Capital markets appetite for infrastructure related assets is at an all time high, yet most project sponsors continue to prefer the relative simplicity of bank debt versus issuing project bonds.

The result has been an imbalance of supply and demand for infrastructure related paper on the global bond markets, but the anticipated rapid growth of balance sheet collateralised loan obligations (CLOs) looks set to address that investor demand.

Investors are hungry for more opportunities to buy into well-diversified portfolios of projects with ratings in the triple-B to double-B area, and synthetic CLOs are filling the gap.

There is naturally some skepticism in the market, since thus far infrastructure CLOs have been very slow to develop as an asset class.

Deals to date

Between 1999 and 2001 there were three cash CLOs backed by project loans originated by Citigroup and Credit Suisse First Boston. It was not until November 2004 that the first synthetic deal appeared out of Europe, when Depfa launched Essential Public Infrastructure Capital (EPIC). This deal achieved regulatory capital relief via a £392 million synthetic securitisation of a portfolio of UK PFI loans, in an offering led by Merrill Lynch.

In summer of 2005 Depfa was back with a second EPIC deal, this time backed by a Eu900 million global portfolio of PFI loans and bonds:This was once more lead managed by Merrill Lynch.

In November 2005 the £383 million Stichting PROFILE Securitisation I, which was also driven by regulatory capital considerations, came to market. The deal was a partnership between Sumitomo Mitsui Banking Corporation and NIB Capital. NIBC acted as lead arranger on the deal.

Late in 2006 Dexia entered the market with its WISE transaction, backed by a portfolio of wrapped bonds. And in March SMBC launched the SMART PFI 2007 deal, which is a synthetic CLO.
An average of one deal a year since 1999 isn't much to get excited about, and has left infrastructure CLOs in the position of being one of those asset classes where there are more conferences than deals. But arrangers believe that the market has reached a tipping point, and are anticipating a rush of issues this year and next.

"I know of five deals that are actively being worked on right now, and there are rumours of another half dozen transactions being considered by project lenders," comments one arranger. Dexia, Depfa and SMBC all feature as potential repeat issuers, but there are also banks working on their debut offerings.

Laying off Middle East project risk

One high profile deal – and a first for the market if and when it happens – again involves Sumitomo Mitsui Banking Corporation. And yet again it is partnering with another bank, this time Calyon, to synthetically lay off risk.
The two banks are planning to offload the risk associated with a $1 billion portfolio of their Middle East project loan books via a synthetic CLO.

Calyon can not only put in some of its project loan book alongside SMBC, but also help arrange the deal and act as bookrunner.

"In the Stichting PROFILE deal NIB sold a portfolio of project loans to us, and we synthetically created a structure where they could still benefit from the ongoing returns from those assets, through their participation in the first loss piece," says Mark Gordon, Head of Securitisation at SMBC in London.

"Both the Stichting PROFILE and SMART PFI 2007 deals involved KfW as a zero risk weighted swap counterparty within the structure," adds Gordon. "But for our upcoming securitisation of a portfolio of Middle East project loans originated by SMBC and Calyon, which we plan to launch in the fourth quarter of 2007, KfW will not be involved, since their participation in these risk transfer deals as a swap counterparty covers property loans, SME loans and PFI/PPP loans, but it does not cover Middle East project loans."

"The structure of our Middle East transaction will involve an offering of Credit Linked Notes out of an SPV, with the proceeds used to collateralise the bonds," he explains.

The motivation for a bank such as SMBC to do joint deals (such as with NIB and Calyon, though SMART was a standalone transaction), is to put together a big enough portfolio to make the high legal, rating agency, trustee and SPV costs of setting up a securitisation worthwhile. But in general the future of the market is viewed as single-originator transactions.

"In the Middle East, a lot of banks are rapidly increasing their volume of project loan exposure, but are now running up against country limits and/or sector limits," comments Khalid Howladar, Vice President, Middle Eastern and Islamic Structured Finance at Moody's Investors Service. "We have seen growing interest from originators of loans in the region looking to offload project risk into the capital markets via CLO type structures.

"Transactions are likely to be synthetic rather than cash CLOs, for ease of execution, especially where multiple jurisdictions are involved," he adds. "It is worth noting that over 95% of the Eu50 billion of balance sheet deals in 2006 were synthetics. Another reason for doing deals synthetically is that banks may wish to preserve their relationships with project sponsors by keeping loans on their balance sheet and just offload the risk."

Project pooling

One key element is that when the rating agencies look at CLOs they assign a probable average recovery rate in case of project bankruptcy. For a standalone project a bank cannot assume a recovery rate of around 70%, but a pool of projects (especially in a tried and tested market such as UK PFI) will get this treatment from the rating agencies. From the bank originator point of view this adds to the efficiency of securitisation as a risk management tool.

At the outset, each individual project is looked at closely by the rating agency. Thereafter the loan pool may be replenished, so the CLO documentation must include a comprehensive list of eligibility criteria. Though there is no independent testing of new loans being added during the life of the transaction, an originating bank will not be able to claim for a loss suffered on a loan that was not eligible when entered into the transaction portfolio.

In addition to project sponsors, the rating agencies will want to make sure that the pool does not have too much concentration in the form of exposure to individual sub-contractors. Definition of credit events is also very important.

Even in synthetic securitisations, project sponsors will see their loans in the pool, and will have to agree to the release of confidential information to the rating agencies. Relationships are very important in the Middle East, much more so than in Europe or the United States, so it generally suits bank lenders to keep their loans on balance sheet, and simply transfer the risk via Credit Default Swaps.

Cash CLOs?

True sale cash CLOs would be even more difficult, as voting rights on restructurings would be transferred to portfolio investors, which might not be to the liking of sponsors.

"Though there may not be full shadow ratings, the initial CLO pool is assessed on the basis of cashflows for each individual project," explains Matthew Job, Managing Associate at Linklaters. "For replenishment, there are strict eligibility criteria, and a variety of concentration tests to ensure that there is not too much exposure to any one project sponsor or sub-contractor. There are also geographical concentration limits."

"The five transactions done since 2004 have all featured a synthetic structure, and achieved regulatory capital relief, but the market is waiting with interest to see if anyone is going to successfully launch a cash CLO which provides the bank issuer with a cheap source of funding, in addition to moving risk off balance sheet," says Job.

"Synthetic risk transfer may not be sufficient to allow additional funding to be allocated within a bank to the project finance department," he says. "But if the projects loans have actually been sold, that would allow project finance departments to re-lend the cash raised and so increase the number of projects that they can lend into."

A cash CLO is likely to happen this year or next, though the market is expected to continue to be dominated by synthetic deals, such as the EPIC transactions from Depfa.

"Banks are continuing with their efforts to phase in Basel II, though both banks and regulators have yet to fully digest all the implications," says Andrew Bride, Managing Director at Depfa Bank.

"In 2004 Depfa pioneered the use of KfW as an intermediary in infrastructure CLOs with the first EPIC deal backed by UK PFI loans, and in 2006 we securitised a broader portfolio that included loans made in Europe, North America and Asia," says Bride. "The zero risk weighting of KfW removes a lot of the complexity of the regulatory treatment of counterparty exposure."

"For triple-B type assets, the capital allocation will probably be around 75% under the advanced ratings based model, instead of 100% under Basel I, so CLOs should continue to be an attractive way of repackaging loans, provided you have sufficient volume and a good mix of assets," Bride says.
"But for monoline wrapped triple-A rated assets, the risk weighting will fall to somewhere between 10% and 20%, so there will be less motivation for securitisation."

The Basel II effect

Securitisation drivers are changing across the board for banks under Basel II. For example, many analysts expect a drop off in the volume of Residential Mortgage Backed Securities (RMBS) offerings, since the regulatory capital relief gained is lower.

But moving project loan risk off balance sheet is going to become much more common, given a risk weighting of around 75% (of the 8% capital set-aside).

The Basel II effect is going to be magnified by the effect of very high levels of project financing being done. And the high geographic concentrations in regions such as GCC, or UK PFI, mean that many banks want to address the economic risk in the Asset Liability Management, in addition to seeking capital relief. For Dexia, its WISE deal was a short-term transaction designed to work during the phase-in period of Basel II.

The WISE 2006-1 transaction was a partially funded balance sheet CLO. The underlying pool of bonds are on the books of the Dublin branch of Dexia Credit Local. Most of the risk has been transferred via a Super Senior Swap. The junior credit default swap is with WISE 2006-1 Plc, an SPV in Ireland. WISE 2006-1 entered into a credit default swap with Dexia Credit Local, Dublin branch. The WISE 2006-1 SPV provides protection on a £63.75 million junior portion of a £1.5 billion portfolio of public infrastructure bonds guaranteed by seven monolines.

The obligations of the WISE SPV are collateralised by a Gage Especes Agreement, under which WISE deposits the proceeds of the sale of the notes as cash collateral.

WISE in turn issued three classes of notes with a 10-year average life. £30 million of bonds rated AAA by S&P priced at a 30bp spread. The £22.5 million AA rated tranche priced at 37bp, while the £11.25 million AA minus rated tranche had a 39bp spread. Lead arrangers were Dexia Capital Markets and Credit Suisse.

The mix of the portfolio was improved by including seven PFI bonds, relating to projects in sectors such as schools, hospitals, and waste disposal plants. There are also 21 bonds that were issued by regulated utilities in the water, electricity or gas sectors. The total size of the bond portfolio is £1.47 billion, and there are replenishment provisions over the life of the deal.

Balance sheet CLOs

It is balance sheet CLOs that are the hot new asset class in infrastructure, though project bonds will continue to feature.

The series of Breeze transactions arranged by HVB have successfully refinanced the debt associated with wind farms. The recent Breeze 3 deal was a Eu455 million offering backed by a 350MW pool of wind farms located in Germany (90%) and France (10%).

Breeze 3 was structurally very similar to the previous Breeze deals, though there were less projects still under construction in the pool. The first two Breeze deals were unwrapped, appealing to specialist investors looking for some yield pickup, but Breeze 3 came as a triple-A wrapped deal featuring MBIA. HVB is already working on Breeze 4, which could include solar plants for the first time in the asset mix.
German wind farm developer Plambeck has also accessed the capital markets for its funding needs, with the Eu102 million Alte Liebe 1 offering led by Dresdner Kleinwort.

But these deals pale in comparison with some of the big deals linked to the current infrastructure M&A mania, where a securitisation take-out using project cashflows has formed part of the strategy from the outset on many deals.

Last year Thames Water was sold by RWE and acquired by a consortium led by Macquarie, which informed water regulator Ofwat of its intentions to move forward with a securitisation backed by water payment cashflows. This deal is expected to close during 2007, and could raise £2 billion.

Also in 2006, a Canadian consortium led by Colonial First State acquired Anglian Water in the UK for £2.2 billion. It too is looking at a securitisation.

And the market is still waiting on the £8 billion securitisation of cashflows from UK airports operator BAA plc, which was acquired last year by Ferrovial (the deal has been held up by changes to the asset mix and negotiations with the regulator and existing bondholders, but is expected to close by the end of the third quarter).

But these blockbuster deals tend to be one-offs, and could make 2007 a very big year for infrastructure securitisation volume, rather than be a sign of a steady pipeline to come.

In contrast, bank balance sheet infrastructure CLOs look like an asset class that will deliver a steady flow of deals to the capital markets in the coming years. "In the project finance market bank lenders remain dominant, with project bonds typically making up 10% to 15% of global project financings. Bonds offer the advantage of very long tenors, but banks are also increasingly lending at long tenors, as well as offering fine pricing. Clearly banks don't want to hold all that long dated exposure, and more lenders are managing their exposure with CLOs. The focus to date has been very much on regulatory capital relief, so deals have been synthetic CLOs, but we may start to see some cash deals to raise funding," says Laurence Monnier, Senior Director at Fitch Ratings.

Securitisation is thus becoming an efficient way in which to manage balance sheet risk exposure, as well as addressing the cost of regulatory capital requirements. And if cash deals start delivering cheap funding then the market will take another big step forward.

There have been some assertions that project lenders that have done securitisations have been responsible for over-aggressively driving down pricing in the global project finance and PFI/PPP market. But given the vast majority of project loans have been syndicated to banks that do not have securitisation programmes that cut in project debt pricing has more to do with high bank liquidity than the offloading of project loan risk.