Estimates of future annual revenue levels are of interest to three types of participants in most P3 projects: equity investors, debt providers, and government sponsoring and oversight agencies. Government agencies may consider debt guarantees; subsidies via availability payments, grants, or other means; procurement methods; pure revenue issues; and a wide range of combinations of the foregoing. All require reliable estimates of future revenues.
Equity investors expect a substantial internal rate of return on their investment. Consequently, they are often willing to accept more risk than others, and P50 forecasts. The business model of equity investors is also relevant. Some accept a high level of risk but look to sell their equity share and position as rapidly as possible. Thus, their risk exposure is limited by the time interval between financial close and sale of their shares. Other equity investors are in "for the long haul" and are often more reliable P3 partners as a result.
Project debt may range from long-term bonds to short-term bank notes and loans, or bridge financing. Lenders expect interest rates/payments that they will receive to be only a few points above treasury bonds or other indices. They thus insist on prudent and conservative T&R forecasts in the P80 to P90 range to minimize their risk exposure. Thus, "risk-reward" ratios for equity and debt providers are quite different.
A government agency may have one or more of several roles and responsibilities in a P3 project, with each requiring evaluation of project revenues and its sources and uses. An agency must first represent the public and ensure that a project, its method of implementation, and its use of public funds and other resources is in the public interest and consistent with prevailing laws and regulations. This is an oversight role common to State DOTs.
A second government agency role can be that of project sponsor. This requires that an agency be completely familiar with all aspects of a project, including projected revenues and expenditures. This could be to advance a P3 project deemed in the public interest without accepting any financial obligations; that is, the P3 concessionaire takes full T&R, construction, and all other risks. Or the government agency could provide various financial and other incentives so that potential P3 bidders would be more likely to respond. Or the government agency could subsidize the project completely, either with or without tolls, and have tolls, if levied, flow directly to the government.
If the last alternative is a series of payments made to the concessionaire it is often termed "availability payments," whereby the government is paying for the "availability" of the facility. Under this situation, the government solicits bids from prospective DBFOM concessionaires, and the qualified bidder requesting the lowest series of availability payments is selected, all else being equal. Thus, all T&R risk is taken by the government, with the toll revenue stream (if levied) covering a portion of the availability payments. Toll revenue could conceivably exceed availability payments, particularly in the later years of a P3 contract if the government escalates toll rates and traffic growth occurs.
Revenue sources for a toll facility typically include toll receipts from traditional toll or managed lanes, service area/concession rentals, special vehicular permitting, and leasing of right-of-way for fiber optic cables or other utilities. All sources other than tolls produce minimal income with respect to total facility revenues, and the overwhelming predominance of toll revenues is clear. This section will hence focus on that revenue source.
The commonly used logic and model consists of four steps: trip generation, trip distribution, modal split, and trip assignment. There are many computer-based model programs, and in the US the models are typically developed by or under the auspices of metropolitan planning organizations (MPOs). These are Federally mandated and used primarily for long-range transportation planning and air quality conformance purposes. These purposes do not produce the ideal model for project evaluation needs, so a regional model must often be modified or re-focused, possibly re-calibrated, and validated for a specific study area within the MPO region.
Model inputs include traffic analysis zone definitions, land use, transportation network, and observed travel characteristics. Zones are relatively small, often based on census tracts, homogeneous land uses, major traffic generators, and understandable geographic boundaries. Land use for each zone reflects permitting zoning/uses and other measures such as population, employment, and floor space by typical categories.
The transportation network consists of a region's highways and roads and public transportation systems. These are represented by a coded link-node network or diagram. The network includes HOT lanes and bus lanes, truck routes, "one way" or restricted access roadways, and (most importantly) toll amounts and locations. Each link is described by length, its travel time to volume/capacity relationship, and free flow average speeds. Zonal data are assumed at a zone's centroid, which is connected to the network with a connector link having specified link characteristics.
Observed travel characteristics include zones of trip origins and destinations; proportions of trips being made for various trip purposes (work, business, social, shopping, school, etc.); traffic volumes; trip-making frequencies; vehicle occupancies; and hourly, daily, and seasonal variations as well as transit usage. Origin-destination information is crucial, and there are two types of surveys utilized. One is "revealed preference," whereby data on actual travel performance in the field are compiled. The other is "stated preference," which ask a statistically valid sample of potential network users what changes in their current travel preferences might occur if a new (carefully described) facility were available which had both trip time savings and toll costs associated with its use. Stated preference surveys typically ask for responses to a series of toll costs and time savings and employ "trade-off" analytic techniques to calculate a representative value of time for subsequent toll diversion analyses.
Outputs of the modeling process are link assignments/traffic volumes for future analysis years and future travel times, speeds, and generalized traffic conditions. These outputs can then be used to calculate fuel consumption and air pollutants in addition to revenues on a toll facility.
There are many other model and toll revenue estimation factors that must be considered. Regional models are developed to identity future traffic and network capacity needs, so that making data and analysis judgments that produce "upside" values is "conservative." For toll estimation purposes, "conservative" means making judgments that produce reasonable (and generally lower) traffic and toll revenue estimates that will make it most likely that revenue and debt service target levels will be met. These two forecasting perspectives may be addressed and reconciled, often on a "case" basis. (Probability analysis methods that accomplish this will be discussed later in this appendix.)
Related to the above, regional models often focus on peak hour conditions, a common "driver" to develop future improvement needs. Toll road forecasts require annual estimates of demand and toll revenues. Thus, factors to substantiate shoulder, off-peak, daily, weekly, monthly/seasonal, and ultimately annual traffic and revenue must be developed.
Truck and commercial traffic can be a major revenue component of a toll road, and many regional models only handle them as a proportion of overall link volumes. If truck volumes and revenues are significant, particularly if truck peaking characteristics do not resemble those for autos, then a separate truck model may be created and used.
The four-step model is based on zonal averages, can produce reasonable replicas of future conditions, and is relatively low cost and technically manageable. As a result, it is widely used. However, a more precise micro-analytic modeling approach which is based on individuals and households is used in some large/populous regions. This approach better addresses travel behavior variations and schedules through the use of activity-based models, and activities are modeled rather than only trips.
Network micro-simulation models are sometimes used to simulate the dynamics of corridor and network traffic flows. These models represent flows as a series of individual vehicles and track each one at a resolution of one second or less. Micro-simulation models are particularly useful in the forecasting of managed lane usage since they can simulate the dynamics of individual lanes and the shifts of drivers between lanes and lane categories.
The professionalism, accuracy, and credibility of traffic and revenue forecasts, and the reports presenting them, are always subject to review. A senior-level peer review, internal and/or external, is therefore necessary. An internal review concurrent with the analyses and report preparation can be very effective (i.e., quality assurance and quality control). An external peer review by an independent third party can greatly improve its credibility with potential investors, lenders, government officials with oversight and approval responsibilities, and others. To improve the credibility of the reviewer, his or her background, contractual charge, timeframe, and budget/cost may be revealed.
The first task of a reviewer is to examine the forecast purpose, study level, and client objectives. "Purpose" may reflect the position of the client and/or a designated government agency, lender, or investor and their preferences with respect to forecast conservatism or lesser prudence (producing an upside forecast). Study levels are typically termed I, II, or III, with Level I being conceptual and based on available information. Level II requires current and comprehensive survey data and a full analysis, while Level III is investment grade with the toll plan and other pertinent factors and assumptions detailed with full support, necessary commitments from others when appropriate, and complete documentation. Client interests and positions may also be summarized.
As noted above, the regional transportation model is the primary forecasting tool and is typically carefully assessed. This assessment considers experience with the model program; regional coverage; zone sizes in the study area itself; network detail, particularly in the study area; network improvements assumed for the forecast years; bases, commitments, and/or funding likelihood for these improvements; future traffic conditions and congested links focusing on the study area and competing/feeder routes; analysis time periods; the extent and currency of traffic data; and socioeconomic data for the region and its zonal breakdowns. How the selected economic information and its projections compare to other projections of public agencies and private organizations is also of importance. The calibration and separate validation of the model, or its subarea model derivative, in the study area may also be requested and examined.
Major traffic generators and special project attributes are typically identified and assessed, including managed lanes, unusual design features, and Intelligent Transportation System measures. Sensitivity analyses of major model inputs are crucial (as noted below), as is the quality and completeness of the report document, particularly if it is investment grade.
Managed lanes (or HOT lanes) fundamentally differ from a traditional toll road because the T&R of the managed lanes depends greatly on traffic volumes and conditions (travel speeds, level of congestion) in the adjacent general purpose lanes. These can make managed lane T&R forecasts more volatile than those for traditional toll lanes; that is, a small change in a key input such as study area population growth will typically produce an appreciably larger change in managed lane traffic volumes than in the general purpose lanes. Whether toll rates for managed lanes are set to optimize revenue, traffic, speed, or some combination is also of interest.
A key step in both risk assessments and probability analyses is to identify possibly significant model input variables and perform sensitivity tests on each. These tests show the relationship of changes in these variables to changes in the "bottom line," annual toll revenues. Initially, judgments regarding revenue effects of changes in these input variables can provide a good perspective on the robustness of the forecast and likely or possible variations in the project's revenue stream.
A better way to determine the quantitative probability of revenue forecasts being achieved or exceeded is called Quantified Probability Analysis (QPA). It is a practical approach permitting both debt and equity participants in a deal, and developers, government agencies, and other interests, to be satisfied that their (possibly differing) revenue forecast needs are properly addressed.
Historically, financial analysts have been giving forecasts completely arbitrary reductions, sometimes considering the results of sensitivity tests of major input variables such as population growth and competitive facilities. But sensitivity tests are not intended to address the probability of a particular input variance actually occurring.
The QPA requires both expert transportation planning and statistical capabilities. It first recognizes and quantifies inherent uncertainties in the modeling and forecasting process. Specific steps are:
Typically the calculated values can be for a 90 percent probability of attainment, or investment grade; a 50 percent probability, often considered to be a "central" or "most likely" case; or other values requested by project participants for their risk assessments and financial models. Traditional reductions can be omitted or based on rigorous analysis.
Numerous input variables for the modeling process need to be screened and selectively analyzed, including:
The methodology relies on sensitivities of project revenues to variations in each of the significant inputs. The preferred source for this is successive model runs which will permit the sensitivity of traffic to individual input variations to be ascertained. These model runs may also be carried out for scenarios combining variables in reasonable combinations, running each scenario for future years of interest. In some cases there may be inputs whose values are highly uncertain; in this case, special methods borrowed from the field of statistical design can be usefully employed.
The development of the combined probability distribution function can use a surface response model, Monte Carlo series, or similar analysis tool to develop and present a combined probability function for possible future year revenue outcomes. The resulting matrix of overall annual revenue levels versus estimated probabilities of attainment or exceedance for selected years is the information of interest to project participants. Financial analysts of debt or equity providers will now know the probability of various levels of future revenues occurring and will be able to compare the project's financial performance (IRR, debt coverage, etc.) with the probability of achieving that performance. "Appetites for risk" will be clear, and decision-making will be improved.
The following are reasons that some traffic and revenue forecasts and reports do not meet their objectives, user purposes, and professional standards.
Extensive data needs and inputs, if not properly conceived, monitored, and checked can lead to problems in quality and credibility of data as well as cost problems. Calibration and validation tasks can be time-consuming. Managed lane forecasts can be particularly "fragile" due to general purpose lane "leveraging" of managed lane traffic demand, either plus or minus.
Well-defined study objectives and scopes of work, budgets, and schedules are essential if misunderstandings and misdirections of effort are to be avoided. Contractual flexibility to modify scopes and terms if necessary is most desirable, and often overlooked.
Pleasing the client (and improving the chances for additional assignments) and making favorable judgments regarding individual inputs may have a compounding effect on study outputs; normal economic dips can be overlooked, historical data may not be indicative of the future (hindcast vs. forecast), and "normal" optimism and an urge to help make "it" happen can contribute to poor forecasts.
The accuracy of the forecasting process may vary substantially, and adding complexity does not assure accuracy or precision. The perspectives and risk "appetites" of debt and equity deal participants are different, developer business models may greatly affect their expectations, and risk management and mitigation is often not understood.
Major variables include toll rates and their escalation bases; values of time, demand elasticity, and willingness to pay; population, employment, and economic cycles; land use patterns and intensities; environmental constraints; roadway network characteristics and improvements, including managed lanes; modal splits; and current and future trip purpose mixes. Particularly in an investment grade report, the assessment and proper representation of variables must reflect commitments of cognizant parties, checks by third parties where feasible, and clear statements and sensitivity analyses of variations and effects on future revenues.
Consultants are utilized to bolster client organization staff size, augment client staff capabilities and pertinent experience, and possibly increase the cost-effectiveness of a study effort. Consultant personnel who have worked on projects throughout the nation and often internationally generally have deep and broad experience to draw upon.
The responsiveness, efficiency, and usefulness of consultants depend on contractual provisions and client directions and management as well as those of the consultant. Steps that may be followed with respect to consultant procurement, project execution, and report approval by the client are outlined below.
A key step is for the client to develop study objectives, scope of work, and schedule and budget parameters. This information clearly identifies the type of toll facility, whether it is a brownfield or greenfield project, and whether it includes managed or HOT lanes alongside general use lanes. Other significant improvements, such as additional interchanges or traffic-carrying capacity upgrades, may also be noted. New users of the proposed facility and major benefitting parties and stakeholders may also be identified and outlined. This leads to the selection of an appropriate study level - 1, 2, or 3, as described earlier.
Project objectives, study scopes, and related information may be summarized in an RFQ sent to a broad field of consultants. A two-step selection procedure may be followed. This includes the issuance of RFQs to a large number of consultants, the creation of a short list of three or four clearly qualified firms, and the issuance of an RFP to just those pre-qualified firms.
RFQ responses emphasize consultant qualifications pertinent to the summarized scope and contract terms. RFP responses include complete work plans/proposed scopes that best meet client needs, as well as staff commitments, schedule, and cost. Time and budget can then also be factors in a final consultant selection. The use of this two-step process maximizes the number of consultant firms responding, and also encourages the "best" consultants to enthusiastically prepare in-depth proposals since they know that competition is limited.
Coordination between client and consultant staffs during execution of the work is essential, as are exchanges of information and task interpretations. Unofficial reviews of preliminary draft text can also be encouraged. The review and approval of the final report document by the client will typically consider:
Properly selected and utilized, a consultant can add substantially to the credibility of the project and its acceptance by the financial community, government officials, and the general public.
This Appendix provides an overview of public sector financing for highway projects. It provides a frame of reference for comparing P3 project delivery to current practice in publicly financed project delivery. "Governmental" project financings (i.e., user charge-backed standalone projects financed with non-recourse debt) are an alternative to project financing through P3s. Examples of government-financed projects are:
The Appendix begins with a summary of the municipal bond market, the main source of financing for transportation infrastructure projects among US state and local governments. Credit enhancement is then addressed, including a focus on bond insurance, one of the most common forms of credit enhancement for bond financing. The Appendix then addresses risk transfer on publicly financed projects, including both construction and O&M risk transfer through contracts. The Appendix ends with a case study of a greenfield toll road, the Triangle Expressway in North Carolina.
The US municipal bond market is unique in the world. No other country in the world features such a diverse array of financing instruments. Maximum tenors on US muni bonds are longer than those of many countries' sovereign bonds. It is now a $3.7 trillion market with thousands of individual bond issuers.
The vast majority of muni bonds are tax-exempt, i.e. the interest income earned by investors from the bonds is not subject to federal income tax. It is also often exempt from state and local income taxes, in the jurisdiction of the issuer. These tax exemptions help to lower borrowing costs for state and local governments because investors can accept a rate that is lower than the taxable rate.
As the graph below illustrates, individuals are the largest group of muni bondholders, typically representing 40-50 percent in any given year. Other significant bondholders are mutual funds (30-40 percent), banking institutions (7-13 percent), insurance companies (11-13 percent) and others (3-6 percent). Since most muni bonds are Federally tax-exempt, they pay lower nominal interest rates than like-rated taxable debt. Certain financial market participants such as pension funds and international investors do not benefit from this tax-advantaged treatment: Pension funds are tax-exempt entities and as such derive no value from the tax-exempt status of the interest income, and international investors that do not have US corporate tax liabilities similarly are not attracted to US tax-exempt investments.
Source: Securities Industry and Financial Markets Association (SIFMA)
While muni bonds are the main source of financing for public infrastructure projects in the US, not all muni bonds are issued for new investments. A high percentage of annual issuance is for refunding (refinancing) previous bond issues. The graph below indicates the level of New Capital and Refunding each year for the 2005-2014 period 37. New Capital reached approximately 65 percent of issuance during the pre-recession boom and also in the wake of stimulus efforts post-recession before dropping to less than 40 percent in 2012 and then rising again.
Source: Securities Industry and Financial Markets Association (SIFMA)
Most muni bonds are not in fact backed by state or municipal balance sheets or General Funds. Instead, most are backed by a dedicated revenue stream. These bonds are referred to as revenue bonds instead of General Obligation bonds which are backed by the "full faith and credit" of the issuing authority. As the graph below illustrates, revenue bonds typically represent 60 percent or more of total muni bond issuance in any given year.
Different bond issues feature different yields and spreads depending on the perceived creditworthiness of the issuer. A large state like New York will typically pay lower yields on its bonds than a small city like Boise, Idaho, even if both are AAA-rated. This is likely not due to perceived creditworthiness based on issuer name recognition. Rather, it may be due to the fact that large numbers of New York investors are willing to pay a higher price for NY State bonds because of the substantial value of the interest being exempt from high New York City and State taxes. Similarly, a revenue bond issued by an individual agency in New York, such as the Buffalo Sewer Authority, will typically pay a higher yield than a NY State GO bond issue. The chart below includes data from the Bond Buyer 2015 in Statistics Midyear Review. It shows average yields for different types of bonds for the first half of 2015 38. As can be seen in the chart, the benchmark (A-rated) 30-year revenue bond offers a yield of 4.52 which is a spread of 156 basis points (or 1.56%) above the 30-year US Treasury bond. The second chart below displays yields by credit rating. As can be seen, bonds rated BBB, the lowest rating still considered investment grade, are on average approximately 100 basis points higher than AAA-rated yields. The highest rating typically assigned to greenfield toll roads is BBB given the risks related to construction completion and traffic and revenue projections. Higher yields help to attract investors that seek higher investment returns. However, higher yields represent increased borrowing costs for bond issuers.
When managing their budgets, state and local governments may also want to isolate themselves from certain risks by transferring these risks to other parties. In fact, that is exactly what they do when they allow individual authorities and projects to issue project-backed revenue bonds (as opposed to special tax revenue bonds such as sales tax or fuel tax-backed bonds). Investors buy revenue bonds because they offer higher yields than GO bonds or US Treasury bonds (on a tax-adjusted basis). Those higher yields come with higher risks since revenue bonds are issued on the basis of individual projected revenue streams and are not backed by the balance sheets of state and local governments. If actual revenues fall short of projected revenues, those bonds may default. Although the tax-exempt market is considered second only to Treasury and Federal agency bonds in terms of safety, muni bond defaults occur every year. Investors must analyze the economics of the projects to understand the risks related to the revenue bonds they purchase. Typically, investors rely on rating agencies for this analysis and buy bonds on the basis of their rating.
Toll revenue bonds represent one such case. Toll revenue bonds typically are backed only by future toll revenues. State and local governments typically do not provide guarantees for toll revenue bonds. In this way, transportation agencies transfer revenue risk to bondholders and they do not need to include debt service in their own budgets, even on a contingency basis. If net toll revenues are insufficient to meet debt service requirements, the bonds default. The government is not obliged to prevent the default or to compensate bondholders in the case of default. The map below indicates the extent of tolling in the US by type of operating entity.
Toll Facilities in the United States 39
Another kind of muni bond used in the transportation industry is the Grant Anticipation Revenue Vehicle (GARVEE bond), which can be issued by a state, public authority or a political subdivision. GARVEEs are beneficial because they can increase access to capital markets and allow a state both to expedite construction timelines and distribute the cost of a transportation facility beyond the construction period (generally over a 10-15 year period). Specifically, GARVEEs are a kind of anticipation vehicle, which are securities (debt instruments) issued when funds are expected from a particular source, to advance the upfront funding of a specific need. In transportation finance, expected Federal-aid grants serve as the revenue source for the anticipation vehicles.
Regarding highways, GARVEEs are used as a term for a debt instrument that has future Title 23 Federal-aid funding pledged to it, in the form of Federal reimbursement of debt service and the related financing costs. Hence, a state can receive Federal-aid reimbursements for a number of debt-related costs incurred in connection with an eligible debt financing instrument (such as a bond, certificate, note, lease or mortgage), and the proceeds provide capital for a project that is eligible for Title 23 assistance. When backed by future Federal-aid highway funding, each of the aforementioned instruments is deemed a GARVEE, with bonds as the debt instrument used most often. Section 122 of Title 23 indicates that debt financing instrument-related costs that are eligible for Federal-aid reimbursement include retirement of principal, interest payments and all other costs that are incidental to the sale of an eligible debt issue.
As indicated in the graph below, transportation has represented approximately 13 percent of total muni bond issuance during the past 10 years. Within the transportation sector, as the next graph shows, highway, toll road and street projects along with bridges represent approximately 46 percent of issuance, or approximately $20 billion of issuance each year. The second largest sub-sector is mass transportation which accounts for approximately 28 percent of issuance. Airports account for approximately 21 percent of issuance, seaports for four percent and issuance for other sub-sectors accounts for one percent.
Source: The Bond Buyer ($ in thousands)
Source: The Bond Buyer ($ in thousands)
As with the overall muni market, the preference within the transportation sector is also for revenue bonds. This is indicated by the data in the table below. Also, revenue bond issues are on average much larger than GO bonds issued for transportation purposes. While toll revenues provide the easiest example of transportation revenue bond financing, other types of revenues can also be pledged as repayment for bonds issued to fund transportation improvements. These include revenues from gas tax, vehicle registration fees and others. They also can include increased property (and potentially other) taxes expected to be raised as a result of transportation investments that yield economic development. Some of these financings are raised on the basis of Transportation Improvement Districts, Transportation Development Districts or other Special Purpose Districts. Under such scenarios, a specific geographic region is identified that is expected to benefit from a specific transportation investment such as an upgrade to a road or interchange. Special taxes or fees are assessed on the district, often in the form of additional property taxes. The additional taxes are then leveraged through a bond issue whose proceeds are used to fund the investment.
| H1 Transportation Muni Bond Issuance | ||||||
|---|---|---|---|---|---|---|
| 2015 | 2014 | |||||
| Type | Volume ($mlns) |
#Issues | Avg Size | Volume ($mlns) |
#Issues | Avg Size |
| Revenue | $17,704 | 132 | $134 | $21,446 | 136 | $158 |
| GO | $2,902 | 110 | $26 | $4,868 | 109 | $45 |
| Source: The Bond Buyer | ||||||
Credit enhancement in general is discussed in Section 3.5 of this Guidebook. This section provides additional discussion specific to governmental project financing. The three most cited types of credit enhancement used on municipal bonds are guarantees, bond insurance and letters of credit, as indicated in the graph below. Guarantees may be extended by the federal government under specific programs such as the Department of Energy's Loan Guarantee Program or the Department of Transportation's Maritime Administration Loan Guarantee Program. State and local governments may also extend guarantees to projects directly or through targeted programs. Bond insurance is discussed in detail below. Letters of credit are generally provided by banks to support debt service in case of credit or liquidity challenges.
In addition to these explicit types of credit enhancement, most bonds also feature reserve accounts, such as debt service reserves, and most projects supported by muni bond issues also feature a range of guarantees, warranties and insurance by contractors and their insurers.
Bond insurance (also referred to as a "wrap") is provided by specialized insurance companies ("monolines") that agree to make principal and interest payments in the event that bond issuers cannot. As shown in the graph below, bond insurance reached a high of 57 percent of the face amount of all new issues before the 2008 financial crisis and subsequent recession. It currently stands at less than 10 percent. Ambac which pioneered muni bond insurance in the 1970s filed for bankruptcy in 2010 along with Financial Guaranty Insurance Co. Assured Guaranty emerged as the market leader. A new insurer, Build America Mutual, was launched in 2012. Whereas the insurers carried AAA ratings pre-crisis, the market leader now carries only a AA rating.
In April of 2014, an A-rated issuer would save approximately 46 basis points on a 10-year bond by purchasing bond insurance from a AA-rated bond insurer. Premiums for muni bond insurance, which are usually paid upfront, range from 25 bps to more than 100 bps of total principal and interest depending on the type of credit, and the corresponding yield cost is roughly estimated at one fifth of the premium. Average premiums for an A-rated GO, for example, would be in the range of 25 bps, corresponding to approximately 10 bps of yield. It would save approximately 50 basis points on a 20-year bond. This compares to savings of 160 basis points on a 10-year bond in 2009. Proponents argue that as interest rates rise, bond insurance will offer greater value and become more popular 40.
Bond insurance reached a post-recession high of 6.45 percent of total issuance in the first half of 2015. The market leader was Assured Guaranty with 61.4 percent of the insured bond issuance market, including insurance provided by its subsidiary Municipal Assurance Corp. Build America Mutual reached a market share of 37 percent and National Public Finance Guarantee, a muni bond insurance subsidiary of the Municipal Bond Insurance Association reached 1.6 percent of insured new issuance 41.
Source: Bond Buyer
State DOTs and other transportation agencies may transfer project risks on publicly financed projects through innovative construction contracting. Typically, design, schedule and cost risks are the most often transferred in this manner. As of 2013, 45 state DOTs have been authorized to use design-build contracting, and some states have implemented Construction Manager/General Contractor (CM/GC) projects. The USDOT has supported the state interest in innovative contracting by preparing sample P3 state enabling legislation and publishing reports and other material, including this guidebook. Note that it is possible that a state may have legislation for one form of innovative contracting, but not for all of the forms. State DOTs can face issues regarding legislation, local finance and market issues, and their own departmental culture. Some innovative contracting methods are described below.
There are several variations on the design-build model, including:
Construction Manager/General Contractor (CM/GC) also known in some states as Construction Manager at Risk (CMR). Under the CM/GC contracting method, the construction manager and owner form a contract for project management and construction that is in line with the construction manager's expertise, construction price goal, and management fees. 42 When the design of the project is being determined, the contractor provides pre-construction services and during the construction functions as the general contractor. The owner is able to receive information from the contractor about possible risk. In some cases the construction manager has the responsibility of finding competitive bids for all of the work packages that are to be subcontracted. The contractor's involvement in the process at an early stage fosters a stronger comprehension of the contract, creation of a Guaranteed Maximum Price (GMP) that can help create a clearer budget, a good relationship among the contractor, owner, and designer, and improved risk allocation. Further, the owner is able to maintain control over the final stage of the design process. Fast-tracking and determining the construction price can be challenging; also, there is possibly less competition for the general contractor. The method might be useful for more significant multi-phase highway projects. Though "CMR" is often used interchangeably with "CM/GC," and there is no difference regarding design issues, CMR can have a self-performance element that differs from those found in some CM/GC deliveries. 43
Indefinite Delivery/Indefinite Quantity (ID/IQ), also known as Task Order contracting or Job Order contracting. Under this arrangement, a contractor will bid on a particular unit of the project where the contract has a specific minimum of work units. 44 Work requirements are clear but there may be labor staffing issues, escalation costs for construction, and larger unit bid prices.
Cost + Time (A + B) bidding. When determining the low bid, both cost and time can be examined. "A" is the dollar amount needed for items in the contract and "B" is the number of days needed to complete the project. This model can help minimize construction time.
Best-Value Contracting A + B + C (also referred to as Multiple-Parameter Bidding or source selection). This method is similar to A + B bidding, with the addition of the "C" factor, an additional criterion such as the existence of a warranty. 45
Alternative Technical Concepts (ATCs). This highway contracting procurement method can be used in combination with other delivery methods, like design-build. It features the solicitation of innovative design concepts. 46
Lane Rental. Under this method, contractors are charged a rental fee for closing lanes while they work, which provides an incentive for timely work, with reduced traffic disruption.
State experience with these innovative contracting methods includes the following:
State DOTs and other agencies transfer risk to the private sector through operations and maintenance (O&M) contracts. These contracts enable public authorities to transfer management and cost risk to the private sector. Through the implementation of asset management plans, these contracts also enable public authorities to transfer lifecycle cost risks to the private sector.
State leaders in performance-based maintenance contracting include Virginia, Texas and Florida. Virginia's 1995 Public and Private Transportation Act authorized not only DBFOM P3s but also maintenance contracts. One of the first unsolicited proposals the state received was for maintenance of 250 miles of highway including parts of I-95, I-81, I-77 and I-381. The five-year, $131.6 million contract was awarded in December 1996. The contractor received much higher grades than publicly maintained portions of roadway. A Virginia Tech study found that the contract saved the state between $6.5 and $22 million depending on the assumptions used. The contract was renewed after its first term.
In 1998, Washington, DC, awarded a five-year, $69 million performance-based maintenance contract for 75 miles of national highway within the District of Columbia. The contract featured incentive and disincentive payments for the contractor. After the first year of the contract, ratings for maintenance rose from the high 20s to the low 80s on a scale of 100.
In 1999, Texas DOT entered into two performance-based maintenance contracts, one for 120 miles of I-35 and another for 60 miles of I-20. The contract included required standards for every type of O&M activity on these two sections of highway. 52
Source: Hyman 2009
Florida began performance-based maintenance contracting (which it refers to as "asset maintenance") in 2000. Between 2000 and 2005, FDOT entered into 22 such contracts with a total value of $672 million. It currently has more than $1.2 billion worth of active O&M contracts. The average contract size is $7 million. Most contracts are for seven years while some are for five and others 10. Contractors include Transfield, DBI, ICA, Jorgensen, TME and Louis Berger. There are four main contract types:
FDOT has claimed many benefits from the contracts, including: long-term fixed pricing; cost savings; risk transfer; streamlined administration; and performance results 53. FDOT has developed a Maintenance Rating Program (MRP) to manage the contracts, which include the following activities:
The Triangle Expressway is an example of a project developed, financed and operated by a governmental entity - the North Carolina Turnpike Authority. The project is North Carolina's first modern toll road. It is approximately 20 miles long, extending from the interchange of I-47 and NC 147 to the NC 55 Bypass near Holly Springs, NC. It completes a portion of the outer loop of Raleigh and provides a six-lane, controlled access alternative to the parallel NC 55. It improves access to Research Triangle Park and other regional employment centers.
The project was delivered through two main fixed-fee, lump-sum, design-build contracts, one awarded to S.T. Wooten Corporation for $137 million and the other to Raleigh-Durham Roadbuilders (Archer Wester Contractors, Ltd and Granite Construction Corp) for $447 million, both in August 2008. Both contractors provided payment and performance bonds equal to their respective contract amounts. The contracts included price adjustments for asphalt, cement and diesel fuel. The combined bonus for on-time completion of both contracts was $4.5 million. The contracts also featured liquidated damages for a combined amount of $35,000 a day for delays in substantial completion and $15,000 a day for delays in final completion.
Roadside tolling equipment and back office systems were procured for $14.2 million and $8 million respectively from Affiliated Computer Services (ACS). URS Corp (now AECOM) was contracted for operations, including 18 months of development and five years of management. Transcore won the $5.9 million contract for transponders 55. NCDOT has agreed to fund O&M on the project in the event toll revenues are insufficient.
The Triangle Expressway was financed with three tranches of toll-backed debt (two series of senior tax exempt Series 2009 revenue bonds and a TIFIA loan) along with a series of state-appropriations-backed Build America Bonds. The Series 2009 bonds include both current interest bonds (a bond whose interest payments are provided semi-annually to the bondholders) and capital appreciation bonds (a municipal security whose investment return on a starting principal amount is compounded at a specified rate until the bond matures). The principal amount of the current interest bonds is $235 million and is scheduled to be paid from 2019 to 2039. The capital appreciation bonds have an initial amount of $35 million and mature (together with compounded interest) in the amount of $206 million from 2030 to 2038. The Series 2009 bonds were insured by Assured Guaranty. This increased their ratings by Moody's, S&P and Fitch at time of issuance from Baa3, BBB- and BBB- respectively to Aa2, AAA and AA respectively. The Series 2009 bonds debt service on the $387 million TIFIA loan was scheduled to begin in 2015 and to end in 2043.
State appropriations in the amount of $25 million per year from the State Highway Trust Fund have been pledged to the project over a 40-year period. They are being used to support the toll revenue bonds during the ramp-up phase and during the final five years of the TIFIA loan. In addition, state appropriations are paying the debt service on $352 million of state appropriation revenue bonds issued under a separate trust agreement by the Authority
Toll revenue was expected to increase from zero in 2011 to nearly $170 million in 2049. It is now exceeding forecasts. For 2014, the projected toll revenue was $17.6 million while the actual was $19.7 million. The Expressway is credited with fueling economic development in the region and the DOT is moving forward with new interchanges and other enhancements 56.
Unusually for project financings, debt service payments rank higher in priority claim on annual cash flows than annual operations and maintenance expenses. That is because NCDOT has agreed to pay O & M costs from the State Highway Trust Fund should toll revenues be insufficient after meeting annual debt service requirements. This represents a form of governmental credit enhancement described in section 3.3 of this guidebook.
| Sources and Uses for Triangle Expressway | |
|---|---|
| SOURCES | US $ millions |
| State Appropriation Revenue Bonds | $352 |
| Series 2009 Bonds | $269 |
| TIFIA Loan | $387 |
| Annual Appropriations for FY09 | $25 |
| Total Sources | $1,032 |
| USES | |
| Construction and Acquisition Costs | $878 |
| Senior Lien Debt Service Reserve Fund | $27 |
| State Appropriation Revenue Bonds Reserve Fund | $8 |
| Capitalized Interest 57 on Series 2009 Bonds | $49 |
| Capitalized Interest on State Appropriation Revenue Bonds | $42 |
| TIFIA Loan Origination Charge | $11 |
| Cost of Issuance | $18 |
| Total Uses | $1,032 |
2015 In Statistics: Midyear Review. The Bond Buyer. August 17, 2015. (See https://www.bondbuyer.com/research-report/2015-midyear-in-statistics). Accessed November 2015.
Ashuri, B., Mostaan, K. and Hannon D. of Georgia Institute of Technology (under contract with GDOT). Georgia DOT Research Project RP 11-21, Final Report: How Can Innovative Project Delivery Systems Improve the Overall Efficiency of GDOT in Transportation Project Delivery? Georgia Department of Transportation Office of Research. April 2013 (See http://www.dot.ga.gov/BuildSmart/research/Documents/11-21.pdf). Accessed October 2015.
Bonello, M. and Renick, O. Bond Insurance Then & Now: The Revival of an Industry. The Bond Buyer. April 30, 2014. New York. (See https://www.bondbuyer.com/news/bond-insurance-then-now-the-revival-of-an-industry). Accessed November 2015.
CDOT. Alternative Project Delivery Methods (Approved by FHWA). Colorado Department of Transportation. (See https://www.codot.gov/business/designsupport/adp-db-cmgc/icac/DEFINITIONS.pdf). Accessed October 2015.
FDOT. Asset Maintenance, Office of Maintenance. Florida Department of Transportation (See https://www.fdot.gov/maintenance/asset.shtm). Accessed November 2015.
FHWA. Construction Manager/General Contractor (CM/GC). Federal Highway Administration. (See https://www.fhwa.dot.gov/construction/contracts/acm/cmgc.cfm). Accessed November 2015.
FHWA. Construction Manager/General Contractor (CM/GC). Federal Highway Administration. (See https://www.fhwa.dot.gov/construction/cqit/cm.cfm). Accessed December 2015.
FHWA. Managing Asset Risks at Multiple Levels in a Transportation Agency, Report 2 in Risk-Based Asset Management: Examining Risk-based Approaches to Transportation Asset Management. Washington, DC: US Department of Transportation Federal Highway Administration. August 2012. (See https://www.fhwa.dot.gov/asset/pubs/hif12050.pdf). Accessed November 2015.
FHWA. Alternative Technical Concepts: What is the Alternative Technical Concepts (ATCs) Method? (See https://www.fhwa.dot.gov/construction/contracts/acm/atc.cfm). Federal Highway Administration. Accessed October 2015.
FHWA. Project Delivery Defined, Design Build Finance. (See https://www.fhwa.dot.gov/ipd/alternative_project_delivery/defined/new_build_facilities/dbf.aspx). Federal Highway Administration. Accessed November 2015.
FHWA. Tools & Programs, Federal Debt Financing Tools. (See https://www.fhwa.dot.gov/ipd/finance/tools_programs/federal_debt_financing/garvees/). Federal Highway Administration. Accessed November 2015.
Hawkins, N. and Smadi, O. Use of Transportation Asset Management Principles in State Highway Agencies: A Synthesis of Highway Practice, Synthesis 439 of the National Cooperative Highway Research Program. Washington, DC: Transportation Research Board of the National Academies. 2013. (See http://www.trb.org/Publications/Blurbs/168489.aspx). Accessed November 2015.
Hyman, W. Performance-Based Contracting for Maintenance: A Synthesis of Highway Practice, Synthesis 389 of the National Cooperative Highway Research Program. Washington, DC: Transportation Research Board of the National Academies. 2009. (See http://www.trb.org/Publications/Blurbs/161949.aspx). Accessed November 2015.
MDOT. Youngs, C. Presentation on innovative contracting. Michigan Department of Transportation & Innovative Contracting. March 26, 2012. (See https://www.michigan.gov/documents/mdot/Design_Build_Chris_Youngs_381837_7.pdf). Accessed November 2015.
Minchin, E. Ptschelinzewv L., Migliaccio, G., Gatti U., Atkins, K., Warne, T., Hostetler, G., Asiamah, S. Transportation Research Board. NCHRP Report 787: Guide for Design Management on Design-Build and Construction Manager/General Contractor Projects. 2014. (See http://www.trb.org/Publications/Blurbs/171479.aspx) Accessed October 2015.
Municipal Securities Rulemaking Board (MSRB) Electronic Municipal Market Access (EMMA) database. (See https://emma.msrb.org/). Accessed October 2015.
New York State Thruway Authority (See http://www.thruway.ny.gov/index.shtml). Accessed November 2015.
Parker, B. Innovative Contracting. Utah Department of Transportation. July 10, 2013. (See https://blog.udot.utah.gov/2013/07/innovative-contracting/). Accessed November 2015.
Samuel, P. North Carolina Pike Say Got Good Prices for All-Electronic Toll Systems - Nearly 20% Off. Toll Roads News. Pine Street Publications LLC. December 19, 2009. (See http://tollroadsnews.com/news/north-carolina-pike-say-got-good-prices-for-all-electronic-toll-systems---nearly-20-off). Accessed November 2015.
Scott, S. and Klei, H. (Trainer Consulting Services, Inc.) and Ferragut, T. (TDC Partners, LTD) for National Cooperative Highway Research Program, Transportation Research Board, National Research Council, requested by American Association of State Highway and Transportation Officials (AASHTO). Innovative Contracting for Major Transportation Projects. January, 2006. (See http://onlinepubs.trb.org/onlinepubs/archive/NotesDocs/20-24(43)_FR.pdf). Accessed October 2015.
Siceloff, B. Two New Interchanges Will Bring More Paying Customers to Triangle Expressway. The News & Observer. Raleigh. July 10, 2015. (See https://www.newsobserver.com/news/local/counties/wake-county/article27000928.html). Accessed November 2015.
Tappan Zee Constructors, LLC. (See www.tappanzeeconstructors.com). Accessed November 2015.
The New NY Bridge. (See www.newnybridge.com). Accessed November 2015.
Workman, S. and Rall, J. Toll Facilites in the United States. National Conference of State Legislatures. February 1, 2013. (See http://www.ncsl.org/research/transportation/toll-facilities-in-the-united-states.aspx#Map_1_Toll_Facilities). Accessed November 2015.
Workman, S. and Rall, J. Toll Facilities in the United States. National Conference of State Legislatures. February 1, 2013. (See http://www.ncsl.org/research/transportation/toll-facilities-in-the-united-states.aspx#Overview). Accessed November 2015.
WVDOT. Policies and Procedures Concerning Design-Build Projects. West Virginia Department of Transportation. (See http://transportation.wv.gov/highways/engineering/designbuild/Pages/default.aspx). Accessed October 2015.
WVDOT. Warranty Contracting. West Virginia Department of Transportation. (See http://transportation.wv.gov/highways/contractadmin/ipd/Pages/Warranty-Contracting.aspx). Accessed October 2015.
| Term | Description |
|---|---|
| Bidder | A respondent to a Request for Expressions of Interest or an invitation to submit a bid in response to a project brief. Typically, a bidder will be a consortium of parties, each responsible for a specific element, such as constructing the infrastructure, supplying the equipment, or operating the business. The government normally contracts with only one lead party (bidder), who is responsible for the provision of all contracted services on behalf of the consortium. |
| Brownfield | Projects that focus on improving, operating, and/or maintaining an existing asset (contrast to greenfield). P3 brownfield projects in transportation typically are long-term operation and maintenance contracts or lease concessions. Blended greenfield-brownfield projects also exist, for example, improving an existing asset by adding new capacity (e.g., more lanes). |
| Concession Period | Total construction and operating periods. |
| Concessionaire | Private entity that assumes ownership and/or operations of a given public asset (i.e., train station, bus operation) under the terms of a contract with the public sector. |
| Contingency | An allowance included in the estimated cost of a project to cover unforeseen circumstances. |
| CPI | Consumer Price Index. |
| DB | Design-build: Under a DB, the private sector delivers the design and construction (build) of a project to the public sector. The public sector maintains ownership and operations and maintenance of the asset. Build refers to constructing the road, which includes reviewing conditions at the building site, providing construction staff and materials, selecting equipment, and when necessary, amending the design to address problems discovered during the construction phase. |
| DBFOM | Design-build-finance-operate-maintain. Under DBFOM, the private sector delivers the design and construction (build) of a project to the public sector. It also obtains project financing and assumes operations and maintenance of an asset upon its completion. |
| Debt Maturity | Maturity date for project bond. |
| Discount Rate | Percentage by which a cash flow element in the future (i.e., project costs and revenues) is reduced for each year that cash flow is expected to occur. |
| Discount Rate, Nominal | Discount rate that factors in the inflation rate. |
| Discount Rate, Real | Discount rate that does not account for inflation. |
| DSCR | Debt service coverage ratio. |
| Finance | Phase or delivery aspect of the project that includes providing capital for the project, which may include issuing debt or equity and verifying the feasibility of plans for repaying debt or providing returns on investment. |
| Greenfield | Projects that focus on developing and/or building a new asset (contrast with brownfield). Many P3 structures are available for greenfield projects, including design-build, design-build-operate-maintain (DBOM), design-build-finance-operate-maintain/manage (DBFOM), and others. Blended greenfield-brownfield projects also exist. |
| Inflation Consumer Price Index | Used as a base rate for inflation assumptions. |
| IPD | The Office of Innovative Program Delivery (IPD), a part of the Federal Highway Administration, provides tools and expertise in use of different public-private partnership (P3) approaches. |
| Leveraging | Degree to which an investor or business is utilizing borrowed money. |
| Maintenance | This phase includes keeping the project in a state of good repair, which includes filling potholes, repaving or rebuilding roadways, and ensuring the integrity of bridges and highways. |
| Net Present Value (NPV) | Present value of the expected future revenues minus the net present cost. |
| Private Activity Bond | New type of financing that provides private developers and operators with access to the tax-exempt bond market, lowering the cost of capital significantly. |
| Public Sector Comparator (PSC) | Represents the most efficient public procurement cost (including all capital and operating costs and share of overheads) after adjustments for competitive neutrality, retained risk, and transferrable risk to achieve the required service delivery outcomes. This benchmark is used as the baseline for assessing the potential value for money of private party bids in projects. |
| Ramp up | Period after opening of a new toll facility during which traffic steadily increases as users become more familiar with the facility |
| Retained Risk | The value of those risks or parts of a risk that government proposes to bear itself under a partnership arrangement. |
| Revenue Leakage | Assumed annual revenue losses for a tolling facility. |
| RFP | Request for proposal. |
| ROW | Right of way |
| Risk Allocation | The process of assigning operational and financial responsibility for specific risks to parties involved in the provision of services under P3. Also see risk transfer. |
| Risk Transfer | The process of moving the responsibility for the financial consequences of a risk from the public to the private sector. |
| Routine Maintenance | Work that is planned and performed on a routine basis to maintain and preserve the condition of the highway system, or to respond to specific conditions and events, that restore the highway system to an adequate level of service. |
| Technical Risk | Risks arising from deviations from the project's original technical assumptions, specifications, or requirements. |
| T&R | Traffic and revenue. |
| Transportation Infrastructure Finance and Innovation Act (TIFIA) | This program provides Federal credit assistance in the form of direct loans, loan guarantees, and standby lines of credit to finance surface transportation projects of national and regional significance (FHWA, 2013). |
| Transferrable Risk | Any risk that is transferrable to the bidder. |
| Value for Money (VfM) | The procurement of a P3 project represents VfM when - relative to a public sector procurement option - it delivers the optimum combination of net life-cycle costs and quality that will meet the objectives of the project. |
37 SIFMA includes bond issues that are a mix of new capital and refunding in its refunding category so their data understates new capital issuance to some extent.
38 The 20-Bond Index consists of 20 general obligation bonds that mature in 20 years. The average rating of the 20 bonds is roughly equivalent to Moody's Investors Service's Aa2 rating and Standard & Poor's Corp.'s AA. The 11-Bond Index uses a select group of 11 bonds in the 20-Bond Index. The average rating of the 11 bonds is roughly equivalent to Moody's Aa1 and S&P's AA-plus. The Revenue Bond Index consists of 25 various revenue bonds that mature in 30 years. The average rating is roughly equivalent to Moody's A1 and S&P's A-plus. The indexes represent theoretical yields rather than actual price or yield quotations. Municipal bond traders are asked to estimate what a current-coupon bond for each issuer in the indexes would yield if the bond was sold at par value. The indexes are simple averages of the average estimated yields of the bonds.
The 10-year and 30-year Treasury yields are the market quotes for those securities at the time that the indexes are calculated. (Excerpt from https://www.bondbuyer.com/broker/bond-buyer-data?details=true)
39 http://www.ncsl.org/research/transportation/toll-facilities-in-the-united-states.aspx#Map_1_Toll_Facilities
40 Bonello, Maria and Oliver Renick. April 30, 2014. "Bond Insurance Then & Now: The Revival of an Industry" in The Bond Buyer. New York.
41 BondBuyer. 2015. 2015 in Statistics Midyear Review. New York: BondBuyer.
42 https://www.fhwa.dot.gov/construction/contracts/acm/cmgc.cfm
43 http://www.trb.org/Publications/Blurbs/171479.aspx
44 https://www.codot.gov/business/designsupport/adp-db-cmgc/icac/DEFINITIONS.pdf
45 http://onlinepubs.trb.org/onlinepubs/archive/NotesDocs/20-24(43)_FR.pdf
46 https://www.fhwa.dot.gov/construction/contracts/acm/atc.cfm
47 https://www.michigan.gov/documents/mdot/Design_Build_Chris_Youngs_381837_7.pdf
48 http://transportation.wv.gov/highways/contractadmin/ipd/Pages/Warranty-Contracting.aspx
49 http://transportation.wv.gov/highways/engineering/designbuild/Pages/default.aspx
50 https://blog.udot.utah.gov/2013/07/innovative-contracting/
51 http://www.dot.ga.gov/BuildSmart/research/Documents/11-21.pdf
52 Hyman, William. 2009. "Performance-Based Contracting for Maintenance: A Synthesis of Highway Practice." Synthesis 389 of the National Cooperative Highway Research Program. Washington, DC: Transportation Research Board of the National Academies.
53 https://www.fdot.gov/maintenance/asset.shtm
54 Source: information largely from the Official Statement available for download at https://emma.msrb.org/ using CUSIP 65830RAK5.
55 Samuel, Peter. December 19, 2009. "North Carolina Pike Say Got Good Prices for All-Electronic Toll Systems - Nearly 20% Off" in Toll Road News. Pine Street Publications LLC.
56 Siceloff, Bruce. July 10, 2015. "Two New Interchanges Will Bring More Paying Customers to Triangle Expressway" in The News & Observer. Raleigh.
57 Capitalized interest is interest paid during construction; therefore, it needs to be borrowed up front and is in fact added to the principal of the debt. Under generally accepted accounting principles, interest expense during construction is considered a capital expenditure and is included in the project's construction costs.
