[Editor's note: Shortly after this article was published, on 27 February, the Mexican transport and communications ministry decided to halt the Farac II sale. It said that neither of the two bids it received matched its minimum price expectations. However, its also said that the SCT, Fonadin and Banobras would continue to work on measures to increase bidder interest in the programme. The following article lists the issues that bidders have with the Farac sale process, their suggested improvements, not all of which the SCT has acted upon, and some of the structures that the SCT, Banobras and Fonadin might offer to refloat the sale process]

President Felipe Calderón has met Mexico's looming downturn with a western-style spending spree, promising to push Mexico's outlay on infrastructure development to 1.8% of GDP, with spending on roads, airports and sea ports of Ps570 billion ($39.5 billion) this year.

For 2009, the Mexican Congress authorised a record budget for highways of $3.68 billion. The Department of Communications and Transport (SCT) is very focused on moving projects forward, according to Pablo Garcia, director of project finance at Santander in Mexico. But with so many projects happening at once and needing government support, resources are stretched and delays and time lags are bound to grow, cautions Gabriel de la Concha, chief investment officer at ICA.

Farac 1 inflates expectations

Perhaps the most significant plank of the road program is the series of packages of roads in different areas of the country known by its Spanish acronym Farac, a programme to re-concession existing roads and build new ones.

Farac 1, which covered the operation and maintenance of 560km divided between four existing highways in west-central Mexico for 30-year periods, was extremely successful. The number of bids, at six, was surprisingly high and attracted some well-regarded sponsors. The winning bidder, announced in August 2007, was a consortium of ICA and Goldman Sachs, which offered roughly $4 billion. To meet the concession fee, and the concession's capital expenditure requirement, they took out a seven-year Ps32 billion ($2.1 billion, at today's rates) short-term, mini-perm financing. The facility's bookrunners were Santander, Dexia and NordLB, with WestLB, ING, Inbursa and Banobras as mandated lead arrangers.
Banks lapped up Farac 1, with many supporting different bids out of different offices, notes Juan Francisco Toro, project finance director in Dexia's Mexico office. The government immediately pushed ahead with Farac 2 with the same expectations of success, adds Garcia. Bidding kicked off at the start of 2008, but the ensuing credit crunch has delayed the process repeatedly. "Market conditions have changed substantially both compared to our expectations and compared to Farac 1," admits Oscar de Buen, under-secretary for infrastructure at the SCT.

The Farac 2 program is more ambitious with more roads – around 850km – and much more new construction. Concentrated on the Pacific coast, in the states of Jalisco, Nayarit, Sinaloa and Baja California Sur, it consists of three brownfield projects and a commitment to build more than 400km of new roads across 10 sections, making it both larger and of a very different composition to Farac 1, says Federico Patiño, managing director of the National Infrastructure Fund, Fonadin. The scale of the undertaking has changed the attitude of participants to the bidding, adds de Buen.

Difficulties in stimulating interest in such a large project at a time of credit scarcity has led to a slew of suggested changes from possible sponsors and financiers, helping explain the numerous delays. One of the ideas with the most traction was to switch the order. The much smaller, brownfield-focused Farac 3, which takes in about 400km of mostly existing roads in the far northeastern states of Nuevo Leon and Tamaulipas, could precede Farac 2. The smaller size of the project would also encourage participation by local construction companies, reckons Toro.

Another suggestion was to break Farac 2 into brown and greenfield components. De la Concha thinks the government could build new roads as a public works contract and sell the completed product as a brownfield asset, obviating key risks and making the deal more appealing to newly-cautious sponsors and banks.

Neither of these solutions has been countenanced by the government, although there is recognition that the ideas have their merits. Farac 3 would be easier to sell as it has a lower risk profile with 70% of the roads brownfield, admits Patiño. But he notes that some sponsors liked the large size of Farac 2 and when the government contemplated dividing the project up, the reaction was mixed: "We asked sponsors if they wanted us to divide up the project, but some said they were willing to take it as it was. Ten new highways is good business for some," he notes. He admits that since then markets have deteriorated, but says the project is well advanced and the government has other solutions up its sleeve.

The SCT's staple

To help close a sale, Banobras and Fonadin have been working to ready a financing for bidders. The proffered debt superficially resembles stapled financing, but will be sweeter than what a sell-side financial adviser might offer. "It is designed to take risk that the market is not willing to take," says Patiño.

Banobras will provide up to some Ps15-18 billion in debt to any successful bidder, with the requirement that sponsors kick in a minimum of 25% in equity. Banobras will then seek to syndicate this commitment down to commercial banks, says Patiño. The new deal does not pose the moral hazard of a guarantee, he says: banks will come in pari passu with Banobras. "They will take the risk. Banks need to take their own decision on a voluntary basis," he says.

At the same time, Fonadin will provide subordinated debt for bidders to reduce the need for equity and more importantly bank loans, says Patiño. The final terms and conditions of the debt were being worked on as Project Finance went to press, but could amount to 30% of the total amount of debt at similar tenors to the senior debt. De Buen says the figure of Ps40 billion cited by some bankers for the amount of support is too high.

These coordinated moves have ensured the participation of key players, says de Buen. They include Carlos Slim's IDEAL, Global Via and Iridium. ICA has put together a preliminary proposal with medium-term loans through Santander and others, confirms de la Concha. If it proves a hit with banks, the model could be repeated for Farac 3 and 4, de Buen adds.

The question is whether banks will follow. The most active international banks in Mexico include BBVA's Bancomer, Santander, HSBC and a second tier of international banks has been active in Mexico including BNP Paribas, Bank of America and WestLB, says Patiño. These banks, however, are usually locked into exclusive mandates, he adds, and will only jump in and participate with Banobras once they know who the winner is.

Bankers are pleased with the government support, but the date for submitting bids has already slid twice and sponsors have requested a further delay to the most recent deadline of 27 February. As Project Finance went to press, De Buen said his department was reviewing a final adjustment to the bid presentation date for Farac 2, but that it would be defined imminently.

The question is whether banks will be willing to provide financing. Toro says there may not be enough banks willing to provide an underwritten offer. "Banks want more commitment, less debt, less stress on the project," he believes.

Long lags between bid acceptance and financing mean deals may collapse, Garcia reckons. The terms written into commitment letters in today's market often offer guaranteed pricing for just 30 days. That could result in banks not being able to sustain financing offers for long periods and having to walk away from deals, leaving sponsors with open-ended pricing, he notes. Banobras and Fonadin say they are very alert to what's happening in the market and keeping a close eye on developments. "We will adjust what we need to as we go along," says de Buen.

What refinancing?

There's another snag. Commercial banks lent freely to the first Farac package in 2007 with the expectation that sponsors would quickly refinance through capital markets. That has not happened.

This is a big concern for ICA. The government priority should be to help banks refinance Farac 1, says de la Concha. It has been reluctant to revisit the issue as Farac 1 is completed, and the point of the Farac program is to transfer risk to the private sector, he says. He suggests that one solution may be for the government to take the subordinated debt tranche, perhaps with Fonadin providing a backstop.

The government recognises the problem. There is concern regarding concentration on Mexican toll road risk, acknowledges Patiño, although he pins the blame on the crisis rather than any intrinsic worries about the highways. Some banks are prepared to participate in Farac 2 as they like the risks and have room for more Mexican exposure; others accept current levels of exposure, but are not willing to participate in new deals; still others are looking for a speedy refinancing of their Farac 1 miniperm commitments.

ICA was talking to investment banks about a refinancing before the crisis. The idea was to create a two-tranche, multi-year operation as the size of the loan precluded a one-shot deal, de la Concha says. The plan was to issue a AAA-senior tranche of debt and subordinated debt of between $500-900 million with a slightly lower rating, probably AA. That will have to wait until credit markets improve, he notes.

The peso-denominated Farac 1 debt deal was widely seen as aggressive at the time and some bankers and sponsors (always off the record) suggested that the winners had overpaid. The debt has a series of step-ups, which should encourage the sponsors to refinance as the capital markets open up, although current conditions make predicting that moment impossible.

Initially the Farac 1 debt was priced at 165bp over TIIE (the Mexico interbank rate) and the step up starts in 2010, when the debt will be priced at 185bp over for one year, before repricing to 195bp for the next two years and eventually reaching 225bp. Even so, right now, the step up looks cheap, since the gap between the bridge and a market-rate loan could possibly be as wide as 250bp, Garcia admits.

Thorough times

If concerns about the refinancing of Farac 1 were not enough, sponsors are also much more thorough on due diligence. The negotiations between government and sponsors have been tense, notes Garcia. The government faced tough questioning over the green-brownfield mix and sponsors have proved very conservative in due diligence.

Sponsors are being particularly demanding with respect to rights of way for Farac 2, as it consists of many greenfield sections. The Mexican constitution favours landowners and the constitutional challenges and demonstrations have tripped up infrastructure projects in the past, including a proposed expansion of Mexico City airport in 2000, notes de la Concha. Many of the roads in Farac 2 pass through dense areas of population and nature reserves, he notes. Delays to the timetable could reach one year thanks to these obstacles, and this might deter foreign players or make commitments uneconomic for others, he believes.
Sponsors are also going to be unusually assertive in their discussions over expected levels of traffic, because sponsor scepticism about government-provided numbers is pronounced.

"It is very difficult to explain the discrepancies between some government figures and reality," says de la Concha. In the year beginning October 2007, ICA's numbers showed an increase of 2% in traffic in one section while the government was claiming an increase in income of close to 35%. When it investigated how limited traffic growth could translate into huge income growth, it found government numbers were unreliable, he says. Indeed, this explains why the firm wanted to have two years of its own numbers before talking to banks about refinancing Farac 1. "If you can't explain numbers, the market will punish you," he says.

The government is well aware that a slowdown in economic activity means traffic will grow slowly, says de Buen. But Mexico has passed through other crises and while traffic growth slows, levels do not fall on an absolute basis, he says. Mexico is very highway-dependent with only 1% of passenger transport by air and 80% of freight travelling by road. Furthermore, income per capita is stimulating car use and Mexico has a young population with 800,000-1 million people reaching driving age per year, he says.

The non-Farac contenders

The Farac program is just part of a raft of stimulus-related infrastructure spending. The SCT has 40 projects with private participation. Of these, 13 are concessions that involve investments of $1.8 billion to build 708.5km of new roads; six private service contracts with $1 billion invested to build or improve 530km of roads; 16 projects (590.8km) are under development as part of the asset utilization model with total investment of $1.2 billion, and five projects (383.7km) are being developed as part of existing concessions.

Other priorities include the suburban train system in Mexico City and the airport Riviera Maya in the southeast, says Patiño. He insists that the bank and fund will have the resources to finance priority projects. The fund generates cash by selling its road assets but can tap the Mexican government and even borrow on the capital markets, he notes.

Bankers agree that the government is willing to finance these undertakings. But Garcia says that the market is concerned that the Mexican government has the capacity to manage all these initiatives. Human resources are stretched and government banks are working flat-out at a time when the market is changing constantly. The appetite from international banks for more Mexican project finance assets even with all the government sweeteners too remains untested. The differences with Farac 1 could hardly be more pronounced.