The private sector has been central to both operating rail lines and leasing rolling stock ever since the UK privatised its rail sector in the 1990s. Perhaps because of this early involvement, UK rail, with the exception of light rail, has not been fertile ground for public private partnerships (PPPs). The UK’s Department for Transport (DfT) tenders franchises for rail routes to train operating companies (TOCs).

Those operators lease trains for the duration of their franchises from rolling stock companies (ROSCOs), with Angel Trains, Eversholt Rail and Porterbrook dominating the market. The TOCs build leasing costs into their models, and tend to assume most of the operating risk attached to the trains. The ROSCOs finance the rolling stock themselves, historically through bank debt. The ROSCOs have been attractive investments for financial sponsors, and their acquisition financings have featured big bank groups.

But the ROSCOs have been less effective at upgrading the UK’s fleet. Indeed, 53% (or 6,666 vehicles) of the 12,691 rail vehicles in service were built before British Rail’s privatisation in 1994, 24% date to the 1980s, and 13% date back to the 1970s. The UK Long Term Passenger Rolling Stock Strategy for the Rail Industry estimates that the UK will require between 13,000 and 19,000 new electric vehicles by 2042, or between eight and 12 new electric vehicles on average per week.

“An alternative financing solution was needed because the fleets required were so large and it was unlikely that a ROSCO could deliver [the billions needed] under the traditional structure,” says one adviser active in UK rail. “The challenge was that such a large investment needed long-term stability but the established TOC-ROSCO leasing model just did not provide that.”

ROSCOs struggle to raise capital for expensive fleet additions when their clients’ franchises only run for up to eight years. Virgin, for example, operated the InterCity CrossCountry franchise from 1997, and procured a 34-carriage fleet of Class 220 Voyager to run on the route, butlost the franchise to Arriva in 2007. TOCs have little incentive to look beyond their franchise period when procuring rolling stock. Hell-or-high-water leases mean ROSCOs receive lease payments regardless of train performance

But political pressure for a new procurement model has grown: “There has been significant growth in passenger numbers and passengers expect trains to be available and reliable,” the adviser adds. “The DfT assumed most of the risk under the traditional procurement model for rolling stock so it was important to help improve services by transferring more risk to the private sector.”

Reassessing risk

The DfT’s approach shifted towards PPP on the larger procurements, including Inter-City Express Programme (IEP), Thameslink and Crossrail. The DfT would tender for a special purpose vehicle (SPV) to design, build, finance and maintain (DBFM) a fleet of rolling stock on a long-term basis, beyond the life of a single franchise. The SPV would own the fleet and its lenders would be assured of lease payments that went beyond the term of the franchise.

The DfT has used two PPP structures: one, which appeared on the IEP first phase, where it was the direct counterparty under an availability-based concession; and the other, which featured on Thameslink, where the DfT creates a framework that assures lenders that the end of a franchise does not disrupt their repayment streams.

“The government was keen for the private sector owners to assume some of the performance risk on the rolling stock,” explains Alex Carver, a partner at Freshfields Bruckhaus Deringer, which advised the DfT on Thameslink. “Under the Thameslink PPP, for instance, the owner assumes a certain amount of liability for poor performance, in the form of reduced rent payments or even resulting in termination. Because of this, a PPP incentivises the private sector owner and its lenders to evaluate rigorously the rolling stock, as well as making it possible to include related depots as part of the procurement.”

Mark Westbrook, head of transport at John Laing, adds: “PPP allows the DfT to tap alternative financing markets and provide greater whole-of-life risk transfer to the private sector party responsible for delivering the assets. I would not say it will be the only solution, however, but it certainly is an option for authorities to consider.”

Creating comfort

The ROSCOs and lenders, however, were cautious. The first attempt at a conventional rail PPP – the £295 million Borders Railway in Scotland – collapsed because of contractual wrangling, including Network Rail’s unwillingness to assume risks such as engineering works or cost overruns on a project built by a third party, meaning that the project company took all that risk. Borders eventually went back to the public sector.

Roneil Thadani, a director for infrastructure and transportation at Standard & Poor’s, says that ROSCOs struggled to cope with the possibility of a TOC default. The collapse of Network Rail’s private predecessor Railtrack, the demise of the London Underground PPP and the ill-starred West Coast mainline franchise tender made these worries far from hypothetical.

“The government has provided strong support, both through supportive policies but also directly through the contracts,” Thadani claims. “As such we believe that it would step in should a [default] scenario occur and ensure payments are honoured, as was seen in the East Coast mainline. This gives the ROSCOs more certainty.”

Carver points out that rolling stock is usually the responsibility of TOCs, whose franchises are shorter than the likely economic life of the units, whereas a PPP tends to run for 25 to 30 years: “So the DfT used a Section 54 agreement to give long-term comfort that the trains would be leased for a longer term.”

Under Section 54, the department agreed that, when a franchise terminates, the department will guarantee that the rolling stock would re-leased on the same terms by any subsequent franchise operators.

The Thameslink project structure


Source: XLT

Some success

IEP’s structure is slightly different, in that the department’s obligations to the project company and lenders are more direct. IEP is a two-stage project, and its first phase involves replacing ageing Great Western mainline intercity trains, and the second stage covers the replacement of the East Coast mainline fleet.

Agility Trains, whose sponsors are Hitachi and John Laing, won the contract to supply the first 57-train IEP batch, and closed its financing in 2012; the first rolling stock contract completed on an availability-based PPP structure.

IEP 1 featured a £2.2 billion ($3.48 billion), 29-year package. JBIC provided a £1 billion direct loan alongside £800 million in uncovered commercial debt, arranged by BTMU, HSBC, Lloyds and Mizuho. There was a £150 million NEXI-covered loan, a £235 million direct loan from the European Investment Bank (EIB) and £300 million equity bridge loan.

IEP includes a no-train-no-pay mechanism based on the daily availability of vehicles, a parent company funding deed and a liability cap. IEP has a master availability and reliability agreement (MARA) and a train availability and reliability agreement (TARA) between the DfT and the winning SPV, which outlines all the DBFM provisions and various guarantees, and the SPV then enters into simpler leasing contracts directly with TOCs.

IEP was the first time a MARA and a TARA have been used and, insiders claim, will also probably be the only appearance of those agreements. The contracts were long and complex – the MARA for IEP 1 was 669 pages and the TARA 348 pages, increasing to 725 pages and 367 pages, respectively, for phase 2. For Thameslink, the department wanted to use a simpler approach, even if it required more credit work from the SPV’s lenders.

Thameslink is a DBFM contract for 1,140 Desiro City EMU train carriages, for which the project company is a consortium of Siemens Project Ventures, 3i Infrastructure and the Innisfree PFI Secondary Fund 2. Its £1.59 billion financing closed on 26 June, just ten days after the formal award of the contract. The Thameslink PPP asks the project company to assume similar types of risks, with poor performance reducing payments.

SMBC, BTMU, Lloyds and KfW IPEX arranged the £1.15 billion 22-year commercial bank financing, with KfW contributing £150 million. The package also had a £425 million 20-year loan from the EIB and £180 billion in equity and sub-debt, split between the sponsors.

The hurdles that the two deals faced in reaching close said more about the state of financial markets than the bankability of the structure. Thameslink’s commercial tranche, for example, was originally meant to have pricing starting at 175bp over Libor, though unsettled debt markets pushed that number up. Siemens’ strong credit rating, plus the involvement of the EIB, resulted in a A3-rated financing that had pricing of about 250bp.

Westbrook says the biggest challenge for IEP 1 was the reduced financing capacity in the aftermath of the 2008 financial crisis: “The benefit of using the PPP model in the funding was that it opened the deal to a wider group of infrastructure debt providers. There were some significant modifications made to the typical PPP framework because rail is a heavily regulated sector that requires bespoke solutions.”

Westbrook adds that Agility is currently finalising the financing for IEP 2. The commercial contract is similar to the earlier deal and is scheduled to close in the first quarter of 2014.

Preferred model?

So does the success of IEP and Thameslink pave the way for further PPP in the UK rail sector? There is some potential. Abellio ScotRail issued a tender in late 2013 for up to 400 new electric trains, including the option to finance the design, manufacture, testing, commissioning and introduction of the fleet. The cost of that order could be anything between £100 million and £600 million. Elsewhere, expectations are minimal.

Suzanne Tarplee, a partner at Stephenson Harwood who advised on West Coast, ScotRail, Essex Thameside, Great Western and Thameslink, says that manufacturers are yet to be convinced by PPP: “From [their] perspective, they are comfortable with the leasing model and the risk profile. The PPP structure has thus far proven more expensive and taken considerably longer to deliver.”

The gap between tender and financial close for both IEP and Thameslink was the best part of five years. “The trains need to be built and tested and this can take years,” says Thadani. “Also the cascading plans for the old stock require extensive planning. As such, I can see why rolling stock orders may need to be made prior to franchises being awarded to TOCs, as was the case with Thameslink.”

Tight deadlines have already resulted in the DfT’s decision to ditch PPP on the £1 billion ($1.5 billion) Crossrail rolling stock contract. The 60-strong fleet for Crossrail, which runs east-west across London between Heathrow and Canary Wharf, was meant to use a DBFM contract. But faced with the possibility of the Crossrail infrastructure being completed before its rolling stock, the government reverted to procure the rolling stock directly.

The other sizable future rail project, of course, is the £42.6 billion HS2. As Carver notes, PPPs are complex and highly bespoke and so tend to suit larger deals. HS2 will be a massive project overall and would not be practical as a single PPP. But direct government procurement of the entire project or using the existing ROSCOs to procure rolling stock will stretch the resources of both.

“That said, it will involve procuring a fleet of new and probably highly sophisticated rolling stock so could possibly suit a PPP mechanism,” Carver suggests. “Even so, I don’t expect to see PPPs being widely used on smaller rolling stock procurements.”

Indeed, ROSCOs enjoy strong financial support, even as they passed from bank ownership into the hands of private equity firms in the late 2000s RBS sold Angel Trains to a group that included Arcus Infrastructure Partners and AMP Capital Investors; Porterbrook went from Abbey to a consortium of Antin Infrastructure, Deutsche Bank and OpTrust; and Eversholt, formerly HSBC Rail, was bought by infrastructure funds 3i, Morgan Stanley and Star Capital Partners.

The ROSCOs are still bringing in new orders. Porterbrook is providing a new £200 million Electrostar fleet for Southern and Eversholt recently completed a £600 million refinancing involving 11 banks.

“The ROSCO market is dominated by three large players and bank finance has been an important source,” says Thadani. “These established leaders, coupled with supportive government policies, lack of available rolling stock in the UK and long-term contracts, make ROSCOs an attractive investment.”

The capital markets are also on the radar for the ROSCOs. Angel Trains recently issued a 10-year £60 million private placement, priced at 145bp over the ten-year Gilt, to retrospectively fund its London Midland/Trans Pennine Express Desiro orders, as well as an 18-year £60 million amortising bond. The spread on the ten-year piece was the lowest on a ROSCO issue to date.

Bond financing has certainly caught the eye of investors, although there are doubts as to whether bond debt is available suitable in larger quantities, whether on PPPs or otherwise.

Westbrook says: “For the larger greenfield schemes it is really a question of whether there is sufficient capacity in the bond market for such greenfield construction risk. The debt funding requirement of the Intercity East Coast, for example, is around £2 billion and the capacity of the bond market to fund that amount of greenfield construction risk is untested.” Refinancings, however, might provide an opening.

As such, a pipeline of small- to medium-sized deals would be within the financing capabilities of the ROSCOs. “I think the DfT will look at each deal on its own merits and assess the right financing model,” the rail adviser says. “PPP may have a future role in UK rolling stock, potentially on the larger deals, but the traditional ROSCO financing will continue for the smaller deals.”