The sale of the Pacifico Norte roads concession package could mark the moment that Mexico fixed its roads programme. The country's transport and communications ministry, the SCT, had failed to attract any compliant bids when the roads were part of a larger package, a follow-up to the richly-valued Farac I portfolio. The smaller package is meant to be easier to digest, and came with the offer of stapled financing from Banobras, Mexico's development bank.

But the winner of the package, Carlos Slim's Impulsora del Desarrollo y el Empleo en América Latina, or IDEAL, is unlikely to use the staple. Shortly after it won the package, it closed a bond financing for an earlier collection of roads, without monoline or government support. IDEAL's chief rival, ICA, together with Goldman Sachs, had conspicuously failed to refinance the bank debt for Farac 1 in this market. But it did complete a sale of equity in the portfolio to institutional investors, using the proceeds to pay down bank lenders.

Mexico's pension funds threw off their nervousness about infrastructure and responded to the sale of the equity in Red Carretaras de Occidente with gusto. These institutions had until recently been nursing heavy losses on monoline-wrapped debt commitments. They now throw themselves into the riskier parts of infrastructure's capital structure with gusto.

Meanwhile, junk-rated monoline MBIA has returned to market with an Opic-backed financing for the state of Mexico's property registry. Can financial engineering and the search for yield do for the country's infrastructure market what well-intentioned government initiatives have been slow to achieve?

Farac's far-out offspring

The thread that links all three developments is the programme to sell toll road concessions from the Farac (Fideicomiso de Apoyo al Rescate de Autopistas Concesionadas, in full) road trust. The trust was a creation of the 1990s, when a previous bout of excessive leverage and rosy economic predictions let to a slew of distressed toll road concessions. Farac was created to take them over, and in 2007, the SCT decided to sell a first package of concessions from the trust.

The first sale brought in Ps44.05 billion (now $3.3 billion, but then $4.1 billion), a much better price than the SCT anticipated, thanks to healthy bank interest in the Mexican road and PPP sectors. Subsequent attempts to sell roads from the Farac trust have failed to do as well, as sponsors and lenders grope towards a way of replacing that cheap foreign bank funding.

In March 2008, as Farac I still lurched through syndication, the SCT moved with haste to get second and third sets of packages through the market. When the second package was delayed, the SCT indicated that the third package might overtake it. The SCT enlisted Banobras to offer the stapled senior debt financing to bidders, to complement an offer of subordinated debt from Fonadin, or Fideicomiso Fondo Nacional de Infraestructura, which was created in 2008.

Banks and sponsors complained that Fonadin money might be best used buying out the banks' underpriced and overleveraged commitments to the Farac 1 financing, although the government, perhaps suspecting that this debt would not be recycled at any price, declined to so. Shortly afterward, the bids on Farac 2 came in lower than the SCT's minimum price, and the process was abandoned.

The SCT rides back

The SCT, after the failure of Farac 2, restructured its concessions process much more deeply. According to Oscar de Buen, subsecretary for infrastructure at the SCT, it took eight steps to reform the bidding process, though some of these initiatives predate the Farac 2 bids: Making concessions smaller, combining greenfield and brownfield sections, offering Banobras and Fonadin financing, developing sections as public works projects before a later sale, restructuring existing concessions to allow additional large-scale works, bringing in additional financing from pension funds, revising bid documents to broaden the bidder pool, and reform the regulations covering rights of way acquisitions, among others.

The first reform, making concessions smaller, had the most immediate results. Farac 2, the Pacifico package, was split into four smaller sections: Norte, focused on the cities of Culiacan and Mazatlan in the state of Sinaloa, Sur, between and around Guadalajara, in Jalisco, and Tepic, in Nayarit, and concessions for sections around San Jose del Cabo, in Baja California Sur, and Puerto Vallarta, in Jalisco and Nayarit.

The Norte package was the first to come to market, and attracted bids from ICA, IDEAL and OHL. It includes the 181.5km existing Culiacan-Mazatlan road, as well as the construction of 22km bypass round Culiacan and a 38km bypass round Mazatlan. IDEAL bid Ps3.2 billion for the road, ahead of ICA with Ps1.7 billion and OHL with Ps210 million. Any financing for the deal is likely to come from Carlos Slim's family of companies rather than Banobras or the small number of remaining active project banks.

The Nordeste package has been split into two, with one in Tamaulipas, and one in Nuevo Leon. Later packages might feature sections completed with government funding and then transfered to the private sector. None of these elements are entirely new, since the SCT has in the past larded technically difficult or little-used concessions with recently-completed sections of road, and the first Farac packages included new capital expenditure. But the SCT hopes that by bringing a series of similarly-sized concession packages with similar risk profiles it can establish a durable rhythm in the market in 2010.

Given the demands that a souped-up concession programme will place on sponsors, and despite IDEAL's demurral, Banobras and Fonadin are likely to be useful in the long term. Fonadin recently formed a $1 billion infrastructure fund with Macquarie to provide equity to struggling sponsors, with the idea that having Macquarie as manager will keep the process arms-length. By getting so close to one manager, however, Fonadin looks like it is tacitly conceding that interest from other large international funds will be limited.

CKDs OK?

It also endorses the notion that domestic equity capital will be the best hope of sustaining the roads programme. Similar thinking lies behind the development of the development certificates, or CKDs, in Spanish. The certificates are designed to bolster the interest of the country's pension funds, the afores, in infrastructure. The timing is, however, very incongruous. The afores are sitting on large losses on monoline-wrapped bond investments, a mixture of infrastructure and structured finance issuers. Listed infrastructure has fallen out of favour in its homeland, Australia, again leaving institutions nursing heavy losses.

But the logic, from the point of view of Mexico's government, is compelling. Domestic equity is more patient, and more plentiful at times of stress. Even during a severe US-spawned recession, the afores have been receiving contributions and are looking for assets. But bankers and sponsors familiar with the market worry whether the new investors are wise enough to the risks of infrastructure equity. "Do they assume that the government will intervene if these concessions get into trouble? I hope not," notes one banker following the market.

There are some restrictions on afores' exposure to the asset class. As Gabriel Ramirez Fernandez, at Mexico's securities regulator, CONSAR, notes, they cannot take more than 30% of any given issuance, and their exposure to the asset class in total is capped. But striking the right balance between giving CKD-holders a measure of ability to control investments and asking them to take on roles unsuited to their skills is difficult.

Nonetheless, demand for the Farac 1 CKDs was brisk. Bookrunner Santander sold Ps6.55 billion in certificates, themselves equal to 100 class B shares in the portfolio, at Ps77 each. The two sponsors promised to match the largest single ticket written, meaning they contributed another Ps2 billion in equity. The result of the process was that a foreign sponsor – Goldman Sachs Infrastructure Partners – ended up with 54.5% of the equity in Red Carreteras de Occidente, ahead of the trust that issued the certificates, with 31.7%, and ICA, with 13.6%. Lenders to the portfolio received most of the proceeds, as a prepayment of their debt.

The country's stock exchange is predictably enthusiastic about the CKDs' potential. Luis Tellez, former minister in charge of the SCT and now head of the stock exchange, has talked of Macquarie launching a fund. The next step in the development of the certificates, because the supply of private concessions is limited, will be to issue them in advance of winning concessions, almost as blank cheque companies. The mechanism is not as outlandish as it sounds, since the blank cheque companies could simply invest proceeds in government bonds (the afores' chosen diet) until needed for concession or construction payments. However, how these bidders might absorb the costs and risks of bidding on infrastructure assets has yet to be decided.

Banks struggle to keep relevant

The scale of the banks' retrenchment was larger in Mexico than anywhere else. The Ps37.1 billion in Farac 1 bank debt, from October 2007, priced at 165bp over the TIIE local interbank rate. As Gabriel de la Concha, ICA's director of concessions notes, financing costs have gone from 250bp to 350-400bp, and banks will only provide tenors beyond 10 years reluctantly. Several of the mandated lead arrangers on the first Farac deal have pulled out of lending to the Mexican market altogether.

Most of the rest have much lower hold levels. Even Santander, which dominates the infrastructure finance market, does not have a limitless balance sheet. It put up half of the cost of ICA's Ps1.896 billion La Piedad bypass as 15-year bank debt priced at 275bp, and attempted shortly after Lehman's collapse to syndicated Ps2.25 billion in debt for ICA's Rio Verde/Ciudad Valles toll road at pricing of 180bp.

The correction, admit bankers was long overdue, and some commitments from 2007 and 2008 will never be able to attract secondary interest. Only 5-6 banks had substantial peso project finance funding capacity, even before the crash. Lenders to the earlier projects will hope that the SCT asks sponsors to expand these concessions and this allows the projects access to debt from Banobras and Fonadin. Between a squeeze in peso funding, banks' insistence on mini-perms and high pricing, and the SCT's willingness to restructure concessions, Banobras has the opportunity to become a dominant force in lending to private infrastructure like its counterpart in Brazil BNDES. It is, for instance, the sole provider of the debt for a hospital in Ixtapa sponsored by domestic construction firm Grupo Gia.

A more unlikely evolution is that of MBIA, the monoline bond insurer now operating in Latin America as a boutique transaction. It has formed a new entity, known as Latin America Capital Advisors, to originate bond financings for infrastructure assets with the bonds benefiting from a guarantee from the Overseas Private Investment Corporation. Opic in turn reinsures some of the risk with MBIA, which is BB-/Ba3 rated, and unable to write conventional bond insurance.

The product invites grousing from Assured Guaranty, which is still rated AAA/Aa3/AA- (S&P/Moody's/Fitch), and is cautiously open to Mexican deals, Since it is domiciled in Bermuda, however, it cannot avail itself of Opic support. However, MBIA's first deal, for the modernisation of Mexico state's property registry, benefits from $100 million equivalent in cover from CAF, on top of $250 million from Opic.

The peso-denominated deal is set to price in mid-December, and has the potential to prod other export credit agencies to produce similar financings for infrastructure assets in Mexico and elsewhere, possibly where sponsor equity comes from the right country. The afores, who miss the comfort of triple-A guarantees, no matter the fate of the monolines, will be suitably enthusiastic.