The coming year will see the toll road public-private-partnership (PPP) business become truly global with new markets in the US, Canada and Latin America looking a reality for the first time.

The concept of PPP in roads – to date a methodology adopted primarily by the UK, Australia, Spain, Portugal (with mixed success), Chile and South Korea – is also picking up in those European markets that have been slow to adopt in a meaningful way – Germany, France, Greece and Italy.

Similarly, some of the EU accession states – Hungary and the Czech Republic for example – have already closed template deals with more expected to follow. And the Russian government has awarded a number of advisory roles for a feasibility study into its first toll road PPP: Coordinated by the Russian design institute GIPRODORNII with Ernst & Young as financial advisor, Freshfields as legal adviser and Ramboll as technical adviser, the project is a $7.5 billion toll highway linking Moscow and St. Petersburg.

Real toll over shadows

There is also a general shift away from shadow tolling to real toll. Portugal, the first to adopt shadow tolling in a big way, has dropped the concept having found itself unable to afford payments on its SCUT programme (for more details search 'SCUT').

Even the UK – a longstanding user of DBFO/PFI schemes for roads but resolutely anti-toll (other than the Birmingham Northern Relief Road) – looks set to follow the tolling mantra with the government's new Transport Innovation Fund designed to promote toll road projects at regional level.

At its simplest, the cost and demand for road infrastructure on a global scale is outstripping hard-pressed government budgets, and while real tolls require strong economics to attract private finance, rates of return and cost of borrowing are falling, making more real tolls a viable proposition.

Shadow tolling will continue – but driven either by political ideology (as in Spain for example) or as the concession method of last resort for roads where real toll economics do not stack up.

All this is good news for sponsors, and banks are enjoying unprecedented levels of advisory and arranging business across an increasingly broad range of debt and equity instruments – bank debt, bonds, stapled securities.

Sponsor IRRs down

But for traditional project lenders, a confluence of pressures on margins – lower sponsor internal rates of return (IRRs), high liquidity in the debt markets, the influx of infrastructure investment funds and hence equity capital – has taken the edge off of what should be a few years of record toll road volume.

Toll roads have become the must-have asset with investors – for example Cintra's share price has upped by around 30% since its IPO in October 2004 and the share prices for Australian operators Transurban, Hills Motorway and Macquarie Infrastructure Group (MIG) have risen 57%, 77% and 35% respectively in the past 18 months.

But traditional sponsors face growing competition, not only from each other as international firms merge or acquire for economies of scale and market share, but also from bank-managed investment funds like MIG (one of, if not the biggest toll road operator in the world). And aggressive bidding for market share has seen internal rates of return (IRR) in markets like Spain and Australia drop from up to 20% in the mid-1990s to around 12% today.

Debt margins pressured

The drop in toll road IRR – due in part to lower interest rates but primarily because of the aggressive behaviour of bidders as they grow more comfortable with the asset class – is being passed on by toll road sponsors to lenders in the form of pressure on debt pricing. And high liquidity among debt providers is making is easier for sponsors to apply that margin pressure.

Spain has witnessed the most extreme drop. Upcoming regional shadow tolls are expected to price at sub-100bp and the precedent set by the Ocana La Roda financing in late 2004 hit real toll margins hard.

The first of the four real tolls in Phase 2 of Spain's second national toll roads programme to reach financial close, Ocana La Roda was an 8.5-year miniperm priced at a surprisingly flat 110bp over Euribor through both construction and operations.

More recent Spanish real tolls – Madrid-Toledo and Cartagena-Vera – have seen some recovery: Pricing on Madrid-Toledo ratchets on ADSCR, paying 140bp over Euribor on 1.4x to a minimum of 100bp at healthier coverages; and on Cartagena Vera pricing is an average 130bp over Euribor.

Although all three deals are too dissimilar to be truly comparable on price difference they share a common element that requires no comparison – they all came in at least 20bp lower than envisaged in their original bids in early 2004.

Will US mirror Europe?

The aggressive sponsor bidding for deals in Europe's developed markets is already spilling over into the fledgling US market. The $1.9 billion paid by Cintra and Macquarie for the Chicago Skyway 99-year lease – $1 billion more than the other nearest bid – has set a highly competitive IRR benchmark for US concessions: the original bank debt deal closed earlier in the year enabled the sponsors to realise an IRR of 10% with the subsequent bond refinancing last month upping IRR by 150bp-plus.

By European standards the project is a standard return and given the predictability of the existing income streams was priced reasonably – Cintra and Macquarie could have paid less but if anything the competing bids were way too low to reflect market reality.

But the risk/reward precedent that Chicago has set leaves little room for improvement in a market that looks set for more aggressive bidding. And the first signal of more risk being taken on is already in the market. Macquarie Infrastructure Group's (MIG) $533 million purchase of 86.7% of the Dulles Greenway in Virginia will be its first road investment where toll prices are independently regulated.

MIG predicts the purchase will have an IRR of 12.4% and produce a cash yield of 5% over five years and 8% over 10 years. The IRR and yield is average but it is questionable how much MIG will be able to raise tolls given independent regulation and a US market where low-cost motoring is politically sacred and where increasing petrol costs are effectively beyond the control of politicians.

MIG has factored in a risk premium for tough regulation and has recourse to appeal to the courts – but the revenue risk is very real as are the chances of any court decision not going Macquarie's way.

The deal may also prove to be the first salvo in an unsolicited bidding competition for the Dulles Toll Road, which is fed by the Greenway. Autostrade and John Laing have put in an unsolicited bid for the road. Transurban is considering bidding and given Macquarie's Greenway acquisition, MIG must also be at the very least considering bidding.

European bidding pressures will almost certainly lead to European-style bank debt pricing in the US toll road sector.

The original bank debt pricing on Skyway was reasonable -125bp over Libor for years 1 to 5, 150bp for 6-7, and 175bp through 8-9 – but not particularly cheap compared with similar European deals and given the predictability of the cashflows. Then again the plan all along was to refinance in the bond markets.
And although pricing on the SR125 San Diego Expressway – the first US road PPP closed for Macquarie by BBVA and Depfa in 2003 – has never been released, the deal took its lead from European toll debt margins at the time: the Dulles Greenway is to be funded primarily by a A$675 million share issue and project banks will therefore not get a look–in.

Are lenders taking too much toll risk?

Given the US market looks set to be dominated by European and Australian sponsors, some of which already have or are in the process of buying up US contractors, the first US greenfield deals to come to market in the coming years will probably be European-style miniperms with a view to bond refinancing post-construction in the deep US capital markets.

Whilst the current race for existing tolls may result in overpayment for some acquisitions, the predictability of income streams will keep those mistakes to a minimum and lenders will have a fairly crystal idea of what they are getting into.

The worry is however that on future greenfield deals, the amount of risk lenders are taking on at cheap rates may be underestimated, especially given the track record of toll roads elsewhere.

According to a 2005 forecasting update just released by Standard & Poor's, traffic forecasts in the first year across 104 toll road case studies (all rated deals and therefore higher than average credits) were on average 20%-30% overestimated. And the bias does not improve up to year 5 of operation (there is very limited data for beyond year 5 because project debt repayments based on traffic flows are a relatively new phenomenon).

Ultimately tolls roads are a fairly safe bet – a fact attested to in the Americas by the growing value of Chile's interurban concessions that have just started being traded on the secondary market, and the popularity of roads with investment funds which view tolls as long-term steady income streams whilst other markets go through cycles.

Road projects in Europe that have gone wrong have mostly been refinanced or resurrected in a different form – even Portugal no longer seems phased by its SCUT payments and although a refinancing will go-ahead, plans for the moment are on hold.

But if risk is not the issue with lenders and sponsors that it is with ratings, reward is. Ostensibly banks are going to have to look for low margin/high volume and quick turnover or long term high margin opportunities – and the US market is unlikely to be the place to get high margin.

A number of smaller roads lenders are already exploring niche markets in central and Eastern Europe (see participation list for M5 and M6 in Hungary on www.projectfinancemagazine.com) where volume is lower but margins are high. The same will be true of Brazil as the first deals in its new PPP programme come to market. Simply put – banks wanting margin should not be looking to the international sponsors.