Portugal is one of the most promising infrastructure markets in the EU. There is a swell of projects in the road, rail and hospital sectors and the location of the new Lisbon airport has finally been settled at a less technically challenging site – to the east of Lisbon at Alochete – which should smooth a path to financial close. All in, infrastructure investment should total around Eu20 billion ($30.1 billion) within the next three to four years.

First on bankers' tables is the new road programme. The Portuguese government has overcome its fiscal constraints to fast-track nine new highways using a bold and innovative structure whereby liabilities and competency for awarding new concessions are passed to Estradas de Portugal SA (EP). Under the new structure, EP has been granted a 75-year concession by the government and in return EP must pay a concession fee to the government.

Bankers are confident about the fast delivery of the road programme. Portugal has delivered before, though its best-and-final-offer (BAFO) system was long and cumbersome, and sponsors and government have still to firm up a solution to the vast state payments necessary under the SCUT shadow toll regime. Nevertheless, international banks are setting aside resources to service the Portuguese market.

There are, however, three problems that could hold things up. The first is the state of global credit markets, which is making banks nervous about accepting any degree of syndication risk. Although this should not affect the speed at which deals get away – banks are arranging deals on a club basis and take and hold around Eu70-Eu100 million each – it will make competition more intense for the liquidity available.

The second issue is the government's handling of the SCUT concessions. Negotiations about the conversion of three shadow toll concessions – Costa da Prata, Grande Porto and Norte Litoral – to real toll regimes could quickly end in stalemate if sponsors consider that a refinancing, rather than an amendment to financing documentation, is the likely outcome: In the current environment refinancings will be more expensive than the original deals even if they are less risky.

Since the December 2007 deadline for the SCUT renegotiations was missed a new proposal has been submitted and a new deadline of December 2008 has been set. If the next deadline elapses, continued postponement of shadow toll payments will almost certainly lead to court action and harm the reputation of Portuguese authorities.

EP risk

The third and probably most important issue is the status of Estradas de Portugal SA (EP). Under the new structure EP acts as a super-concessionaire, and awards sub-concessions to private consortiums. The government's rationale for the new system is to deconsolidate highways debt from its balance sheet as RAVE does with its rail infrastructure. For EP's debt to remain off balance sheet in the eyes of Eurostat at least 50% of EP revenues must come from third-party sources, which is why under the new financing model EP collects real tolls. EP is a wholly-owned government company but by decree it is not guaranteed by the state and theoretically could go bankrupt. Logic dictates that the Portuguese government would never allow the insolvency of EP, but there are doubts about its long-term status and whether it will always be 100% owned by the government.

The model for the nine upcoming concessions seeks to avoid the unsustainable liabilities built up under the SCUT regime through the natural hedge of real tolls. Although the programme has been billed by the government as a real toll regime, EP collects the tolls and the private consortium gets paid through a mixture of availability payments and a service component based on the amount of traffic across a 30-year concession. In all but one of the forthcoming concessions – Baixo Tejo is the exception – the toll revenues are not expected to cover EP's liabilities to the concessionaires. EP, a government entity, bears most of the traffic risk, with the market widely acknowledging that private consortiums are taking less risk than under the SCUT regime.

The real tolls are set annually by government and annual increases cannot exceed 90% of benchmark inflation. Under the service component, which effectively acts like a shadow toll, concessionaires get paid per unit of traffic on given stretches of road, with no distinction made between the different classes of vehicles, such as heavy goods vehicles, even though they pay higher tolls. Minor stretches of road built by the concessionaires but handed over to EP after construction are not measured.

The first test of bank sentiment toward EP counterparty risk is the syndication of Duoro Litoral and the Tunel do Marao projects. One market participant questioned why banks had accepted the proposal of the authorities for Duoro Litoral without requiring any assurances on the future of EP. The source says the deal has set a precedent in the market and now "the authorities hold all the cards".

Syndication of the senior debt portion of Brisa's Eu1.2 billion Douro Litoral real toll project facility was launched in March. Caixa BI, Millennium BCP, Royal Bank of Scotland and Santander are bookrunners, and lenders are being offered tickets of Eu120 million with fees of 70bp on the senior tranche and 20bp on the equity bridge. The 24-year Eu372.5 million senior loan is priced at 100bp over Euribor during the four-year construction and then rises 5bp every five years to 125bp at maturity; the four-year equity bridge of Eu285 million is priced at 50bp. ADSCR is 1.4x.

The lead arrangers on the debt backing the Eu435 million Tunel do Marao concession in Portugal are bringing aboard another mandated lead arranger to bring the club to seven. CaixaBI, HBoS, Fortis, RBS, WestLB and La Caixa have held a beauty contest for another bank to ease the pressure on a general syndication that will launch imminently. Whereas Douro Litoral is backed by real tolls, the Tunel do Marao concession is backed by availability payments and has heavy capex requirements, and so is more vulnerable to EP counterparty risk.

Although the financing documentation does not include the possibility of a market flex, it was included in presentation documentation. The grantor, EP, has come round to the notion that if it can be shown that the market has moved significantly, particularly if there is a significant amount of time between priced commitments and financial close, a modest flex is acceptable. Still, with margins at 110bp to 130bp and a further bank joining the club, the mandated lead arrangers (MLAs) are confident the deal can negotiate a tricky syndication market without flex.

The status of EP has split the banking community. Of the bankers canvassed by Project Finance, those participating in the Douro Litoral and Tunel do Marao projects were confident that the wider banking community would accept EP's status while accepting that they faced a difficult syndication market. However, one banker active in Portugal and not involved on the projects says, "perhaps the MLAs would be assisted if EP made a presentation to clarify its position."

From a cashflow perspective EP is most vulnerable in the near future. EP has assumed the shadow toll liabilities of the SCUT regime but can offset these liabilities against the concession fee it pays government. EP's balance sheet has been substantially boosted by a decree that gives it a proportion of an oil tax that is expected to bring in around Eu500 million per year. EP's income is expected to be around Eu900 million for 2008, boosted by upfront payments from the granting of concessions. The Duoro Litoral project, for instance, involved an upfront payment to EP of Eu207 million. The new wave of road projects is expected to start bringing in toll receipts in 2010.

In the long term EP will have extremely healthy cashflow as it will assume control and collect tolls from the existing concessions which expire, with the largest – Brisa's original concession – ending in 2032.

Three new roads in bidding

Of the nine new road concessions, seven have been launched and three have had bid submissions. The first project, the Eu800 million Transmontana Highway is likely to be in the market around year-end, with the next projects finishing sequentially over the following few months, and with all projects expected to be complete by the general election in October 2009.

All the concessions are for 30 years and share the same template, with bidders asked to submit capex figures, anticipated net present value (NPV) of toll receipts accruing to EP based on their own traffic assumptions, the NPV of a service component linked to traffic and the NPV of availability payments minus any lump-sum payments made to EP (the Esforço Financeiro Líquido Concedente). The bidders are invited to propose a structure of payments paid back to EP as initial or annual lump-sum payments that the concessionaire may wish to make to the EP during the over-profiting years of the concession.

As is typical of the Portuguese PPP process, EP will then assess the bids and whittle the shortlist down to two for the best and final offer (BAFO) stage, although a set timetable has not been disclosed. While the weighting of the various technical and financial details is known by the market, with the most critical metric the capex bid, the NPV calculations will vary from bidder to bidder, so EP will seek uniform comparisons at the shortlisting stage. It is likely that most of the concessions will feature an upfront payment to EP to help its cashflow in the near term.

The favourite for Transmontana is a Soares Costa-FCC-Global Via consortium advised by BancoBPI with a capex of Eu577 million and a NPV of concession income of Eu740 million and a service component representing 36% of the income. A Somague-MSF consortium posted a NPV of income of Eu806 million, and the AENOR consortium is a close third with Eu814 million.

AENOR and Soares Costa consortiums are favourites to be shortlisted for the second tender, Douro Interior, posting an NPV of income of Eu816 million and Eu882 million respectively and a service component of around 13%.

Somague and a Dragados-led consortium are the two frontrunners for the third road scheme under Portugal's new model, Baixo Alentejo. Somague posted the lowest NPV of income at Eu561 million with a Dragados-Edifer-Tecnovia-Conduril consortium posting Eu589 million, and AENOR third with Eu640 million. There were divergent bids on the service component with the Dragados consortium posting 59.5% against a Soares Costa bid of 35.8%.

Bids will be opened on the Baixo Tejo concession on 14 May. The bids are expected to feature the highest service component across the new programme. At around Eu300 million, it is the smallest concession in the new wave, and it is likely to be the only road busy enough for the real tolls to fully meet EP's obligations to the concessionaires.

Of the remaining five concessions, Centro and Litoral Oeste have a bid submission date of 7 July, and Algarve EN125 late July. The Alto Alentejo and Pinhal Interior concessions are awaiting environmental approvals and will be launched imminently.

Early indications are that these concessions in the current market will get financed with margins of around 120-130bp, with an average debt service coverage ratio (ADSCR) of 1.10-1.15x, for a mostly availability-based payment regime, rising to around 1.3x for a concession with a 50% service component.

Hospital rethink

The Portuguese PPP health sector is also seemingly on its way to a full recovery, following the launch of the tender for Lisbon's Eu370 million Todos-os-Santos (All Saints) hospital. The tender signals the demise of the concept of separate provision of clinical services (clinicco) within the PPP framework in Portugal. The 789-bed hospital will be procured along traditional UK PPP lines with a design-build-finance-operate (DBFO) project company, and the public provision of clinical services.

But there are still three 'Wave 1' hospitals still to reach financial close that feature a 10-year clinicco concession parallel to a 30-year DBFO concession. The most advanced is the 700-bed Braga University hospital sponsored by Somague, José de Mello Saúde and Edifer. The negotiations are now at an advanced stage, with the project waiting for clearance by the new finance minister, which participants believe should be a formality. The lead arrangers on the deal are BNP Paribas, Caja Madrid and SG.

The same sponsor and bank consortium are shortlisted for the 200-300 bed Vila Franca de Xira hospital and are pitted against a GPS-Ferrovial consortium backed by ING, BBVA and Banco Efisa. A preferred bidder on the Eu60-70 million project is expected in the summer.

The third hospital to involve a clinicco is the re-tendered Eu75 million Loures hospital. A Somague, José de Mello Saúde and Edifer consortium with Caja Madrid and RBS are shortlisted against ESS and Mota, backed by BES. The Loures project has been a long-winded affair, dogged by legal challenges from losing bidders that challenged aspects of the public sector comparator, which indicated that different proposals were impossible to compare under the old tender procedures.

Given the amount of effort put into development of the clinicco template, some market participants question whether the government will come to regret its decision. One source says the decision "was just a bone to throw the Left". Now constructors rather than healthcare providers will lead the bidding for future Portuguese PPP hospitals.

In mid-May a tender will be launched for the Algarve hospital with a capex of around Eu230 million. Two more purely DBFO hospitals are expected to be bid in 2008. BancoBPI is advising the government.

High-speed rail

On 2 June RAVE, the authority responsible for the high speed rail network, will tender the first portion of the six-phase high-speed rail programme. Bankers are confident that the sections will be launched before the elections in October, given the expertise of the grantor.

The first tender is the Poceirão–Caia high speed rail link which will entail around 170km of track and a total investment estimated at Eu1.7 billion ($2.67 billion). The project will complete the rail link between Lisbon and Madrid.

In June 2007 RAVE proposed and publicly presented a business model for the project. Under the model, the project is likely to use around Eu1.5 billion of EU financial support.

The tender is the first of five high speed track infrastructure projects, which split into:
• Two along the Lisbon-Porto link: Lisbon-Pombal (Eu2.1 billion) and Pombal-Porto (Eu1.7 billion)
• Two along the link from Lisbon to Caia, on the Portuguese border (connecting to Madrid): Lisbon-Poceirao (Eu1.6 billion) and PoceiraoCaia (Eu1.7 billion); and
• One between Braga and Valenta, on the Portuguese border, connecting to Vigo (Eu800 million).

The sixth tender is for the signaling and telecommunications for the entire high-speed network. RAVE undertook much research into the procurement of the system and found that one signaling concession avoided the integration problems of other rail systems, such as that in the Netherlands. The sixth tender is said to correspond to 5-10% of the total investment (up to Eu900 million).

The track infrastructure concessions will be 40 years long, and the signalling concession 10 years. RAVE estimates that the project will generate operating cash flow covering 45% of the investment, with the EU contributing 19%, leaving the state to fund the remaining 36%. The European Investment Bank is expected to be heavily involved.

Next on the block is the tender for the Lisbon-Poceirao PPP scheme, including the Tejo River crossing, with its conventional rail track component expected in November. The launch of tenders for the Lisbon-Pombal and Pombal-Porto PPP schemes is expected in the first half of 2009, with Pombal-Porto likely to be first. The Braga-Valenca PPP scheme is scheduled to follow in the second half of 2009. And the IT and communications tender is due in the first half of 2010.

The tenders will follow a similar pattern to the road concessions, with a shortlist and then a BAFO stage. However, the high-speed tenders will be the first to be conducted under a new public procurement procedure, CCP (Código dos Contratos Públicos), which is an enactment of an EU directive (17/18). The regulations should promote transparency during negotiations. RAVE has the option of shortlisting three bidders if they are close, rather than the usual two.

New Lisbon airport – at last

Like the high-speed rail project, the new Lisbon airport project has a long history of false starts. But the Portuguese authorities have finally settled on a location for the Eu4.9 billion airport at Alcochete, on the south side of the Tagus River, and not the original, technically more challenging site at Ota. Environmental and technical studies are ongoing. BancoBPI and Citi are advising NAER (New Airport Company) but their mandate has been suspended while the studies are being complete. NAER mandated Banco Efisa, ABN Amro and PwC in 2001 but their mandate elapsed in 2004.

If EU cash from the 2007-2013 round of structural funds are to be used, the Commission has said that a decision needs to be taken as quickly as possible and the various reports must be complete promptly in order for it to make its decision in time.

A public tender could be launched late 2009, but of all the infrastructure projects to come to market the airport seems most likely to be delayed. The deal is likely to be structured as a government sale of a majority stake in operator ANA plus the development and operation of the new airport. ANA's stake in Madeira airport needs to be resolved before a privatisation can progress.

A frontrunner for the airport investment is a consortium comprising Brisa (24%), Mota (24%), Sacyr (15%), and three banks, Banco Espirito Santo, CaixaBI, and Millennium BCP, each advising and with a 9% stake. The remaining portion will be taken by small local construction companies.