There are many superlatives that could be used about the Tours-Bordeaux high-speed rail project, and many flattering adjectives: prestigious, gigantic and sophisticated spring to mind. But is ‘final’ one of them?

The participants of this mega project, which at €7.8 billion is a likely contender for the biggest project finance deal to close in 2011, succeeded in overcoming considerable financial, economic and political obstacles [Transactions Database]. They demonstrated that the world’s biggest, most complex and multi-party traffic risk concession could be closed despite being tendered in the middle of the crisis. This is no mean feat; yet some have suggested that another project in France or Western Europe will not be procured in quite the same way as Tours-Bordeaux (or Sud Europe Atlantique ['SEA'] as its participants invariably call it) again.

That is not because anybody did a bad job, but for a handful of other reasons. On a simple level, there is unlikely to be a demand for a high-speed rail project of this size in a developed country again (and the UK’s High Speed 2 will not be project financed). The Tours-Bordeaux project was procured during a financial crisis that demanded extraordinary assistance to guarantee its viability, and then again, the challenges encountered by the project gave some of the parties lessons to take with them for future projects. Anyway, how often is it that a PF deal gets entangled in the presidential aspirations of an ambitious politician, much less the race to become managing director of the IMF?

Background and Structuring

LGV Tours-Bordeaux alignment. Black line indicates classic rail line and blue lines indicate links between HSR and classic lines. Source: RFF

The issue with Tours-Bordeaux was really how it would be done, not if. The line has been planned since the 1990s as the first section of the Sud-Europe Atlantique rail corridor that will eventually stretch all the way to the Spanish border, connecting at its northern end with the existing LGV Atlantique Paris-Tours line which fully opened in 1990. In the mid-2000s, rail authority Réseau Ferré de France set its mind to considering how to procure 340km of new line, including 303 km of high-speed line.

As Europe’s high-speed rail evangelist, France had used conventional procurement for all previous rail lines besides the much smaller Perpignan-Figueres concession. Yet when the time came to settle on the financial model, RFF chose not a contrat partenariat (availability-based PPP) but a concession contract, which had indeed been used extensively in the transport sector, but almost exclusively on toll roads. Why?

In the absence of an official explanation, one can look to recently closed toll road deals, such as the A41 and A65, which had reached financial close without a public sector subsidy, contrary to previous expectations. This may have filled the transport ministry with confidence that SEA could also be a money-spinner, in spite of the glaring differences in complexity between a toll road and a high-speed railway.

Another factor was likely to have been the great newness of the contrat partnariat in France. The concession contract had been used in France for decades to procure infrastructure, whereas the Partnerships Ordinance that brought in a streamlined availability-based contract system had only been passed in June 2004 and the Ministry of the Economy’s PPP unit, MAPPP, had only been established in May 2005.

A contrat partenariat was however chosen for the smaller GSM-R project which hit the market in 2006. In retrospect, the reduced interfacing offered by the legislative framework for PPP, in which the state generally contracts with the private partner for all aspects of the contract, might have helped SEA be procured more quickly.

Before the project came to market and the concession option was decided upon, Natixis was hired as financial adviser and carried out a value for money study, in an appointment announced by RFF in July 2006. In the light of the global financial financial crisis that was to hit, Nicolas Bourgouin, director at Natixis’ project finance division, says: “Phasing the project was considered not after the crisis but at the early stage of the project – it was not decided if SEA was one or two projects.” Although it was clear then that the project would require substantial subsidy from the government, the exact scale needed would only become clear later.

Public enquiries and the declaration d’utilite publique necessary to advance the project were carried out in two stages, first for the southern (Angoulême-Bordeaux) section and then for the Tours-Angoulême section. From 2001 to 2003, the conceptual design for the first section was carried and from 2003 to 2007 for the second. Angoulême-Bordeaux received its DUP decree in July 2006, but that for Tours-Angoulême was not issued until June 2009, fully two years and two months after the project was brought to the market.

Procurement

It was in March 2007 that RFF published a contract notice for SEA and sought expressions of interest. Capital expenditure for the project was bracketed between €4.5 and €5.5 billion. A Spanish consortium was initially talked about in the early stages, but by late 2007, it became clear that no such team had been prequalified.

By November, three consortia had been shortlisted. Eiffage on its own had been prequalified, while the other two prospective bidders had teamed up with infrastructure funds. Bouygues had joined forces with HSBC, Barclays, NIBC and Meridiam, and Vinci had picked AXA and the French state bank, the Caisse des Dépôts et Consignations via its fund CDC Infrastructure. CDC thus participated in the competition from the outset, unlike on the Balard PPP where it was imposed on the preferred bidder after selection.

The first round of offers went in in September 2008 and a preferred bidder was expected to be picked in 2009. But meanwhile, trouble was brewing on the global financial markets.

Danger signals

“The size was the point,” Nicolas Bourgouin of Natixis explains. “The first problem and the first thing we had to think about with RFF was the size, and the project was launched before Lehman Brothers’ insolvency. It was not easy before Lehman Brothers. It was impossible just after. When I say impossible, that’s with a pure private financing on the whole financing plan.”

The credit crunch that followed the bankruptcy of Lehman Brothers in September 2008 affected all financial markets, and project finance was not immune. As 2009 dawned, the public sector and Natixis realised to their dismay that bank liquidity had to a great extent dried up, and the cost of it had increased. Recklessly priced PPP deals were soon to become a far from fond memory.

It was not merely SEA which was the problem. The French state, living up to its reputation for grands projets, had launched no less than three multi-billion transport PPP projects to the market – Tours-Bordeaux, the €3.4 billion LGV Bretagne-Pays de la Loire and the €4 billion Canal Seine-Nord Europe, plus the €1.2-1.6 billion Nîmes-Montpellier high-speed rail bypass. To say nothing of smaller projects in social infrastructure. The moment of maximum strain in the project finance market was also the moment when the French were demanding the most liquidity from it.

In the late winter of 2009, IJ representatives crossed the Channel and gathered at Louis XIV’s Trianon palace in Versailles to pay their respects to the Sun King of PPPs, François Bergère, secretary general of MAPPP since its inception in 2005. Far from panicking, the government with the backing of President Nicolas Sarkozy had drawn up a multi-point action plan to carry its PPP projects through the crisis. The projects were seen as vital for economic stimulus, and Sarkozy had singled out SEA in a speech.

Bergère told IJ at the time, “The government realised that the whole programme was running the risk of falling down if nothing was done to ensure that projects that had been prepared over the last two-to-three years were not now given that essential bit of extra support - that could take the form of funding support.”

High-speed remedies

In early 2009, the Ministry of the Economy, Finance and Industry set up a €10 billion loan guarantee scheme to cover up to 80 per cent of project debt on PPPs. Meanwhile, an €8 billion fund was established through the CDC.

At this stage, MAPPP, which was to oversee the loan guarantee, appointed Compagnie Benjamin de Rothschild as financial adviser. Overseeing the implementation of this weighty commitment became the responsibility of their head of project finance Jean-Francis Dusch.

“There was limited liquidity in 2009 and there was, among others, a deal coming in for potentially three billion of debt, so it was really a trigger [for deploying the state guarantee],” Dusch explained. “As soon as the tender was ready we worked on the term sheet for the state guarantee with MAPPP.

“Because there were a number of large deals in the French PPP market it made even more sense to make sure such a guarantee was in place, as some involved several consortia to provide committed financial offers, hence effectively tapping the bank market several times for the same deal.”

MAPPP also introduced more flexibility in financing requirements for projects in bidding. Specifically, bidders would be allowed to submit final bids with only around 50 per cent of financing committed instead of 100 per cent as previously expected. For Natixis as financial adviser, this was critical to ensuring that the project could be procured on a genuinely competitive basis without driving away any of the bidders.

Bourgouin commented, "I think the most important point for us was to adjust the level of financial commitment for each bidder, because after the crisis it was impossible to have three times 100 per cent on final bid like it was before. So we had to change the financing requirements to have three competitive bids and to have a real competition between the bidders. We worked a lot on that, on what we could ask for."

The greatest threat to the project, it seems, loomed in late 2009 as the three bidders were preparing their final bids, and the challenge was to ease the passage to that stage without handing over so much public money and risk that the PPP model lost its purpose. Bourgouin adds, "The biggest challenge was to define financial requirements in line with market conditions. It was really a challenge for RFF to decide what could be done by the three bidders for the late stage of tender process."

Not that the palliative measures were leapt on that gratefully by the markets. Dusch recalls, “Banks, rightly or wrongly, considered the state guarantee as a difficult instrument to live with – despite the straightforward nature of this credit enhancement, the state had to put in a few conditions so that banks would still be a responsible party to the debt finance package. One has to admit it created additional inter-creditor issues to an already complex deal. Overall, we had to find the balance between the banks’ and the state’s respective constraints.”

Looking back on SEA in June 2011, Bergère says, “The message channelled to the market that the government was serious in its will about seeing them come to signing and implementation. This package exemplified or supported the political will to take the necessary steps to ensure these projects found the required funding.

“We were reasonably worried that we would have to make [do] with a limited amount of candidates, industrial sponsors or on the financial side… this didn’t happen.”

So how precisely did the global financial crisis change the financial structure of SEA? First of all, most of the debt became covered by one of three separate guarantees. Seeing the commercial banks taking up MAPPP’s loan guarantee, the EIB, which was preparing to lend more than it had ever done to a single project, also sought cover. €400 million of its €600 million senior debt became guaranteed by MAPPP.

Secondly, the state became a lender to the project, in addition to its roles as grantor, equity investor in a competitive bidder and loan guarantor. The CDC, as a state lender under the crisis remedial programme, also wanted a guarantee for its loan, except that this was to be guaranteed by RFF, who in turn hired Compagnie Benjamin de Rothschild to advise on this too. Thus a third guarantee arose, similar in form to the EIB and commercial guarantees but legally separate.

Thirdly, and to a lesser impact, the balance between debt, equity and subsidy altered somewhat. With rising liquidity costs, a gap emerged and was plugged by slightly increased subsidy.

The measures, and their careful implementation, did the trick. Three bidders made it to the final bid stage in December, and in March 2010 RFF announced the Vinci-led LISEA consortium as preferred bidder. Financial close, IJ News confidently predicted at the time, was set to “follow before the summer”.

The long railroad to close

Vinci had a majority stake in the project, with the remainder split between the CDC (25.82 per cent) and AXA Private Equity (19.48 per cent). As laid out in the documentation, LISEA had about 50 per cent financial commitment, and was supported on the lending side by Crédit Agricole, BNP Paribas, Sanander and Société Générale. The consortium and its financial advisers, CA CIB and SocGen, were seeking five commercial banks to provide a total of about €1.7 billion senior commercial debt.

Before long, it became clear that the close was going to take longer than the predictions of early 2010. Summer came and went. In mid-September, the consortium finally held their kick-off meeting and MLAs began to sniffed out. CA CIB, BNP Paribas and SocGen handled financial documentation.

By the autumn, IJ was reporting that the consortium and RFF had “sent out a strong message about timing” and set a deadline of 1 December, but this too proved optimistic. It was not that the project was unattractive to lenders; far from it, the demand risk concession offered attractive pricing on the project risk debt and as markets calmed down, appetite was returning. But the sheer size of the project, unprecedented in transport project financing, was magnifying problems that on a smaller project might have been resolved more quickly.

Jean-Francis Dusch notes, “Issues, when they are on a large scale, make it much more difficult. Swapping a three billion underlying asset is not the same as swapping 500 million. Everything was big and making issues more obvious to a certain degree.” Others agree.

But by the start of 2011, LISEA had its nine banks. The core bank group had been joined by BBVA, Dexia, Mediobanca and Unicredit. BBVA, Dexia and SMBC accepted project documentation roles. Atkins was on hand to provide technical due diligence, while Steer Davies Gleave offered traffic and operational advice. Under the tender, the sponsor was to bear traffic risk, with access charges paid by each train using the line.

Some parties to the deal blame the public side for causing unnecessary delay. For example, AXA was keen for the consortium to receive yield payments during construction in order to make shareholder payments. Although the scheme was unprecedented for a French deal, the banks were apparently not opposed to it. Eventually an agreement was hammered out. Then RFF put its foot down and the deal was abandoned.

National and regional politics also intervened. Ségolène Royal, former presidential candidate and a hopeful for the 2012 election, was president of the regional council of Poitou-Charentes. In February, Royal began announcing that she would not contribute the €257 million grant due from her region – the third largest after Aquitaine and Midi-Pyrénées – because it was the central government’s responsibility to provide infrastructure.

In April, Royal said she had authorised a €95 million loan instead. She said, “Every region has the freedom and responsibility to use its public funds in the best way possible and I have decided to make a loan… Since the LGV is subcontracted to a private enterprise, it is normal that when the enterprise makes a profit, it pays back the contribution.”

Once again, the name of SEA was being exchanged at the highest levels of the French state. The government was not prepared to sign the concession contract without all financing, private and public, committed. Bordeaux mayor and political opponent Alain Juppé thundered that “the fierce resistance of Poitou-Charentes could block everything”, while transport secretary Thierry Mariani pointedly called Royal “the only problem” preventing the project from closing.

However, though the theatrics were colourful, the actual effect was minimal. The subsidy withheld amounted to just over three per cent of the total project cost, and were to be provided in all cases by RFF, not the regions. By mid-May, Prime Minister François Fillon had relented and said that the contract should be signed “on the basis of the finances available”.

The state would have to pick up the bill, though Fillon hinted that the region would have to lose out on future funding. In a public statement, he added that RFF was to sign the contract “before the end of the month of June”. All parties had been given their marching orders. But the fact that the Minister of the Economy and Finance, Christine Lagarde, was jetting around the world on a campaign tour for the position of IMF managing director, did not help.

Apart from this sideshow, the final few months also saw a lively discussion between RFF and lenders over the timing of a swap on the bank debt. Lenders apparently felt that the authority was being too precipitate and exposing them to interest rate risk as a result.

June came, and two more challenges were overcome – finding a room at White & Case’s Paris office large enough to hold 60 people, and a safe large enough to hold several thousand pages worth of documentation. Financial documentation was finally signed around the midnight hour over 14/15 June, and the concession contract was signed by Vinci and RFF on 16 June. At the same time, Meridiam Infrastructure finally agreed to buy the 21.31 per cent stake in the SPV allocated to Vinci’s SOJAS vehicle, after months of discussions with Vinci.

Financing

Nine commercial banks are providing €1.7 billion in debt in two tranches, comprising €612 million in uncovered project risk debt and €1.06 billion in a tranche guaranteed by MAPPP, the finance ministry’s PPP unit. They are:

• BBVA
• BNP Paribas
• Crédit Agricole
• Dexia
• SMBC
• Mediobanca
• Santander
• Société Générale
• UniCredit

The largest tickets are being taken by the French banks, Dexia, and BBVA (full breakdown below).

The debt is fully amortising and has a 27-year term with a minimum DSCR of around 1.6x. Pricing on the project tranche is 300bps over Euribor with step-ups until 430bps on completion of construction, and 145bps rising to 175bps on the state guaranteed tranche. Full cash sweeps are due from the end of year 10.

The EIB is also in for €200 million of project risk debt, a €200 million gurarantee under the LGTT (Loan Guarantee for TEN-T projects) and a €400 million loan guaranteed by MAPPP.

Commercial banks and the EIB are each contributing €386 million in equity bridge loans, to be repaid with €772 million of sponsor equity. A €200 million construction (subsidy) bridge loan is also signed by the commercial banks. The equity bridge loan is priced at 225bps and the subsidy loan at 200bps.

The French government is providing 25 per cent of debt financing, amounting to €757 million, through the Direction de Fonds d’Epargne (DFE) of the Caisse des Dépôts et Consignations, the state investment bank. This is the largest single investment ever made by the CDC, and is guaranteed by RFF.

The remaining €4 billion is funded through grants: around €1 billion from RFF, about €1.23 billion from the regions and around €750 million from the French state.

Project Description and Roles

The 50-year DBFOM concession will see the development of a new 303km high speed line along the Sud-Europe Atlantique route that will transport up to 20 million passengers per year. It also includes 39km of connections with existing lines, involving the construction of 40 overpasses and 390 bridges.

Construction is expected to begin in 2012, following enabling works, and last four years. The line is due to enter service at the end of 2016 or the start of 2017. When complete, trains will travel on the line at 300km/h (186mph), later rising to 320km/h and putting Bordeaux within two hours’ reach of Paris.

White & Case was legal adviser to the commercial lenders, Steer Davies Gleave was operations and revenue adviser and Atkins was technical adviser. Wilkie Farr & Gallagher was legal adviser to the EIB and Herbet Smith was legal adviser to the CDC.

RFF and MAPPP, the PPP unit of the Ministry of the Economy and Finance, were advised by Compagnie Benjamin de Rothschild on their guarantees to the EIB/DFE and commercial loans respectively. Natixis was financial adviser to RFF, Hogan Lovells was legal adviser in its grantor capacity and Ingérop was technical adviser. Paul Hastings was legal adviser to MAPPP on financial documentation and the guarantee, while Freshfields Bruckhaus Deringer was legal adviser on project documentation.

For the sponsors, Crédit Agricole and Société Générale were financial advisers, Allen & Overy was legal adviser, Inexia was technical adviser, while PwC and Landwell were providing tax and accounting advice and insurance advice came from SIACI.

Cegelec, BEC, NGE, TSO, Ineo, Inexia, Arcadis and Egis Rail are also involved as D&B contractors. Post-completion and testing, operations and maintenance will be entrusted to MESEA, a consortium comprising Vinci Concessions (70 per cent) and Inexia (30 per cent).

Conclusion

The successful close of Tours-Bordeaux showed the ability of experienced project finance banks to deliver a very large and highly complex deal through the most difficult financial crisis in living memory, and the resilience of the market as a whole when dealing with well-structured projects. Taken as a whole, the project was unique.

As Jean-Francis Dusch observes, “If you take each element, nothing is that innovative. It was the accumulation of them which made this deal unique to a certain extent : you have project risk, interface between project parties, you had the EIB also involved as a lender and LGTT provider, commercial banks, the DFE, the state guarantee , several financial sponsors alongside Vinci... it’s the combination of all these things, plus the size element, which made the deal a market precedent.”

Nicolas Bourgouin agrees,“I think the most innovative thing is to have three different public guarantees on the project. There was on the debt provided by the Direction des Fonds d'Epargne of CDC, one on commercial banks and one on the EIB. It’s definitely this kind of guarantee and all the inter-creditor issues that was the most challenging issue for the financial close.”

The size of the project and the large amount of risk carried by the private side required the state to stack up formidable amounts of guarantee and support to provide bankability. This in turn added to the project’s size and complexity, and undoubtedly to the time it took to close the deal. There will be no more PPPs in France under the state loan guarantee, since the offer is now withdrawn.

From the point of view of MAPPP’s François Bergère, the concession model was not in retrospect the best one. Speaking a fortnight after close, he commented, “If it were to be done again, we would proceed differently. It’s probably too large a deal. Apart from the fact it set a record in the category of European PPPs, when you have to get every possible bank active on the market in the field, you are exposed to a number of difficulties, and we went through all of them. We had very long negotiations with 60 people in the room including twelve lawyers. I can tell you, it makes for very exhausting negotiations.”

Noting the effect of market risk, he added, “This deal would have been easier to push if it had been contrat partenariat or if it had been segmented,” he said, but he added that a contrat partenariat was the more likely route since splitting the project into several contracts would have created interface risk.

However, not everybody agrees with Bergère on this one - incidentally, MAPPP does not have the last word on whether a PPP project should go forward - and the contrat partenariat relies on confidence in the state’s ability to pay – not something that would have been discussed pre-crisis, but in the current eurozone turmoil an all too hot topic.

Tours-Bordeaux looks set to make headlines next month when syndication begins on the loans, and bank appetite should be healthy given the pricing, especially given its attractiveness relative to recently closed PPP deals such as Balard.

There will be no more mega HSR projects procured on PPP lines in France for some time. The next may be the Paris-Orleans-Lyon line, which is planned to enter service in 2020, for which a contrat partenariat is already being talked about. The prolonged and rather fraught discussions over Tours-Bordeaux already seem to have taught the authorities a lesson: the LGV Bretagne-Pays de la Loire PPP took only five months to go from preferred bidder to financial close, compared to 14 for SEA.

Project at a glance