Export credit agencies and multilateral development banks agree they need each other for the biggest projects in emerging markets, particularly the global South. Three practitioners told TTP the obstacles are as much about who arrives first, who takes the first loss, and who holds the risk afterwards as about money.
When the Berne Union gathers in Tokyo for its annual general meeting from 16 to 19 November, its president, Yuichiro Akita, expects to publish a paper on blended finance. It is the written output of the association’s blended finance hackathon in London in March, where export credit agencies (ECAs), multilateral development banks (MDBs), development finance institutions, commercial banks and private insurers worked through how public money can pull private money in behind it at scale.
Ahead of Tokyo, TTP spoke to Akita, who is also a senior general manager at Japan’s Nippon Export and Investment Insurance (NEXI); to Christopher Marks, managing director, blended finance, at MUFG; and to Alexander Otten, senior project finance manager at Atradius Dutch State Business (Atradius DSB), the Dutch export credit agency. They sit at different points along the same chain and do not define the term the same way.
What blended finance means here
The Organisation for Economic Co-operation and Development (OECD) defines blended finance as “the strategic use of development finance for the mobilisation of additional finance towards sustainable development.” In practice, that usually means a public institution takes enough risk, through a guarantee, insurance, concessional loan or first-loss tranche, for private capital to follow on market terms.
The three practitioners TTP spoke to use the label differently. Marks, who worked at two multilaterals before joining MUFG, said the MDBs generally mean mixing commercial and concessional money “to lower the cost of borrowing”. His own aim is to use guarantees to lift single-B and double-B sovereign risk towards investment grade, so that insurers and pension funds will follow. Otten said EU state aid rules prevent Atradius DSB from pricing its domestic innovation cover below market levels, so in its domestic work cheaper money is not what it offers. “It does not necessarily bring down the pricing, but it attracts more capital,” he said. Otten treats ECA cover itself as a form of blending, because commercial banks keep an uninsured slice while public balance sheets take the rest. Akita focuses on the newer task of layering MDB, ECA and private-insurer risk so ticket sizes that no single institution can carry still close. “For me, blended finance is really about risk orchestration: bringing together institutions with different risk appetites and determining who is best placed to take which risk, and when,” Akita said.
Why it matters: public balance sheets alone cannot fund the infrastructure, climate and industrial projects governments want done, and private investors will not enter single-B and double-B markets without credit enhancement they trust. The fight is less about inventing new products than about who originates the deal, who takes which layer of loss, and whether a guarantee can travel once written.
Three years of pressure have moved the MDBs, the ECAs and the rulebook
Akita traced these changes to the G20, which in recent years has pressed MDBs to use their capital more efficiently. That pressure was crystallised in the G20’s 2022 Independent Review of Multilateral Development Banks’ Capital Adequacy Frameworks, which urged MDBs to stretch existing balance sheets and expand risk-transfer tools, including guarantees. With direct lending and official development assistance both limited, the MDBs have turned to guarantees to mobilise private capital. The ambition is clear: the MDBs set out to mobilise capital “from billions into trillions” when dialogues first started, in 2015. For a long time, Akita said, progress was relatively limited. “We have seen much greater efforts from the MDB side over the past three years.”
He pointed to the African Development Bank’s Africa Investment Forum, which he has attended twice, as an example of an MDB working to bring private insurers and ECAs into its deals. Alongside the Forum, the bank has run dedicated Insurer and ECA days to deepen risk-sharing, and Akita said AfDB has been using private insurers for portfolio guarantees and for cover on single transactions. That matters because private underwriters can sit behind an MDB book without the pari passu and pro rata arguments that have long complicated MDB-ECA structuring.
The ECAs have moved towards them. Marks said that for most of his career in development finance the agencies were “the awkward child in this conversation”, seen as “too mercantile, too nationally tied”. He believes the two worlds are converging as the investment needed for resilience pushes ECAs onto ground the development community considered its own. The major exporting nations, he said, have “a great interest in positioning for whatever shape the energy transition, adaptation preparedness and the embrace of digitalisation/IT takes in the future”, and plenty of infrastructure exists where their commercial interests and borrowers’ development priorities overlap.
Otten described the same drift from inside an agency. Atradius DSB now serves what he called “more strategic political motives”: the Dutch decision, from 1 January 2023, to stop new export credit support for fossil fuel projects; cover for critical minerals and sustainable fuels based on imports rather than exports; a Ukraine facility; and soon, defence. Akita listed economic security and critical minerals among the newer ECA priorities. The rulebook has loosened with them. OECD participants agreed in 2023 to extend export credit support for climate-friendly projects, and Otten said the boundaries have “loosened a lot lately”. Atradius DSB still applies many OECD limits to domestic products the Arrangement does not govern. Asked whether that was right, he said: “No, but this is a reference we and the banks we cover know well.”
Otten said he now sees development banks stepping in alongside his agency “more and more”. Akita said private insurers already underwrite around 20 per cent of Berne Union members’ medium and long-term new commitments, and that amount rises each year.
Whoever sets up the deal writes the rules
Origination is one friction point, Akita said. MDBs typically open talks with a host country’s finance ministry early. On large African infrastructure deals, ticket sizes have grown beyond what any single MDB or ECA can carry, so MDBs need ECAs, and ECAs value MDBs’ negotiating weight with host governments in high-risk countries. The ECAs also want some say over structure to meet their own policy requirements, which cover national interest as well as environmental and social standards. Once terms are settled between an MDB and a host government, he said, it becomes very difficult for a bilateral lender to renegotiate. He would not name transactions, but said there were “a lot of big transactions in Africa” where an MDB wanted ECA participation and the ECAs, with their commercial banks, found the inherited structure hard to work with. If project origination included discussion between MDBs and ECAs from the beginning, he argued, the outcome might be different.
Otten’s remedy was practical: ECAs should compare pipelines with the multilaterals and “sit together in a room with the banks more often,” he said, “and get out of our silos. There are so many instruments and institutions out there. I think many opportunities are now not picked up because of this.”
Layered deals and portfolio guarantees are making co-financing workable
The structures that help are the ones that let each institution take only the risk it can stomach. Risk appetite differs among ECAs by national policy, Akita said, and more sharply between ECAs and MDBs. ECAs work within the OECD Arrangement and alongside commercial banks that keep an uninsured portion of 2.5, five or sometimes ten per cent, so those banks must be comfortable too. MDBs lend directly and carry no such passenger, yet the tier-one MDBs must protect their triple-A ratings and can offer first-loss cover only with trust funds whose size depends on sponsor governments. “It can be difficult for ECAs to take the first-loss position, particularly where commercial banks retain an uninsured portion of the risk and need to be comfortable with the overall structure,” Akita said. His answer is layered risk allocation, with first, second and third loss assigned to whoever can bear each. “I think it takes time, but this is an area where we need to go further.”
Portfolio guarantees from private insurers have made the MDBs’ side easier. Private underwriters can cover a book of deals rather than negotiating each one, rely on the MDB to arrange the transaction, and avoid the pari passu and pro rata arguments that have long complicated MDB-ECA structuring. From an MDB’s point of view, Akita said, a private insurer “can be more flexible and relatively straightforward for MDBs to work with in risk-transfer structures”.
Stacking public balance sheets helps at the riskiest end. Atradius DSB’s innovation cover supports Dutch companies scaling technologies such as photonic chips, where the borrower may still be EBITDA-negative and about to build a first-of-a-kind plant. For those deals, the agency has agreed with its ministries a cap of around €30 million per transaction; with the European Investment Bank alongside, Otten said, the two might reach €80 million. Is that better for the borrower? “Yes, it’s better because there’s more money available. But it’s more cumbersome, of course,” he said. The cap protects a political mandate that would fall away “if we lose loads of money on these projects”. Every public balance sheet in a blended structure answers to a taxpayer somewhere, and taxpayers rarely volunteer for the bottom tranche.
Marks stacks guarantees for a different reason. To maximise capacity and deploy the differential power of each instrument against investor appetite, he combines ECA and MDB products across tranches in a single large-volume transaction. Debt guarantees from UK Export Finance (UKEF) or Sweden’s EKN can sit alongside equivalent World Bank Group products that, in simplest terms, cover the risk of a government not delivering on its commitments. In this way, an investor’s credit committee can look at the paper and say: “Against the specificities of my own balance sheet, I am confident that we have minimised or eliminated unfamiliar country risk. I’m okay with that.”
Guarantees are scarce and slow to travel
Marks said guarantees remain a very small share of the big multilaterals’ deployment and, in many cases, those issued must be reinsured from finite capacity. In West African jurisdictions that are active destinations for such products, such as Côte d’Ivoire, Benin and Togo, he said, lenders like him buy political risk insurance while MDBs also buy reinsurance at Lloyd’s for their own guarantees, drawing on the same market. New guarantors are arriving. Marks chairs the Green Guarantee Company, which uses an initial $100 million of capital to provide up to $1 billion of investment-grade guarantees for emerging market borrowers. He called it “a climate positive and recalibrated descendant of the monoline”, a reference to the US bond insurers active before the financial crisis.
Distribution is the second constraint. Not every ECA is comfortable seeing its guarantee passed to institutional investors whose behaviour in a default it cannot predict. “They still want the banks to hold it on balance sheet,” Marks said.
Capital is waiting on the seating plan
Set the three accounts side by side and the problems are as much about coordination as capital. Marks can combine guarantees, but not every ECA will let its cover pass to investors. Otten has products and a willing partner in the European Investment Bank, held back by caps and state aid rules. Akita describes MDBs that need ECAs, and ECAs that struggle to join a structure finished before they were invited. Each problem traces back to sequencing: who originates, who takes which layer of loss, and whether a guarantee can travel once written.
That is the test for the paper Akita takes to Tokyo. The instruments exist; the order in which institutions use them is still being worked out.
Marks described what success would look like. Once the country risk is mitigated, he said, an investor should look at the project and see something familiar. “If it’s an airport in Indiana or an airport in Budapest, it should also be an airport in Addis Ababa.” The investors who fund airports in Indiana should be able to fund one in Addis Ababa too. First, public institutions have to agree early and in the right order on who carries which part of the risk.
