No easy wins in development bank power-up

Last year six of the biggest multilateral development banks paid out between them $89bn to their borrowers, for needs from keeping girls in school to providing clean energy and building roads.

That’s a tidy sum, but set against the needs it is tiny — just $13 of investment for each of the 6.8bn citizens in the developing world.

Progress in economic development is slow. The Covid pandemic raised the share of people living on less than $2.15 a day — one definition of extreme poverty — to 9.3%. But even without that, it would only have come down from 11% in 2015 to about 8%.

When the world considers how to tackle problems like these, the multilateral development banks (MDBs) are front and centre.

Not mere channels for distributing aid, they can act as fulcrums. Their capital — and expertise — can be levered to produce benefits multiple times the outlay.

High on the wish list of every policymaker, in rich and poor countries, is “do more with the MDBs”.

The prize is not just what the banks could do, but what they could unlock in co-financing from the private sector.

“There are $400tr of assets under management in the world,” says Alfonso Garcia Mora, vice-president for Europe, Latin America and the Caribbean of the International Finance Corp. “The gap to achieve the Sustainable Development Goals is $4tr. If we could mobilise 1% of the AUM, it’s a done deal.”

But for decades, that has stayed tantalisingly out of reach. The MDBs are growing, but can still only scratch the surface of development needs.

Wealthy countries are urged to stump up more money, but overseas aid does not win many votes. Caught in this impasse, donor governments are asking questions about the MDBs.

Are they trying hard enough — or are they complacent? Their balance sheets are the strongest of any banks in the world — do they really need to be so conservative? The financial world has produced a river of new ideas: why are the MDBs still basically just issuing triple-A rated senior bonds?

Such questions have swirled for years, but were codified into a clear agenda in 2022, with the publication of a report on MDBs’ Capital Adequacy Frameworks by an independent panel reporting to the G20.

The CAF report has won the backing of successive G20 presidents: Italy, Indonesia, India and this year, Brazil. “We decided to reinforce the debate within the G20 on what is the way forward on MDB reform,” says Ivan Oliveira, deputy secretary for sustainable development finance at Brazil’s Ministry of Finance. “We are working on a roadmap for MDB reform. There should be a medium term agenda for the MDB system — where the G20 want it to go over three to five years, taking into account three pillars: capital, operations and impact assessment. There will be some specific deliverables we are going to be presenting in our presidency.”

One will be a template for MDBs to report their progress on implementing the CAF report recommendations.

The G20 presidency is engaging closely with the informal group of MDB chief executives. “Many of the banks were not very happy with the way the CAF review was made,” says Oliveira. “There was very little consultation with them. It’s very important for countries to push them, but also to hear what they think is important. That could help them to implement reforms faster, while not losing the push. The banks have started to flex their muscles together, so their views can be taken into account.”

Crucially, the CAF report came when an internationalist, pro-development Democratic administration was in power in the US.

Prodded by Janet Yellen, US treasury secretary, the MDBs, initially sceptical of the CAF proposals, have realised they need to act.

The African Development Bank, long at the forefront of financial innovation because of its need for capital, has been joined in experimenting by all the others, including the World Bank, which published in December 2022 its Evolution Roadmap, subsequently updated.

The CAF panel made five core recommendations: redefine the MDBs’ risk appetite; incorporate callable capital into CAFs; use innovations to increase lending headroom; assess rating agencies’ methodologies; and improve the governance of capital adequacy.

Implementation has been patchy at best, with little movement on some issues.

Nevertheless, there is enough going on to generate enormous noise in the MDB sector. For now, the issues attracting most attention are hybrid capital — one of the financial innovations the panel suggested — and callable capital.

Crucial, too, though outside the scope of the CAF report, is how the MDBs can lever in more co-lending from the private sector — something they have always tried to do, but which has never fulfilled its theoretical potential.

To describe hybrid capital as “easy” would sound bitterly ironic to those who worked for four years to bring about the African Development Bank’s inaugural $750m issue on January 30 this year.

A world first, it required painstakingly winning over the board, shareholders and the three major rating agencies, none of which had fully developed criteria for MDB hybrids.

Nevertheless, of all the CAF recommendations, hybrids are one of the simplest to make progress with, because they can be bolted on to a bank’s capital structure without requiring fundamental reform.

In principle, it ought to be a no-brainer that MDBs could raise extra capital by issuing subordinated debt to investors willing to take on some risk, without receiving voting rights.

Making this work in practice is much more difficult. But the AfDB and some smaller African development banks have succeeded in carving out a structure for MDBs that achieves 100% equity credit from Fitch, Moody’s and S&P, as well as under IFRS accounting and the banks’ own capital adequacy models. Investors will lose all their principal in the unlikely event that the issuing MDB has to call on its callable capital. This structure can be rated a mere three notches below senior debt.

This means that, as long as they can find investors, MDBs can now raise what is effectively equity at will — something they have never been able to do before.

The AfDB reckons it can do at least $2 of lending for every extra $1 of equity, and its actual balance sheet ratio is 2.2. This means the inaugural $750m deal will support some $1.65bn of loans it would not otherwise have been able to make — half as much as its total disbursements last year.

It took 18 months after the AfDB received board approval for the structure before it could line up everything to its satisfaction — above all, finding a market window in which it could launch the hybrid at a suitable price.

The sale was successful, landing the perpetual non-call 10.5 year bond rated AA- by S&P at a yield of 5.75% or 157.5bp over US Treasuries. The AfDB reckoned that was 134bp wider than it would have had to pay for a senior bond — within what GlobalCapital believes was its target range, of 100bp-150bp.

With a peak book of $6bn from 275 investors and final allocation to 190 with $5.1bn of orders, the deal appeared to prove that investors existed for MDB hybrids at a viable price.

Four months on, that proof is not looking so certain. The hybrid has bounced up and down in the secondary market — more down than up.

The hybrid was originally priced at par. On June 6, Tradeweb displayed one quote for the bond from a named bank, BNP Paribas, at 95.25/96.00 for a size of $1m, equivalent to a Treasury spread of 201bp/191bp.

“Nobody talks about it, but now there is zero market for an MDB hybrid,” says one capital markets specialist. “It’s totally illiquid and it’s being marked even yield to a Citibank hybrid, which is six or seven notches [lower rated]. The deal got done, but it had nothing to do with hybrid investors. It was achieved by selling to hedge funds and momentum buyers during a month when anyone was buying anything that had duration.”

The AfDB’s book composition does lend some credence to the doubts. Some 55% of the bonds went to hedge funds and specialist credit funds. Without them, it would have been a very different deal — and they may not have bought it from a fundamental desire to be long-term holders of MDB capital.

Keith Werner, manager of capital markets and financial operations at the AfDB in Abidjan, says the hybrid’s trading has experienced volatility, like other hybrids, partly because of interest rate moves. But although the AfDB hybrid has somewhat underperformed others, “This product is still in its infancy. There’s only one deal, which does not make an asset class. When there is more issuance this should be conducive to further investor focus. Our trade was over six times oversubscribed. I don’t think you get that sort of demand if investors don’t think it’s a viable asset class. So it seems to be more of a secondary market issue. When other issuers come into the market it should feed through and create a more effective secondary market.”

A syndicate banker who worked on the deal agrees, adding that investors were used to considering extension risk on hybrids. He says a number of investors had been interested in the deal but did not want to buy the first one. And he says that although hybrid specialist investors would be “an important buyer base, the aim should be to develop traditional SSA investors to go further down the capital structure”.

The only way to test this would be to bring another deal, but that is unlikely for some time.

It will be a while before the AfDB needs more capital. The World Bank and European Bank for Reconstruction and Development have both said they will bring pilot hybrid issues to the public market. But the EBRD has just had a general capital increase, while the World Bank does not want to pay up as much as the AfDB and is busy placing hybrids privately with its shareholders.

Other MDBs are sceptical of publicly placed hybrids — as they were of the AfDB’s groundbreaking $1bn securitization in 2018 — fundamentally on grounds of price.

For commercial issuers, hybrid capital achieves an alchemy: it puts equity on the balance sheet, more cheaply than issuing real equity. But public sector development banks do not pay dividends. For them, equity from hybrids is more expensive than the ordinary kind.

That has two implications. If an MDB pays a spread on a hybrid, above that of senior debt, it must find this money from somewhere. On the AfDB’s calculations, this was 134bp, on $750m of debt, or $10m a year.

The second implication is more insidious. Hybrids are perpetual, but investors price them on the assumption that the issuer will call them at the first opportunity. They accept there is some risk of a non-call, but regard this as slight.

In an unfavourable market, it might be cheaper for a commercial bank to let an old hybrid go uncalled than to call it and issue a new one. But it would probably still call the bond, to keep hybrid investors sweet.

Some believe an MDB could not do that. Obliged to guard every penny for the sake of its mission, it would have to take the cheaper option. For that reason, sceptics say, specialist hybrid investors will not trust the call dates on MDB hybrids and will demand more spread than they would even from a commercial bank, despite the MDBs’ far higher ratings. Asked about this, the AfDB says it is “committed to the long term success of this instrument as a permanent part of our capital structure. While we can’t comment on any decision in regard to calling the instrument, we can note that any decision taken closer to the time of the call will incorporate a variety of options, including calling the instrument, not calling the instrument or substituting the hybrid with an equivalent form of capital from our shareholders. Any decision would of course be governed by the best long-term interests of the Bank.”

Until the next deal comes, experts can only estimate its pricing. MDBs — and their shareholders — will have different views on an acceptable cost.

The World Bank lends only to sovereigns at deeply concessional rates — including to many middle income countries with large government bond programmes. They may be reluctant to see lending rates rise to cover the credit spreads on hybrids.

The African Development Bank’s situation is different. From its main balance sheet, it lends to the 26 most creditworthy countries in Africa. For nearly all of these, MDB financing is the cheapest and most attractive they can get. Having to pay a little more is probably insignificant, compared with the advantage of more lending capacity.

Omar Sefiani, the AfDB’s treasurer, explained at a GlobalCapital event in February that even if the AfDB increased its hybrid issuance to the maximum allowed by the rating agencies, a third of its equity, this would only add single digits of basis points to the cost of its loans — which have a base spread of 80bp for sovereigns. And single digits is similar to the ordinary variability in senior funding spreads caused by market conditions.

Another argument is that the $10m extra the AfDB will pay annually to issue a hybrid that lets it do $1.65bn of additional lending, over an ordinary senior bond that finances $750m of lending, would knock just 4.2% off its 2022 net income — ignoring the increase in revenue from the new loans — while enlarging its loan book by 5.8%.

Asked how the AfDB planned to pay for hybrids, Hassatou N’Sele, its chief financial officer, says: “We have a dedicated treasury investment portfolio which recoups a portion of the cost. The hybrid capital will also be complemented by a hybrid capital transaction for shareholders and [development partner] friends at below market costs, and complemented by short dated private placement transactions.”

Since the hybrid, with its 10.5 years till it becomes callable, has quite a long duration, the AfDB can counterbalance this by issuing short dated senior debt, and hence reduce its funding cost.

Since reliable insight can only be gained from real deals, it is valuable that the World Bank and EBRD have committed to issuing, even if they are in no hurry. But it is clear that although the MDB hybrid market has been opened, it is far from being established. We do not yet know if enough investors will buy hybrids tightly enough to make this a viable way of raising capital.

N’Sele remains confident, however: “The hybrid capital is not a one-off transaction, it’s part of our funding toolkit. We plan to issue on a regular basis, and we have a structure to manage the cost.”

In the field of MDB hybrids, most attention this year will go to placement with shareholders.

The World Bank announced in April that the UK, Denmark, Italy, Latvia, the Netherlands and Norway had joined the original investor, Germany, in agreeing to buy individual private placements of hybrid capital for a total of about $1bn. Germany is putting in €305m, the UK £100m and the Netherlands about €50m.

Each dollar will support about $8 of new lending over 10 years, if the country takes the coupon, pegged at the World Bank’s senior debt cost, or $10 if it forgoes the coupon.

Other MDBs including AfDB are also talking to their shareholders about private hybrid issues, but have not struck any deals yet.

Arguably still more promising is the World Bank’s new Portfolio Guarantee Platform.

The US is contributing $9bn, Japan $1bn and France and Belgium about $500m between them to collectively cover the World Bank against loan losses, wherever they might arise in its portfolio. This complete flexibility makes the guarantee exceptionally powerful — and the shareholders will ensure the World Bank receives timely payment of interest and principal. Until the $10.5bn is exhausted, the World Bank cannot suffer loan losses as long as the countries honour their promises.

Losses at the Bank are minimal anyway, but the value of the guarantee is enabling it to lend more. So far, the Bank is counting on $6 of extra loans for every $1 of guarantee.

But that may come to be seen as too conservative. Unlike hybrids, the money is not paid in. But in other ways, the guarantee is superior. It could absorb loss in a near and realistic future, rather than the distant and bizarrely hypothetical Armageddon scenario when a hybrid might be written down. That makes it much more tangible and credible.

The lack of cash up front is also a great deal for the investors — one reason, perhaps, why the World Bank has managed to raise 10 times more of this kind of risk capacity so far, than of hybrids.

Another high hope is that shareholders will invest in hybrids using some of their Special Drawing Rights — the reserve asset issued to member central banks by the IMF. Rich countries have hundreds of billions of SDRs which they rarely use.

The AfDB and Inter-American Development Bank proposed in 2023 an arrangement to allow this, modelled on the IMF’s Poverty Reduction and Growth Trust and Resilience and Sustainability Trust. Those established the principle that SDRs can be used for development finance and still qualify as a reserve asset for central banks, as long as other central banks agree to provide liquidity to any participating country that suffers a balance of payments crisis.

Hybrids are more challenging, because unlike the PRGT and RST they combine elements of grant and loan, are permanent capital and involve money leaving the IMF.

However, the Fund’s board declared on May 10 that it would accept channelling into MDB hybrids, up to a maximum of 15bn SDRs ($20bn), which it will later review. The Asian Development Bank is now working on the plan too and the World Bank is open to participating.

That does not mean central banks are willing to contribute. “To lend SDRs essentially means lending your reserves, and that has never been central banks’ mission,” says the capital markets expert. “Who is going to tell them that ‘you are going to lend it permanently to this new thing called development’?”

To allow SDRs invested in hybrids to still count as a reserve asset, at least five developed countries must agree together to invest, though they can each choose which MDBs’ hybrids to invest in. They must form among themselves a liquidity support agreement, so that if any country has balance of payments problems and needs to sell the hybrid, the others will buy it. Each country would agree to buy up to 25% of the largest contributor’s holding.

The European Central Bank forbids eurozone central banks to invest SDRs in MDB hybrids because it believes this would count as monetary financing, breaching its founding treaty. That knocks out a lot of the likely investors.

The US supports the idea, but has already assigned its SDRs to other purposes, including its Exchange Stabilisation Fund. The UK has recently said it cannot channel its SDRs, but it may in future be able to provide a sterling guarantee to the liquidity support agreement.

The AfDB and IADB have found a partial way round the ECB problem. “We have developed a second layer of liquidity, for countries bound by ECB rules to be able to participate,” says Max Magor Ndiaye, director of the syndication, co-financing and client solutions department at the AfDB in Abidjan. “They could participate in the liquidity support agreement by providing a guarantee, in case there are multiple balance of payments issues at participating investing countries. We have a commitment from France and strong interest from Spain and are working with them.”

This could help in securing Japan’s involvement, since it wants the panel of liquidity support providers to be eight countries. Brazil is also considering investing. Having obtained the IMF’s approval, the MDBs will soon re-engage with governments about it.

On paper, the biggest prize is callable capital — money shareholders have promised to contribute, should the MDBs ever be at risk of defaulting on debt.

Five of the six MDBs summarised in the infographic below — the World Bank, AfDB, IADB, Asian Development Bank and EBRD — have between them $813bn of callable capital, dwarfing their $55bn of paid-in capital.

Yet callable capital is essentially left out of the internal capital models that govern how much the MDBs can borrow. The rating agencies give some credit to it, but much less than to paid-in capital.

The CAF panel, developing ideas put forward by one of its members, Chris Humphrey, an economist at the ETH Zurich university, argued that callable capital ought to be worth something. If even a modest share of it could be leveraged, it would support a huge expansion in lending.

The barrier was that neither the rating agencies, nor shareholders, nor even the MDBs themselves had a clear idea what callable capital really was. Even if one party thought it knew, it was not sure if its counterparties agreed.

The AfDB takes into account callable capital in one risk management ratio, but does not leverage it like capital. “We have a conservative and robust risk management framework,” says N’Sele. “We already use our paid-in capital to the fullest and cannot afford at this stage to introduce callable capital into our internal capital adequacy limits, unless credit rating agencies give more weight and consideration to callable capital in their rating criteria.”

The CAF panel urged the MDBs, shareholders and rating agencies to investigate callable capital and talk to each other about it. This recommendation is being carried out thoroughly.

On April 12, five MDBs published a coordinated set of reports, based partly on surveys of shareholder governments, about how callable capital worked at their institutions, what would have to happen for it to be called, and how shareholders would respond.

Shareholders and MDBs hope this new transparency will enable the rating agencies to reassess callable capital and give the MDBs more headroom to expand lending.

“We need to keep strong capital buffers to withstand shocks and play our countercyclical role when the need arises,” says N’Sele. “We cannot endorse at this stage the introduction of callable capital into our internal capital adequacy metrics. However, should the major credit rating agencies agree or recognise that the recent callable capital analysis, conducted by multilateral development banks and their shareholders, strengthens MDB capital positions, we will follow suit.”

However, anyone banking on a substantial expansion in MDBs’ balance sheets resulting from the rating agencies rethinking how they treat callable capital is likely to be disappointed.

As Moody’s, S&P and Fitch told GlobalCapital after the five MDBs published their reports, they found them interesting, but the new information largely confirmed their existing impressions.

On May 30 Moody’s gave its considered reaction. It said the disclosures on how much callable capital would be likely to be available to the MDBs quickly if needed had strengthened its view of shareholder support for those banks.

As a result, it had decided to give these five MDBs a higher score for contractual support from members. This forms 25% of Moody’s assessment of member support, along with 25% for non-contractual support and 50% for ability to support.

Moody’s has raised this score for those MDBs to the same higher level already enjoyed by the European Stability Mechanism.

Established in 2012 to act as a lender of last resort to eurozone countries, the ESM deals with a very real risk of serious, large defaults. It therefore has strong arrangements for member governments to supply callable capital quickly and securely.

Both Moody’s and S&P regard the ESM’s callable capital as superior to that of most other MDBs.

Moody’s also considers the Nordic Investment Bank and a couple of smaller supranationals to merit this high score for ‘strong enforcement mechanisms’ on callable capital.

However, while this change will improve the five MDBs’ rating metrics, and does provide some more protection against a weakening of other credit metrics, Moody’s has already given them the maximum grade for member support of ‘very high’. That earns them a potential uplift of three rating notches, meaning they could be rated as low as Aa3 on the basis of their intrinsic financial strength, and still be rated Aaa.

As things stand, they have not gained any extra room to weaken their intrinsic financial strength through more leverage, as a result of Moody’s reassessment.

Fitch gave its response on April 24, and was rather less impressed than Moody’s. Reminding market participants that it already counts 10% of callable capital from highly rated shareholders in its assessment of MDBs’ standalone credit profiles — which the other agencies do not — Fitch nevertheless pointed out: “the bulk of callable capital has not been appropriated and most shareholder countries would need to go through budgetary processes to approve funds for disbursement. This could take time and adds political uncertainty.”

Having considered all the issues, Fitch will not be changing its assessment of callable capital.

Oliveira at the Brazilian Ministry of Finance supports exploring how to use callable capital more fully. But he sees it as just one among many tools. Brazil wants to propose giving each MDB a regular capital review to establish its needs and how they could be met, including from a capital increase.

In fact, some MDBs are moving to rely less on callable capital.

N’Sele points out that in 2010, 28% of the AfDB’s capital was from shareholders rated triple-A. Now it is 10%. “Although our shareholders approved last week in Nairobi a general callable capital increase of $117bn, we cannot continue to rely on a shrinking pool of triple-A countries,” she says. “We have been working on our standalone credit profile, to mitigate that risk.”

The basic problem is that callable capital is a claim on governments, which are not 100% reliable, to hand over money in circumstances when an MDB was failing.

Considering that MDBs are some of the strongest institutions in the world, rated better than nearly all governments, and very conservatively managed, any such circumstances would be likely to be dire. Even governments that are well rated now might be ailing.

And if they could afford to pay, would they be willing to? In the age of Donald Trump and Vladimir Putin, it is hard to rely on governments to honour their international obligations.

Three possibilities emerge, all of which depend on shareholders — since only they can make callable capital more reliable.

One is to make callable capital more robust. More countries could seek parliamentary pre-approval to disburse it. They could earmark the money in government accounts and restate their commiments. But ultimately, the MDBs would still have to trust whatever politicians were in power at some unspecified stressful time in the future.

A second is to develop a new, more useful instrument. MDBs have begun working on this, calling it ‘enhanced’ callable capital. “We have to work more on producing things that are more similar to, or improvements to, the callable capital the ESM has,” says a person familiar with the talks.

Introducing a new kind of callable capital that could be drawn, with strong reliability, at a much earlier stage of distress — even as mild as a danger of losing triple-A ratings — could be a powerful support to the MDBs’ credit quality.

Oliveira expects the banks to present their enhanced callable capital ideas to shareholders. “Some banks are desperately trying to do something new,” he says. “They like to push the shareholders to create some contingent reserves that would represent at least partially the callable capital of the bank.”

The third possibility is for shareholders to forget about callable capital and support MDBs in other ways.

One obvious example is the World Bank’s Portfolio Guarantee Platform — which after all, in essence, is callable capital, supplied by a small knot of willing shareholders.

“In most countries, MDBs are not the superheroes that are going to solve all problems,” says Oliveira. “What we need is to use strategically all the [methods] to deliver impact.”