Mitigating currency mismatch in development finance: MDBs and local currency lending

The Sevilla Commitment, adopted at the Fourth UN Conference on Financing for Development in 2025, calls for strengthening public development banks – including multilateral development banks (MDBs) – to scale up climate and sustainable development investment (especially in lower-income countries) because private finance has not delivered at the necessary pace or scale. Yet MDB finance will not produce durable climate and sustainable development impact if it leaves borrowers carrying foreign-exchange (FX) risk by default. The currency of lending is not a technical footnote: it is a core design choice. Many of the investments that matter most – clean energy, electricity grids, transport, water and sanitation, housing, and small and medium enterprise finance – generate cash flows primarily in local currency. When these projects are financed in dollars or euros, FX risk does not disappear: it is reallocated, often amplified, and, in episodes of currency stress, it frequently ends up on the public balance sheet – especially when “de-risking” strategies are poorly designed. 

The question, then, is not whether local currency lending is a “nice to have”. It is whether MDBs and their shareholders are willing to treat it as a central pillar of the development mandate – planned from the outset, properly funded, and governed with clear accountability measures. Indeed, the persistence of FX risk is one of the main reasons the ‘billions to trillions’ agenda has so far fallen short, while also helping to explain the growing emphasis on de-risking private investors and financiers. 

Why the currency of lending matters 

The currency mismatch problem is simple but powerful. Many projects earn revenues in local currency but must service debt in dollars or euros. When the exchange rate depreciates, the local currency cost of repayment jumps, sometimes overnight. Even fundamentally sound projects can be pushed into distress by currency movements unrelated to their operational performance. 

This is not only a borrower-level issue. Currency mismatch can migrate across balance sheets: from firms to banks, from banks to the public sector (via guarantees, bailouts, or emergency liquidity), and ultimately to the macroeconomy through reserve losses, higher risk premia, and rising fiscal strain. In that sense, hard-currency lending to local-currency projects is not merely a pricing decision; it can amplify volatility and crisis risk. 

The stakes rise sharply when such lending is done at scale. Once volumes move from ‘billions to trillions’, the problem stops looking like a micro hedging challenge and becomes systemic: hard-currency repayment needs can strain the external position and FX liquidity, turning green and sustainable development project finance into a potential balance-of-payments issue. 

Local currency lending solves one problem, and can create another 

It is tempting to think the fix is simple: tell MDBs to lend in local currency and the mismatch disappears. Borrowers receive local currency and repay in local currency, so the risk seems to vanish for private borrowers. But the real question is what happens on the MDB’s own balance sheet. 

Most MDBs still fund themselves mainly in hard currency through international bond markets. If they lend in local currency while their liabilities remain in dollars or euros, FX risk is not eliminated – it is shifted. The MDB is effectively holding local-currency assets against hard-currency obligations. In periods of depreciation or FX stress, that mismatch can translate into valuation losses, higher measured risk, and pressure from rating agencies. A downgrade – or even the threat of one – raises funding costs and can quickly weaken the leverage model that supports MDB finance at scale. This helps explain why many MDBs place such a high priority on preserving their AAA credit rating. 

This is the currency-mismatch constraint. Local currency lending can “de-risk” private borrowers, but without a coherent funding and risk-sharing strategy it can leave MDBs, and the broader system, exposed to FX stress, limiting the scale of climate and development finance rather than expanding it. 

Not all projects face the same FX logic 

Hard-currency lending is not always inappropriate. Projects that generate foreign exchange, directly or indirectly, can service hard-currency liabilities more naturally. Export infrastructure, tradable sectors, and other FX-earning activities can often repay in the same currency they generate. 

Moreover, if a country has a sustained external surplus – or if the investment programme contains a large share of FX-generating projects – those inflows can help meet hard-currency debt service more broadly. In that case, hard-currency lending even to some domestically oriented projects may be feasible without immediate balance-of-payments stress, because the system as a whole is generating the foreign currency needed to meet external obligations. 

The problem is that this is not typical for many lower-income and externally constrained economies, and it is unlikely to hold at the scale implied by climate and sustainable development investment needs. A large share of climate- and development-relevant investment is domestic by nature: projects sell into local markets, charge tariffs in local currency, and rely on domestic demand. 

Three ways to scale local currency lending 

There is no single solution. Scaling local currency lending requires a strategy that fits domestic market depth, macro conditions, and MDB balance-sheet constraints. In practice, MDBs will need a mix of approaches across countries and over time. But the menu of options for MDBs is clear – and each comes with a distinct trade-off between scale, cost, and who ultimately bears FX risk, as discussed in my recent work on MDB local-currency lending. 

Option 1: Borrow in hard currency and manage the mismatch 

The most common approach is to continue borrowing in global hard-currency markets while lending in local currency, and then manage the resulting FX exposure through diversification or hedging (see Observer Autumn 2024). The attraction is scale: international bond markets can mobilise large volumes quickly, at long maturities and often at relatively low spreads. 

Diversification reduces exposure to any single currency by spreading local-currency lending across multiple countries and currencies, so that adverse movements in one market are less likely to dominate the overall portfolio. Large, geographically diversified MDBs are best placed to benefit from this. Hedging relies on available FX derivatives (such as forwards and cross-currency swaps) to cover part of the exposure – typically where markets are sufficiently liquid and hedging costs are acceptable – while leaving some positions unhedged where instruments are unavailable, too short-dated, or prohibitively expensive (which, in many lower-income and smaller markets, is the norm). A useful illustration is TCX, a specialised currency-risk provider that enables lenders to offer local-currency terms by absorbing exchange-rate risk on its own balance sheet.  

But the constraint is equally clear. This model still leaves the MDB holding local-currency assets against hard-currency liabilities, so it must actively manage a structural currency mismatch. It is most workable in larger and more liquid markets, where hedging is feasible and where the MDB’s local-currency operations remain small relative to the host country’s FX market. It becomes far more challenging in smaller or more volatile economies, where hedging is limited and FX liquidity tends to evaporate precisely when exchange-rate pressure and refinancing needs intensify. In those moments, what looks like manageable “portfolio risk” in normal times can become a binding constraint on the volume and reliability of local-currency lending. 

Option 2: Issue local currency bonds 

A second approach is to fund local-currency lending with local-currency liabilities. MDBs can issue bonds in domestic markets in local currency, match the currency of assets and liabilities, and thereby avoid currency mismatch on the MDB balance sheet. In principle, this is the cleanest way to scale local-currency lending without shifting FX risk elsewhere. 

Local issuance can generate positive spillovers for domestic financial development: it can extend yield curves, increase the supply of high-quality local-currency assets, and broaden the investor base. But it is not universally feasible. Many markets are too shallow to provide long-tenor funding at scale, and issuance costs can jump in periods of macro stress. Even where feasible, volumes and maturities may still fall short of what a major climate investment push requires. A further constraint is often overlooked: many green projects rely on imported capital goods and specialised inputs. In countries without a current-account surplus, local-currency bond funding can cover the domestic component, but it does not by itself provide the hard currency needed for imports. 

Finally, local issuance raises a strategic question about MDB comparative advantage. In several countries, national development banks (NDBs) and sovereign treasuries already have access to the domestic bond market and may be better placed – politically and institutionally – to mobilise local currency. MDBs often add the most value precisely when they can provide hard currency at long maturities, at scale, and sometimes at lower cost than domestic issuers. If MDBs simply replicate local-market funding, the incremental benefit must be clear – whether through credibility effects, or risk-sharing features that domestic institutions cannot readily provide. 

For these reasons, local-currency bond issuance is an important tool, but not a standalone solution. It is most effective as part of a broader financing package, and it is less suitable where projects rely heavily on imported inputs and the economy faces tight external constraints. 

Option 3: Expand risk-bearing capacity through recapitalisation and new hybrid capital instruments 

If MDBs are to lend in local currency at scale without taking on destabilising mismatches, they need more risk-bearing capacity. In practice, that means recapitalisation and greater use of modern hybrid capital instruments – tools that strengthen MDB balance sheets so they can hold local-currency assets without mechanically expanding hard-currency liabilities. 

Paid-in capital is especially valuable because it is genuinely loss-absorbing and does not create fixed hard-currency repayment obligations. With fresh paid-in capital, an MDB can convert part of the injected hard currency into local currency – often via the domestic central bank – and use it to extend long-term local-currency loans. This avoids the core mismatch problem: the MDB is not funding local-currency assets with hard-currency debt, which can reduce pressure on its risk profile and, in turn, lower the risk of rating downgrades. It also supports longer maturities because lending is not tied to a bond refinancing cycle. As loans are repaid, the MDB can recycle the local currency into new lending and build a revolving “local currency fund” that finances successive rounds of climate and sustainable development investment. Even without new bond issuance, one capital injection can underpin multiple lending cycles as repayments revolve. 

This approach also has an immediate benefit: the initial conversion of hard currency into local currency provides foreign exchange to the country, helping finance imported equipment and inputs for investment projects. But the benefit is largely front-loaded. Once the initial conversion is done, continued lending from the revolving fund does not keep generating new hard currency. In that respect, it shares a limitation with local-currency bond funding: it can finance the domestic-currency component well, but it does not by itself solve import-heavy components in externally constrained economies. 

The obvious drawback is political. Recapitalisation requires shareholder agreement – often from high-income countries – and political will (and sometimes fiscal space) is frequently limited. Where full recapitalisation is not feasible, hybrid capital instruments can help bridge the gap. These are long-dated bonds (often perpetual) designed to be treated partly as equity because they have equity-like features: they absorb losses in stress (for example through write-down or conversion) and allow coupon deferral or cancellation without triggering default. 

To make this route work at scale, robust institutional arrangements are essential. That includes standardised principles for how hybrid instruments are treated by rating agencies, so MDBs can rely on predictable capital recognition. It also requires credible channels to mobilise long-term public resources – potentially including IMF Special Drawing Rights (SDR)-based mechanisms (see Inside Institutions, What are Special Drawing Rights?)  – and strong coordination among MDBs, National Development Banks, and central banks. Crucially, central banks must also have clarity on reserve-asset eligibility: if MDB hybrid instruments are to be widely adopted, they need to be treated as high-quality, liquid official assets – compatible with reserve management frameworks and prudential constraints. The good news is that this agenda is advancing: in May 2024, the IMF authorised its members to use SDRs to acquire hybrid capital instruments issued by prescribed MDBs. Yet the operational follow-through remains weak (see Observer Summer 2024).

This option is politically harder than launching yet another hard-currency lending facility. But it is often economically cheaper – and potentially more developmentally effective – than repeatedly absorbing the crisis dynamics produced by currency mismatch and excessive external indebtedness, which ultimately show up as reserve losses, fiscal stress, and abrupt investment contractions in the poorest and most externally constrained borrowing countries. 

Conclusion: this is ultimately a shareholder decision 

Whether MDBs can scale local currency lending is not mainly a technical question; it is a governance question. As long as shareholders demand rapid scale-up while maintaining balance sheets anchored in hard-currency funding and minimal risk absorption, currency mismatch will keep resurfacing – first on borrowers, then on domestic financial systems, and, in crisis moments, on the public sector. 

The way out is to align mandates with resources. That means embedding a coherent financing strategy that combines (i) greater support for export-enhancing investments that strengthen the economy’s FX-generating capacity, with (ii) local currency lending as the default for domestically oriented projects. It also means financing that make use of a realistic mix of market instruments and explicit public risk-bearing capacity – capital, hybrids, and well-governed liquidity backstops – and being transparent about where FX risk ultimately sits. 

Climate and sustainable development finance will not scale if its currency architecture is built to fail when exchange rates move. 

About the author

Alfredo Schclarek Curutchet, CONICET

Alfredo Schclarek Curutchet is Independent Researcher at the National Scientific and Technical Research Council, Argentina (CONICET) and Associate Professor at the National University of Córdoba, Argentina, where he teaches macroeconomics for undergraduate and PhD students. He has published in, among other journals, Journal of Banking and Finance, Journal of International Financial Markets, Institutions & Money, World Development and Journal of Financial Stability. Since 2018, he is Ford Foundation Visiting Research Fellow at the Institute of New Structural Economics at Peking University, China. Alfredo holds a Ph.D. in Economics from Lund University, Sweden. He also earned a Masters and a BA in Economics from Lund University. He worked at the European Central Bank, Germany, the United Nations, Denmark and the National Agency for Investment Development, Argentina.