Supranationals to copy African Development Bank’s securitization breakthrough
Private investors are reluctant to co-invest with development banks in poor countries. A potentially exciting alternative is securitization, which allows investors to take slices in whole portfolios and could expand the MDBs’ lending firepower.
It is a truism in development circles that engaging the private sector is critical to achieving the scale of investment needed to meet the Sustainable Development Goals.
The numbers of trillions in various estimates tend to become a blur but the UN’s Sustainable Development Solutions Network reckons $5tr-$7tr a year until 2030. That compares with $11.8tr of total global bond issuance in 2017, according to Dealogic — making it look a tough target.
But compared with the multilateral development banks’ capacity to lend, even $1tr a year is off the scale. The two largest MDBs, the European Investment Bank and World Bank Group, disbursed €58bn and $44bn last year, respectively.
No wonder progress towards the SDGs has been mixed. The proportion of families living on less than $1.90 a day has fallen since 2000 from 27% to 9%, according to the UN’s latest progress report on the SDGs. But after a long decline, the number of undernourished people actually rose in 2016, from 777m to 815m, due to conflict, drought and disasters linked to climate change. More than 2bn people still lack even basic sanitation.
The ‘From Billions to Trillions’ vision paper put out by seven leading MDBs in 2015 laid heavy emphasis on crowding in private investment. That is indeed growing. Standard & Poor’s, the rating agency, reported in September that the MDBs had claimed $164bn of private co-investment mobilised in 2017, above all by the EIB.
The trouble is, 97% of that went to high and middle income countries. “We believe it’s because private sector investors find it too risky to go to low income countries,” says Alexander Ekbom, analyst for MDBs at S&P in Stockholm. “The investment climate is tough when it comes to government concessions, the rule of law, transparency, regulation. You typically find risk sharing structures being less available or costly and official co-financings scarcer.”
These problems will take time to solve. But there is another, potentially very promising, route for the private sector to support MDBs without needing to invest directly in specific deals in poor countries — by taking slices of their risk through securitization. This enables the investor to benefit from a diversified pool of assets and to choose a certain level of risk.
The African Development Bank completed a trailblazing $1bn securitization in September, the first in the sector to involve private investors. Its peers are likely to follow suit.
The potential to make the MDBs’ existing capital go further by recycling assets has barely been explored but the AfDB has shown that, although the first transaction was difficult, it can be done.
It has shed the mezzanine layer of risk on a $1bn portfolio of 40 loans to banks and project finance transactions, freeing up 65% of the capital so that it could make $650m of fresh loans. The assets stay on its balance sheet but are partially guaranteed by a fund managed by Mariner, a US hedge fund, and by Africa50, an infrastructure fund owned by African governments. The European Commission has taken a senior mezzanine slice of risk above them.
“I believe we will see a follow-on transaction within 12 months,” says Andrew Hohns, CEO of Mariner Infrastructure Investment Management in New York. “There will be a very significant amount of interest on the part of other MDBs to unpack the deal, identify the best practices and specialisations for their balance sheets. Though you may only see one or two transactions in the next six to 18 months, you could well see five to eight in the next 24 months as the big wheels of these organisations turn to internalise this technology.”
Some MDBs are already working on specific portfolios. “The potential to expand it is huge,” says Juan Carlos Martorell, co-head of structured solutions at Mizuho, which arranged the AfDB deal. “For other MDBs this is a very interesting technique. They are transferring risk at a price which is competitive to enhance their capital metrics. The fact that the rating agencies have been involved and assessed the merits of the transaction means that the first mover has created a path for other MDBs to follow.”
The rating agencies have decided how much capital relief to give AfDB, according to their models, which determine whether it can still be rated triple-A.
S&P, which has the most quantitative approach, took a long time to decide how much capital relief to give the AfDB. But Ekbom says: “When we developed the process we did it with the intention of being replicable. We don’t treat one institution differently from another.” Whether a second deal could be processed more quickly would depend on how complex it was but Ekbom says: “I would hope that, as with anything we do in life, doing it the second time would be quicker than the first time. Not necessarily every MDB will do one but there will be a couple more coming.”
Hohns says he believes the other MDBs are now at the stage of willingness to explore securitization, even if not at the stage of readiness.
“We are looking to see to what extent we could securitize part of our portfolio,” says Alain Godard, director-general of risk management at the EIB in Luxembourg.
The influence of the AfDB’s transaction
need not be confined to development banks. “It’s something that can maybe be
replicated across the continent,” says Olivier Eweck, director of AfDB’s
syndication and technical services department in Abidjan. “We hope it will open
the door for commercial banks in Africa to structure this type of transaction.”
The AfDB itself is determined to go further, hoping to scale up securitization and try other techniques. It was already leading the way on risk transfer in 2015, when it engaged in a three-way deal with the World Bank and Inter-American Development Bank to exchange $4.5bn of each of their sovereign risk exposures. Each got a better diversified portfolio and could hence make more loans.
Deals between MDBs and other public sector entities make sense because all involved understand the issues to begin with and are willing to collaborate. Since then, a few other transactions have been attempted and a smaller number completed.
A couple of years ago, the EIB guaranteed some of the World Bank’s exposure to the Ukrainian government. “They wanted to do more lending in Ukraine but were up against their limit,” says Godard. The EIB and World Bank are now working on some further bilateral exchanges of sovereign exposures that have become inconveniently large.
But involving private investors is a bigger prize. Asked how much in future balance sheet optimisations could add to an MDB’s balance sheet capacity, Tim Turner, the AfDB’s chief risk officer, says: “I think 40%-50% is not unreasonable. The rating agencies have cautioned us that we need to move at a prudent pace and learn from experience so we are experimenting with several pilot transactions.”
The AfDB will sign a credit insurance deal this year on loans to the financial sector.
PIONEERS TO ENCOURAGE OTHERS
The sovereign assets held by MDBs often yield too little to be attractive to private investors, who can buy most countries’ government bonds if they want to anyway.
“I believe the most fertile area is in the non-sovereign lending book,” says Hohns. “It’s hard to get precise figures but I would estimate the MDBs have $200bn-$300bn of non-sovereign loans. Private investors are best suited to take exposure to seasoned risk rather than construction risk so broad securitization of their non-sovereign, non-construction loans could create perhaps 60%-80% of further capacity for the MDBs to lend.”
So far, this is theoretical. Few investors are willing to do what Mariner and Africa50 have done, which requires not just familiarity with synthetic securitization but a hearty risk appetite and willingness to spend months working on a deal, including weeks of asset-by-asset due diligence.
But Martorell thinks bees will start to arrive at the honeypot. “You probably need to have investors who can invest in size like Mariner, that can do $150m to $200m,” he says. “There are five or six investors who can do that and regularly do synthetic deals. Initially they will assess the asset class and issuer, but if they like the issuer and do due diligence I think they will be interested.”
The due diligence will also become less intense the more investors build up experience with the asset class and the larger and more granular securitized portfolios become.
Getting private investors comfortable with these risks is something Godard at the EIB has been building up to for a decade since he started a database called GEMS to track the credit performance of emerging market development loans. About 20 MDBs and national agencies are now sharing their information through GEMS and they are discussing how to open it to the private sector.
Borrowers’ names are already encrypted in the database and searches are not very precise, to protect borrowers’ privacy. But Godard is still convinced GEMS could be a crucial guidebook for investors, letting them know the territory is safe.
The EIB is working with Dutch pension funds on setting up a new fund to co-finance loans with MDBs including the EBRD, IFC and AfDB.
“Data is one of the three barriers investors have before investing,” says Godard. “They will be using the GEMS data to assess the risk/return. That’s why these pension funds are interested. They see that what they thought would be costly in terms of risk in these markets is not that costly in view of the experience we have.”