% average default rate corresponds to a BB or B+ rating. This is not a projection. It is thirty years of realised outcomes across the emerging and frontier universe.
Run the loss arithmetic. Private lending across all these markets defaults at 3.54% and recovers 72.9%, an expected annual loss under 1%. Even sub-Saharan Africa, the highest-default region at 6.05%, recovers more than 78%, giving roughly 1.3%. Emerging-market renewables sit at 3.27% with 64.5% median recovery roughly 1.2%.
One finding deserves particular attention: a country's sovereign rating overstates the default performance of its corporates. The sovereign ceiling that keeps emerging-market infrastructure out of institutional portfolios is not supported by the lenders' own data.
The IFC's risk directorate puts it plainly: these statistics challenge the conventional view of emerging markets as high-risk destinations, and such investments should sit within the risk appetite of a broad range of investors.
One caveat is worth naming, because it is rarely stated. The high recovery rates on sovereign exposure reflect in part the preferred creditor status of multilateral institutions, a convention that does not transfer to an unaccompanied private lender. The private-lending figures are the relevant ones for an allocator.
Where the money sits
A blended vehicle stacks its investors. Returns flow from the top down; losses strike from the bottom up. The investor at the bottom absorbs the first losses so that the layer above becomes investable for institutions that could never underwrite the asset directly.
Public money does not fund the project. It buys the risk that lets private money in. The entire mechanism depends on where that money sits.
As of 2021, 42% of blended finance deployed by development finance institutions went to senior debt, the least effective instrument for mobilising private capital, and 11% to subordinated debt.
The OECD's tranche-level work shows what that allocation costs. Junior tranches are used sparingly and still produce the highest mobilisation of any tranche type, with average leverage of 4.35 in high-risk markets. Meanwhile, the sector-wide ratio sits at roughly 4:1, of which only about $1.10 comes from genuinely private investors.
The sector has spent a decade asking for more concessional capital. The data suggests the more productive question is where the existing capital is placed.
What the structure is for
It is worth pausing on what all this machinery is meant to deliver, because the mechanics can obscure it.
Part III established the asset. Six hundred million people without electricity. Solar resources among the best on earth, almost entirely unbuilt. A continent constructing its first energy system at scale, and constructing it clean because clean is now the cheapest new generation available across most of its irradiance profile. African heads of state have committed to raising renewable capacity from 56 GW in 2022 to at least 300 GW by 2030.
The economics of that build-out work. The engineering is settled. What's missing is a capital structure that lets institutional money participate.
That is precisely the gap blended finance was designed to close, and it is why the instrument matters more in renewable energy than almost anywhere else.
Consider what makes these assets unique. A solar plant generates contracted, long-duration, inflation-linked revenue the profile every European pension fund says it is hunting. Its technical risk is minimal, and its operating costs are predictable. What blocks the allocation is not the asset but its surroundings: whether the national utility pays, whether the currency converts, whether the tariff holds.
Those are exactly the risks a junior tranche absorbs and a guarantee transfers. The structure does not need to improve the project. It needs to reposition the investor's exposure to risks that sit outside it.
And the development logic runs in the same direction. An energy system is not one investment among others it is the precondition for the rest. Industrial capacity, cold chains, digital infrastructure, hospitals and schools all depend on reliable power. Capital that unlocks generation unlocks everything downstream of it.
Which is why the seniority question is not a technical footnote. Every junior commitment that becomes a senior one is generation capacity that does not get built, in markets where the gap between what is needed and what exists is measured in hundreds of gigawatts.
A familiar promise
Anyone who worked through the public-private partnership era will recognise the shape of this.
PPPs carried the same proposition twenty years ago: public money attracts private money, private discipline delivers public assets, the state multiplies its means. The vocabulary changed. The proposition did not.
They are not the same instrument. A PPP is a contractual arrangement for delivering an asset, with the state as counterparty. Blended finance is a capital structure arrangement, with the state as subordinated investor. The Ivorian solar concessions described in Part III are PPPs; a tranched fund financing them would be blended finance. Two floors of the same building.
But the difficulties rhyme as mirror images. PPPs were criticised for leaving too much risk with the state, which bore the real exposures while guaranteeing private returns. The observation here is the opposite: the public sector bears less than the structure was designed for it to bear.
Opposite patterns, one common cause. PPPs proliferated partly because they moved obligations off the public balance sheet, where fiscal rules could not see them. Guarantees were under-used because they sat outside ODA accounting, where donor performance could not count them.
In both periods, the financial structure followed what public accounting measured.
Guarantees: the instrument the data recommends
Which brings us to the clearest illustration in this series.
Guarantees account for roughly 4% of multilateral development bank commitments. Loans account for around 70%. Yet the OECD finds guarantees among the most efficient blended instruments on cost, capital efficiency and mobilisation, and its 2025 guidance describes them as effective and still under-utilised. Across 2021–2024, guarantees were in fact the single largest channel of private finance mobilisation in Africa, accounting for 31% of the total.
The gap between how well the instrument performs and how often it is used has an explanation, and it is an accounting one.
Until 2023, a guarantee was not ODA. Under DAC rules, it was registered only when called only when the project failed and the donor paid out. A guarantee that worked exactly as designed, mobilising private capital and never being invoked, counted as zero.
Consider the choice from a donor's desk. A loan disburses, counts, and appears in the annual figure. A guarantee mobilises more private capital per unit of risk borne, and registers as nothing unless something goes wrong.
The measurement system favoured the less effective instrument. The distortion was significant enough that specialised guarantee providers such as GuarantCo were established precisely because guarantees fell outside ODA.
The DAC reform of 2023 made guarantees ODA-eligible. A decade of institutional practice does not reverse on a rule change but the obstacle has been removed.
What is being built
The structures that work are not theoretical, and they are not confined to one region.
Portfolio guarantees. In 2025, MIGA executed framework terms of $495 million with CrossBoundary Energy Holdings covering distributed renewable projects across up to twenty African countries, with a first-loss participation of up to $37.6 million. One guarantee, up to a hundred projects.
Scale frameworks. In 2026, MIGA committed up to $1.48 billion in political risk guarantees supporting roughly $1.65 billion of investment across up to 23 AMEA Power renewable and storage projects, taking MIGA's lifetime issuance past $100 billion.
Rated senior tranches. Lendable's MSME Fintech Credit Fund II obtained an investment-grade rating for its senior debt, supported by a capital structure in which 60% of the capital is subordinated through junior tranches that absorb losses first. A rating opens the door to mandates that otherwise cannot participate.
Balance sheet recycling. The West African Development Bank's XOF 160 billion multi-tranche securitisation of performing non-sovereign loans, structured in senior, mezzanine, and junior tranches, transferred portfolio risk to private investors, freeing capacity to lend again.
Country platforms. Singapore's Financing Asia's Transition Partnership shows the model operating outside Africa, at national scale.
Scale. Median transaction size rose from $38 million across 2020–2023 to $65 million in 2024, while deals above $1 billion accounted for 2.5% of activity and 27% of total financing mobilised. Below roughly $50 million, an institutional allocator cannot justify the diligence cost.
One constraint is worth naming: MIGA guarantees countries rated BB- and above, which excludes part of the frontier universe and helps explain why the least developed countries' share of blended finance fell in 2024 while middle-income flows rose.
📄 The expanded section on guarantees and a primer on blended finance structures are set out in the full 16-page report. Request your copy below.
What this means for allocators
Underwrite against loss history, not country rating. GEMs publishes thirty years of realised defaults and recoveries by region, sector and seniority. The sovereign ceiling is an assumption; this is evidence. Ask which one the position was sized against.
Ask which tranche the public money holds. A development institution on the capital table tells you little on its own. Junior means you are protected. Senior means you are lending alongside a subsidised lender.
Ask what is guaranteed, and for how long. A fifteen-year breach-of-contract cover on the off-take agreement addresses the risk that actually blocks the allocation.
Judge mobilisation, not participation. The question is not how much concessional capital a vehicle contains, but how much private capital would not be there without it. Ask for the number.
Closing the series
Four instalments ago we argued that sustainable finance measures the wrong thing.
The pattern held at every stage. The taxonomy certifies disclosure. The carbon market certifies an offset. ODA counted disbursement. The blended vehicle certifies the presence of concessional money.
None of them certifies that a transformer was installed, a grid was connected, or a household that had no electricity now has some.
That outcome remains measurable. It is simply not what the system was built to count.
The full Part IV analysis, including the expanded section on guarantees and a primer on blended finance structures, is available for download here.
📖 The complete series, available on this blog: Intro · Part I: The ESG Labelling Method · Part II: Carbon Markets in 2026 · Part III: Africa , Leading the Green Energy Transition · Part IV: Blended Finance
Sources: GEMs Risk Database Consortium, default and recovery statistics 1994–2024; Convergence, State of Blended Finance; ODI; OECD, Private Finance Mobilisation Report 2026, DAC Blended Finance Guidance 2025, tranche-level mobilisation analysis (DCD/DAC/STAT 2026) and The Role of Guarantees in Blended Finance; DFI Working Group on Blended Concessional Finance; MIGA and World Bank Group Guarantee Platform, 2025–2026; PIDG; BOAD; Lendable.
This article is published by Alfinas Alternative Investment Advisers for informational and educational purposes only. It does not constitute investment advice. © Alfinas Alternative Investment Advisers, October 2026.
Marie-Laure Mikkelsen PhD., C.A.I.A | Founding Partner, Alfinas Alternative Investment Advisers | info@alfinas.com | www.alfinas.com
📊 Bespoke Research Services: Alfinas works with private banks, family offices, professional and institutional investors who wish to share independent, institutional-grade research in a customised format.→ Investment Research & Writing → Product Research & Due Diligence → Strategy Analysis → Client NewslettersIndependent. Unconflicted. Institutional grade. 📧 info@alfinas.com | 🌐 www.alfinas.com
Comments