Private developers of toll roads face growing and organized opposition from public toll authorities to PPP and concession agreements. The Texas Department of Transportation (TxDOT) and a number of related regional agencies are due to decide whether the SH121 toll road in North Texas will be controlled by a public toll authority or by a private concession. Public toll authorities in New Jersey and Pennsylvania are at the forefront of opposition to those states' selling off their turnpikes.

The authorities are pushing back against private takeovers that will result in their potential obsolescence, or at least a drastic cut in their responsibilities. These public, though autonomous, agencies may be able compete with private equity concessions, with serious implications for the structure of future toll projects. This opposition follows a period in which the not-for-profit model for toll roads was all but left for dead.

The most notable recent sign of life is the counter-bid by the North Texas Toll Authority (NTTA) for the SH121 concession in Texas, which had already been awarded to a Cintra/JP Morgan-led consortium, but has not yet reached financial close. The bid came as a result of the political wrangling in Texas over a proposed moratorium bill on private toll roads.

The background to the NTTA bid is somewhat convoluted: TxDOT refers to its numerous toll road concessions as comprehensive development agreements (CDAs), and Representative Lois Kolkhorst and Senator Robert Nichols proposed legislative bills to ban such CDAs for toll roads featuring private equity for a period of two years. The bill, in the modified form of SB 792, eventually passed in the Texas legislature, and has been signed by Texas Governor Perry, but is shot through with so many loopholes and exceptions, that its enforcement will be patchy.

Before the SH121 was offered for tender, the NTTA signed a protocol agreement with TxDOT stating that it would not bid for the concession. TxDOT completed the procurement process and awarded the concession to a Cintra-led team. The NTTA has attracted criticism for its subsequent late-stage intervention.

But Susan Buse, chief financial officer for the NTTA, says that the Texas legislature asked the authority to submit a letter at that stage, when the legal viability of the Cintra concession was under dispute. The NTTA responded, initially with a letter and then a "public delivery proposal" at the request of TxDOT. Buse was keen to assert that the NTTA bid, for which RBC Capital Markets is financial adviser, is not a CDA. TxDOT waived the protocol agreement in May.

Super discount?

The NTTA bid is higher in nominal terms than the Cintra bid, since its net present value (2007) is $3.33 billion, it involves an upfront concession fee of $2.5 billion, and annual payments to the state of $833 million. Cintra's accepted bid has a 2007 NPV of $2.8 billion, a concession fee of $2.1 billion, and annual payments of $700 million. RBC has lined up two further underwriters, and is ready to arrange a $3.5 billion bond issuance should the bid be accepted. The NTTA is also promising faster completion and more money for construction ($698 million as opposed to Cintra's $560 million), and lower maintenance fees ($560 million to Cintra's $1.7 million).

Though ostensibly more attractive to TxDOT, a number of concerns have been raised about how sound the bid is financially.
Following the leak of a TxDOT memo, the purpose of which is still unclear, the initial bid from NTTA was discredited on a number of fronts. The memo seized upon the NTTA's choice of a discount rate of 5%, since the Cintra bid factored in a discount rate of 8.5%. Market sources suggest that the 8.5% figure is a realistic one.

Discount rates are used to assess future cash flows; they reflect the cost of capital, or expected rate of return, expressed as the US Treasury rate plus a premium that reflects various risk factors. According to market sources, the standard deviation of future cash flows is currently set at around 8% or 9%, depending on the assessment of risk premiums. Though this is a relatively low discount rate, it is considered acceptable in market terms. Not a single market source could justify the 5% calculation, which is close to the yield on US Treasurys.

According to the memo, the Cintra bid results in the same total cost, $5.06 billion, using both models. The NTTA proposal's projected additional revenue sharing is between $2.057 billion and $4.817 billion at 5%, but falls to indicate a negative cash flow when the 8.5% discount rate is applied. The NTTA bid currently on the table has been modified since this memo was circulated.
RBC's Rebecca Heflin was involved in putting together the NTTA bid, and claims that "the discussion about discount rates is a red herring". She views the SH121 as a low-risk toll road, as it is financially robust and would form part of the NTTA's greater roads system, and is already an established corridor.

Under the NTTA's proposal, SH121 would be financially interdependent with the other tolled roads in the 4-county metroplex (including Collins, Dallas, Denton and Tarrant counties), and would not be a stand-alone entity financially. The risk is spread across the transport network, which could subsidise any unexpected losses. Susan Buse also notes that any profit would go directly back to fund public sector transport development.

However, critics of the NTTA bid have said that if the NTTA were to have difficulties sustaining the project, as regulator it can raise toll charges without a cap; whereas if a private concessionaire were to default, the sponsor would shoulder the burden without affecting the user. If the obligation were to fail, it would result in increases of $0.13 to $0.15 per mile, representing a 3x increase.

Moreover, the Federal Highway Administration (FHWA) has warned TxDOT that federal and TIFIA funding would be jeopardised if the NTTA bid were to be accepted, as is would violate fair competition practices in the procurement process. Despite the circumstances, the bid was submitted after the procurement process had closed, and may have benefited because the details of the Cintra bid were disclosed.

The outcome will be decided in the third week of June; other regional transport authorities in Texas will vote on the proposal, as SB 792 devolved much of the authoritative powers on such matters away from TxDOT. In addition to the NTTA, two other large toll systems in Texas enjoy substantial influence.

The first, the Central Texas Turnpike System (CTTS), controls the SH130 (sections 1-4), the SH45 North, and Loop 1, all to the north of state capital Austin. Also under its remit are sections 4 and 5 of the SH130, for which a Cintra/Zachry consortium holds a for-profit CDA concession, which features revenue sharing with the state. The CTTS is properly described as a part of TxDOT, although it has its own toll collection system – TxTag – and its own project-level bond debt of $1.4 billion. The four sections of SH130 not subject to a CDA were the subject of a design-build contract with the Lone Star Infrastructure Consortium: Fluor, Balfour Beatty and TJ Lambrecht.

The second, the Harris County Toll Road Authority, manages a network of toll projects around Houston. It has roughly $1.3 billion in toll road senior lien revenue bond debt outstanding, and runs 166km of toll roads. Under the SB792 compromise bill, Harris County has received the right to develop six road projects in TxDOT-owned rights of way. Harris County's attitude to the bill was clear – it wanted the extra funding that might result from a CDA programme, but wanted a much greater degree of control over projects in its vicinity.

Timely about-turns

Texas is not the only state where non-profits have faced down a for-profit solution. The first was Orange County's acquisition in 2003 of the private SR-91, the result, market observers agree, of a concession contract that was unusually generous to the private sector. There is a more recent precedent for the public authority overturning a concession proposal. In November 2005 the Virginia Department of Transportation (VDOT) received an unsolicited bid from a Macquarie-led consortium for a concession on the Dulles toll road.

Under state competition law, VDOT then issued the project for tender, and pre-qualified four bidders: The Dulles Corridor Mobility Initiative (Macquarie Infrastructure, Autostrade, John Laing and IIG); Cintra Dulles Proposal; Dulles Express (Franklin Haney); and Dulles SmartLink (Transurban and Goldman Sachs). The Transurban/Goldman proposal was considered to be the frontrunner at the time; it offered an upfront fee of $1.2 billion for a 50-year concession and $2.6 billion for a toll system, variable with inflation.

The following month, during the evaluation process for the received bids, VDOT received another proposal, from the Metropolitan Washington Airports Authority (MWAA) to take over the road and to raise money against it through bond issuance to fund a rail link between Washington DC and Dulles airport. In March 2006, VDOT signed an agreement with the MWAA to transfer ownership of the road for a 50-year period.

The other bids did not include any provision for the rail link. Sources close to the procurement process suggest that the motivation for VDOT agreeing to the MWAA proposal was political rather than financial. The rail project was considered technically problematic, and another private concession was not being considered for it at that time. In exchange for the road, MWAA was willing to take on the perceived burden of the rail project. VDOT has still not transferred the road to the MWAA; both parties are awaiting the approval of the designs for the rail line by the Federal Transport Administration.

One developer familiar with both Texas and Virginia believes that public toll authorities hold a significant role in the market, but that the rules have to be well defined in advance. He continues; "if TxDOT re-award the SH121 concession, it may sour the PPP market to some extent. The success of allowing municipal authorities to bid alongside private consortia is dependent on how the state handles the situation from the outset. It does not matter if the rules are equal or unequal, as long as they are pre-determined". He also notes that the public sector comparator is particularly important in such a situation.

One difference between private concessions and public toll authorities is that the authorities tend to own larger sets of assets. In theory, 63-20 companies, which are essentially hybrids of the private concession and the municipal transport authority's operations, might bridge the gap.So far, these companies have only been formed for smaller single projects, and have not enjoyed a glorious history, but they might come into their own as operators of larger assets.

The 63-20 not-for-profit model was used in the original Pocahontas Parkway deal south of Richmond, Virginia, and was run by Fluor Daniel (now Fluor Corporation) and Morrison Knudsen (now part of Washington Group). The road was just 14km long, but included some technically tricky components, notably a bridge over the James river and a ramp into Richmond. By 2004 it was evident that traffic revenue projections had been overestimated. Transurban took over the project as a for-profit entity in 2006.

Similarly, the not-for-profit Northwest Parkway Public Highway Authority (NWPPHA) in Denver, Colorado is in the process of selling its only asset, an 18 km toll road north of the city, as a 99-year concession to a Brisa/CCR consortium for $603 million. The deal involves defeasing the $500 million debt that the road has accrued, and is struggling to support.

Is there a wave of hybrids coming?

But in Pennsylvania a hybrid of public authority and private sponsor may well be the answer to the state's transport funding requirements. The state Governor, Edward Rendell, and the Pennsylvania Department of Transportation (PennDOT) have commissioned a report from Morgan Stanley to examine possible financing solutions. Rendell has made it clear that he favours a concession. However, the state's toll road authority, the Pennsylvania Turnpike Commission (PTC), is opposed to a private concession, as are Rendell's Republican opponents.

The report was released on 21 May 2007, and identifies a shortfall of $1.725 billion in funding annually, needed for maintaining and upgrading the transportation system in the state. This sum includes a requirement of $965 million for highways and $760 for transit. Three possible alternative solutions are outlined. Whether the 63-20 model could be applied to such a large project remains to be seen, but could be considered if the second option listed is chosen.

The first option would involve a long-term lease of the highways, with an estimated concession fee of between $12 billion and $18 billion for either a 75 or 99 year tenor, which invested at a rate of 7% to 9% would provide $1.3 billion to $1.6 billion against the current shortfall. The single concession would cover approximately 756km of the highways, excepting a section through Monfayette.

The second option is for a public corporation that would raise debt. This option would involve the formation of an agency with the authority to enter into short-term operations and maintenance contracts of up to 10 years. It would involve the use of tax-exempt debt and an increase of tolls after 2010 at the greater of the increase in nominal or per capita GDP or inflation, subject to a 2% per year floor.

The third option is that the PTC continues to run the highways, and implements a 25% increase in tolling by 2010, a congestion tax of $1 in Harrisburg, Pittsburg, Scranton and Philadelphia. It would also introduce a toll on the currently free Interstate 80 at the same rate as the turnpike.

The first option has already excited criticism in political circles, and the third option is unlikely to be popular with the public, who would be subject to the taxes and fee increases. A decision is due imminently from Rendell, and the second option is the favourite according to bankers contacted by Project Finance.

New Jersey Governor Jon Corzine is in a similar situation, in considering options for the New Jersey Turnpike, which is estimated to have around $30 billion in debt. The 63-20 option might well be implemented in this case too. Corzine has recently referenced a role for public benefit corporations and asset monetisation, as opposed to a full concession. An official decision is expected imminently.

This development follows the proposal of bill S2539 in February, which would have allowed the New Jersey Turnpike authority to offer a counter-bid to any concession agreement within 180 days. The bill was later revised. In this instance, the unequal playing-field would have been pre-established.

With public opinion increasingly against private toll concessions, but with large deficits on certain state-run highway systems, the hybrid partnership may become more commonplace. Comparing discount rates and risk premiums has made the debate over the SH121 project livelier, and more informed, than normal. But the private sector has not, so far, needed to make a case that private operators achieve greater revenues, and thus funnel more cash into transport projects, as a result of the profit motive. In this respect, the US is much more receptive to its case than many European societies.