Around 85 to 90 percent of financing from multilateral development banks, or MDBs, in developing countries is hard currency debt. After a decade of mobilization milestones spanning funded and unfunded portfolio programs, B-bonds, securitizations and synthetic risk transfers, that number hasn’t moved. We got better at mobilizing capital. We never questioned what we were mobilizing it into.
Throughout this time, mobilization has become the defining metric of MDB ambition, repeated in annual reports, spring meetings and conference panels as the measure of progress. The programmatic innovations that followed were breakthroughs: Managed Co-Lending Portfolio Program (MCPP) at IFC, Room2Run at AfDB, Scaling4Impact at IDB Invest, FIPP at EBRD. Capital moved. Structures scaled. But every one of these programs mobilized capital into hard currency instruments, while the borrower’s revenues remained in local currency.
The currency problem behind the pipeline problem
Many emerging markets borrowers with local currency revenues face an ugly choice. They can borrow in hard currency and carry open foreign exchange, or FX, risk on their balance sheet, which is fine until it isn’t; and when it isn’t, it tends to be catastrophic, precisely because currency crises and debt distress are correlated. Or they can try to borrow in local currency, but long-tenor local currency financing is scarce, expensive and in many markets simply unavailable from commercial sources.
MDBs have been the primary answer to that dilemma, and their answer, almost universally, has been hard currency. Programmatic mobilization delivered on what it promised. But the currency problem was simply never part of the brief.
Part of the reason is structural. Most MDBs are constrained in their local currency lending by the very policies designed to protect their balance sheets. Their statutes and treasury policies preclude open currency positions, meaning they can only lend in local currency if they can fund or hedge themselves in that currency. In many frontier markets, that hedging simply doesn’t exist. No swap market, no non-deliverable forward market, no willing counterparty for long-tenor local currency positions.
Where onshore funding exists, internal policies often still require matching hard currency liabilities, creating a ceiling on how much local currency lending is actually possible. The Currency Exchange, or TCX, was created precisely to fill that gap, providing currency swaps where derivative markets don’t exist. But TCX could not handle the necessary volumes if the World Bank, regional development banks and other DFIs decided to offer local currency loans as part of their standard product range. The derivative dependency doesn’t just constrain individual transactions. It caps the entire model.
Not every MDB faces this constraint equally. A handful of institutions have spent decades building genuine onshore local currency funding capability, raising liabilities through domestic bond issuance, local bank relationships and central bank access in the markets where they operate. For these institutions, local currency lending doesn’t require a swap market. The asset and the liability are matched directly. The derivative dependency simply doesn’t apply.
The field has caught up to the diagnosis. Convergence, the global network for blended finance, has identified portfolio-level foreign exchange risk as the central problem that its latest mobilization framework cannot yet solve. Their twelfth priority model, known as PIMM12, is dedicated specifically to local currency foreign exchange risk mitigation, and it is the only one of twelve they cannot yet standardize. Solving it requires connecting a treasury capability that exists in specific institutions to a programmatic mobilization architecture that has never been built around it. Blended finance, deployed not at the transaction level but at the portfolio level, is what could build it.
Traditionally, blended finance has been deployed at the project level, almost universally for credit risk. First loss tranches, partial credit guarantees and subordinated debt are deployed to reduce the probability of loss for senior investors and have been the backbone of the blended finance toolkit for a decade. Currency risk has been addressed, but only marginally. And while the donor appetite for local currency solutions exists, it has simply been underutilized. Several instruments have attempted to fill the gap. For example, TCX absorbs FX exposure for individual transactions where derivative markets don’t exist, funded by concessional capital from donors and MDBs. The European Investment Bank´s ACP Trust Fund deploys local currency lending into African, Caribbean, and Pacific markets backed by European Commission funding.
The architecture that would fix it
What has never been built is a structure that connects that concessional appetite to a programmatic mobilization platform at portfolio scale, absorbing the FX conversion mismatch not deal by deal, but across a delegated portfolio of local currency loans originated by an MDB with genuine matched-funding capability.
Institutional investors, including pension funds, insurers and sovereign wealth funds, have been one of the main targets of MDB mobilization efforts over the past decade, and the architecture built to attract them has been compelling.
B-loan structures offer a powerful package: preferred creditor status that compresses default risk, tax exemptions in most jurisdictions, and a track record of managed portfolio programs that have consistently delivered on risk-adjusted returns. But their participation has a ceiling, and that ceiling is currency. Lifted to the portfolio level with a targeted donor facility absorbing the FX conversion mismatch, that ceiling disappears. FX volatility across multiple currencies, sectors, and tenors doesn’t move in perfect correlation. That imperfect correlation transforms an unbankable standalone exposure into a manageable portfolio risk. It also allows a donor facility to be sized against expected shortfall rather than individual worst cases, dramatically reducing the concessional capital required per unit of investor capital mobilized.
The structure works in two directions simultaneously. A multilateral development bank with this treasury capability originates a diversified portfolio of local currency loans and offers hard currency participation to institutional investors, sharing credit risk on a pari passu basis, with the foreign exchange conversion shortfall absorbed at the portfolio level. This converts the currency constraint from a ceiling on mobilization into the next frontier for scaling it. The mechanism doesn’t depend on swap markets that don’t exist. MDBs with genuine onshore local currency treasury operations can convert cash flows synthetically through spot markets and internal operations built over decades, migrating that conversion gap to the portfolio level where a targeted donor facility absorbs it, delivering clean hard currency returns to investors without the currency risk that has kept their larger capital pools on the sidelines.
The economics work for both sides of the table. Applied at portfolio scale, blended finance generates better donor leverage and better investor risk-adjusted returns, and it eliminates the market risk capital charge that currently makes local currency exposure structurally uninvestible for regulated institutions, regardless of how attractive the underlying spread might be.
The tools exist. The math works. The question is why the product hasn’t been built. The three teams that would need to co-design it exist in most major MDBs: “Treasury” knows how to fund in local currency. “Mobilization” knows how to structure programmatic co-lending. “Blended finance” knows how to deploy concessional capital. But they have never been in the same room because of separate mandates, separate metrics, separate floors. The knowledge exists in the institution. The architecture doesn’t, because nobody’s mandate requires building it. That is the institutional answer to what looks like a product problem.
The trust between MDBs and institutional investors has already been built. They share credit risk. They share the same seat at the table when a borrower restructures. That’s not a relationship problem. It’s not a pipeline problem. It’s a product problem, and the reason the product doesn’t exist is institutional, not structural.
The treasury capability exists in specific institutions. The investor appetite exists. The donor precedent exists. What’s missing is the architecture that connects them. And institutional problems, unlike structural ones, can be solved.
Alejandro Muñoz-Alonso Merigó is a development finance professional with experience in structured finance and syndications at IFC. He holds a Master’s in Development Finance from Johns Hopkins SAIS.
Guest posts on ImpactAlpha represent the opinions of their authors and do not necessarily reflect the views of ImpactAlpha.
