U.S. Department of Transportation
Federal Highway Administration
Urs Greiner, Inc.
in association with
Public Financial Management, Inc.
Federal Highway Administration
U. S. Department of Transportation
The purpose of this report is to explore the policy implications to states of financing highway construction and maintenance using "shadow tolls." Shadow tolls are per vehicle amounts paid to a facility operator by a third party such as a sponsoring governmental entity and not by facility users. Shadow toll amounts paid to a facility operator would be based upon the type of vehicle and distance traveled.
Shadow tolls can be an element of a highway finance approach whereby a public or private sector developer/operator accepts certain obligations and risks - such as construction, operations and most specifically traffic - and receives periodic shadow toll payments in place of, or in addition to, real or explicit tolls paid by users. Funds for shadow tolls can come from diverse (and multiple) government and/or private sector sources, including State Highway Funds, special assessments on nearby properties and regional dedicated tax streams.
Shadow tolls automatically spread periodic or annual payments to a facility operator over a concession or franchise period; this can place the initial financing responsibility on the developer/operator rather than placing this burden on the public sector agency sponsoring the project. The reasons that shadow tolls may appeal to governmental units are that:
The traffic risk given to a developer/operator need not be an all-or-nothing proposition. There are several methods by which traffic risk may be dampened by thresholds or guarantees. For example, if traffic is significantly greater than specified, a portion of the additional shadow toll revenues resulting therefrom could be withheld or shared with the government sponsoring entity. Conversely, if the traffic is less than specified in the concession agreement, a portion of the revenue shortfall could be made up by the Government.
Shadow tolls are not a financing source in themselves, but rather a payment approach which can employ a range of financing methods, innovative or traditional, and can permit a viable financing structure that fits the characteristics and needs of certain projects. The concept of shadow tolls is, therefore, particularly applicable to public/private partnerships.
A project is typically defined as a public/private partnership or venture when two or more phases of the overall project development/operations process are the responsibility of the private sector. These phases of a typical process are: identification/initiation; planning; financing; design; construction; operation; ownership; and program management. If only one phase or element is the responsibility of the private sector, then it is often termed "privatization" or "outsourcing."
Recent years have seen growing federal support for public/private partnerships ranging from the landmark 1991 Intermodal Transportation Efficiency Act (ISTEA) legislation to the 1995 National Highway System Act and the current re-authorization discussions. This support has led to the expansion of the State Infrastructure Bank Program and to various other innovative financing initiatives. Shadow tolling can be an integral part of this trend.
Preliminary research in the area of shadow tolls was conducted in 1995 by URS Greiner. The report presenting this work was entitled "The Applicability of Shadow Toll Concepts in the United States" (FHWA P.O. No. DTFH61-95-P-00499). This follow-up study to that work includes capital markets analysis as well as transportation policy analysis. URS Greiner Consultants, Inc., in association with Public Financial Management, Inc., is carrying out the following scope of services for FHWA:
Sections addressing these subjects follow, together with a summary of significant conclusions (Section V).
Eight shadow toll contracts have been signed with private consortia in the United Kingdom, and one in Finland, though thus far no projects with shadow tolls have actually begun operation. However, these contracts, which will result in operating highway projects shortly, demonstrate how the private sector in these countries has reacted to these innovations and what terms it has found acceptable.
The use of shadow tolls in Great Britain has been part of a larger program developing public/private partnerships - the "Private Finance Initiative" - which has extended across many departments of the Government. In the case of transportation, the PFI has taken the form of "DBFO" concessions whereby a single private consortium develops, builds, finances and operates the road for a set number of years. DBFO-type concessions have long been used to develop projects in various parts of the world, but in all cases thus far the concessionaires' revenues have been earned through real, user-paid tolls. Even in the U.K., the Queen Elizabeth II Bridge across the Thames and a second crossing of the River Severn have been developed as DBFOs with user- paid tolls. But because real tolls continue to face strong political resistance in the U.K. (except on the occasional bridge or tunnel), the Government views shadow toll schemes as the next-best alternative. The U.K. Government had two major objectives in using shadow tolls: (1) to obtain better value for money by incentivizing the DBFO company to consider life-cycle costs, and (2) to cultivate a private sector highway operating industry that will be prepared for real tolls when (and if) they are implemented.
It is widely held that use of the private sector can lead to innovations in construction and financing that would create savings, mainly through speeding up the construction schedule and allowing for earlier openings. The Government has also been eager to use the DBFO program to shift economic risks to the private sector. And in fact, the U.K. Highways Agency claims that the contracts signed thus far will shave an average of 15 percent off of total costs compared to what the public sector would have paid under traditional arrangements. The Agency also claims to be very satisfied with the final arrangements to transfer risk - in particular the risk of delays from protests and the risk of latent defects - to the DBFO companies. However, at this early stage any quantification of savings is necessarily arbitrary as it involves attaching net present monetary values to unknown risks. These savings have also since been questioned by the U.K. National Audit Office because they are very sensitive to changes in the discount rate used.
For future projects, the Conservative Government was aiming to involve the private sector in earlier stages of project development. It also planned to move away from pure shadow toll schemes and adopt a greater variety of incentives to factor in other objectives such as safety and environmental quality. However, the new Labor Government is currently reviewing the entire DBFO program, though the first eight projects will be unaffected.
The DBFO program was launched in August 1994. As of March 1997, two groups of four concessions each had been awarded for eight separate DBFO projects totaling £567 million ($947 million), and a third group of seven others was under development. In selecting the first eight, the Government included a variety of projects in order to identify those types for which shadow toll schemes would be most beneficial. Among the projects are brand new roads, upgrades of existing roads to "motorway" standards, and the maintenance of existing roads with only minor improvements. The estimated capital value of these projects ranged from £9.4 to £214 million ($15.8 to $360 million).
Due to their varying nature and location, each of the eight DBFO contracts has unique features. Nevertheless certain terms and structures have been consistent throughout the entire program, and a clearer picture of what burdens the private sector is generally willing to bear has emerged from discussions and negotiations with bidders.
Payments by the Government to the operator will be based primarily on actual traffic levels, as measured in vehicle-kilometers. However, it was up to the bidding consortia to propose a precise formula for determining the payments. Since payments do not begin until traffic begins, there is a powerful incentive for the DBFO company to open the road as quickly as possible. But, since payments are pegged to traffic for the duration of the concession, the company also has reason to ensure that the road will need a minimum of disruptive repairs during that time.
Payments are not based simply on the product of traffic volumes and a single shadow toll. Instead traffic is divided into two to four "bands" representing different levels of annual traffic volumes with different per-vehicle payments attached to each. The lower bands have higher per- vehicle payments, while higher bands have lower per-vehicle payments. In all cases, the top band must be zero so that the government's liability is capped in the event of higher-than-expected traffic. This also addresses concerns that a shadow toll scheme would create incentives for the DBFO to vastly increase travel demand. The bands themselves may be increased over time to match anticipated growth in traffic. Separate bands are constructed for two vehicle classes - vehicles less than 5.2 meters in length, and vehicles greater than 5.2 meters. The DBFO performs continuous traffic counts to calculate annual vehicle-miles, which are verified by the Government every 90 days. Payments are indexed to inflation. An abstracted banding scheme is illustrated on the next page. Note that the increasing slope over time does not represent inflation but the change in bands to match traffic growth.
This banding structure determines the final cost to the Government, and as such constitutes the heart of the DBFO proposals. Most bidders have chosen to split traffic into four bands. The lowest band has typically represented a conservative forecast, with shadow tolls sufficient to cover debt service but with no return on equity. Obviously traffic forecasts are the major input to developing any banding scheme as there is no assurance that even the lowest band of traffic volumes will be achieved. Interestingly, while the Government maintains its own in- house traffic forecasts for each project, it has chosen not to disclose them to bidders. According to official documents, this was to foster a private domestic forecasting industry.
For the current round of projects, there are a few additional financial incentives for performance: bonuses for reduced accidents and deductions for lane closures. The DBFO may propose measures to increase safety. If the Highways Agency approves, then the DBFO builds and pays for the safety-related improvements. The DBFO will earn back this investment by additions to total annual payments of 25 percent of the economic cost of each personal injury avoided. Accidents avoided will be calculated by comparing accidents in the three years prior to implementation of the safety scheme (U.K. Highways Agency 1997).
Similarly, deductions from total payments will be made for lane closures (not including those required by the police or utilities). The deduction will vary depending on the number of lane-miles closed, the length of time, and the amount of expected traffic inconvenienced.
One of the major justifications for the British PFI program as a whole has been to transfer risk to the private sector and thus control costs. The DBFO contracts place all risks related to delivery of the road on the consortium, unless explicitly assumed by the Government in the contract. Thus, any unforeseen risks will be the responsibility of the private sector.
Risks typical of any private sector development (such as in real estate) apply to the DBFO company. These include higher than expected construction and operating costs, delays in completion, extra costs incurred due to inadequate design, and changes in law (except for those which specifically address PFI companies).
Risks which are specific to transportation projects include the risk of lower-than-expected traffic levels, protestor actions, and latent defects in existing roads. The risk of low traffic is assumed by the DBFO through the shadow toll mechanism.
For protestor actions, the Highways Agency asked bidders to compose three separate bids (i.e., three proposed per-vehicle shadow toll schedules) for three different risk-sharing scenarios: one where the DBFO assumes all costs associated with delays from protest, one where the costs are shared, and one where the government covers all extra costs. For most of the eight concessions, a risk-sharing arrangement was negotiated.
Bidders were responsible for inspecting the condition of existing roads for structural problems. But, in some cases, certain problems may not be detectable. The Government again asked the bidders to develop separate bids for three risk-sharing scenarios. It turned out that the private sector was generally willing to assume all of the risks of latent defects (U.K. Highways Agency 1997).
An average of eight consortia expressed interest in the projects put out to bid thus far. To control costs, the Government limited the competition to four bidders. Using its own traffic forecasts, the Government determined which bid provided the best "value for money," as measured in net present value.
In order to measure the savings generated by the DBFO bids, the Agency developed "public sector comparators" - estimates of what the state would pay to build the same project and operate it for 30 years using traditional contracting methods. These costs were expressed as a net present value using an 8 percent discount rate. The comparator includes a value to represent the various risks noted above. The Government project team and a team of risk consultants independently estimated these risks based on historic cost overruns in traditionally procured projects. The two teams came up with similar results.
It was by using such comparators that the Agency derived its average estimate of 15 percent savings for the eight projects. However, no documents are available to ascertain how risks are converted into monetary terms, or what proportion of the total value is made up of these risks (U.K. Highways Agency 1997). Moreover, the U.K. National Audit Office has questioned the precision of these estimates because they are very sensitive to the discount rate used to convert future payments into net present value. The Agency used a rate of 8 percent, but the U.K. Treasury recommends using a lower rate of 6 percent which is closer to the Government's actual cost of borrowing. If the lower rate had been used, one of the first four projects could then have been developed for a less costly amount by the Government, and the savings on the other three would have been significantly less than initially estimated.
All equity has come from project sponsors - there has been none from third parties, perhaps due to restrictions on the transfer of equity. Thus far, debt has been raised through commercial banks, funding from the European Investment Bank, or bond issues. The bank loans were for terms of 15 to 20 years and a margin of 120 to 140 basis points over the comparable term U.K. Treasury "gilt." A single £165 million bond was issued for two projects awarded to two subsidiaries of the same company. Because the bond was guaranteed by the AMBAC Indemnity Corporation, it received an AAA rating. The coupon was 9.18 percent. The Government requires that the consortia guarantee their performance by obtaining a bank guarantee for an amount commensurate with project construction costs (U.K. Highways Agency 1997).
Before the 1997 elections the Conservative Government announced a third group of seven road projects to be included in the DBFO program, some of which had already been put out to bid. However, the new Labor administration has launched a review of the DBFO program, killing two of the new projects, delaying three and allowing two to proceed. The original eight projects have been basically unaffected by the change in administrations ( Public Works Financing September 1997).
The previous administration had wanted to make some modifications to the DBFO program, most notably to incorporate additional incentives other than shadow tolls - deductions for lane closures, payments based on "reliability" of travel time, accidents as measured against national trends, and the proportion of high-capacity transport to low capacity transport. Other criteria may include the quality of information and signing, the condition of driver facilities and noise and air pollution.
Other European countries have long used private concessions to develop roads, but none with shadow tolls. However, Finland has recently followed in Britain's footsteps and recently closed financing on a project that includes shadow tolls - adding a second lane to a 70 km. section of the road linking Helsinki with Lahti. If the state's traffic forecasts hold true, it will pay 1.2 billion Finnish Marks over the 15-year concession period (or about 200 million US dollars). The Finnish government estimates that this represents a ten percent savings over traditional construction methods.
Terms of the concession are very similar to those in Britain. But, unlike the British concessions, the Finnish concession will last only 15 years (though bids for longer concessions were considered). The sharing of risks is also structured similarly to the British projects, although the Finnish government will assume all risks derived from protestor actions (Raitanen 1997). The bidding process also differed slightly in that government traffic forecasts were made known to the bidding consortia.
This project was financed locally, but it is likely that future projects in Finland will require the involvement of international finance due to the modest size of the Finnish banking community ( Privatisation International 1997).
Elsewhere, a major privatization initiative was recently launched in Germany. Thirty transportation projects have been earmarked for DBFO development, 14 of which are under active discussion, but only one of which has been financially closed. While plans are to use explicit tolls on these roads, pressures to keep tolls at "socially acceptable" levels may result in the use of shadow toll mechanisms (Bonar 1997). The German Government has shelved any further DBFO schemes. In France, Spain and Italy, plans to introduce public/private partnerships have been based on a more traditional concession model, with fees paid by users. Thus there are no plans to introduce shadow tolls in these countries.
Elements of financing include: (1) the types of revenues used to pay for the project; and (2) the types of financing instruments used, such as tax-exempt debt versus private equity. The following graphic should not be read to imply that, for instance, user fee revenues result in taxable debt. Rather, the graphic indicates that there is a range of pledged revenues which differs between a public and a private financing and there is a range of financing instruments available to each as well. It is important to note that purely private capital is only derived when the developer's balance sheet supports the issuance of taxable debt and/or developer equity. That is, the private sector participant is putting its full financial resources, and not just project derived revenues, behind the financing instrument. Purely private capital is very rare with public purpose infrastructure projects. Typically, third-party passive investors provide necessary capital via their purchase of municipal bonds, corporate debt, and corporate stock. If all capital costs are assumed equal, it can be shown that publicly issued tax-exempt debt will practically always result in lower project cost.
Where a project falls along the public to private continuum largely determines the project's benefits, risks, control, costs, and accountability. The financing issue is often at the core of project development and private sector participation discussions. The public or private entity that finances a given project in all likelihood will own the project and will heavily influence other development issues. The importance of project financing is probably why the term "privatization" is often used to mean the investment of private sector capital to finance public use transportation infrastructure. The ability to finance - whether public or private - is largely dependent upon the adequacy and stability of project dedicated revenues to support the debt service on bonds and/or the return on equity. The capital markets provide the mechanism which bring together the project sponsor trying to raise funding and third-party investors willing to purchase project debt or equity. Increased private sector participation in the design, construction, ownership, and/or operation of a public purpose transportation project does not necessarily increase or significantly impact the capacity to finance the project.
It is important to this analysis to understand how the private sector raises capital and what it costs. Generally, the private sector raises capital in one of two forms: capital investment in private sector taxable debt and private sector capital equity. In the first case, a private sector entity raises project capital by selling bonds or some similar product which usually provides a lien upon some specified revenue source and pays interest at a taxable rate. Purchasing the private sector entity's debt makes the investor a general creditor who has certain legal remedies to force repayment and who has priority over an equity investor. Private sector equity investment essentially makes the private investor a project owner, although the ownership could be either minority or majority interest, and subjects the private investor to the risk of loss of capital with no or only limited recourse to secure repayment. The ownership and risk associated with equity investment generally require a higher rate of return on capital in order to induce the investment. Whether private sector taxable debt or private sector project equity, one distinguishing characteristic of purely private capital is that a private sector entity is creating an investment opportunity for which a private sector investor is expecting a return on capital (Haugen, 1986).
Two points of interest concerning private capital are particularly relevant to this critical analysis. One is that the ability for a private sector entity to raise debt or equity capital is dependent upon a revenue stream, usually derived from or related to the transportation project being financed, that is sufficient to provide the required return on capital. This truism is no different than the public sector's ability to raise debt capital by issuing tax-exempt bonds, i.e., a revenue stream sufficient to allow the issuing entity to meet debt service payments is required. The second point of interest is that the return on private capital, whether the taxable interest rates paid on private sector debt or the internal rate of return on equity capital, is typically higher than the return required to attract investment in public sector debt. Put simply, a public entity's cost of capital for a given transportation project - all things being equal - is lower than a private sector entity's cost of capital.
In many municipalities throughout the United States there is a major effort to involve the private sector to build and operate transportation systems. The theory driving the use of these approaches is that business can more efficiently build and operate a transportation project than government. The private sector is not constrained by the project approval process (other than legally required approvals such as an environmental impact statement), does not have to follow government procurement practices, can build the project on a "privatized" basis (design and build), and can provide more efficient construction management. On the operating side, the private sector does not have to follow civil service employment requirements and can avoid government procurement procedures for spare parts and material supplies (Curry, 1992). These advantages may be more significant than the major financial fact confronting capital intensive transportation projects in the US, i.e., debt service levels for a privately financed project (not using tax-exempt debt) are higher than those for a publicly financed project of comparable cost.
There is a distinction between what will be referred to as the participant investor and the passive investor. The participant investor is a private sector entity that is willing to contribute capital to a given transportation project if they are to be awarded a franchise, a construction contract, etc., i.e., they will invest as long as they are a participant in project development in some additional capacity other than just financial. The passive investor is a private sector entity seeking an acceptable risk-return relationship from a purely financial perspective. The passive investor is generally, although not exclusively, a general creditor of the project seeking to buy taxable bonds or some similar form of structured debt yielding in the range of 10 percent to 20 percent interest on principal. As a general creditor, the passive investor will be in a stronger position with respect to ability to receive project cash flow. The passive investor may also invest in project equity, but does so solely on the basis of adequate cash flow return.
Generally, the participant investor is willing to contribute equity, possibly in the form of paying for up front development costs, and will expect a return on equity of 18 percent to 30 percent over the long term. A participant investor may also purchase some form of structured debt, such as deferred or subordinate construction notes, as a means to stimulate the project's financial feasibility. As an active participant in the project, the participant investor's risk as an equity contributor or subordinate note holder is offset by the profits received from project management fees, construction profits, operations contract fees, etc. These are profits all of which (1) occur if the project moves forward and (2) are in addition to any return on equity or interest earned. From this perspective, and taken to a simplified extreme, one can view the "equity" that a participant investor is willing to contribute as the resources that same entity would have to commit to respond to a public request for proposals or competitive bid for a project contract. From the participant investor's perspective, why not expend those resources developing a project concept in a less competitive process and call them "equity" for which an 18 percent to 30 percent return might be achieved, in addition to the award of a contract and the associated contract profits.
The participant versus passive investor analysis can be taken one step further by considering the source of backing for capital contributions to a transportation project. Private sector project proposals solicited by state DOTs under "privatization" or public/private partnership programs often include promises of private capital for the project. However, the proposed private capital is almost never contributed directly from the private developer/contractor/operator but is raised from third party institutional investors via a "shell" project corporation and supported solely by project revenues such as explicit tolls or shadow tolls. The third party institutional investors, who are the passive investors described above, include hundreds of mutual funds, bank trusts, insurance companies, pension funds, etc. (there is no shortage of institutional investors for either taxable or tax-exempt financial instruments) who are seeking an appropriate risk-return relationship. The distinction is that the private sector project participant is not risking its own capital by "pledging" its balance sheet and income statement to provide the return on equity. It is simply creating a conduit to raise passive investment capital based solely on the financial merits of the project itself.
These distinctions are important ones, and several relevant points can be derived. First of all, the profit motive clearly stands out for the participant investor. Also, the popular notion that says a huge untapped pool of private capital is available and waiting for public use transportation infrastructure begins to break down. Project development and construction companies may be willing to commit relatively small amounts of "equity" capital toward the development of transportation projects, but typically only if they are awarded lucrative project management/ construction/operations contracts. The true investors in transportation infrastructure, i.e., the passive investors looking solely at the financial risk-return relationship, largely consist of those institutional investors which have become comfortable with the various development risks associated with user fee transportation infrastructure projects, namely toll expressways. These financial companies, and many similar ones, are just as likely if not more likely to be passive investors in publicly developed transportation projects, especially if a municipal entity is willing to mitigate some project risks with credit support. (As suggested later in a case study, shadow tolls could be used as credit support to assist project financing.) A public entity can create a conduit financing enterprise such as a 63-20 not-for-profit corporation and pledge project-derived revenues to debt issuance just as easily as the private sector, and at a lower tax-exempt cost of capital. The statutory and legal complications associated with public issuance of debt are no more difficult to overcome than the procurement, management and public equity issues associated with privatization techniques. In short, "privatization" does not mean that a new source for capital investment immediately becomes available where it otherwise was not.
There have not been many recent examples of successful privatized financings. Two of the first much heralded "private" capital financings, the TCA San Joaquin Toll Road and the Osceola Parkway, were publicly owned and used tax-exempt financing together with a public/ private development partnership (Miller 1990; Reinhardt 1993a; Osceola County 1992). Probably the two most notable transportation project financings which have used privatized capital are the Caltrans SR 91 Toll Road and the Dulles Greenway Toll Road. The various development issues for each of these innovative projects are summarized by the following table. These projects have been greatly praised. Public Works Financing (Reinhardt, 1993b) wrote about SR 91, "California, even wealthy Orange County, doesn't have enough money to build all the infrastructure it needs. Every transportation dollar raised by the private sector contributes to the public goal of improving mobility." The fact this statement ignores is that these private sector contributions were raised largely from third party investors and are being repaid from the tolls paid by the citizen and business users of the facility. No private capital has been raised that could not have been raised on a public basis and at a lower tax-exempt cost. Such is the fallacy of most "privatized" capital techniques.
|CALTRANS SR91||DULLES TOLL ROAD|
|Genesis||Private Proposals--Public Selection||Private Proposal/Private Initiation|
|Determination of Service Characteristics||Private Consortia||Private Certificate Holder|
|Principal Objective||Expand Capacity and Private Profit||Private Profit|
|Right of Way Assembly||Private Lease from State||Private Proffer|
|Financing Instrument||Taxable Debt, Subordinated Debt, & Sponsor Equity||Taxable Debt, Subordinated Debt, & Sponsor Equity|
|Reversion to Public Sector||Yes||Yes|
|Toll Regulation||Regulated Return on Equity||Regulated Return on Equity|
Whether private capital or public capital, these projects embody the basic concepts of project revenue financing which are the same in either case. And the project revenue which is the ultimate financing source is the user fee revenue paid by citizens and businesses. The users of these facilities will now be forced to pay high tolls in the future in order to cover the private sector's higher (than the public sector) cost of capital. Both of these projects utilized taxable debt with yields in excess of 10 percent and equity with expected returns in excess of 20 percent. Although the return on equity has yet to be achieved, it can be reasonably assumed that the overall cost of capital for the private Dulles and SR 91 Toll Roads will be almost double the 6 percent to 8 percent tax-exempt cost of capital for the San Joaquin and Osceola Parkway Toll Roads. Differences in capital markets conditions were relatively minor compared to the differences in cost of capital.
The one argument for privatized capital that seems to sometimes be true is the private sector's willingness to fund the early stages of development, when project risks are at their greatest, if future participation in project management/construction/operation is included. These early development costs were privately funded in the Dulles Greenway project. However, going back to previous discussion of participant investors versus passive investors, it seems that the private sector's willingness to fund early stages of development comes primarily from participant investors who stand the most to gain from a successfully developed project. Passive investors, bound by the basic fundamentals of project revenue financing (whether the project is public or private), will not commit capital until a project is better defined and some development risks mitigated.
This issue brings forth an important question of how to fund early project development costs, such as preliminary design and environmental assessment. If private (participant investor) capital is not cost effective, then what is available that is cost effective? For years many states have been financing wastewater facilities through mechanisms called revolving loan funds. In concept, a state finances - through grants, taxes and/or bond proceeds - a loan pool that is then used to make low cost and/or subordinate loans to municipalities needing to finance wastewater facilities. The State of Florida brought this concept to highway financing when it created a Toll Facilities Revolving Trust Fund ("TFRTF") initially funded by State gas taxes and highway funds. The TFRTF makes low cost or even interest free loans to municipal entities trying to start up the development of a Toll Road project. The TFRTF loans are then repaid on a subordinated basis from any toll revenues derived from the fully developed project and in excess of operations and debt service. The revolving loan fund concept for public surface transportation has now been introduced at the Federal level via US DOT's state infrastructure bank ("SIB") pilot program. At the time of the writing of this report, legislation which makes permanent and expands upon this concept is pending in the next transportation act re-authorization. In addition to direct loans, it has been suggested that the revolving fund concept could be modified as a form of credit enhancement, which would lower tax-exempt borrowing costs and aid project financial feasibility (Feldman, 1993).
These ideas could be modified to include the shadow toll concept or a variation thereof. For example, rather than loan money to a project for construction costs, a State SIB could pledge shadow toll payments to a project on top of explicit tolls so that the bonding capacity of the project is increased. A similar variation that would offer credit enhancement would be an "inverse" shadow toll, i.e., any shortfalls of actual traffic to projected traffic could carry an associated shadow toll payment from a SIB or some other governmental entity. The inverse shadow toll would, in effect, provide a line of credit which would "guarantee" the projected revenue stream and should result in a lower project cost of capital. When combined with a revolving loan or SIB line of credit concept, any shadow toll payments to the project would ultimately be repaid from explicit tolls once they became sufficient.
Private sector project development techniques embody much more than private capital investment and, as discussed, most "privatization" techniques in the US generally work better without private capital investment due to the lower cost of capital resulting from governmental tax-exempt status. Public/private partnerships, defined previously on page 2, can incorporate the best ideals of "privatization" without a loss of public control over a transportation project and without the excessive cost of capital and profits associated with private capital investment. The many benefits of public/private partnerships have been well debated and documented.
Public/private partnerships using shadow tolls can also be implemented with a government entity or "not-for-profit" organization as the sponsor or administrator receiving funds and disbursing construction moneys and shadow tolls. This would permit the dual advantages of tax- exempt financing and private sector efficiencies to be realized. As an example of how public/ private partnerships might work with shadow tolls, consider value capture techniques. Such techniques include benefit assessments, tax increments, and even direct payments from private entities (with no associated repayments). The concept of value capture is that the private entities which stand to gain the most from transportation infrastructure improvements - such gains being derived from increased access, increasing property values, etc. - help to pay for such infrastructure via one of the capture techniques. Any of the value capture techniques could be combined with shadow tolls, i.e., the assessment or increment paid would vary according to actual traffic. Such an approach may be more acceptable to private participants who may have difficulty estimating the real economic impacts a project might hold for them. For credit reasons to be discussed, value capture shadow tolls would probably best be used as revenue enhancements but not as the primary backing for project financing.
The previous sections have indicated that: (1) a transportation project's financing can largely be dis-aggregated from many other elements of overall project development: and (2) private sector financing is probably not desirable, all other things being comparable, in that it carries a higher cost of capital. Furthermore, access to the capital markets is far more dependent upon the financial merits of the project to be financed, including any dedicated revenues, than upon any particular form of private sector project development participation. These conclusions provide a more focused lead into an analysis of shadow tolls as a method of financing. More specifically, shadow tolls can be compared to explicit toll revenues and to tax-backed revenues as sources to support a project debt financing. (A comparison of cash funding via grants and capital reserves versus debt financing would make another useful analysis but is not the focus here as it is assumed that grants and capital reserves, if available, would be used before debt financing.) In making this comparison, it will be assumed that a public purpose transportation project will be publicly owned and publicly financed to avoid unnecessary confusion related to differences in cost of capital.
There are numerous factors which are evaluated by capital markets participants such as credit analysts and institutional investors when making decisions regarding credit ratings and investment yield requirements. Many of these factors are uniform for all public infrastructure financings without regard to the specific revenues pledged to the repayment of debt. The uniform credit considerations include socioeconomic data, entity administration and management, and the legal contracts and bond covenants. Socioeconomic data considerations focus on the regional trends in population, employment, property values, retail sales, building permits, etc. Administration and management considerations include the political structure, stability of management, long- term capital planning, and public need/support for the project. Legal and bond considerations include the strength of contracts with funding partners, the ability to raise additional revenues, the projection of revenues including feasibility reports, the flow of pledged revenues pursuant to the bond indenture, and debt service coverage requirements (S&P, 1997). All of these factors should generally impact the creditworthiness and marketability of a proposed transportation project financing equally, except for the projection of revenues and ability to raise additional revenues. It may be more difficult to estimate future revenues with shadow tolls and especially with explicit tolls, as compared with straightforward tax-backed revenue sources, due the difficulties in estimating traffic.
Toll road financings have their own unique issues which require additional capital markets analysis. One of the more important distinctions between toll roads and most other financings is that toll roads face competition from toll free roads (S&P, 1997). Toll roads must offer meaningful time savings to travelers in order to cause them to voluntarily pay a toll to use the facility. Some utilities, especially public power and solid waste, also face competition although typically to a lesser degree. But most public infrastructure financings are backed by revenue streams which are not subject to competition. The typically easy access which drivers have to toll-free roads puts toll revenues among the most difficult to predict, especially in urban areas where there are often many route alternatives and roads are built or improved almost continuously. The competition issue is one area where shadow tolls may offer a better credit than explicit tolls and therefore improve access to financing when compared to explicit tolls. Because usage is "free" to travelers using a shadow toll facility, traffic should be easier to predict as it will be based upon times savings and not a time savings/toll cost analysis. Other credit factors unique to toll road financings include the toll rate structure, trip characteristics, rate setting procedures, toll collection procedures, and maintenance standards (S&P, 1997). Some of these factors only apply to explicit tolls and not shadow tolls. A summary table of explicit toll and shadow toll credit risk considerations is provided by the following table. Due to these unique credit issues, toll-revenue-backed financings typically have a lower credit rating and a higher cost of capital than tax-backed financings for the same geographic region and/or issuing entity.
New or "start-up" toll roads, especially those which are not an addition to an existing toll facility system, present even more difficult credit issues which differentiate toll credits from most other public infrastructure credits. First of all, traffic usage is much more difficult to project when there is no existing traffic to use as a historical basis. In addition, the non-recourse nature of toll road financings combined with the increasingly high costs of development and construction typically result in bonding capacity from projected tolls being fully utilized. Therefore, the project has no ability to absorb capital cost increases resulting from construction risks or other risk factors, such as those related to right of way or environmental issues. Capital market participants are increasingly focusing on terms and features of the construction contract, insurance coverages, and contingency reserves available in assessing the creditworthiness of a start-up toll financing. It is quite difficult to achieve an investment grade credit rating for start-up toll facility financings, and a rating above "BBB" is literally without precedent in the past decade or more. Whether or not these same start-up features would apply to a shadow toll facility would probably depend upon the ability to adjust shadow toll rates to reflect any increased capital costs. Of course, such a toll adjustment mechanism would shift risk to the sponsor entity paying the shadow toll.
|COMPARISON OF CREDIT RISKS FOR EXPLICIT TOLL AND SHADOW TOLL FINANCINGS|
|FACTORS APPLICABLE TO BOTH||UNIQUE TO EXPLICIT FINANCINGS||UNIQUE TO SHADOW TOLL FINANCINGS|
General Traffic Risk, (Not Attributable to Driver Willingness-to-Pay) -
Toll Revenue Risk - Determined By:
Explicit Toll-Related Traffic Projection Risk, Directly Attributable to Driver Willingness to Pay - Affected by:
Toll Agreement Risks - Risks specifically associated with terms of tolling agreements, including:
|Shadow Toll Revenue Risk - Determined by:
Underlying Revenue Risk - Underlying risk associated with revenue source(s) pledged as shadow tolls, including the existence or absence of full faith and credit pledge, and all the traditional risk factors used in analyzing municipal revenue risk.
Shadow Toll Agreement Risk - Risk specifically associated with terms of shadow toll agreement, including:
1 Note that availability of competing facilities/comparative time savings will affect explicit toll revenues more significantly because explicit toll facility users will have a stronger preference for non-toll facilities, whereas users of shadow toll facilities should be neutral.
Given the above discussion, it is reasonable to conclude that an issue based upon a stable tax-backed revenue stream should achieve a higher credit rating and lower cost of capital than one based only on toll revenue, all else being equal. Although there is no U.S. public finance precedent, it is also probably reasonable to conclude that some of the risks unique to toll facilities, such as the difficulty in predicting traffic, would apply to shadow toll facilities; however, not all risks inherent to explicit tolls are applicable or as onerous when applied to shadow tolls. Nonetheless, it cannot be determined that a shadow toll revenue credit is better than an explicit toll revenue credit without knowing the initial source of the funding for the shadow toll payments.
The credit ratings and cost of capital of a shadow toll revenue source is going to be dependent in large part upon the credit rating and cost of capital associated with the underlying revenue source(s). If stable and predictable revenue derived from a general obligation property tax, sales tax or excise tax (e.g. a gas tax) is the underlying source of shadow toll payments, then the shadow toll revenue will likely represent a better credit and lower cost of capital than explicit toll revenue for the same roadway. Alternatively, if an unpredictable revenue source such as a tax increment or special assessment is used, then the shadow toll may represent a weaker credit than an explicit toll.
The more important conclusion to draw is that a shadow toll adds an increased element of variability and unpredictability (and possibly appropriation risk) to the underlying source of revenues and thus will probably result in a weaker credit than the underlying source by itself. Shadow-toll-backed bonds would probably represent a higher cost of capital than simply issuing bonds backed by a direct pledge of a single underlying source. Of course, the revenues required to service the bonds would then be fixed - assuming fixed rate bonds - without regard to the usage of the facility.
Credit ratings are project and financing specific, making any generic cost comparison impossible. Furthermore, market conditions change constantly, including interest rate differentials between higher credit quality and lower credit quality financings. Nonetheless, given the above assessment of how risks differ among tax-backed versus shadow toll versus explicit toll credits and given historical interest rate spreads between financings of different credit ratings, we can estimate within a reasonable range of accuracy the financing cost differential.
For purposes of analysis, assume a $100 million new highway project to be financed over thirty years under each of the following revenue scenarios:
A historical estimate of the interest rate differential, assuming no additional credit enhancement, would be 40 basis points (0.40 percent) between the "AA" rated bonds and the "A" rated bonds and also another 40 basis points between the "A" rated bonds and the "BBB" rated bonds. Actual interest rate spreads will vary depending upon the market level of interest rates and the steepness of the interest rate yield curve. (For instance, in the historically low interest rate and flat yield curve environment of early 1998, actual spreads may be as low as 20 basis points from one rating category to the next.) Assuming general credit market conditions such that the interest rate on the "A" rated bonds is 6.0 percent, we get the following results from this simple example.
|Bond Rating||Interest Rate||Annual Debt Service||Aggregate Debt Service||Present Value @ 6%|
As shown, the cost differential from one rating category to the next is approximately $4.3 million in present value dollars or 4.3 percent of the total project cost. In aggregate future value dollars, the cost difference over the financing term is approximately $9.3 million.
It may be possible to mitigate the cost differential due to the above credit concerns. One alternative would be to back a DOT shadow toll security with the state's full faith and credit, the state's overall strongest credit. It is common for states to place a full faith and credit pledge in support of self-supporting revenue bond programs, including highway/gas tax financing programs and explicit toll revenue financings. The analyses presented above may create an argument for using tax-backed revenue bonds to fund highway construction costs as opposed to shadow tolls backed by the same tax revenues because the financing costs would be higher for the shadow toll option. However, if the state has a history of backing bonding programs with a full faith and credit pledge, then financing costs become a moot point as the state's general obligation credit rating is substituted. Given this alternative, the decision to pledge tax revenues outright as opposed to using shadow tolls becomes one of preferences toward traffic risk.
The question of pledging revenues to fixed rate debt (or providing full faith and credit backup to shadow tolls) versus an obligation to make shadow toll payments (with no credit enhancement) becomes, to a certain extent, a question of long-term traffic risk and long-term cost. If using solely shadow tolls to support a project financing, traffic lower than projected means the sponsor pays less as the bondholders take the risk (passed through from a DBO if the franchise model is used). However, if traffic is higher than projected, the sponsor pays more (and the DBO gains a large profit if the franchise model is used). Fixed debt service versus variable shadow toll payments thus may be either an advantage or disadvantage depending upon the risk profile of the sponsor entity.
Capital markets participants have studied and evaluated the long-term credit risk associated with start-up toll roads from an investment perspective. This analysis can be useful for public project sponsors considering shadow tolls. In brief, the capital markets analysis suggests that new toll roads often can not cover both amortized construction costs (via debt service) and M&O costs during the early years of operation and sometimes into the intermediate (10 to 12) years. However, in the long term (15 years or more) almost all toll roads receive enough traffic and revenue to more than pay for construction amortization and M&O costs. This researched fact is the reason that many start-up toll roads can receive investment-grade "BBB" category long- term credit ratings. Historically, investment-grade-rated municipal bonds have a very low default rate. Although the credit analysis was based upon explicit tolls and not shadow tolls, it is still instructive for guidance. Public project sponsors may conclude that it is in their best long-term financial interest to use tax-backed fixed-rate debt to finance a highway project as opposed to variable shadow toll payments to back debt. Not only does tax-backed fixed-rate debt have a lower cost of capital, it may also result in lower overall payments for the financed highway due to the elimination of long-term upside traffic potential.
Functions under a project arrangement utilizing shadow toll concepts will include:
Non-user funding sources can be based on government contributions, including:
Funding sources can, and even should, include "value capture" techniques whereby beneficiaries of the project contribute in rough proportion to the benefits they receive. Value capture funding sources may include:
The administering entity, generally but not necessarily a governmental unit, could be an office or agency of a state department of transportation, a state infrastructure bank, an economic or industrial development agency, a local government, or a non-profit, private sector organization. It may be possible and advisable for this administering entity to have the power to issue debt; clearly it would be most beneficial if this debt could be tax exempt. Payments to the DBO could be on the basis of shadow tolls to cover life-cycle costs - construction plus operations - or shadow tolls to cover only maintenance and operations ("M&O") costs. Under the latter scenario the bonding capability of the administrator could be utilized to cover construction costs, and annual income from the various funding sources could pay construction debt service and M&O- based shadow tolls. This is illustrated on the accompanying graphic entitled "Alternative Shadow Toll Schemes."
The DBO entity could be a private sector consortium, but not necessarily. An established toll agency (already possessing bonding capabilities) could also fill that role, as could a state DOT operating under contract to or a special arrangement with the project administrator. A particularly interesting approach would be to have public and private sector organizations compete to be the DBO.
The fundamental questions to be asked in assessing a possible shadow toll structured project are:
For most projects considered meritorious but for which normal governmental program financing is not available, the first assessment is whether they can be financed as a pure (real or explicit) toll project on a non-recourse basis. Factors in this phase of the decision-making process typically include:
If "real" tolls are not feasible, the next phase in the decision-making process may be comparing shadow toll concepts to government administration of the project using comparable revenue sources. Considerations in this phase of the process may include:
Project debt may be issued to cover construction costs only, or construction and major maintenance costs. Debt may be issued by the project administrator or by the DBO consortium. As noted above, both the administrator and the DBO could be either public or private sector entities, though it will be more likely that the former is from the public sector and the latter from the private. Depending on this, either entity may be able to issue tax-exempt debt - this should certainly be a factor in the formulation of the overall project structure.
In a typical (real) toll-financed non-recourse project, the variability of traffic and the fact that revenue derived therefrom is subject to elasticity as well as political concerns (will they really be permitted to increase tolls if necessary?) directly influences interest rates and overall project viability. Shadow-toll-based debt issued by the DBO would certainly have its terms influenced by traffic variability, but not elasticity or political reactions to proposed toll increases. Debt issued by the project administrator would be completely independent of all of these concerns, and instead would reflect the creditworthiness of the underlying funding sources.
The potential number and diversity of these funding sources can improve the credit worthiness of an issue dependent on them. Further, the project administrator could be granted the power to increase income from some or all of these sources. Agreements with government funding sources could include provisions for adjustments if construction difficulties or variable traffic made this necessary. Agreements with developers and assessment districts could be similarly structured, though limits would seem to be an appropriate subject of discussion. From the perspective of financial markets, this ability to increase revenues, if required, could take the debt out of the pure "non-recourse" category and improve its marketability and interest rate.
Many toll authorities today, particularly those in urban areas, are often requested to institute congestion relief measures or modification/improvements to achieve environmental or social objectives. These measures may include: the reduction of tolls during off-peak periods; reductions of tolls to benefit commuters, local resident or other categories of users; the conversion of an existing traffic lane to high-occupancy vehicle usage; the implementation of intelligent transportation system (ITS) techniques; and the construction of new high-occupancy vehicle or other special use lanes.
Many toll authorities are bound by trust indenture agreements which specifically prohibit any actions on their part which will impair net toll revenue levels. Thus many of the foregoing measures, while socially, environmentally or politically most desirable, cannot be implemented within the present legal arrangements under which the toll authorities must operate.
The calculation of the revenue loss per vehicle were such measures to be implemented can be the basis for shadow toll payments - or more precisely in this case, shadow toll supplements to the real tolls charged by the operating toll agency. The shadow toll payments could be met from recurring annual payments from a federal or state program, or other sources. Alternatively, a federal grant or single state appropriation could be used to establish a fund that could permit necessary construction to be subsidized and on-going operating revenues to be supplemented via shadow tolls.
The toll authority or agency would typically have the resources to carry out all of the necessary project implementation functions ranging from planning and design through construction and operation. They would also be the owners of the improvement, and could in many cases finance it themselves via a tax-exempt issue.
Shadow toll concepts can be most applicable to the needs of toll authorities requiring such improvements where net revenue levels may be adversely affected.
In some instances, the need for a new highway project or improvement can be based upon potential community benefits. These can include: access roads to permit economic development; by-passes of a central business district to reduce congestion or truck traffic; other types of congestion relief measures, including ITS procedures; and highway improvements to preserve the character of landmarked, historic or scenic areas.
It is appropriate to identify categories of beneficiaries, and specific benefitting parties within each category. Through such an approach, and normal governmental/political/institutional processes, sources of funding may be identified. As indicated above, these could include grants under various federal programs, state appropriations, county/municipal contributions, special tax assessments, developer fees, and tax increment financing, among others.
To achieve the benefits of tax-exempt financing, and to ensure local/community control of the process, the project sponsor or developer could be a local governmental entity or a non-profit organization established for the specific purpose of advancing the project. This sponsoring entity could then utilize the resources of existing government departments or consultants to carry out necessary planning, design and construction management services. Funds would be accumulated in a special reserve, and construction contractors and, if necessary, facility operators could be paid via the disbursement of shadow tolls. Shadow tolls are not essential to this process; funds could be received and properly disbursed by the sponsoring entity without the employment of shadow toll concepts; nevertheless, if some of the contributions were of a recurring nature, and a private sector operator was to be given the traffic risk, shadow tolling could be appropriate. Alternatively, if a local government were given the facility maintenance responsibility, and the level of required maintenance was roughly proportional to traffic usage, recurring shadow toll payments to the government entity maintaining the facility could be reasonable.
Periodically there are projects proposed of national significance due to location, historic importance, or other factors. In some cases shadow tolling may be an appropriate vehicle to transfer resources from the funding sources to the developer, contractor or operator providing project services.
Depending upon the characteristics of the project, funding sources could include federal grants, state appropriations and local governmental contributions. Having a sponsoring entity possess the ability to issue tax-exempt bonds would clearly be advantageous; a properly constituted commission or authority could be suitable.
Professional services required through construction of the project could be paid for directly by the sponsor; these might include project planning, design, construction and construction management. The continuing operation of the project could be reimbursed using shadow tolls - particularly if the attraction or attainment of specific traffic levels were a responsibility of the operator. A significant portion of the operator's maintenance expenses would be proportional to traffic usage, and shadow tolling would be responsive to this.
As noted previously, major advantages of shadow toll concepts include: ready adapt ability to multiple funding sources; the cost to the sponsoring government entity over the concession period can be largely known; the operator is given the traffic risk, or a portion thereof; and competition for a concession or franchise can be based primarily upon a shadow toll schedule and hence can be very "open" and clear cut. Major strengths of public/private partnerships include the use of private sector funds and the extensive support already provided by various legislative actions and mandates. The strengths of both concepts can be most complementary.
Generally accepted considerations with respect to project implementation include:
The foregoing factors guiding project development must be reflected in tasks and pronouncements over the course of the project development process. Each is necessary to obtain and hold support, and create the broad consensus that is essential for a project to advance. The length of the development time frame is a major concern in the implementation guidelines suggested below. It is crucial to minimize uncertainties and durations of involvement on the part of private sector parties if their interest is to be maintained.
Phase One of an implementation process could include the following:
While a potential project could be identified by virtually anybody, the other first steps noted above often have a local orientation, though state agencies' support will be necessary in practice. A private sector developer or a local interest could be a catalyst to pull these initial pieces together. The output of Phase One should be a set of documents demonstrating the merits of proceeding further. The identification of a potential sponsor - and the availability of tax- exempt financing - should also be explored.
Phase Two of a possible implementation process could be under the sponsorship of a state agency using legal, financial, and engineering consulting services as required. Major elements of a Phase Two effort could be:
The development of a financing package should consider "who benefits." The public sector administrator/sponsor/banker could be a state infrastructure bank, or a properly staffed affiliate of a state infrastructure bank or another state agency. The capability to issue tax-exempt debt could be advantageous if the sponsor or developer is not a purely "for-profit" private sector organization.
A third and final implementation phase could include:
The determining factor among qualified responders is simply a shadow toll schedule. A term of years could be specified, and the developer proposing the lowest toll schedule could be selected. A combination of term and toll at a given interest rate permitting economic comparisons to be made could also be the primary selection criterion.
Public sector advantages associated with the use of shadow toll concepts and public/ private partnerships include:
Public sector administrator duties include the collection of revenues from diverse sources, remitting the shadow-toll-based payments to the developer/operator, and monitoring performance.
Advantages to the private sector include:
There are many private sector developers who have been badly hurt by project schedules that go on without end. The procedure outlined above limits that duration and has government- sponsored studies and support precede formal private-sector commitments. The minimizing of front-end funding requirements will encourage more private-sector organizations to consider infrastructure development and stimulate competition.
Highway infrastructure development in the United States has traditionally followed a limited number of models which have generally been based upon Federal Aid Highway programs, formulas and procedures or upon toll authority powers. But the fundamentals of highway development are changing, and due to funding limitations, new Federal policies and legislation (such as state infrastructure banks and other initiatives), and the increasing acceptance of public/ private partnerships, there is an increasing diversity of highway development models now being considered throughout the nation.
Shadow tolls have been successfully used abroad, and can be applied in the United States under the new highway development "models." This is increasingly appropriate as a wider range of project sponsors, legal frameworks and potential financing sources are explored and utilized.
Shadow tolls are a funds disbursement method to a highway developer/operator (or DBO or DFBO) which is directly keyed to actual, achieved traffic levels. It thus passes all or a portion of the traffic risk to that entity. In a broader sense, however, shadow tolls can sometimes permit a development process to be structured which can be more responsive to institutional, financial and political realities and acceptable project development approaches. Shadow tolls may be appropriate if:
Project debt may be issued to cover construction costs only, or construction and specific recurring major maintenance costs. Debt may be issued by the project sponsor or administrator or by the DBO. While the administrator and the DBO could be either public or private sector entities, it will be more likely that the former is from the public sector and the latter from the private. Depending on this, either entity may be able to issue tax-exempt debt - which can reduce interest costs to the project. This should be a major consideration in project formulation and the selection of sponsor and developer/operator entities and the determination of their roles and responsibilities.
In a typical (real) toll financed, non-recourse debt project, the variability of traffic and the fact that revenue derived therefrom is subject to demand elasticity, as well as possibly political concerns at times of proposed toll increases, directly influences interest rates and overall project viability. Shadow-toll-based debt issued by the DBO would have its terms influenced by traffic variability, but not elasticity or political resistance to toll increases. Debt issued by the project administrator would be completely independent of all of these concerns, and instead would reflect the creditworthiness of the underlying funding sources.
The potential number and diversity of these funding sources also can improve the credit worthiness of an issue dependent on them. Further, the project administrator could be granted the power to increase income from some or all of these sources under specified conditions. From the perspective of financial markets, this ability to increase revenues, if required, could take the debt out of the pure "non-recourse" category and improve its marketability and interest rate.
Many toll authorities today, particularly in urban areas, are often requested to institute congestion relief measures or modifications/improvements to achieve environmental or social objectives. These measures may include: the reduction of tolls during off-peak periods; the conversion of an existing traffic lane to high-occupancy vehicle usage; and the construction of new high-occupancy vehicle or other special use lanes. Trust indenture agreements may specifically prohibit any actions which will impair net toll revenue levels. Thus the foregoing measures, while socially, environmentally or politically most desirable, cannot be implemented within the present legal arrangements under which many toll authorities must operate.
The calculation of the revenue loss per vehicle were such measures to be implemented could be the basis for shadow toll payments - or more precisely in this case, shadow toll supplements to the real tolls charged by the operating toll agency. The shadow toll payments could be met from recurring annual payments from a federal or state program, or other sources.
In summary, shadow toll concepts can be selectively utilized in the United States. Considerations influencing their viability for a particular set of circumstances include:
Bonar, Mary and Stephen Fuss. The Use of Private Finance for Infrastructure Investment in Western Europe. Wilde Sapte, London. 1997.
Curry, Keith. Introduction to Public Finance And Public Transit. U.S. Department of Transportation, Federal Transit Administration, Office of Technical Assistance and Safety, Washington, DC. 1992.
Feldman, Roger D. Reinventing Transportation Infrastructure Credit Enhancement. McDermott, Will & Emery, Washington, DC. 1993.
Haugen, Robert A. Modern Investment Theory. Prentice-Hall, Englewood Cliffs, New Jersey. 1986.
Miller, John. Innovations in Transportation Development sponsored by Public Financial Management, Inc., Philadelphia, PA. 1990.
Osceola County. Official Statement for the Osceola County, Florida Transportation Improvement Bonds (Osceola Parkway Project), July, 1992.
Public Works Financing. "U.K. Shadow Roads Review," 110. 1997.
Public Works Financing. "U.K. Shadow Toll Savings Suspect," 115. 1998.
Raitanen, Nina. Private Finance for Main Road 4. Ministry of Transport and Communications, Finland. 1997.
Reinhardt, William G. "San Joaquin Toll Road Funding Builds A Strong Base For New, Urban Infrastructure In The U.S." Public Works Financing, 61. 1993a.
Reinhardt, William G. "Kiewit Moves SR 91 Financing To Closure Launching A New Era In U.S. Toll Roads." Public Works Financing, 65. 1993b.
Reinhardt, William G. "First U.S. Private Startup Toll Road Is Financed And Under Construction." Public Works Financing, 70. 1994.
Standard & Poor's Corporation. Public Finance Criteria. McGraw Hill, New York, NY. 1997.
United Kingdom Highways Agency. DBFO - Value in Roads: A Case Study on the First Eight DBFO Road Contracts and Their Development. 1997.