)
First-loss capital has struggled to scale in affordable housing for two reasons: much of the cost of building and monitoring a blended structure is fixed, so it weighs hardest on the small deals the sector is made of, and the commercial capital that concessional capital actually draws in is modest. Across 340 blended deals, mostly in developing markets and across all sectors, Convergence found each concessional dollar mobilised about $4.1 of commercial capital on average, only $1.8 of it private. A shared register of tranches, distributions and losses can lower the cost of operating and monitoring these structures; it does not remove the legal documentation, the valuation judgments or the regulatory perimeter.
The case for blended finance starts from the size of the gap. The European Commission's Affordable Housing Plan, adopted in December 2025, states that the EU needs more than two million homes a year to match current demand, which means adding about 650,000 homes a year to the roughly 1.6 million currently built, at a cost of about EUR 150 billion a year (European Commission, COM(2025) 1025, December 2025). Philanthropic and concessional capital cannot fill a gap of that size on its own. The logic of blended finance is to use it differently: not as the main source of funding, but as a protective layer that draws in far larger commercial flows.
How is first-loss capital supposed to work?
A foundation, public agency or development finance institution provides a tranche of capital that absorbs the first losses in a portfolio. Commercial investors sit above it, with priority on income and on recovery of principal. Because their downside is bounded by the layer beneath them, they can accept a lower return for the risk than the asset class would otherwise demand, and capital reaches a sector that it would otherwise avoid.
- Senior tranche commercial capital, paid first and protected by every layer beneath it
- First-loss tranche concessional capital that absorbs losses up to a defined threshold before anyone else is affected
- Rental income collected by the vehicle and paid out in a fixed order of priority
- The housing portfolio the affordable homes the structure exists to fund, held in a dedicated vehicle
Source: Structure defined in this article
How much protection is "enough" varies widely, and there is no reliable typical figure. In the five case studies examined by the Global Impact Investing Network, the first-loss protection ranged from a little above 5 percent to 37.5 percent of total capital, and the GIIN itself cautioned that "given the small sample size of just five cases, it is difficult to draw broad conclusions about drivers behind protection levels" (GIIN, Catalytic First-Loss Capital, October 2013).
Why does measured mobilisation fall short of the arithmetic?
The arithmetic of a first-loss structure suggests that a thin protective layer can support a senior tranche many times its size. The measured record is more modest. Convergence's data cover blended finance mostly in developing markets and across all sectors, not European housing, but they are the most systematic measure available. Across 340 blended transactions in its database, Convergence found that every dollar of concessional capital had mobilised on average $4.1 of commercially priced capital, "of which just under half ($1.8) on average has been sourced from private sector investors" (Convergence, February 2023). An earlier Convergence analysis of 72 blended funds found an average leverage of 4.0 but a median of 2.7, and the gap between the two suggests that a minority of high-leverage funds pulls the average up (Convergence, November 2018).
At the level of public development finance as a whole, the ratio is lower still. ODI found that each dollar invested by multilateral development banks and development finance institutions mobilised on average $0.75 of private finance in developing countries, falling to $0.37 in low-income countries (ODI, Blended finance in the poorest countries, April 2019). That is a broader measure than first-loss capital alone and should not be read as one, but it points the same way: the private money that public and concessional capital actually draws in is modest relative to the amounts committed.
Source: Convergence, Blended Finance and Leveraging Concessionality (February 2023); ODI, Blended finance in the poorest countries (April 2019). The two measure different things: Convergence counts all commercially priced capital alongside concessional tranches, ODI counts private finance mobilised by all multilateral and development finance
| Convergence, 340 blended transactions, commercially priced capital per concessional dollar | $4.1 |
| Of which from private investors | $1.8 |
| ODI, private finance per dollar of development finance, developing countries | $0.75 |
| ODI, the same measure in low-income countries | $0.37 |
The market that does close is real but small relative to the need. Convergence counted 123 blended finance deals reaching financial close in 2024, totalling $18 billion across all sectors and regions (Convergence, State of Blended Finance 2025, May 2025). Many take far longer, or never close: in its 2021 report, Convergence noted that "only 16% of the fundraising project transactions" captured in its previous report had reached financial close in 2020 (Convergence, State of Blended Finance 2021).
Where does the friction actually sit?
Both sides of the transaction meet obstacles that have little to do with the soundness of the concept.
The first-loss provider is usually an organisation built for grant-making rather than for managing a financial instrument. Once committed, it has to monitor the underlying portfolio, reconcile reports from several operators, confirm that the impact conditions attached to its capital are being met, and explain the position to its own board. Much of that effort follows from the number of relationships and reporting cycles rather than from the amount committed, which is why it weighs more heavily on a small tranche.
The commercial investor faces the mirror problem. Pension funds, insurers and impact funds, the natural holders of a senior tranche, look for deal sizes, track records, reporting standards and legal structures that most affordable housing operators, especially community-scale ones, cannot provide. The operators that can meet those standards have often already attracted conventional finance, which weakens the case that the blended structure is adding anything.
Underneath both sits the structuring burden. A small blended vehicle needs the same categories of legal work as a large one: tranche documentation, a loss waterfall, governance arrangements, an administrator, a reporting framework. The fees that can justify that work for a manager scale with the size of the vehicle; the work itself does not.
What does a shared, automated record change?
In a tokenized structure, the portfolio sits in a dedicated vehicle and the interests in it are recorded on a shared register, with the first-loss and senior positions represented as distinct classes. The order of priority that the legal documents define, income to the senior class first and losses to the first-loss class first, can then be executed automatically each time income arrives or a loss is recognised, and every permitted party sees the same state of every position at the same moment.
That changes the operating relationship between the two sides. A first-loss provider does not have to wait for a quarterly report to know how much of its tranche has been absorbed; it can read the position directly. A senior investor does not have to rely on the operator's account of the buffer protecting it; it can see the buffer's current size. Reconciliation between parties working from different spreadsheets largely disappears, because they are working from one record.
What it does not change matters as much. The tranche priorities, the definition of a loss, the valuation of the properties and the governance rights still have to be written into legal documents within a regulated structure; code executes what those documents define and cannot replace them. Whether a property is impaired, and by how much, remains a judgment made off the register and entered into it. A transferable class of interest is only transferable where a compliant venue exists to trade it, and early markets are thin. The cost of structuring falls where the structure is standardised and reused across many small vehicles, not because any single vehicle stops needing lawyers.
Does a shared record help a fund that discloses under SFDR?
It helps with part of the evidence, not with the label. A fund disclosing under Article 9 of the Sustainable Finance Disclosure Regulation must explain how its investments pursue a sustainable objective, which can be social, and how they avoid significant harm to other objectives (Regulation (EU) 2019/2088). In affordable housing, much of the supporting evidence has come from operators' own reports. A shared record lets a senior investor read part of it directly: income collected and distributed, the size of the first-loss buffer, and whether conditions written into the structure, such as a rent cap, were respected when each payment was made. Tenant eligibility and household outcomes remain attestations made off the record, and the regime itself is in flux: the European Commission proposed in November 2025 to replace the current structure with three voluntary product categories, and the proposal was still being negotiated in October 2026 (European Commission, November 2025; Arthur Cox, October 2026). The measurement side of this question is examined in detail in a companion piece on impact measurement.
What does the public side of the market look like in Europe?
Europe's public side has its own risk-absorbing instrument at scale. The InvestEU programme is built on an EU budget guarantee that absorbs part of the risk of eligible financing by partner institutions; its social infrastructure scope explicitly covers "social and student housing" (Regulation (EU) 2021/523). The guarantee was set at about EUR 26.2 billion and is expected to mobilise more than EUR 372 billion of additional investment; in December 2025 it was raised to about EUR 29.1 billion (Regulation (EU) 2025/2005).
Source: Regulation (EU) 2021/523, Article 4(1) and recital 40. The guarantee was raised to about €29.1B by Regulation (EU) 2025/2005 of 16 December 2025; the mobilisation estimate shown is the one stated for the original amount
| Item | Value |
|---|---|
| InvestEU EU budget guarantee, as adopted in 2021 | €26.2B |
| Investment the programme is expected to mobilise | more than €372B |
The European Investment Bank Group has been scaling its own housing lending alongside it: it aimed for more than EUR 4 billion in 2025 and is targeting EUR 6 billion in 2026 (EIB, December 2025). The Commission's plan also records that national and regional promotional banks aim to invest EUR 375 billion in social, affordable and sustainable housing by 2029 (European Commission, COM(2025) 1025). The large public layers of the market are being built. The gap that persists is at the small end, where community-scale operators and the foundations that back them meet the same structuring burden as everyone else.
The bottom line
Blended finance in affordable housing has not failed as an idea. It has underdelivered against its own arithmetic, because the measured ratio of commercial to concessional capital, on the most systematic data available, is well below what the structure's arithmetic suggests, and because the fixed cost of building, documenting and monitoring a blended structure falls hardest on the smallest deals. A shared, automated record attacks the second problem directly, by making the waterfall execute itself and every position visible to every permitted party. It does not touch the first problem on its own, and it does not dissolve the legal and regulatory work that a structure involving other people's capital requires. Foundations considering a first-loss role should treat shared records as a way to lower the cost of operating a well-designed structure, not as a substitute for designing it.
This article is part of DeReal Perspectives on real-estate access, liquidity, and tokenization. It is analysis of the market, not investment advice or an offer of any kind.
Sources
European Commission, European Affordable Housing Plan, COM(2025) 1025 (16 Dec 2025): data.consilium.europa.eu
Global Impact Investing Network, Catalytic First-Loss Capital (Oct 2013): thegiin.org
Convergence, Blended Finance and Leveraging Concessionality (Feb 2023): convergence.finance
Convergence, data brief on leverage in blended funds (Nov 2018): convergence.finance
Convergence, State of Blended Finance 2025 (May 2025): convergence.finance
Convergence, The State of Blended Finance 2021: convergence.finance
ODI, Attridge and Engen, Blended finance in the poorest countries: the need for a better approach (Apr 2019): odi.org
Regulation (EU) 2021/523 establishing the InvestEU Programme: eur-lex.europa.eu
Regulation (EU) 2025/2005 amending the InvestEU Programme (16 Dec 2025): eur-lex.europa.eu
Regulation (EU) 2019/2088 on sustainability-related disclosures in the financial services sector (SFDR): eur-lex.europa.eu
European Commission, Commission proposes improvements to the SFDR (Nov 2025): finance.ec.europa.eu
Arthur Cox, SFDR 2.0 update (16 Oct 2026): arthurcox.com
European Investment Bank, EIB Group doubles its financing to €6 billion for homes (Dec 2025): eib.org
Frequently asked questions
What is first-loss capital?
Capital that agrees to absorb the first losses in a portfolio or fund before other investors are affected, usually provided by a foundation, a public body or a development finance institution. Its purpose is to improve the risk-return profile for commercial investors so that more capital flows to an objective, such as affordable housing, than the concessional provider could fund alone.
How much commercial capital does first-loss capital actually mobilise?
Less than the arithmetic of a thin protective layer implies. Across 340 blended transactions in its database, mostly in developing markets and across all sectors, Convergence found in February 2023 that each dollar of concessional capital had mobilised on average $4.1 of commercially priced capital, of which only $1.8 came from private investors. ODI found in 2019 that each dollar of multilateral and development finance mobilised on average $0.75 of private finance in developing countries, a broader measure than first-loss alone.
How large is a typical first-loss tranche?
There is no reliable typical figure. In the five case studies examined by the Global Impact Investing Network in October 2013, the first-loss protection ranged from a little above 5 percent to 37.5 percent of total capital, and the GIIN cautioned that with only five cases it is difficult to draw broad conclusions about what drives protection levels.
How big is the affordable housing financing need in Europe?
The European Commission's Affordable Housing Plan of December 2025 states that the EU needs more than two million homes a year, about 650,000 more than the roughly 1.6 million currently built, and that delivering those extra homes would cost about EUR 150 billion a year.
Can a smart contract replace the legal structure of a blended finance vehicle?
No. The tranche priorities, loss definitions and governance still have to be set out in legal documents within a regulated structure. What a shared, automated record can change is execution and monitoring: distributions and loss allocations follow the documented order automatically, and every permitted party sees the same position at the same time instead of reconciling separate reports.