Scaling Securitisation in Development Finance

Acronyms

1. Executive Summary

Context

This MOBILIST research report assesses the potential for securitisation to be scaled in the development finance sector, presents case studies of pioneering transactions to date, and draws lessons on promising routes to scale and the roles that multilateral development banks (MDBs) and bilateral development finance institutions (DFIs) can play.

Securitisation is a financing technique in which a pool of financial assets (such as loans) or risk is packaged and transferred to a third-party. Investors in securities or derivatives backed by the asset pool receive cashflows generated by underlying portfolio. A basic true-sale securitisation process involves three steps as shown in Figure E1.

Securitisation offers a powerful, scalable route to mobilise private investment into emerging market and developing economies (EMDEs), by converting illiquid assets into familiar, tradable securities. Securitisation can generate cash, free up capital for new lending, and enhance risk management. When assets and risk pass from public sector MDBs and DFIs to the private sector, securitisation can be a major accelerator of private capital mobilisation for sustainable development.

Securitisation has two main variations: true-sale securitisation, in which the underlying loans are sold outright from the originator’s balance sheet to a third-party; and synthetic securitisation, in which a specific share of the credit risk is transferred to a third-party using credit derivatives, while the asset remains on the initial originator’s balance sheet.

Building Momentum

Pioneering transactions show that both true-sale and synthetic securitisation have the potential to increase annual direct lending volumes by development finance actors to EMDEs, without compromising credit ratings and while mobilising private capital. The transparency and standardisation of public markets mean that securitisations resulted in the issuance of listed debt instruments can offer particularly compelling scope for replication and private capital mobilisation.

Precedents in the market show that momentum is building: not only are development finance actors executing and proposing transactions at meaningful scale, but private sector supply of (and demand for) sustainable securitisations is growing. These private sector transactions set benchmark terms that development finance originators and investors should understand and align with if they are to see early deals scaled and replicated.

Yet, securitisation is not in itself a strategy. It is a technology, and like any other technology it creates opportunities and risks, the balance of which depends on one's objectives and constraints. Development finance actors must be clear what they are trying to achieve through securitisation, must understand the true-sale and synthetic structures available and the range of roles they can play in each, and must consider relative costs and risks. This paper provides a brief introduction to each aspect and draws lessons from pioneering transactions and from the dedicated professionals and champions that executed them. The paper appraises five models through which securitisation can be scaled in development finance against MOBILIST’s investment criteria: feasibility, commercial viability, additionality, scale, and replicability (see Box E1).

Routes to Scale

Securitisation in development finance can be scaled through at least five models (see Figure E2). Recent transactions highlight their relative feasibility, commercial viability, additionality, scalability, and replicability for development finance actors:

1. Model 1: Single development finance actor serves as both originator and issuer
2. Model 2: Assets from multiple development finance actors are pooled to back issuance by one member of the pool
3. Model 3: Assets from multiple development finance actors are pooled by an independent, specialist development finance securitisation platform
4. Model 4: Development finance assets are contributed to collateral pools including private sector assets and managed by fully private securitisation platforms
5. Model 5: Private sector assets are pooled and managed by private securitisation platforms, with development finance actors serving as investors, including to build new markets and asset classes

Development finance actors can play multiple roles in each of these models, acting as originators, investors in securitisation platforms, equity investors in individual transactions, guarantors of specific tranches, and debt investors in asset-backed notes. Yet very few MDBs and bilateral DFIs have the requisite volume of assets to sustain an active standalone securitisation programme. Therefore, over time MDBs/DFIs may need to move towards solutions that involve asset pooling and standardisation of transactions within the community and potentially with private sector counterparts.

While all five models are technologically feasible, no one of these models is superior in terms of commercial viability, additionality, scalability, and replicability. The optimal steady state is likely to include a mix of these models across true-sale and synthetic structures operating in parallel, offering originators and investors a range of tools to meet their diverse objectives and manage varied constraints. As the head of one pioneering institution puts it:

“When people ask how they can do my transaction, I tell them: ‘Don’t. Do your transaction.’”

Market Building

Tailoring to each institution’s objectives must not be at the expense of alignment with market standards. With each new transaction, the development finance community sets and reflects precedents in the market. To reach scale and trigger replication, future transactions should continue to respond to the high degree of standardisation on which securitisation thrives and align wherever possible with market norms. This will require collaboration with market participants, rating agencies, and regulators, to understand expectations and requirements. In doing so, the development finance community can transition from individual transactions, to programmes of repeated transactions, and ultimately to intentionally building markets for securitisation of development finance and impactful private assets in EMDEs.

High-level recommendations for the development finance community’s consideration include:

• Strategy: Transitioning now from a strategy focused on pioneering transactions, to establishing securitisation programmes, and to building securitisation markets, can help to accelerate private capital mobilisation and risk transfer. Proposal 1: Develop a unifying ambition in terms of dollars mobilised or transactions executed by 2030 to encourage a shift in strategy towards market-building and extended securitisation programmes.
• Operations and policy: Sharing operational experience, developing standardised/template documents, preparing harmonised loan data templates, and so forth within the development finance community will help to align future transactions with market standards and norms. Coordinated external operational engagement with rating agencies, investors, and (where relevant) regulators will build awareness of specificities of development finance assets and ensure that calls for reform and recalibration benefit from the community’s collective insight and advocacy. Proposal 2: Deepen collaboration between MDBs/DFIs to share lessons, data, and documentation, building on recent roundtables and convenings. Prepare shared messaging – even if high-level – for policymaking fora and identify other opportunities to reiterate and refine joint messaging on the securitisation agenda.
• Asset pooling: There are several routes to scaling securitisation in development finance, including potential to pool assets (i) among those institutions that do not have sufficient scale to sustain a programme of transactions independently and (ii) with private sector assets, to enhance scale and diversification. Proposal 3: Commission a joint feasibility and structuring study to develop and appraise the pooled structures identified in MOBILIST scoping and agree an action plan for feasible options.

2. Introduction

The G7 and G20 have underscored the urgent need for trillions of dollars in private capital to finance sustainable development, the energy transition, and green infrastructure, particularly in emerging market and developing economies (EMDEs). Securitisation offers a powerful, scalable route to mobilise private capital, by converting illiquid EMDE assets into tradable securities. This process transfers assets and/or risk from original lenders (‘originators’) to third parties (see Figure 3) and in doing so can generate cash, free up regulatory or economic capital (defined below), enhance risk management, and reduce concentration, enabling the originator to lend more. When these originating institutions or third parties are development finance actors, securitisation can increase direct lending and mobilise private capital for sustainable development in EMDEs.

This MOBILIST research paper assesses the potential for securitisation to be scaled in the development finance sector, presents case studies of pioneering transactions to date, and draws lessons on promising routes to scale and the roles that multilateral development banks (MDBs) and bilateral development finance institutions (DFIs) can play. Box 1 provides an introduction to the MOBILIST programme. The paper benefitted from insights and perspectives shared by the pioneers that championed and executed innovative securitisation transactions. These insights were shared through bilateral consultation and a series of roundtables on the topic hosted by the MOBILIST programme.

2.1 What is Securitisation?

2.1.1 Overview, process, actors

Securitisation is a financing technique in which a pool of financial assets (such as loans) or risk is packaged and transferred to a third-party. Investors in securities or derivatives backed by the asset pool receive cashflows generated by underlying portfolio. A basic true-sale securitisation process involves three steps as shown in Figure 3:

1. A company (often a bank or a group of banks) – the ‘originator(s)’ – identifies the assets to transfer from its balance sheet and pools them into a collateral portfolio.
2. The portfolio is sold to the ‘issuer’, often a Special Purpose Vehicle (SPV) created solely for the transaction in question. The originator can recycle the cash and capital freed-up to finance new projects.
3. The issuer finances the acquisition of the underlying assets by selling tradable, interest-bearing securities to capital market investors. The holders of these asset-backed notes receive interest and principal payments funded by the underlying portfolio.

A hallmark of securitisation is the creation of tranches – portions of the transaction with different risk-return profiles. The underlying asset pool’s cashflows are allocated to investors based on a waterfall structure: senior tranches have first priority on payments from the pool and are shielded by subordinate mezzanine and equity tranches that absorb losses first. This structuring benefits the issuer by allowing the senior tranches to attain high credit ratings, even if the underlying loans are lower-rated. It also benefits investors, whose risk-return preferences can be matched to the varied risk profiles of tranches in the structure. Figure 4 shows this tranching in the context of a true-sale transaction in which the MOBILIST programme invested in 2023.

Successfully transforming a pool of assets into securitised instruments that meet investor preferences and regulatory standards requires coordination between several actors. The key players involved in a basic true-sale transaction are summarised in Table 1. This list underscores the central role played by credit rating agencies in assigning independent risk scores to each tranche of the structure. It is the comparability of these ratings across tranches within a securitisation, and the consistency of similar ratings for similar risk across transactions, that have enabled the securitisation market to scale and transaction costs to be minimised. In turn, comparable ratings are enabled by highly standardised structures, legal documentation, and marketing materials in mainstream securitisation markets. Perhaps more than anything, successfully and sustainably scaling securitisation in development finance will depend on the sector’s ability to align with this degree of standardisation expected in the market.

2.1.2 Types of securitisation: True-sale and synthetic

Securitisation has two main variations: true-sale securitisation, in which the underlying loans are sold outright from the originator’s balance sheet to a third-party; and synthetic securitisation, in which credit risk is transferred to a third-party using credit derivatives or guarantees, while the asset remains on the initial originator’s balance sheet. The full taxonomy of securitised products is shown in Figure 5.

True-sale securitisation, for example through collateralised loan obligation (CLO) structures, has existed for decades. Early applications emerged in the mortgage market and over time the same structure has been used to securitise anything from consumer credit to corporate loans to utility receivables and project finance. This structure allows originators to sell credit risk from their balance sheets and ultimately to a wide range of institutional investors familiar with interest-bearing, rated securities backed by the underlying asset pool in ways described in Figure 1. True-sale securitisation is a standard tool used by private sector banks and other originators to generate cash, reduce risk, expand lending capacity, and to generate attractive returns by offering investors exposure to loan portfolios that they would otherwise find difficult to access.

Synthetic securitisation enables the issuer to transfer credit risk associated with a ‘reference portfolio’ of loans on its balance sheet, by purchasing guarantee cover or selling the right to loan principal and interest to investors through credit derivatives, like credit default swaps (CDS) or through the issuance of credit-linked notes (CLNs). Issuance of tranched credit-linked notes can broaden the appeal to a more diverse set of institutional and capital market investors, and to investors that would be unwilling or unable to offer credit insurance directly. Whether via CDS or CLNs, while synthetic securitisation can enhance risk management, capital adequacy, and so an originator’s credit rating, it does not generate cash that can be deployed into new loans.

Synthetic securitisation has proven particularly attractive for private sector institutions since the introduction Basel III regulations that required banks to hold more capital in reserve for some assets than historic default and recovery rates would suggest was needed. Regulatory acceptance of synthetic securitisation since 2020/21 has seen uptake accelerate, with underlying assets tending to be corporate and SME loans. Much like development finance assets, these loans are often subject to confidentiality and transfer restrictions, which can complicate true-sale securitisation.

2.2 The potential role(s) of securitisation in development finance

MDBs and DFIs have been encouraged by their shareholders to use risk transfer of the kinds just described as a tool in Balance Sheet Optimisation (BSO – see Box 2), shifting their business model from an originate-to-hold to an originate-to-share or originate-to-distribute model. The 2015 Toronto G20 declaration recommended that MDBs consider securitisation as a means of sharing risk on their Non-Sovereign Obligor (NSO) portfolios. In 2022, the G20 Capital Adequacy Framework Panel Report recommended that MDBs use securitisation for transferring both NSO and Sovereign Obligor (SO) risk.

In parallel, MDBs and DFIs have been encouraged to use their scarce resources to be more catalytic in accelerating private capital mobilisation (PCM) to EMDEs, rather than focusing solely on lending on their own account. This emphasis on PCM reflects the relative magnitude of institutional investor firepower relative to official development finance.

While they do not face regulatory capital requirements, MDB-lending is constrained by these institutions’ strategic decision to protect currently high credit ratings. World Bank Group lending arms and major regional MDBs are all AAA-rated by the three global credit rating agencies. These ratings are at the core of the traditional MDB business model, which involves borrowing at very low rates afforded by their AAA status and on-lending to EMDE sovereigns at modest spreads. This model allows EMDE sovereigns to borrow from MDBs at much lower interest rates than they would secure from commercial lenders. The traditional business model is broadly sustainable for the MDBs because sovereign borrowers accord de facto seniority to MDB debt over commercial debt in the form of Preferred Creditor Treatment (PCT), leading to lower default rates and higher recovery rates than EMDE sovereigns’ standalone credit ratings would imply.

In the context of this business model, securitisation can play an important role in accelerating MDB lending, by generating cash, enhancing risk management, and strengthening credit rating metrics. Credit rating agencies evaluate MDBs in part by comparing equity resources to risk-weighted assets (S&P and Fitch) or by assessing leverage ratios and weighted average ratings of an MDB’s loan portfolio (Moody’s) (see Table 2). They also consider concentration risk, for example in terms of the ratio of the five largest exposures to the total banking exposure; and liquidity and funding, for example in terms of liquid assets relative to net cash outflows. By transferring risk and/or assets to third parties through securitisation, MDBs can improve these rating metrics by improving capital adequacy, enhancing liquidity, and reducing concentration risk. In turn, securitisation can allow MDBs to provide new loans in EMDEs without capital increases while protecting their credit rating. When considering BSO through securitisation, an MDB compares the potential gains it will obtain in these rating metrics to the cost of the risk transfer, expressed in forgone spread on underlying assets or insurance premiums.

In addition to accelerating direct MDB and DFI lending through enhanced balance sheet management, securitisation can accelerate PCM in several ways. First, securitisation can contribute directly to PCM to the extent that risk and/or assets are transferred to private sector actors. Recent recommendations encourage consideration of this form of mobilisation in official PCM statistics. Second, PCM could be even greater when an MDB or DFI makes possible a securitisation transaction that combines their assets and assets originated by private institutions. Third, MDBs and DFIs can act as co-investors with private sector actors in asset-backed notes, in the equity tranche, or in risk transfer, with anchor and higher-risk positions likely to prove particularly catalytic. Perhaps most scalable of all would be PCM strategies and transactions that saw MDBs and DFIs help to build a market for securitisation of asset classes with development impact.

2.3 Costs and risks associated with securitisation in development finance

The 2007-08 financial crisis highlighted significant real and perceived risks associated with securitisation. These, and several unique risks facing development finance sector, must remain front and centre as debate around the role of securitisation in development finance evolves.

Box 3 highlights four categories of risk facing investors in securitisation transactions, including credit, counterparty, legal, and market risks that differ markedly from risks associated with traditional lending. As originators, development finance actors also face additional risks to project impact, particularly in true-sale transactions where a third-party is responsible for loan servicing. Finally, securitisation has been associated (particularly since the 2007-08 crisis) with systemic risk in the financial sector, particularly given risks of (i) potential misalignment of incentives between originator, sponsor, and investor; and (ii) the potential for synthetic risk transfer to amplify the impact of losses on underlying assets.

Misalignment of incentives is possible because of imperfect information:

1. Adverse Selection – If originators have better risk information than sponsors and investors, they could be incentivised to offload their worst-performing assets; and
2. Moral Hazard – Once risk or assets have been transferred, the originator may act less diligently when collecting principal and interest, enforcing covenants, and collecting recoveries.

Variations of these risks have underpinned significant regulatory reform since the 2007-08 financial crisis. For example, European CLO regulations stipulate that CLO managers must retain a minimum of 5% of the highest-risk (equity) tranche, to ensure that their interests align with those of other investors in the structure. In one of the case studies discussed below, the sponsor retained a meaningful equity position, while originating banks were required to hold at least 30% of their original exposure. This two-tiered retention structure aligned interests of both the sponsor and the originators with those of the investors.

Their development impact mandate may also create a misalignment of incentives between MDBs/DFIs and private sector actors. For example, development finance actors may be willing to take higher risks or to be more accommodating to borrowers when enforcing covenants and pursuing recoveries than private sector originators. Despite this theoretical risk, the relatively high-quality nature of their assets and track record in recoveries is demonstrated in recent data releases, which shows MDB and DFI assets performing multiple notches above their similarly rated private peers.

More broadly, securitisation can contribute to systemic risk. The 2007-08 global financial crisis highlighted how complex securitised products, like collateralized debt obligations (CDOs), can obscure underlying asset quality and spread risk throughout the financial system. Synthetic securitisations can amplify these risks by creating layers of exposure that are difficult to assess. In MDB and DFI securitisations where the total loan pool is limited, there is the potential that this complexity can lead to a lack of transparency and increased interconnectedness among financial institutions, potentially exacerbating financial instability during periods of market stress.

In addition to these distinct risks, securitisation creates material costs against which any strategic and operational benefits should be assessed. For example, if improving access to capital is an MDB/DFI’s objective, securitisation should be compared with other available options such as pursuing capital injections from shareholders or considering hybrid capital transactions. While securitisation can free up capital, it may lead to higher funding costs relative to MDB and DFI senior debt and may involve higher transaction costs, including legal fees, rating fees, arranger fees, and potential discounts on asset sales, which can erode the financial benefits. Additionally, if the securitised assets underperform, MDBs and DFIs may need to offer higher yields to attract investors in future issuances, further increasing funding costs. In synthetic transactions, any capital relief is also dynamic as it changes with the quality of the underlying assets and in turn with the amortisation of the reference portfolio. In this context, the cost-benefit advantages of a synthetic securitisation can change over its lifetime.

3. Securitisation in Development Finance and the Global Markets

The total market for asset-backed securities reached US$ 6.9 trillion in 2024 and is projected to grow to US$11.7 trillion by 2034. In the development finance sector, pioneering deals have overcome initial technological barriers and expanded market awareness, creating an opportunity to transform one-off large transactions into an established asset class that benefits future originators and investors. Investor interest in structured products with a sustainability focus outside the development finance sector is also growing, supported by greater familiarity with labelled CLOs.

Recent feedback from investors, rating agencies, and legal advisers suggests that future deals will be particularly attractive and scalable if participants standardise documentation, processes, and disclosures. The opportunity is potentially transformational: accelerating capital deployment in the development finance sector and presenting investible development finance assets to private sector structured credit investors in a familiar format.

This section discusses the market context and recent innovations in the development finance sector that demonstrate the scale of the opportunity presented by securitisation structures. Following a brief overview of growing and somewhat divergent securitisation markets in the US and Europe, this section presents five models through which securitisation could be scaled in the development finance sector. Each model is described drawing on case studies in the market where they exist and qualitatively appraised against MOBILIST’s investment criteria.

3.1 Market Context

Markets for both true-sale and synthetic securitisation have grown significantly over the past decade. True-sale CLOs have seen a substantial increase in issuance, particularly in the United States. The market grew from $308 billion in 2015 to $617 billion in 2024, ultimately comprising 38% of the U.S. ABS market. The European CLO market has also expanded, with issuance expected to remain high at €135 billion in 2025. This growth is driven by a broadening base of originators and sponsors, improved outlooks for underlying lending, and increased market engagement from bank originators motivated by funding and risk management requirements.

Investors have been drawn to CLOs for their potential to enhance income in a low-interest rate environment, without significantly increasing default risk. Additionally, the introduction of CLO-focused Exchange-Traded Funds (ETFs) has made these instruments more accessible to a broader range of investors. In Europe the growing interest in green securitisation (see Box 4) alongside the global demand for geographically diversified issuers provide potentially strong tailwinds for MDB and DFI securitisations.

In recent years, particularly from 2023 through early 2025, the cost of financing CLOs in Europe has experienced significant changes, driven primarily by investor demand, loan supply dynamics, and market stability. Initially, strong investor appetite and ample loan supply pushed CLO issuance in Europe to record levels. Between 2023 and 2024, financing costs notably declined, with AAA tranche spreads falling by 43 basis points, from 172 bps to approximately 129 bps, and the overall CLO weighted average cost of capital (WACC) decreasing by around 70 bps, down to 215 bps. Typically, these reductions would strengthen arbitrage opportunities between CLO loan yields and their financing costs.

By 2025, however, market volatility caused CLO tranche costs to decline further, dropping below 200 bps for the first time in three years. Specifically, CLO WACC reached a new low of 182 bps in February 2025, although AAA tranche costs only modestly improved by 4 bps to 123 bps. Concurrently, new loan spreads tightened to 356 bps, significantly narrowing the arbitrage opportunities. Consequently, the full arbitrage (new-loan discount margins minus CLO WACC) fell to 174 bps, with coupon arbitrage similarly compressing to 171 bps, far below previous years (450 bps in 2022 and 350 bps in 2023). This reduction in arbitrage to historically low levels implies limited CLO issuance volumes in the near future, indicating potential constraints compared to recent years.

Synthetic securitisations, particularly through Significant Risk Transfers (SRTs), have also gained popularity as banks seek to manage credit risk and optimise capital requirements following the implementation of Basel III standards. Since 2016, over $1.1 trillion in assets have been synthetically securitised globally, with Europe accounting for nearly two-thirds of this activity.

3.2 Securitisation Pioneers in Development Finance

3.2.1 Synthetic Securitisation by AfDB and IDB Invest

Singly originated, in-house synthetic securitisations by large MDBs have been the most common format undertaken by MDBs and DFIs to date. There have been at least three case studies from: the African Development Bank’s (AfDB) Room-to-Run NSO and sovereign transactions and IDB Invest’s Scaling4Impact. These have typically been bespoke transactions, tailored to the specificities of the originator’s balance sheet and BSO objectives.

AfDB pioneered the use of synthetic securitisation of MDB assets, with its two Room-to-Run (R2R) transactions. In 2018, AfDB entered in a US$1 billion synthetic securitisation for NSO exposures. Then, in 2022, AfDB implemented a second US$2 billion synthetic securitisation of sovereign exposures. More recently, IDB Invest completed a synthetic securitisation of NSO exposures in 2024. A comparison of the transaction structures is shown in Figure 7.

The AfDB was motived in the R2R transactions by a strong incentive to respond to the 2015 G20 Action Plan for MDB balance sheet optimisation, balancing the objective to maintain the MDB’s AAA credit ratings with the ambition to increase lending. The initial R2R transaction enhanced AfDB’s rating agency-assessed capital adequacy, thereby creating room to originate new loans to the private sector. Similar to commercial bank synthetic securitisations that free regulatory capital, MDB synthetics free economic capital linked to rating methodology (see Table 3 and Box 5). AfDB’s objectives also included diminishing earnings volatility arising from the increased complexity of models relating to IFRS 9 provisioning requirements and ensuring that, by retaining a portion of the risk, it provided investors with comfort that incentives remained aligned.

R2R’s design balanced rating agency requirements, investor pricing demands, and AfDB’s need to limit costs. The junior mezzanine investors, Mariner and Africa50, demanded yields in excess of 10%. The amount of private capital mobilised in the structure is the thickness of the mezzanine tranche as the rest of the structure was retained by AfDB, and therefore a thicker mezzanine tranche equates to more private capital mobilisation. AfDB was constrained in the thickness of the mezzanine by balancing the cost of the funded and unfunded guarantees with the capital unlocked by S&P’s risk weight adjustment. Beyond a 10% thickness, the additional risk-adjusted cost of capital protected would have been greater than the cashflows from the tranche’s collateral or the capital unlocked. By contrast the senior mezzanine tranche was cheaper, benefiting from the European Commission (EC) guarantee.

The requirement to align incentives between AfDB and other parties to the transaction was largely satisfied through the Bank’s retention of the junior and senior tranches of the structure. For the equity portion, the AfDB retained a thin 2% tranche, which was below the expected loss on the portfolio (2.73%), but avoided higher retention that would have inflated its risk weights. The tranche thicknesses ultimately allowed the senior retained portion to obtain an A-equivalent rating, which significantly contributed to the US$ 650m in capital relief under S&P’s methodology.

Building on the success of its inaugural transaction, in 2022 AfDB synthetically securitised $2 billion of sovereign loans to 11 African governments. The UK’s Foreign Commonwealth and Development Office (FCDO) took the senior position and three private insurers from the Lloyd’s market the junior position. The transaction at inception improved the Bank’s Risk Capital Utilisation Rate by 5.1% and strengthened S&P’s RAC by 1.5%. In totality, the AfDB estimated that this created $2 billion in additional capital headroom.

The sovereign and non-sovereign R2R transactions differ markedly in how the rating agency treated the initial RWA, the loan margins, and investor appetite. First, under S&P’s methodology sovereign exposures are treated as less risky than NSO, therefore portfolios start at a lower risk weight relative to NSO. This means that the RWA reductions which the sovereign transaction could achieve depended heavily on guaranteeing the senior portion, so there was a greater difference between the pre-securitisation and post-securitisation RWA and so the amount of capital unlocked was meaningful. Second for junior positions, sovereign loans have tighter margins than NSO so any protection cost, even if lower in percentage terms, can still represent a substantial share of the total margin limiting the thickness of the junior positions. Finally, because of the lower margin, the cost of protection on the senior tranche needs to be kept low to maintain the transaction’s cost-benefit ratio. Therefore, sovereigns or multilateral guarantee agencies that can satisfy counterparty risk requirements and price guarantee fees concessionally may be required in these transactions. In R2R sovereign, the credit protection from FCDO as a low-risk counterparty was both low cost and reduced the AfDB’s RWA on the senior tranche, drastically improving the transaction’s cost-benefit ratio.

Building from AfDB’s precedents, in 2024 IDB Invest identified $1 billion in assets out of a $10 billion portfolio, composed of vanilla products rated B or higher across 10 sectors and 17 countries. As shown in Table 4, there are significant similarities in terms and outcome between R2R and IDB Invest transactions. Key differences between R2R and Scaling4Impact were:

• IDB Invest’s collateral pool was more highly rated, at BB- compared to AfDB’s at B+ for the original R2R NSO transaction.
• This difference in risk profile of the reference portfolio meant that IDB Invest was able to employ a thicker senior tranche and thinner mezzanine tranches.
• IDB Invest’s transaction attracted almost exclusively private capital.

This comparison underscores how originators with relatively high-quality corporate exposures (like IDB Invest) can construct reference portfolios with relatively low PLRs. A lower PLR means that the senior tranche can achieve a higher rating (and so enhance economic capital) with more limited mezzanine protection. From a BSO perspective, a thicker mezzanine tranche than is necessary to achieve the intended economic capital or rating uplift would not be cost-effective, as fees or foregone spread would unnecessarily consume a larger portion of the transaction’s benefit for the originator. However, from a PCM perspective, every dollar of risk transferred to private holders is a dollar of private capital mobilised into EMDEs. This highlights something of a potential tension between MDBs’ strategic BSO and PCM objectives that must be systematically resolved based on the relative importance of these objectives and the relative benefits, costs, and risks involved.

3.2.2 Hybrid and True-Sale transactions by BOAD

In line with its strategic vision (known as the "DJOLIBA" plan for 2021-2025), the Banque Ouest Africaine de Développement (BOAD) initiated one of the first MDB securitisation programmes to enhance its financing capacity for development projects across the West African Economic and Monetary Union (UEMOA) region. BOAD views securitisation issuance as a long-term strategy: (i) to diversify from their traditional source of funding, (ii) to increase its funding footprint by 50% by moving BOAD through an originate-to-distribute model, and (iii) to develop West African capital markets in the process. Ultimately this securitisation programme has facilitated long-term resource mobilisation for BOAD, mobilisation of private capital into development finance assets, and diversification of the regional capital market instruments by offering a new, AAA-rated, fixed-income instrument to UEMOA investors.

The focal point of this initiative is the creation of a “Fonds Commun de Titrisation de Créances” (FCTC), a common securitisation fund structured under UEMOA’s regulatory framework, which was specifically developed for the first issuance under BOAD’s programme. BOAD’s hybrid securitisation is structurally distinct from AfDB’s R2R, and IDB Invest’s Scaling4Impact transactions. As shown in Figure 8 the transaction had features of both a synthetic securitisation and a true-sale CLO:

• Synthetic – BOAD was not comfortable selling their loans fully off-balance sheet and needed to be the lender of record to retain PCS for a credit uplift on the portfolio. Therefore, the principal and interest were synthetically sold to the SPV using pass-through notes linked to the underlying collateral. These are similar to a form of credit default swap.
• True-sale – The SPV then issued securities which were listed, similarly to a CLO.

The full structure is shown in Figure 8 and in several ways mirrors proposals for Sustainable Development Certificates developed by Eighteen East in 2020.

With its inaugural 2023 transaction, BOAD sought to generate capital relief and liquidity while demonstrating a proof of concept. BOAD generated capital relief by freeing up RWA under Moody’s and Fitch’s methodologies, as with the synthetic transactions, and generated immediate cash from a portion of its existing sovereign loan portfolio by selling pass-through notes. The transaction served as a pilot, proving that high-quality sovereign loans could be securitised within the regulatory frameworks of the WAEMU region and that investors in the region had appetite for such assets.

The collateral pool was formed from 24 of BOAD’s sovereign loans from Senegal, Ivory Coast, Benin, Togo, and Niger. Only Sovereign loans were included as these are considered safer and more transparent, thereby giving the market confidence for the proof-of-concept. A pass-through note structure transferred credit risk without assigning loan ownership, thus preserving confidentiality around sovereign lending and maintaining BOAD as lender of record, so as to maintain the 3-notch credit uplift afforded by PCT. The pass-through notes were structured on a loan-by-loan basis, allowing any potential issues to be contained to a single underlying loan.

The first structure saw a 99% senior tranche (rated AAA locally) transferred to the market and a 1% junior tranche retained by BOAD, raising XOF 150 billion. This tranching structure was rated by Global Credit Rating West Africa, an affiliate of Moody’s. The international market rating equivalent would have been BBB. To further protect the junior tranche, a 5% reserve, funded by the spread between the total funds raised (XOF 150 billion) and the amount sold to investors (XOF 143 billion), provided additional protection for the junior tranche. Furthermore, the transactions opened a short-term liquidity facility from NSIA Côte d’Ivoire to cover timing mismatches, and absorb temporary liquidity shortfalls and credit losses. Overcollateralization was discarded due to IFRS constraints, making the cash reserve a central enhancement mechanism.

Although it was the first securitisation for BOAD, existing regulation was sufficient, and no major regulatory overhaul was required. The initial 12-month setup period focused heavily on harmonising internal approvals and aligning with IFRS, particularly regarding over-collateralisation rules.

Ultimately the transaction closed within a few weeks, demonstrating strong appetite, mostly from regional banks, pension funds, insurance companies, and some retail investors. There was also some international appeal with roughly 5% of interest came from international investors, despite the local currency denomination.

For their second transaction BOAD wanted to move beyond pure de-risking to actively improve the quality and diversification of BOAD’s portfolio, while ensuring strong performance. There was a continued focus on freeing up additional rating capital by reducing RWAs, allowing BOAD to redeploy funds into priority sectors such as infrastructure, housing, and energy. Moreover, BOAD aimed to improve the regulatory sophistication, incorporating aspects of European regulatory frameworks to attract a broader set of global investors.

DOLI-P 2024 was a XOF 160bn transaction with three tranches: 75% senior, 18% mezzanine, 7% junior. The junior was retained by BOAD to demonstrate shared interest with investors. There were two key differences from the first transaction:

• The collateral pool contained private and SOE loans across infrastructure, energy, and housing, rather than strictly sovereign assets;
• The transaction introduced for the first time in the region a mezzanine tranche through a structure entirely aligned to international standards, both of which were to capture more international interest.

While the first transaction closed within weeks, DOLI-P 2024 closed in just 3.5 hours, illustrating a clear surge in demand and the success of BOAD’s long-term market building approach. Table 5 provides a comparison of the 2023 and 2024 transactions.

3.2.3 IFC-MOBILIST Emerging Market Securitisation Programme

The Emerging Markets Securitisation Program (EMSP) was launched to tackle one of the toughest challenges in development finance: mobilizing private capital at scale in emerging markets. Public resources alone cannot meet the financing gap, and global capital still flows mainly to advanced economies even though approximately 85% of the world’s people live in emerging markets.

The program is directly aligned with the UN’s Sustainable Development Goals (SDGs) and responds to the G20 Capital Adequacy Framework (CAF) Panel’s recommendations for multilateral development banks to recycle their balance sheets. It is also consistent with the World Bank Group Evolution Roadmap. At its core, EMSP is about transforming illiquid bilateral private loans into standardized securities that global institutional investors can access in a format they know. By doing so, IFC can recycle its capital to support more projects, while giving investors a familiar, market-standard way to invest in emerging market credit in line with their risk-return preferences.

MOBILIST’s contribution was critical to getting the inaugural EMSP transaction off the ground. As a co-equity investor alongside IFC, MOBILIST took part on the riskiest tranche of the structure, absorbing potential losses and giving private sector investors the confidence to participate. Without this catalytic equity tranche, the transaction would not have been possible.

What makes the IFC’s EMS 2025-1 transaction unique?

The first transaction, EMS 2025-1, raised $510m by pooling IFC loans to 57 companies diversified across sectors and regions. The portfolio spanned Latin America and the Caribbean (33%), Africa (24%), Europe and Central Asia (15%), South Asia (14%), East Asia and Pacific (8%), and Middle East and North Africa (6%). More than 30% of the portfolio was exposed to low-income countries supported by the World Bank’s International Development Association (IDA), or fragile and conflict-affected states (FCS). Sector exposure was broad, with over 85% of portfolio allocated to manufacturing, agribusiness and services, and infrastructure, and the remainder in financial institutions. It brought together a breadth of emerging market credits that are rarely available to investors in a single, standardized instrument.

The securitization was structured with senior notes of $320m that were sold to private investors and rated Aaa by Moody’s, supported by a rigorous review and mapping of IFC’s internal credit processes to market standards. The mezzanine tranche of $130m was insured by a consortium of insurance companies. The $60m equity tranche was co-invested by IFC and MOBILIST. MOBILIST’s anchor role in the equity tranche was a central element that helped make the transaction possible, creating the first-loss buffer that gave confidence to the rest of the capital stack. This equity enhances the risk profile of the issued notes making them investment grade and hence enables institutional investors to participate in the transaction.

The listing of the senior notes on the London Stock Exchange enhanced visibility and has the potential to broaden the investor base over time.

Impact and demonstration effect

The EMSP shows that emerging market loans can be transformed into investable securities that resemble familiar products for investors. This is not about reinventing finance; it is about adapting proven structures to open new opportunities in emerging markets that can be scaled up. IFC’s and MOBILIST’s co-equity investment sent a strong signal: that development-focused capital can play a catalytic role in making innovative structures work for the private sector. By standing at the bottom of the capital stack, IFC and MOBILIST helped create space for institutional investors, particularly asset managers and insurance companies, to come in at the top and mezzanine tranches. That balance is what made the deal successful.

The broader impact goes beyond a single issuance. By offering investors exposure to a diversified portfolio of emerging market credits across sectors and regions, EMSP challenges assumptions about the risk profile of these markets. Emerging markets credit, when structured and managed properly, can be more resilient than many assume. And when combined with developed market credit portfolios, it can diversify risk and improve the overall risk-return profile of such portfolio. By proving this point in practice, EMSP has created a template for how to scale institutional capital flows into markets where financing gaps are deepest.

Looking ahead

This first issuance is just the beginning. Over time, IFC aims to build a regular cadence of transactions, with the ambition of establishing benchmarks and deepening liquidity. Other issuers may follow suit. The long-term vision is to build a new asset class of emerging market loans that institutional managers can access with confidence, just as they do in developed markets.

The program is also designed to be replicable, paving the way for other multilateral development banks and development finance institutions to adopt similar models. If replicated widely, the impact could be transformative: mobilizing large-scale private capital into emerging markets, supporting jobs, growth, and resilience where they are needed most. At the heart of this effort, MOBILIST’s role working with IFC in its inaugural transaction stands out. By stepping in as a co-equity investor, MOBILIST demonstrated how catalytic investors can unlock markets for others to follow.

3.2.4 MDB/DFI participation in Bayfront

Rather than developing and executing their own securitisation transactions and programmes, development finance actors could collaborate with partners already active in the public and private securitisation markets. Such a collaboration could entail MDBs and DFIs pooling assets alongside private sector originators into portfolios managed by private sector sponsors, and/or MDBs and DFIs acting as investors, lenders, or guarantors to enable securitisation of impactful assets.

Bayfront Infrastructure Management (Bayfront) exemplifies this approach. Bayfront is a securitisation platform formed in 2019 as a joint venture between Clifford Capital Holdings (70%) and the Asian Infrastructure Investment Bank (AIIB) (30%) to develop infrastructure financing channels through capital markets. In July 2019, the AIIB invested $54 million in Bayfront’s equity. Debt instruments issued by Bayfront to acquire and warehouse loans from banks benefits from a guarantee provided by the Government of Singapore. This established a dedicated structure to originate, securitise, and distribute infrastructure-backed debt to institutional investors, thereby expanding lending capacity for originating banks and enhancing private capital mobilisation.

Bayfront acquires infrastructure and project finance loans from commercial originating banks, pools them, and then issues infrastructure asset-backed securities (IABS) as bonds or notes representing claims on the underlying cash flows. Since inception, Bayfront has launched six IABS issuances (BIC 2018, BIC II 2021, BIC III 2022, BIC IV 2023, BIC V 2024, BIC VI 2025), each backed by a diversified portfolio of infrastructure loans across Asia and the Middle East. Beyond its equity participation in Bayfront itself, AIIB acted as an anchor investor in issuances following the first. In BIC IV, the UK Government through its MOBILIST programme acted as an anchor investor by purchasing US$ 5m of the equity tranche. Additionally, GuarantCo provided a guarantee for the mezzanine (Class D) tranche, which was fully subscribed by funds managed by Apollo Global Management. This approach supports deeper liquidity, fosters investor confidence, and enhances the profile of infrastructure debt as an investable asset class in emerging markets.

AIIB’s involvement in the project was primarily motived by the desire to contribute to a track record of primary IABS issuances and provide a valuable Asian market benchmark, which are critical elements of the market- building process for any new asset class. The growth of this market will lead to virtuous cycles on (i) the supply side, whereby banking partners can rely on Bayfront as a take-out partner for qualifying project finance and infrastructure loans, leading to greater origination appetite, and (ii) the demand side, as the increased deal flow will help establish market benchmarks that enhance institutional investors’ familiarity with and demand for Asian infrastructure assets, ultimately addressing Asia’s infrastructure financing gap.

3.3 Scalable Securitisation Models

Pioneering transactions and a wider review suggest that MDBs and DFIs have at least five options when seeking to scale securitisation of impactful assets, ranging from in-house solutions to selling loans into commercial platforms (see Figure 8):

• Model 1 – Single development finance actor serves as both originator and issuer
• Model 2 – Assets from multiple development finance actors are pooled to back issuance by one member of the pool
• Model 3 – Assets from multiple development finance actors are pooled by an independent, specialist development finance securitisation platform
• Model 4 – Development finance assets are contributed to collateral pools managed by fully private securitisation platforms
• Model 5 – Private sector assets are pooled and managed by private securitisation platforms, with development finance actors serving as investors only

Figure 8 also highlights the range of roles that MDBs, DFIs, and donors can play in securitisation transactions, including as originators, equity investors at the platform or SPV level, debt investors in asset-backed notes, and guarantors or credit insurance providers.

The following section appraises these options in terms of MOBILIST’s investment criteria:

• Feasibility: Is securitisation of target assets technically feasible, particularly given the often bespoke nature of underlying contracts? Can incentive structures be aligned within the model for long-term feasibility?
• Commercial Viability: Does the model offer requisite risk-adjusted returns and diversification to achieve competitive pricing on asset-backed notes and adequate returns for equity investors? (see Box 6)
• Additionality: Does the involvement of development finance actors facilitate impactful transactions that otherwise would not have occurred, either by mobilising private capital or relieving capital constraints/generating cash to enable new development finance projects?
• Scale: Does the model enable access to a pool of assets of requisite scale to sustain a growing programme of securitisations?
• Replicability: Does the model enable replication of comparable transactions in the market by peers and ideally with less or no development finance involvement over time?

4. Comparative Analysis of MDB/DFI Securitisation Models

This section synthesises learnings from the models and case studies above in terms of MOBILIST’s investment criteria: Feasibility, Commercial Viability, Additionality, Scale, and Replicability. Precedent transactions collectively clarify how each model navigates technical complexities, market constraints, demonstration effects, and strategic trade-offs. Ultimately, the choice of model will depend on both the specific objectives of the MDB or DFI, including both BSO and strategic PCM considerations. Several of the key objectives highlighted by those involved in prior transactions and potential priorities going forward are outlined in Table 6.

4.1 Feasibility

The five models described in the previous section face varied technical, legal, and structuring hurdles. For true-sale transactions, the central challenge is the bespoke nature of MDB and DFI loan contracts, potential constraints on transferability, and the need to align documentation with rating agency and investor requirements. For both true-sale and synthetic structures, development finance originators have also articulated internal concerns over data sharing with third parties. MDBs and DFIs are used to conducting detailed due diligence but are much less used to being subject to such exercises, particularly at the level of underlying assets.

A broader feasibility consideration across all models is institutional capacity and knowledge: securitisation is complex and non-traditional skill set for most MDBs and DFIs. Those involved in the pioneering transactions discussed above have hired external experts, invested in in-house staff training, and engaged external advisers to successfully navigate the specificities of each transaction. Mobilising internal resources to build this capacity has only been possible with the support of senior champions that are willing to back technical teams looking to innovate.

Capacity constraints are relevant across all models, whether MDBs and DFIs act as originators or investors. However, models that centralise issuance – for example through specialist platforms – could prove to be more efficient in the near-term, while each institution builds a minimal level of expertise and minimal systems required to participate in pooled structures. BOAD’s decision to establish BOAD Titrisation as a specialist subsidiary to develop and manage the bank’s securitisation programme is one potential solution that could be replicated by others. ADB and AIIB’s participation in Bayfront Infrastructure Management also demonstrates the opportunity to outsource relevant expertise and operational capabilities.

Model 1 (In-house): Feasibility is high for MDBs or DFIs with large, standardised asset portfolios and well-established internal systems. AfDB’s synthetic R2R programme and IDB Invest’s Scaling4Impact synthetic transaction exemplify how a single originator can collaborate with advisers to identify a reference portfolio and structure a transaction, navigate rating agency methodologies, and engage with investors to market the deal. This in-house approach simplifies coordination and contract uniformity but limits the underlying portfolio to the originating entity’s assets.

Operational complexity remains a limit, as shown with BOAD’s decision to establish a new specialised subsidiary to operationalise its securitisations. Moreover, MDBs/DFIs building parallel systems and advanced expertise required to operate in-house securitisation programmes independently may be inefficient. Legal complexity remains a major consideration and constraint on structuring options in the near-term. However, this complexity would likely be greater for multi-originator models, as the multiple actors involved in each transaction or programme would only be required to familiarise themselves with one set of underlying contracts in a single originator model.

Model 2 (Multi-originator, single MDB/DFI securitisation platform): Feasibility is lower due to complexity in harmonising documentation from multiple MDBs/DFIs, each with unique contracts, covenants, and credit approval processes. While incremental onboarding of new institutions is possible, centralising the securitisation under one lead MDB/DFI raises potential conflicts of interest. The lead institution could be perceived as prioritising its own assets or skewing valuation and so pricing of underlying loans in its favour. Overcoming these perceptions and establishing robust governance and transparency is key. Legal due diligence, ratings, and operational tasks escalate in complexity when integrating multiple institutions.

Model 3 (Specialised independent MDB/DFI securitisation platform): No independent MDB/DFI securitisation platform exists today. Fully appraising the feasibility of such an entity – effectively a new, commercially oriented development finance institution specialising in securitisation – would require significant further analysis. However, at a minimum, feasibility would require a critical mass of participating MDBs and DFIs as originators, harmonisation of contracts across multiple institutions, and institutional willingness on the part of participating MDBs and DFIs to share requisite data with the platform and in turn with rating agencies and investors.

These considerations similarly affect Model 2 and, once formed, the independent platform’s arm’s-length structure reduces conflicts of interest and would enable it develop a range of structures and strategies tailored (within market standards) to the diverse needs of all participating institutions. Its independence would allow the new entity to serve as a centre of excellence on securitisation for the development finance community.

Models 4 and 5 (Engaging with commercial platforms as originators and/or investors): Partnering with existing commercial securitisation platforms that combine development finance and private sector assets (Model 4) would require investors, rating agencies, and sponsors to consider the risk-return profile and contract terms of individual development finance assets on a case-by-case basis, enhancing feasibility. However, educating commercial platform sponsors and advisers about MDB/DFI-specific legal and credit characteristics would involve significant upfront investment of technical bandwidth within originating institutions. While no private platform currently specialises in MDB/DFI assets, the success of Bayfront’s IABS issuances, at least one of which included MDB B-loans, suggests that, over time, commercial platforms can integrate MDB/DFI assets into larger, more diverse CLO portfolios and programmes.

Model 5 is perhaps the most straightforward to further operationalise. MOBILIST’s equity investment in the BIC IV transaction, ADB and AIIB’s initial equity participation in Bayfront Infrastructure Management itself, and AIIB’s successive investments in senior IABS tranches demonstrate the feasibility of this approach. MDB, DFI, and (in the case of MOBILIST) donor systems and capacities are already well-tailored to accommodate such direct and indirect equity and debt operations. MDBs/DFIs do not need to build in-house operational capacity, alter standard contract terms, or share more data than they are used to. The securitisation itself is managed by private-sector sponsors using familiar legal and rating frameworks.

4.2 Commercial Viability

Commercial viability hinges on achieving portfolio quality, geographic and sector diversification, pricing efficiency, and investor-friendly structures. Across all models, non-sovereign private sector loans, priced closer to market rates, are more conducive to cost-effective securitisation than sovereign lending at subsidised rates. The question is whether the chosen model can deliver sufficient credit enhancement, scale, and portfolio variety to attract investors while maintaining commercial viability.

Model 1: MDBs and DFIs with extensive, well-diversified non-sovereign portfolios could achieve commercial viability while directly securitising in-house loan portfolios. AfDB, IDB Invest, and BOAD’s experience indicate that single-institution transactions can attract investors and deliver BSO objectives if risk tranching is well-structured. AfDB’s transactions effectively transferred risk at a cost overshadowed by the capital relief it gained, making the net outcome worthwhile. Capital relief from R2R Sovereign was even more significant, though in part enabled by likely sub commercially priced guarantees from donor partners. BOAD’s DOLI-P found strong demand in the regional market with the 2024 transaction being placed in just 3.5 hours, though local currency issuance can limit global investor participation. Unlike these multilateral institutions, bilateral DFIs and national development banks may struggle to achieve commercially competitive pricing in a single-originator model due to limited diversification and higher funding costs compared to unsecured borrowing.

Models 2 and 3: Pooling assets from multiple MDBs/DFIs can create larger and more diversified asset pools than are feasible in the context of a single originator model. Taking the weighted average rating as given, greater diversification can enhance risk profile at the portfolio level and in turn improving pricing. Establishing a new platform (Model 3) would be associated with higher startup costs that would need to be recovered over time; however, the purpose-built nature of the new platform means that its operations and governance could be more streamlined and robust, enhancing longer-term efficiency.

Under either model, the lead MDB/DFI (Model 2) or specialist platform (Model 3) may decide to compromise on commercial viability for strategic benefit. For example, a AAA-rated sponsor optimising for commercial viability would likely retain the AAA tranche of a securitisation, because its senior debt would price more attractively than a AAA tranche backed by only subset of its overall asset base. Conversely, the same issuer optimising for PCM would likely sell/transfer the AAA risk, as this tends to be the largest tranche and the most attractive to more risk averse institutional investors.

Model 4: Blending MDB/DFI loans with private sector assets in a commercial platform offers the greatest potential for competitive pricing. Investors are familiar with standard CLO formats, and the added diversification of both development and commercial assets can stabilise performance. For example, Bayfront’s IABS pools have a lower weighted average spread than typical U.S. and European CLO portfolios thanks to their (on average) more highly rated assets, as shown in Figure 10 left-hand panel and right-hand panel respectively. Bayfront’s experience shows that pricing improves as platforms establish a track record and demonstrate resilience. However, achieving a degree of comfort with MDB and DFI assets will require further market education and track record proven by pioneering transactions.

Model 5: Because the sponsor and underlying collateral are wholly private, transactions would typically follow standard true-sale or synthetic formats. These structures are well-understood by mainstream investors, which fosters robust market demand and liquidity. MDB/DFIs acting as anchor investors in certain tranches may also help private issuers secure more attractive pricing (lower spreads) by signalling confidence in the structure. This, in turn, may increase the commercial viability of the overall securitisation. On a purely commercial basis, participating MDBs and DFIs could gain exposure to a diversified set of asset classes and geographies beyond their typical mandate, in turn enhancing the quality of their own balance sheets.

4.3 Additionality

In the present context, we take additionality to mean the capacity of development finance actors to increase the flow of impactful capital to EMDEs and affect systemic change through their participation in the various securitisation models described above. Additionality hinges on the extent to which BSO operations enable additional development finance projects in EMDEs by relaxing MDB/DFI capital constraints; and the extent to which PCM operations mobilise impactful private capital that would not otherwise have flowed into EMDEs. PCM can be delivered at the transaction level, the portfolio level, or through systemic market-building impact.

Model 1: By freeing capital and improving rating metrics, in-house securitisation can accelerate lending for institutions relative to a counterfactual without securitisation. However, this will only be the case if the limiting factor on an institution’s lending is capital adequacy (taking as given its target credit rating). AfDB and IDB Invest case studies show how synthetic securitisation can generate additional direct lending flows by major development finance actors that would otherwise have been capital constrained. As a result, the barriers to additionality and scale (see below) are the same: the thickness of the unretained tranches and the delta between risk weights pre and post securitisation.

If other factors (such as pipeline quality) are binding, securitisation will not increase an institution’s lending volumes, though it will lead to the institution incurring additional funding and transaction costs. When capital and rating constraints are not binding, these costs can be justified by other strategic benefit, including if securitisation provides a cost-effective route to additional PCM. For NSO, AfDB and IDB Invest demonstrated that MDB significant risk transfer can crowd in significant private capital, with private investors serving as funded and unfunded guarantors

In addition to unlocking additional development finance and directly mobilising private capital, several of the pioneering transactions presented in the previous section have contributed to additional systemic impact. For example, the R2R transactions led to significant updates of rating agency methodologies, helping to build the asset class. In the R2R transactions, S&P’s RACF initially produced conservative risk weights for the retained senior tranches. This was partly because S&P’s CDO Evaluator’s model was calibrated mainly for developed markets and overestimated loss-given-default and probability of default for the AfDB’s relatively low risk portfolio. To address this, the AfDB worked with S&P to devise a methodology for treating the secured senior tranche somewhat like a stand-alone lending entity. This adjustment allowed R2R’s senior tranche to achieve an A-grade equivalent rating. The subsequent treatment of IDB Invest by S&P, as well as their efforts to contribute to Fitch’s methodology, highlight how pioneering transactions can generate additionality through systemic impact and reduce barriers to entry for future transactions.

Models 2 and 3: Multi-originator platforms can spread the burden of asset contributions, potentially delivering incremental capital relief and liquidity to multiple MDBs/DFIs simultaneously. This lighter touch on each individual institution’s balance sheet could offer greater flexibility than Model 1 and so could be more likely to create net new capacity for development lending. Over time, these platforms may catalyse broader investor interest and lead to replication (see below).

In terms of PCM, the use of a warehouse has advantage that the structure enables immediate mobilisation of private capital from a range of investors, including commercial banks and sophisticated market participants, at the start of new loan origination. This early external involvement not only diversifies risk but also ensures a steady flow of capital. By freeing up capacity within the warehouse, these securitisations enable ongoing origination of new loans, whether extended to non-sovereign or sovereign obligors, without breaching the MDB/DFI’s capital constraints.

Throughout the warehousing process, the MDB or DFI originator retains its role as lender of record, thereby preserving the benefits of PCT for both the initial warehouse phase and subsequent securitisations. If necessary, the MDB or DFI can establish distinct warehouse and securitisation structures for sovereign and non-sovereign portfolios, tailoring each mechanism to investor preferences and ensuring the most advantageous allocation of resources.

Model 4: Pooling development finance assets with private sector loans through commercial platforms can mobilise large pools of private capital that would not have engaged with development finance assets otherwise. By drawing on existing market infrastructure, the incremental capital freed can be substantial, yielding higher additionality. If MDBs choose to reinvest proceeds in priority sectors that lack traditional financing, this channel becomes a potent lever for generating additional developmental impact. However, there is also a risk that development finance assets could be used by private platforms to enhance the perceived impact of other assets in the pool and so a transaction that lacks genuine developmental benefits. This concern arises because the inclusion of MDB/DFI assets could lend credibility to projects that do not align with sustainable development and climate objectives.

Model 5: Because the securitisation is fully private, the MDB/DFI must ensure its participation delivers genuine additionality by unlocking transactions that would not otherwise have happened or by contributing systemic impact. If the transaction can easily proceed without development finance support, then MDB/DFI capital may not be truly additional. MDB investment in Bayfront Infrastructure Management demonstrate how Model 5 can generate market-building additionality, by helping to establish a platform that in turn has built a new asset class in IABS. BOAD Titrisation’s planned transaction series also shows how MDBs/DFIs can also generate ‘value additionality’ can by working with private financial institutions and corporates to prepare for inaugural securitisations and by enhancing impact management and disclosures. Both market-building and value additionality has potential to allow MDBs/DFIs to generate systemic impact far beyond their own balance sheets.

4.4 Scale

On the supply-side, securitisation in development finance can reach meaningful scale in several ways. Perhaps the most efficient route to scaling capital flows into EMDEs are transactions and strategies that build securitisation markets, by setting standards, establishing benchmarks, and generating market information. However, on the demand-side, scalable investor appetite for MDB/DFI securitisations is the most practical constraint. While interest is growing, sustainable securitisation is still a somewhat specialised market and the number of investors with experience of dedicated development finance transactions remains low.

Inaugural deals often relied on a handful of pioneering investors, including Mariner/Newmarket in AfDB and IDB Invest deals and a small number of regional banks in BOAD’s case. Unless the investor base can be broadened, scaling the securitisation of MDB/DFI assets will remain constrained by the availability of capital managed by specialist investors. In several transactions examined, credit enhancement was needed to entice investors: for example, the EFSD guarantee was crucial in R2R to attract investor comfort for the senior mezzanine risk; GuarantCo’s wrap was needed to place BIC IV’s unrated tranche with private investors. This suggests that pure private sector appetite for the riskiest pieces of emerging market securitisations is still limited, unless the tranche yields are high or the tranche is made safer via guarantees. Absent private investor demand, scaling such structures will be constrained by the availability of development finance and donor guarantees.

Conversely, demand in the private CLO markets remains strong. In 2024 and the first months of 2025, CLOs in Europe absorbed nearly all net new-money loan supply, approximately €44 billion, resulting in issuance volumes of €46.4 billion. Looking ahead to 2025, Bloomberg projects CLO issuance of around €45 billion, buoyed by investor demand for yield and equity models that assume at least a 175 bps arbitrage.

Model 1: BOAD has shown that by issuing two transactions in the space of a year it has a technically feasible and repeatable structure. By using pass-through notes without overcollateralization, the structure setup is simple and straightforward to manage, thereby reducing complexity and associated costs for BOAD. Investors’ returns are linked directly to the performance of the underlying loan portfolio, making the transaction transparent and directly aligned with asset performance.

BOAD’s framework could be particularly appealing for an issuer aiming to bring a new securitisation quickly to market at scale, leveraging the ease of execution and reduced operational burdens. Denominating the underlying loans and the securitised notes in West African CFA francs, BOAD eliminates currency mismatch risk, supports building the local market, and likely facilitates the extreme thickness of the senior tranches.

Yet few MDBs/DFIs can muster the US$400-500 million issuance volumes needed for sustained market presence on a standalone basis (see Figure 11). This threshold precludes all but the very largest institutions globally from launching a solo-origination programme. IFC, with its large, diversified private sector portfolio, stands out as a potential exception. Others, including ADB or IDB Invest, would need to securitise over 10% of their non-sovereign loans each year. IDB Invest’s 2024 transaction transferred risk associated with 10% of its portfolio, having screened assets below B-, set minimum tenors, and prioritised USD assets. Bilateral DFIs and regional development banks are too small individually to achieve the consistency and volume required.

Local currency exposure and the bespoke structure of BOAD’s initial transactions will likely limit demand to regional markets, constraining scale. A hard currency securitisation rated by Moody’s, Fitch, or S&P would likely require significantly more subordination and equity retention by the originator than was the case in BOAD’s structure. Moreover, the transactions thus far, at US$ 250 – US$ 260 million, are approximately half the size of commercial transactions in the US and EU markets.

Models 2 and 3: Pooling assets from multiple MDBs and DFIs can achieve greater scale than any one institution could alone. Although asset sourcing and coordination are more complex, the combined asset base of several MDBs and bilateral DFIs could yield two or more securitisations per year, sustaining market presence and spreading fixed costs. Over time, a dedicated platform or MDB/DFI lead originator could institutionalise the securitisation process, enabling continuous pipelines of new deals.

Under Model 2, the risk of (perceived) conflict of interest discussed above will be particularly acute if the lead MDB/DFI provides a significant majority of assets into the pool. Maintaining a more balanced portfolio across originators to mitigate this risk would limit scale relative to an independent platform, which would not be affected by such conflicts and so could draw a greater share of assets from the largest originators.

Model 4: Commercial platforms have greater potential for scale and platforms pooling only MDB/DFI assets. On the supply-side, pooling development finance with private sector assets enhances both the volume and diversification of assets that issuers can access. In turn, this increases the likelihood that feasible portfolios can be constructed on a scale and cadence required to maintain market presence. On the demand-side, complying with market standards and transferring risk/assets through established private sector platforms increases access to large pools of liquidity. This scale advantage ultimately supports more stable pricing and better market access for asset pools that include development finance assets.

Model 5: With an average annual issuance volume of nearly EUR 2 trillion in European markets alone, Model 5 taps into the vast private-sector CLO/ABS pipeline, which often runs into hundreds of billions of dollars in annual issuance globally. MDBs/DFIs can selectively allocate capital to multiple platforms and deals without the need to assemble large asset pools from their own balance sheets. As illustrated by Bayfront’s repeated issuances, commercial markets can absorb large volumes if the product meets mainstream CLO criteria. However, in isolation Model 5 does not provide a route enhance MDB/DFI balance sheets.

4.5 Replicability

This section gauges whether the chosen model can generate market precedents and serve as a blueprint for replication by future transactions, with less ongoing MDB/DFI involvement or sub-commercial participation over time.

Model 1: Pioneering in-house, single-originator transactions have seen some replication within the development finance community and to a certain extent these pioneering transactions in development finance replicated features of the growing private sector SRT market. AfDB’s 2018 R2R transaction was the first securitisation of MDB loans in recent times and garnered significant coverage among the development finance community, among market participants, and in the financial press.

Despite this extensive publicity, the next comparable AfDB transaction was not launched until 2022 and evidence of replication of the R2R series beyond AfDB was not found until IDB Invest’s 2024 Scaling4Impact transaction. Scaling4Impact benefited significantly from structuring and rating precedents created by AfDB’s R2R transactions and its main investor also participated in the R2R series. However, IDB Invest continued to face an investor base that was largely unaware MDB/DFI assets and their unique credit risk profile.

Publicly listing asset-backed securitisations would be one way to significantly accelerate investor education. Their transparency makes public markets the ideal venue for demonstration effects and replication, with publicly available pricing, risk ratings, and analyst reports containing vital information for those considering follow-on transactions. Listed transactions that help mitigate investor misperceptions of EMDE assets in general, and MDB/DFI portfolios in particular, would be a particularly scalable route to impact.

On the supply-side, the major barrier to replication under Model 1 within and outside the development finance community is scale. As discussed in the previous section, very few institutions have NSO portfolios of requisite scale, quality, and diversification to launch a standalone transactions comparable to R2R and Scaling4Impact. On the demand-side, the private nature of these transactions leads to relatively limited publicly available information at launch and (in particular) on asset performance post-launch. This limits the extent to which private synthetic securitisations can genuinely trigger replication in the market or by development finance peers.

Models 2 and 3: Although a single-MDB/DFI platform drawing assets from multiple originators could repeat its own transactions, it is less likely to serve as a template for other development finance actors to replicate independently due to the scale considerations discussed above. A specialised independent platform may be more replicable than models tied closely to a single originator, as by standardising legal frameworks, credit processes, and ratings methodologies, such platforms can, in principle, be adapted for different sets of MDBs, DFIs, or regions. Yet, the total addressable market for MDB and DFI assets would remain limited, so while replicable, the model would be unlikely to trigger replication to the same extent as models that include pooling with/of private sector assets.

Model 4: Integrating MDB and DFI exposures into existing commercial CLO platforms is the most easily replicated route to securitising development finance assets. Replication is facilitated by standard CLO and SRT formats with which investors and analysts are deeply familiar, and by the global scale of commercial loan markets. As markets become familiar with MDB and DFI assets and with MDBs and DFIs as originators, other commercial securitisation vehicles can follow suit one asset or one originator at a time. This model can eventually create a broad ecosystem of structured products featuring MDB and DFI assets but not relying on them, normalising their presence in mainstream capital markets. The transparency and standardisation of public markets means that listed true-sale securitisations offer particularly compelling scope for replication using Model 4. Triggering replication through demonstration effects that mitigate investor misperceptions of EMDE assets may be the most scalable route to PCM of all.

Model 5: Serving as investors rather than originators is inherently replicable because it relies on standard private platforms, market practices, and large, diversified pools of private sector assets. Any MDB/DFI with the mandate and capital to invest can replicate the approach. Publicising successful investments in frontier markets or less familiar underlying assets within private sector securitisations can encourage other impact-minded investors to follow suit, further enlarging the universe of demand for and supply of sustainable assets. Ultimately because MDB/DFIs are “takers” of private deals rather than the sponsors or arrangers, their direct influence on establishing new structuring frameworks is lower compared to in-house or multi-originator platforms. Replication thus occurs more organically through private market dynamics than through direct MDB/DFI leadership.

5. Conclusions and Recommendations

This paper sought to explore whether, and if so how, securitisation should be scaled in the development finance sector. Drawing on pioneering MDB and DFI transactions in addition to market precedents, the paper highlighted the varied ways in which diverse securitisation structures can enhance MDB and DFI balance sheets and mobilise private capital. The five models discussed show that MDBs and DFIs can play varied roles in the pursuit of feasible, commercially viable, additional, scalable, and replicable securitisation structures, yet no single model and no single role is optimal for all.

For institutions with sufficient scale and in-house capacity, a single-originator programme (Model 1) can achieve capital relief and partial additionality but has limited replicability. A single MDB platform with multiple originators (Model 2) can enhance scale and diversification but faces potential conflicts of-interest. A specialised, independent development finance securitisation platform (Model 3) provides stronger governance and broader participation, improving commercial viability, scalability, and additionality, though at higher upfront complexity and cost. Commercial platforms offer the greatest long-term potential for scale, competitive pricing, and replicability, with development finance actors serving as originators (Model 4) or investors (Model 5). Yet when serving as one originator or investor among many in commercial transactions, development finance actors may struggle to establish and maintain additionality.

The preferred routes to scale will reflect the priorities and constraints of individual MDBs or DFIs: immediate capacity building versus long-term ecosystem development; control versus independence; and simplicity versus mainstream market integration. Over time, demonstration effects from early movers like AfDB, BOAD, IDB Invest, and those involved in Bayfront Infrastructure Management’s IABS series can pave the way for more standardised, widely replicable structures that deepen markets, mobilise private capital at scale, and sustainably support global development finance objectives.

Above all else, development finance actors should continue to learn from one another and collaborate with market participants and regulators to intentionally build a market for securitisation of development finance and impactful private assets in EMDEs. Letting a thousand flowers bloom will undermine the standardisation upon which securitisation thrives. Based on preceding sections, high-level recommendations for consideration include:

• MDBs/DFIs should learn from market participants and regulators to align with best practices and investor expectations in terms of consistent docs, structuring and risk management, terms, data templates, rating methodologies, and underlying contracts. Yet collaboration on market standards should seek to aim enable competition and innovation, and not undermine it. The emerging development finance community of practice on securitisation should now engage more systematically with their regulatory and private sector counterparts, to build awareness of development finance assets in the market and to deepen understanding of investors' and regulators' expectations.
• Donors as MDB/DFI shareholders should equip themselves with an understanding of securitisation and its potential advantages, risks, and costs, and should encourage prudent consideration of securitisation as one route to accelerate direct MDB/DFI lending and private capital mobilisation in EMDEs. As shareholders, donors should focus on setting strategic direction and not seek to drive technological or technical direction on securitisation specifically.
• Private sector rating agencies, legal advisers, and market participants should be ready to engage with and learn from development finance and donor communities, to understand objectives, opportunities, and constraints, with the aim of establishing and scaling a sustainable market for securitisation of impactful EMDE assets. Credit rating agencies are especially fundamental to the securitisation process. Many times, emerging market securitisations don’t come to fruition because they haven’t been able to get a highly enough rated structure. Having accurate rating methodologies, tailor-made to the needs of MDBs/DFIs operating in emerging markets, to rate the collateral and the structures rather than trying to fit emerging market debt data into existing developed market boxes, will be the game changer.