This appendix provides detail on revenue risk sharing mechanisms and related P3 programs in Chile, Brazil, and Korea to illustrate the context in which the mechanisms were derived. As shown in Table 6, the appendix also reviews mechanisms in Canada, Spain, and the U.K., and those for transit in various countries. Lastly, this appendix also discusses some mechanisms used in the U.S.
Table 6: Summary of Revenue Risk Sharing and Related Mechanisms By Country
| Country | Present Value of Revenues | Minimum Revenue Guarantee | Contingent Finance Support | Availability Payment & Revenue Risk | Revenue Distribution Mechanism | Internal Rate of Return Balancing2 | Innovative Finance Programs | Shadow Toll | Price Cap | Rate of Return2 |
|---|---|---|---|---|---|---|---|---|---|---|
| Chile | ● | ● | ● | |||||||
| Brazil | ● | ● | ||||||||
| South Korea | ● | ● | ||||||||
| U.K. | ● | ● | ● | ● | ● | |||||
| Canada | ● | |||||||||
| Portugal | ● | |||||||||
| Spain | ● | ● | ● | ● | ||||||
| Transit Cases1 | ● | |||||||||
| U.S. | ● | ● | ● | ● | ||||||
| 1 In Canada, Columbia, Spain, Sweden, and the U.K. 2 In Section 3, these two mechanisms are discussed under the term "Regulated Return Mechanisms." |
||||||||||
Chile serves as a case study to understand the working of three risk sharing mechanisms: the least present value of the revenues (PVR), minimum revenue guarantees (MRGs), and Revenue Distribution Mechanisms (RDM). Its experience with more than twenty highway and infrastructure concessions spans over two decades. 3
Starting in 1992, Chile embarked on an ambitious P3 program for highways and other infrastructure. Many of the highway P3s were improvements of Highway 5, Chile's north-south road. They were implemented by the Chilean Ministry of Public Works (MOP). Table 7 shows the concessions awarded from 1992 to 2004.
| Year of award | Concession | Highway | Investment | Total Investment | Main Economic Tender Variable |
|---|---|---|---|---|---|
| 1992 | Túnel del Melón | Interurban tunnel | $38M | $38M | Mix* |
| 1994 | Camino de la Madera A. Norte Concepción |
Interurban Interurban |
$26M $196M |
$222M | Mix* Tariff |
| 1995 | Santiago S. Antonio Nogales Puchancavı ́ Acceso a AMB Talca Chillán |
Interurban Interurban Suburban Interurban |
$180M $12M $10M $169M |
$371M | Tariff Tariff Tariff Tariff |
| 1996 | Los Vilos Santiago Santiago Los Andes La Serena Los Vilos |
Interurban Interurban Interurban |
$250M $152M $246M |
$648M | Tariff Tariff Tariff |
| 1997 | Chillán Collipulli Temuco Río Bueno Río Bueno Pto. Montt Collipulli Temuco |
Interurban Interurban Interurban Interurban |
$241M $200M $249M $255M |
$945M | Duration Upfront fee Upfront fee Upfront fee |
| 1998 | Santiago Talca Santiago Valparaíso |
Interurban Interurban |
$575M $340M |
$915M | Upfront fee LPVR** |
| 1999 | Costanera Norte | Urban | $384M | $384M | Upfront fee |
| 2000 | Norte-Sur Red Vial Litoral Central |
Urban Interurban |
$442M $104M |
$546M | Upfront fee Subsidy |
| 2001 | Vespucio Sur Vespucio Norte Talcahuano-Penco Variante de Melipilla |
Urban Urban Suburban Suburban |
$280M $240M $19M $19M |
$558M | Upfront fee Upfront fee Subsidy N/A |
| 2002 | Camino Internacional Ruta 60 | Interurban | $180M | $180M | N/A |
| 2003 | Acceso Nororiente a Santiago | Suburban | $165M | $165M | LPVR** |
| 2004 | El Salto-Kennedy | Urban tunnel | $70M | $70M | Upfront fee |
| * Mix: Several economic variables are employed. N/A: Information not available. **LPVR is the same as the term "PVR" used elsewhere in the Discussion Paper. | |||||
Chilean law required that concessions be financed with no more than 70% debt. The intent was to ensure that Developers had enough "skin in the game" so that they were focused on the project throughout the concession.
As was common in early P3 programs, Chile experienced poorly designed competitions and concession contract designs, volatile traffic demand, and political pressures to renegotiate failing projects (Vassallo, 2006). During the 1998-2002 recession, many concessions experienced difficulties with significantly lower than expected revenues. Many Developers sought to renegotiate with MOP. Engel et al estimate that 50 Chilean concessions were renegotiated in total 144 times, or more than three times per concession (Engel et al, 2014). These renegotiations put into question the credibility of the entire P3 competition process.
Chile has been closely associated with development of the PVR mechanism in 1998, although the U.K. first pioneered a similar technique for projects in the Dartford Bridge and the Second Severn crossing projects (Engel et al, 2014, p.67). Under a PVR, the MOP sets the following parameters:
The Developer prepares the construction and O&M cost estimates, secures debt and equity commitments, and submits its bid for the lowest PVR that is acceptable to it. The winning bidder receives toll revenues during the concession. The revenues are discounted each year by the discount rate set by the MOP. The concession ends once the bid PVR amount is reached.
While Chilean law establishes a maximum concession period of 50 years, MOP can terminate PVR contracts after twelve years, allowing flexibility should project circumstances change, including the need for enhancements that the original Developer would not be able to or willing to take on, or if there were any other issues with the Developer (Engel et al, 2014). This is not dissimilar to a "put and call option" suggestion of Quiggin (2005), allowing an Agency or a Developer to terminate a concession contract through a pre-determined payment structure. Some U.S. P3 contracts also have such a mechanism in which an Agency must pay a portion, but not all, of outstanding senior debt if it terminates the P3.
The first highway PVR was Route 68 in Chile (Engel et al, 2014). The project was procured through a PVR process and received a minimum revenue guarantee (MRG), as discussed below. The winning bid came in lower than MOP's estimates, possibly because the MOP had underestimated how much the PVR and MRG mechanisms had reduced project risks.
The advantages of the PVR mechanism for Chilean P3s were as follows:
A disadvantage of PVR was the perception that the Developer would earn a return regardless of the Developer's quality of customer service. They might not adequately manage O&M and/or devote too few resources to it. Demand is inelastic in response to service quality, as can be the case for highways, ports, water facilities, and airport runways (Engel et al, 2014). However, Engel et al argue that this is less an issue in road concessions, where service quality can be easily monitored through contracts with clear performance measures.
Academics have suggested a remedy to this is to have Developers specify key O&M costs, such as annual upkeep costs or major overlays, allowing for these services to be bid separately. This may make negotiating these costs between Developer and Agency easier, in those circumstances when project changes result in much different O&M costs, such as a tunnel or a road in an area subject to greater snow removal (Rus and Nombela, 2000).
PVR contracts were first used in Europe in Portugal's first highway concession, the Litoral Centro Highway. In that P3, the contract would end if one of the following events would occur:
The Litoral Centro Highway utilized the 12 month Euribor rate as the discount rate, providing a natural interest rate hedge. Later Portuguese P3 transactions, however, did not follow a flexible-term concession PVR (Engel et al, 2004, p.118).
Chile used the PVR mechanism in conjunction with other risk mitigation measures, including both direct and contingent subsidies (Engel et al, 2014). Of the 26 Chilean highway concessions between 1992 and 2004:
MOP offered MRG bands as high as 80%-85% of expected revenue with MOP paying the MRG if it fell below that year's guarantee level, for which bidders would pay a guarantee fee of 0.75% of the MRG amount. In one competition, two of four bidders did not want the MRG as the PVR adequately mitigated risk. Furthermore, Developers who opted for the MRG requested them to be weighted to the early concession years, reflecting Lenders liquidity concerns (Gómez-Lobo, A., 2000).
MRG payments were treated as another form of revenue in PVR bid projects and therefore included in the PVR amounts that the Developer received (Engel et al, 2000). This meant that MRG and PVR mechanisms could effectively be combined. PVR was both a bidding criterion as well as a method to equitably share revenue risk. MRG played an important role in providing short-term liquidity essential to cover debt service.
In non-PVR projects with a MRG, revenue sharing was triggered when the return of cumulative revenues reached an internal rate of return (IRR) of 15% (Gómez-Lobo, A. 2000). When that trigger was reached, the MOP would share with the Developer 50% of those revenues that exceeded the IRR trigger band (Vassallo, 2006). This cumulative IRR approach, a revenue "cap," is similar to many U.S. P3 provisions. In contrast, MOP paid the MRG on an annual basis when revenues were below the minimum, a revenue "floor," reflecting Lenders' needs. This is an "asymmetric" arrangement - cumulative for the cap and annual for the floor - but necessary to secure financing (Gómez-Lobo, A. 2000).
Table 8 shows that there were only four attempted Chilean PVR projects out of a total of 26 projects, two of which were successfully awarded (Vassallo, 2006). In the Talcahuano-Penco road (Ruta Interportuaria) concession, a PVR was offered, but bidders sought a subsidy instead.
| Project | Year of Tender | Investment1 | PVR2 | Maximum Term | Number of Bidders3 | Situation |
|---|---|---|---|---|---|---|
| Santiago-Valparaiso | 1998 | $340M | $381M | 25 years | 4 | Successfully awarded, in operation |
| Costanera Norte | 2000 | $384M | - | 30 years | 0 | Not awarded, tendered again a year later under other economic variable |
| Talcahuano-Penco | 2001 | $19M | - | 31.5 years | 2 | Awarded, bidders requested a subsidy instead of PVR |
| Acceso Nororiente | 2003 | $165M | $346M | 40 years | 1 | Successfully awarded, in operation |
| 1 Investment predicted by the government. 2 Present value of the revenues offered by the granted bidder. 3 Bidders in the last stage of the project. |
||||||
Due to the Chilean economic crisis of 1998 to 2002, many Developers sought to renegotiate concessions. To handle these renegotiations, MOP developed "Revenue Distribution Mechanisms (RDM)," which required Developers to make expansions to those projects in financial distress in return for one of three revenue guarantee alternatives, reflecting different annual revenue growth rates, of 4, 4.5, and 5 percent. Essentially, MOP required higher investments in return for guarantees of more revenues. These RDMs had a PVR element to them with the concession finishing early if certain revenue marks were achieved. It appears that most of these RDMs were carried out through bi-lateral negotiations, instead of open competitions, which raised the question whether MOP was receiving market-based bids. The renegotiations and the introduction of the PVR and RDM mechanisms are part of the evolution of the Chilean P3 program towards improved risk sharing arrangements, a theme common to many P3 programs.
With the economy stabilizing and projects maturing, the ownership of the original Chilean P3s has changed. Chilean Developers have sold some or all of their concession interests to pension funds and other institutional investors once projects were completed and exhibited steady cash flows (Engel et al, 2014, p.88). This practice began in 2010 and as recently as July 2015, Hochtief sold a 50% share in the Tunnel San Cristóbal toll highway (Hochtief, 10/4/2012).
Brazil evolved from providing internal rate of return balancing or regulated return mechanisms (RRM), contract extensions, relaxation of investment triggers, beneficial toll rate adjustments, to full MRGs. This evolution recognized the need to improve project financeability in light of overly optimistic T&R forecasts, yet it also increased the fiscal impact on Brazil's budget. 4
Brazilian P3s were authorized under two laws passed in 1995 and 2004: The "Toll Concession Law" (N.o8987/1995) and the "Public-Private Partnership Law" (N.o11.079/2004), respectively. Under the first, "Concessions" are projects that are fully funded by user fees, whereas under the second, "Public-Private Partnerships" are concessions that require a public subsidy (such as a capital grant or a service payment). Two entities under the Ministry of Transportation (MOT) manage P3s. The Brazilian Planning and Logistics Agency (EPL) manages national P3 policy, planning, and knowledge advancement and the Brazilian National Surface Transportation Agency (ANTT) directly manages concessions.
There are currently 28 highway contracts at the federal level, about 15 of which have been signed since 2000. A further 15 projects are to be tendered by 2016. Brazil experienced four distinct waves of P3 development, again demonstrating how P3 approaches evolve.
The first wave of Brazil P3s occurred in the 1980s and 1990s, during which the Brazilian government aggressively promoted P3s by offering heavily subsidized project debt from the National Bank for Economic and Social Development (BNDES). The debt tenor was generous, often matching the length of the concession of 25 to 30 years, similar to the TIFIA program's tenors of up to 35 years after substantial completion. Lenders typically required 30% equity contribution.
User demand during this first wave was very high, so actual revenues often exceeded forecasted revenues. In addition to strong demand, the more limited need for capital investment in these brownfield projects allowed Developers to earn healthy returns.
During this time, revenue risk was mostly assumed by the Developer. Concessions were awarded to the Developer offering the highest upfront fee, assuming a fixed tariff indexed to inflation that was determined by the Developer's business plan, which incorporated the Developers forecasts for traffic, revenue, operations costs, equity returns, and taxes. However, the Developer's demand risk was somewhat mitigated through adjustable tariffs. ANTT would track the project's actual rate of return against the Developer's forecast rate of return and ensure they "balanced" by adjusting the tariff appropriately. Known as the "balancing equation," this mechanism indirectly protected the Developer from revenue changes.
As a result of Brazil's high inflation during the 1980s and 1990s (Stratfor Global Intelligence, 2015), risk free returns were very high, driving up Developer IRRs to as high as 45%. Once macroeconomic conditions stabilized, equity IRRs settled to around 10%. ANTT introduced other mechanisms during this period to ensure performance, such as penalties and incentives for safety, construction delays, accident rates, etc.
During the second wave of P3s, ANTT required higher levels of investment. Because it was imposing higher capital investment requirements on Developers, ANTT decided to assume more demand risk and alleviate the financing risk, making the following changes:
Figure 10: Illustration of Brazilian Investment Trigger Mechanism
View larger version of Figure 10
Following a period of failures experienced during the second wave which were exacerbated by the global financial crisis, ANTT introduced additional modifications to the "balancing equation:"
By 2014, ANTT decided it needed to absorb more demand risk, as toll roads in this wave were not located in areas with as strong demand as those in the previous waves. ANTT introduced formal revenue sharing bands, with 50% of revenues in excess of ANTT's revenue forecasts shared with ANTT and shortfall of revenues below 40% of the forecasts paid as a subsidy. Through this mechanism, ANTT effectively provided a MRG, reducing the revenue risk.
Table 9 shows the evolution of the Brazilian federal P3 program's revenue and financing mitigation measures.
| Risk Management Mechanism | Focus of Mechanism | Wave | Function and Benefits |
|---|---|---|---|
| Balancing equation | Revenue | 1st | Tariff adjusted so that actual revenues matched forecasted revenues. Provided downside protection but no MRG. |
| Investment trigger | Financing | 2nd | Capital investment triggered by demand milestones. Not a revenue guarantee mechanism, but reduced financing risk. |
| Toll rate revisions | Revenue | 2nd | Tariffs could be revised every five years. Not a revenue guarantee mechanism, but did reduce revenue risk. Same as IRR Balancing. |
| Relaxing of investment trigger | Financing | 3rd | Further reduced financing risk by allowing demand to increase beyond original trigger's threshold. |
| Concession extension | Revenue | 3rd | Extended revenue collection period. Has not been used to date. |
| Marginal cash flow mechanism | Revenue | 3rd | Provided a subsidy in the event of a revenue shortfall. |
| Revenue bands | Revenue | 4th | 50% of excess revenues are shared with ANTT, whereas the shortfall below 40% of the forecast are paid as a subsidy. |
As can be seen in the table above, the early measures allowed the Developer to adjust tariffs or defer capital expenditures, addressing revenue and financing risks. The later measures provided subsidies in downside situations through a limited and then more comprehensive MRG.
South Korea's experience best illustrates how a MRG mechanism evolved over time from being generous to one that appears to have a better balance of benefits between Agency and Developer. South Korea also used of innovative financing through government loan guarantees and upfront capital subsidies. 5
Spurred by rapid economic growth in the 1990s, South Korea initiated a number of P3s with the signing of the 1994 PPP Act and subsequent amendments in 1999-2000 that included a MRG (Park 2014). The South Korean government's institutions dedicated to promoting P3s included: 1) PIMAC (South Korea's key P3 agency), 2) the affiliated South Korean Development Institute (KDI), South Korea's leading think tank, and 3) South Korea's Credit Guarantee Fund, established to support P3s in 1994.
Under these acts, most projects were eligible for a 20-30% construction subsidy. The lowest required construction subsidy became the primary basis for award. Concession terms were fixed at 30 years but could be shortened or extended.
The initial MRG mechanism guaranteed 80% to 90% of revenues for the entire operations period. For example, the Soojungsan Tunnel had a 90% guarantee and 110% cap on forecasted revenues, while the Incheon Grand Bridge had an 80% guarantee and a 120% cap (Macquarie 2015).
Project debt was frequently guaranteed by South Korea's Credit Guarantee fund or provided by the government, so that the MRG served effectively as a "double guarantee" for debt, which helped in case the payments from South Korea's Credit Guarantee fund were delayed.
The Seoul Incheon International Airport was procured as a P3 at end of the Asian Financial Crisis in 1999 with the selected Developer bidding an IRR of over 20%. At the time the project was considered a national priority in preparation for the 2002 World Cup. PIMAC has had to pay a large amount of revenue guarantees.
Because of projects such as the airport, PIMAC realized it was over-exposing itself to revenue risk, so in 2003 it changed the scheme to cover just 15 years of operations with its exposure progressively reduced over time: 90% of the revenue was guaranteed for the first five years of operations, followed by 80% for the next five years, and 70% for the final five years, as shown in Table 10.
One of PIMAC's chief concerns with regard to the MRG mechanism was its continued exposure to most revenue risk while higher returns were all but assured to the Developer. Moreover, the scheme incentivized Developers to overestimate future demand, thereby allowing them to make more claims on revenue shortfalls. In fact, according to Kokkaew and Chiara (2011), PIMAC significantly exposed itself to this contingent liability because in most cases, actual revenues were falling well short of the guarantee levels.
In 2006, the revenue guarantee period was reduced from 15 years to 10 years, with 75% of the revenues guaranteed during the first five years and 65% guaranteed during the final five years. The MRG would also now apply only to solicited (but not unsolicited) proposals. The updated MRG scheme addressed PIMAC's concern regarding overly optimistic Developer forecasting by stipulating that the MRG was only valid if actual revenues were greater than 50 percent of the forecasted revenue ("50 percent feasible exercise condition"). The Seoul-Chuncheon Expressway, completed in 2009, followed this model (Macquarie 2015). This unique feature meant that Developers had to be highly confident of their forecasts, since there was no support below 50% of forecasted revenues.
The degree of revenue sharing by PIMAC varied by project, but only occurred when actual revenues exceeded the Developer's forecasted revenues. In the case of OOO Urban Railway PPP project, completed in 2002, PIMAC collected revenues exceeding 120% of forecasted revenue from operations years 0 to 5, and exceeding 130% from years 6 to 10 (Park, 2013).
In 2009 PIMAC abandoned the MRG scheme and replaced it with a "New Risk Sharing Scheme." In addition, since September 2015, PIMAC has been developing a standard concession agreement, financial model, and RFP tailored to this new scheme, which works as follows:
According to PIMAC, the new scheme has the following benefits:
Table 10 shows how MRG levels progressively decreased from 80%-90% of forecasted Developer revenues for the entire concession period to 75%-65% in the first and second five years respectively, and to the latest scheme which was based on PIMAC's forecasts.
| Period | Years 0 to 5 | Years 5 to 10 | Years 10 to 15 | Whole Period | Additional Requirement |
|---|---|---|---|---|---|
| 1999 to 2003 | 80% to 90% | No minimum revenue requirement. | |||
| 2003 to 2006 | 90% | 80% | 70% | N/A | Actual revenues must be at least 50% of forecasted revenues. |
| 2006 to 2009 | 75% | 65% | N/A | N/A | Actual revenues must be at least 50% of forecasted revenues. Only solicited projects eligible. |
| 2009 to present | N.A. (New Risk Sharing Scheme introduced) | Actual revenues must be at least 50% of forecasted revenues. The revenues are forecasted by the Agency based on investment costs and risk free interest rate. | |||
Even with the 50 percent feasible exercise condition, past MRG schemes were subject to public criticism for excessive Developer profits and the disproportionate PIMAC revenue risk transfer, forcing PIMAC to substitute its own forecasts for the Developer's in the latest scheme. It is too early to know if this creates more VfM since new projects are not yet operational. However, competition under this scheme continues to be strong, so it does not appear to be a deterrence.
The P3 programs in Chile, Brazil, and South Korea have evolved over the last two decades as these and other Agencies struggle to find the right revenue risk sharing balance. In order to understand why countries affected by the same revenue risk sharing issues take different approaches, it is important to acknowledge the differences in experience and issues each country encountered. In the case of Chile, frequent renegotiations led the government to conclude that Developers should be better protected from revenue risk. As a result, Developers were offered various mechanisms including PVR and MRG. Similarly, in Brazil the mechanisms offered to Developers provided increasingly more revenue risk protection.
In the case of South Korea, the rationale for changing their approach was different. South Korea started with a very lucrative revenue sharing mechanism, from the Developers' perspective, that resulted in returns perceived as excessive by the general public. In the face of public criticism, the Agency scaled back and reformed its generous MRG. Furthermore, it introduced a clause that would keep Developers from receiving public support if less than 50% of the projected revenues were realized. This unique clause exposed Developers to significant downside revenue risk, which is the opposite of what other countries struggling with revenue risk sharing have done. However, taking into consideration the previous experience with perceived excessive returns, the South Korean approach is understandable.
Programs in other countries have evolved as well, including in Colombia and Mexico. For instance, the short concession terms and low actual traffic in Mexico forced the government to bail out many road concessions.
Extrapolation from the experience of these countries to the U.S. must be done carefully. The general takeaways for the U.S., discussed in Section 3, are that:
This section presents the experience in other countries that have utilized revenue risk sharing mechanisms in toll road and transit financings. Since there is less of a toll road tradition in these countries, these are more "one-off" examples rather than descriptions of evolving revenue risk approaches or programs.
The Autoroute 25 concession in Québec, Canada is a unique example of an AP P3 coupled with a revenue risk sharing mechanism. The project includes a four-lane 4.5 mile road and six lane 0.75 mile cable-stayed bridge between Montréal and Laval. Operations commenced in 2011. A dynamic toll system allows tariffs to increase beyond the maximum rate when the traffic reaches a certain threshold, serving as a congestion management mechanism (Parsons, 2015).
The 35-year concession includes four years for design and construction activities and 31 years for operation, maintenance, and rehabilitation activities. The Developer, "Concession A25," is responsible for design, construction, financing and operations. The Ministry of Transport has responsibility for toll collection and remittance of the toll revenues via APs. The APs are subject to availability and performance deductions (APEC 2014).
The revenue sharing mechanism works as follows:
The AP component of this transaction is effectively a MRG. By including both downside protection for the Developer and upside revenue sharing, the Agency helped reduce the risk profile of the project while also potentially securing its own future revenues.
Some AP mechanisms in Spain allow the Developer to charge higher tolls depending on the Agency's road condition evaluation. This is similar to the remedy that Brazil allowed Developers in that country's IRR balancing mechanism. Such an approach only works if the demand is sufficiently inelastic, which has not been the case for Brazil and U.S. P3s. This approach contrasts with most U.S. AP mechanisms, which impose a penalty for poor road conditions (Vassallo, 2006).
In transit concessions, Agencies can shift some of the revenue risk of bus, light rail and rail operations, and intermodal facilities to Developers. These are primarily arrangements in which the Developer is incentivized to increase ridership while still receiving a minimum payment for providing the service in the form of an AP. Some examples are listed below.
Under the shadow toll approach, an Agency pays a Developer based on the traffic on a non-tolled highway or on a tolled highway whose revenues the Agency retains. The shadow toll concept was developed in the U.K. to incentivize Developers to complete construction projects more quickly and/or carry out capital improvements in a way that minimized traffic impact.
While shadow tolls eliminate tolling risk - the risk that users will find the cost or inconvenience of using the toll road too onerous - they do not mitigate traffic risk for the Developer. Furthermore, revenue risk remains with the Agency if the road is tolled. Shadow tolls complicate financing and create contingent fiscal liabilities. The U.K. has moved away from shadow tolls and instead uses the AP method extensively. U.S. Agencies have considered shadow tolls, but no major projects have employed this approach.
Florida DOT (FDOT) considered an innovative approach to shadow tolling in the case of a managed lanes project, the I-595 express lanes project in Broward County, Florida. In that project FDOT would have entered into an agreement with a Developer to manage both the managed lanes and general purpose lanes, incentivizing it to maximize traffic on both facilities, within the constraints of the managed lanes free flow requirements, "thus better aligning the concessionaire's compensation with FDOT's policy emphasis of maximizing throughput for the overall facility" (Florida Department of Transportation, 2009, p.11). This was intended to address poor managed lanes use when general lanes were full and toll rates or technological issues became a barrier to their optimal usage. Such an approach could provide greater public benefits (DeCorla-Souza and Barker, 2005, p.65), while still taking advantage of the Developer's superior management capabilities to operate the facilities. Such an approach might be achieved through a two part payment, consisting of an AP sufficient to cover debt service and O&M and a shadow toll payment based on peak period vehicle throughput. This payment structure could incentivize concessionaires to maximize use of managed lanes during periods of congestion (thereby reducing congestion on regular lanes).
In comparing shadow tolls with AP and design build finance (DBF) approaches, FDOT rated the shadow toll inferior to both approaches, with FDOT ultimately deciding to award the concession on the basis of an AP. FDOT's rationale was that Developers perceived the payment mechanism as riskier. Compared to the AP, FDOT's shadow tolls alternative assumed:
The shadow toll mechanism FDOT considered was intended to align the Developer's financial incentives with the Agency's broader societal goals such as maximizing efficient use of the managed lanes facility. Such societal benefits could also be achieved under an AP, in which the payments to the Developer could, for example, be subject to 1) lane availability, 2) performance standards (including minimum speed on the managed lane), and 3) throughput on the managed lanes, provided that the Developer maintains full control over the dynamic tolling. Moving from a shadow toll to an AP transaction would potentially reduce the perceived risk while still allowing the Agency to pursue its broader economic goals.
The Rate of Return (ROR) model is used extensively to regulate electric utilities, which are natural monopolies. It allows firms to recover costs and earn a "fair" return by setting a regulated price, which is calculated by establishing the rate base and the value of all fixed assets used to produce the infrastructure at the agreed upon a rate of return. As discussed, South Korea provides its MRG on a similarly calculated rate base, but does not regulate the setting of toll rates.
Public utility commissions or their equivalent state regulatory agencies generally set prices on a periodic basis (such as every five years). The price base must guarantee financial feasibility, be lucrative enough to attract investment, and provide companies with returns similar to others with comparable risks (Buckberg, Kearney, and Stolleman, 2015).
ROR has rarely been used to regulate U.S. P3 toll roads. One private toll road, the Dulles Greenway, in the Washington, D.C. area, is regulated on the basis of ROR by the Virginia State Corporation Commission (VSCC), Virginia's public utility commission. The project entered technical bankruptcy in the 1990s, was restructured, and sold to Macquarie. Because of this loss of equity, VSCC has allowed Dulles Greenway to set its tolls as necessary, which Dulles Greenway has done subject to elasticity constraints.
Price Cap regulation gained popularity as a result of the deregulation of infrastructure in the U.K. in the 1980s and 1990s. Just as in the ROR model, the price cap mechanism protects consumers from excessive prices, while allowing the project sponsor to transfer the demand risk entirely to the private partner. The company is subject to price ceilings; yet the amount of revenues it receives are not regulated. The company maximizes its profits by fostering cost efficiency mechanisms. Price caps may also be set on a periodic basis, such as every five years.
As with ROR, U.K. price cap regulation has mostly been applied to large monopolies, such as utilities or commercial airports. It has generally not been applied to P3 toll roads.
The United Kingdom (U.K.) is rethinking its approach to P3s under its new P3 program, PF2. This has included a new policy allowing the national government to serve as a minority investor in future P3 projects. The purpose of this approach is to:
This approach is somewhat new for the U.K. and many other countries. However, some countries, such as Mexico, have required partial public ownership in infrastructure project companies, such as airports. Several questions will need to be answered in assessing this approach as a revenue risk transfer mechanism:
This approach may be another way for Agencies to better share risk and is an area that needs to be researched once relevant cases and data become available.
Determining the cost of a revenue risk sharing mechanism for an Agency can be challenging. In particular, mechanisms that include contingent liabilities, such as MRG and CFS, are difficult to evaluate quantitatively. However, in order to compare different mechanisms, an Agency may want to estimate the cost of different revenue risk sharing mechanisms. This Appendix provides a number of key considerations for evaluating revenue sharing mechanisms.
Besides valuing the fiscal impacts of the alternative revenue risk sharing mechanisms, the Agency must also consider the mechanisms' impact on Agency accounting and budgeting. For example, the rules for accounting for contingent liabilities vary by state making it difficult to budget for liabilities beyond the current planning period, often no more than five years. The same is true for the credit rating agency recognition of toll revenues. As discussed, credit rating agencies will generally fully count APs on Agency balance sheets if they are reliant on toll revenues for at least the first three years of a new project.
| Term | Definition/Explanation |
|---|---|
| Agencies | DOTs and other public transportation agencies that provide highway and/or transit infrastructure and/or services. |
| AP | Availability payment. |
| B | Billion. |
| Concession | Used interchangeably with P3, a long-term contract between a Developer and an Agency in which some or all of the following services are provided: design, construction, financing, operations, maintenance. |
| CFS | Contingent finance support |
| Developers | Usually organized as a special purpose vehicle, a company or a group of companies that provide some or all of the following services in a highway or infrastructure P3: design, construction, financing, operations, and maintenance. Developers are usually dominated by Strategic Investors, yet may include Financial Investors (see definitions). |
| DOT | Department of transportation at a local, state, or federal level. |
| DSCR | Debt service coverage ratio. |
| Equity Investors | Strategic and Financial Investors. Most Developers include Equity Investors. |
| Financial Investors | Private equity funds, pension funds, and other institutions that invest in infrastructure projects, independent of "strategic" motives, such as those of contractors, suppliers, or operators. See Strategic Investors. |
| IFI | International Financial Institution (such as the World Bank, Inter-American Development Bank, European Investment Bank). |
| Lenders | Financial institutions and their intermediaries that provide debt in the form of loans, bonds, and private placements. These can include commercial banks, credit agencies, investment banks, insurance companies, and government lenders (such as the USDOT TIFIA program or state infrastructure banks). Providers of deeply subordinated capital would be consider Equity Investors in this Discussion Paper. |
| M | Million. |
| MRG | Minimum revenue guarantee. |
| O&M | Operations and maintenance. |
| PVR | Present value of revenues. |
| P3 | Public-private partnership, used interchangeably with concession. |
| Secondary Market | The market of primarily Financial Investors that buy into concessions once the project has been completed and is operating successfully. |
| RRM | Regulated return mechanism. |
| Strategic Investors | Strategic investors are firms that make equity investments in a P3 with the goal of obtaining strategic benefits, such as a related construction, O&M, and/or supply contracts. Strategic Investors usually are the primary shareholders of most Developers. Once a project is completed and demonstrates stable cash flows, the strategic investor may sell some or all of its ownership in the P3 to Financial Investors, depending on the ownership requirements of the P3. |
| T&R | Traffic and Revenue |
| TIFIA | Transportation Infrastructure Finance and Innovation Act |
| VfM | Value for Money |
| WACC | Weighted Average Cost of Capital |
From July 2015 to October 2015, in-person and conference call discussions were held with more than 25 representatives of the following types of institutions:
| Respondent Type | Respondent Entity |
|---|---|
| Agency | Brazilian National Surface Transportation Agency (ANTT) |
| Former P3 executive with the State Government of Minas Gerais, Brazil | |
| Colorado High Performance Transportation Enterprise | |
| Maryland Department of Transportation | |
| North Carolina Department of Transportation | |
| Texas Department of Transportation | |
| Virginia Office of Public-Private Partnerships | |
| Developer, Strategic Investor | ACS Dragados |
| Cintra | |
| Macquarie | |
| Fluor | |
| Former U.S. executive of US-based Developer | |
| Kiewit | |
| Plenary Group | |
| Transurban | |
| Financial Investor | Meridiam |
| Law Firm | Nossaman LLP |
| Lender | Bank of Montreal |
| Barclays Capital | |
| Fitch Ratings | |
| Moody's | |
| Standard and Poor's | |
| Sumitomo Mitsui Banking Corporation | |
| University, Think Tank, Policy Office | Cornell Program in Infrastructure Policy |
| Polytechnic University of Madrid | |
| Reason Foundation | |
| South Korea's Public and Private Infrastructure Investment Management Center | |
| University of Minnesota, Department of Civil, Environmental, and Geo- Engineering | |
| U.S. Department of the Treasury |
The academic literature is rich on major revenue risk sharing mechanisms as shown below. Engel et al have written extensively on PVR and on the challenges of P3 negotiations and economics worldwide. Vassallo has also written on Chile's PVR method as well as on other risk sharing guarantees and mechanisms used in P3s. A number of economists have written on estimating the costs of guarantees, such as Aldrete, R., A. Bujanda, G. Valdez-Ceniceros. The World Bank has published a number of case studies on country-specific P3 programs and guides on guarantees and P3 tools. The Discussion Paper relied on trade publications for data on more recent transactions and tweaks to MRG, CFS, and AP + Revenue mechanisms.
In Brazil, Chile, and South Korea, P3 programs and their respective revenue risk sharing mechanisms have evolved over the last two decades as Agencies struggle to find the right risk sharing balance, seek VfM, and ensure that Developers maintain interest in their P3s. Other countries, including Colombia, Mexico, and Spain, have undergone similar evolutions. Engel et al, 2014 has a comparative chapter and the USDOT FHWA, USDOT FTA, and trade groups like APTA have commissioned international case studies that provide analysis comparing country approaches.
The greatest literature challenge is to find analysis that gets "into the weeds" of specific transactions to evaluate how respective mechanisms actually worked - for Agencies, Developers, Lenders, and others. This analysis would require confidential forecasts and financial models from both Agencies and Developers. For some projects, a conclusive analysis of the value of the revenue risk sharing mechanisms cannot be made until the concession contracts are complete or at least nearly complete - in twenty plus years.
Acosta, C. (2015). Interview with Sasha Page in Bogota, Columbia, TransMilenio headquarters, November, 2015.
Adler, T. and R. Tillman (2015). Quantified Probability Assessment of Revenue Forecasts. 2nd. International Conference on Public-Private Partnerships. RT Consultancy / RSG.
Albalate, D. and G. Gel. (2009). Regulating Concessions of Toll Motorways: An Empirical Study on Fixed vs. Variable Term Contracts. Transportation Research Part A: Policy and Practice, 43, 210-229.
Aldrete, R. and M. Bujanda (2012). Valuing Public Sector Revenue Risk Exposure in Transportation Public-Private Partnerships. Transportation Research Record: Journal of the Transportation Research Board, No. 2297,2012, pp. 88-96.
Aldrete, R., A. Bujanda, G. Valdez-Ceniceros (2010). Valuing Public Sector Risk Exposure in Transportation Public-Private Partnerships. Texas Transportation Institute.
American Public Transportation Association (APTA) (2013). Millennial and Mobility: Understanding the Millennial Mindset. October, 2013. Retrieved 11/22/2015 from: https://www.ssti.us/wp/wp-content/uploads/2013/10/APTA-Millennials-and-Mobility.pdf.
American Public Transportation Association (APTA) (2015). Final Report: APTA Innovative Funding And Financing Study Mission, London, Stockholm And Munich. April 19-24, 2015.
Amorelli, L.C. (2009). Brazilian Federal Road Concessions: New challenges to the regulatory framework. George Washington University Institute of Brazilian Business & Management Issues, the Minerva Program.
Asia Pacific Economic Cooperation (APEC) (2014). Infrastructure Public-Private Partnership Case Studies of APEC Member Economies. 21st Finance Ministers' Meeting, Beijing, China, October 21st to 22nd 2014. Retrieved 10/05/2015 from: http://mddb.apec.org/Documents/2014/MM/FMM/14_fmm_019.pdf.
Aziz, A. and K. Abdelhalim (2015). Comparative Analysis of P3 Availability Payments in the USA and Canada. 2nd International Conference on Public-Private Partnerships. Department of Construction Management / University of Washington.
Baeza, M., A. Ortega and J.M. Vasallo (2012). Risk Allocation in Toll Highway Concessions in Spain. Lessons from Economic Research. Transportation Research Record: Journal of the Transportation Research Board, No. 2297, 2012, pp. 80-87.
Bain, B. (2009). Error and optimism bias in toll road traffic forecasts. Springer Science+Business Media, LLC. February 28, 2009. Retrieved 11/22/2015 from: http://www.pppcouncil.ca/pdf/S&P_tollroads_2009.pdf.
Brandao, L. and E. Saraiva (2008). The option value of government guarantees in infrastructure projects. Construction Management and Economics. p 1171-1180.
Buckberg, E., O. Kearney, N. Stolleman (2015). Expanding The Market for Infrastructure Public-Private Partnerships: Alternative Risk and Profit Sharing Approaches to Align Sponsor and Investor Interests. U.S. Department Of The Treasury, Office of Economic Policy, April 2015.
Carbonara, N., Constantino, N., Pellegrino, R. (2015). Revenue guarantees in PPPs: A win-win option-based model. 2nd International Conference on Public-Private Partnerships, Austin.
Cheah, C. and J. Liu (2006). Valuing governmental support in infrastructure projects as real options using Monte Carlo simulation. Construction Management and Economics, p: 545-554
Chiara, N. and M. Garvin (2007). Using Real Options for Revenue Risk Mitigation in Transportation Project Financing. Transportation Research Record: Journal of the Transportation Research Board, No. 1993, 2007, p. 1-8.
Ciommo, F., J. M. Vassallo, and A. Oliver (2009). Private Funding of Intermodal Exchange Stations in Urban Areas: Case of Madrid, Spain. Transportation Research Record: Journal of the Transportation Research Board, No. 2187, 20-26.
DeCorla-Souza, P. and W. Barker (2005). Innovative Public-Private Partnership Models for Road Pricing/BRT Initiatives. Transportation Research Record: Journal of Public Transportation, Vol. 8, No. 1, 2005, 57-78.
Diu, F. C. PPP renegotiations in Portugal: Motorway PPP case study. Public Private-Partnership Congress. Lima, June 4, 2014. Retrieved 5/11/2016 from: www.idbdocs.iadb.org/wsdocs/getdocument.aspx?docnum=38843098.
Engel, E., R. Fischer, and A. Galetovic (2000). Least-present-value-of-revenue auctions and highway franchising. Working Paper, University of Chile.
Engel, E., R. Fischer, and A. Galetovic (2002). A New Approach to Private Roads. Regulation, Fall 2002, 18-22.
Engel, E., R. Fischer, and A. Galetovic (2003). Privatizing Highways in Latin America: Is It Possible To Fix What Went Wrong? Economic Growth Center Yale University, Discussion Paper No. 866, July 2003.
Engel, E., R. Fischer, and A. Galetovic (2014). The Economics of Public-Private Partnerships: A Basic Guide. New York: Cambridge University Press.
Engel, E., R. Fischer, and A. Galetovic (2015). Soft Budgets and Renegotiations in Public-Private Partnerships: Theory and Evidence. Serie De Dcoumentos De Trabajo, Facultad Economia Y Negocios, Universidad De Chile, Departmento De Economia, SDT 408.
European PPP Expertise Centre (EPEC) (2011). State Guarantees in PPPs: A Guide to Better Evaluation, Design, Implementation and Management. (EPEC).
Feigenbaum, B. (2011). Risks and Rewards of Public-Private Partnerships for Highways. Reason Foundation, Policy Brief 98, December 2011.
Fitch Ratings (2015). Fitch Assigns High Speed Rail Finance (1) PLC's Tap Final "A-" Rating. April 22, 2015, 10:31 AM (EDT).
Florida Department of Transportation (2009). I-595 Value for Money Analysis. Prepared by Jeffrey Parker and Associates, Inc., June 2009. Retrieved 11/28/2015 from: http://www.transportation-finance.org/pdf/funding_financing/financing/i595_vfm_0609.pdf.
Flyvbjerg, B. in association with COWI (2004). The British Department for Transport Procedures for Dealing with Optimism Bias in Transport Planning: Guidance Document.
Flyvbjerg, B., M.K. Skamris Holm, S.L. Buhl (2005). How (In)accurate Are Demand Forecasts in Public Works Projects? The Case of Transportation. Journal of the American Planning Association, Vol. 71, No. 2, 131-146.
Gao, O., R. Geddes, and O. Rouhani (2015). Revenue Risk Sharing Options for Highway Facilities. Cornell University.
Garvin, M. and E. Gonzalez (2013). Fiscal Support Mechanisms For Public- Private Partnerships. Engineering Project Organization Conference (EPOC). Colorado: EPOC.
Gómez-Lobo, A. and S. Hinojosa (2000). Broad Roads in a Thin Country: Infrastructure Concessions in Chile. Washington, D.C. World Bank, Policy Research Working Papers.
Guasch, L. (2004). Granting and Renegotiating Infrastructure Concessions. Washington, D.C.: The International Bank for Reconstruction and Development.
Hecht, J. (2015). Are Availability Payment Obligations Debt? Public Works Financing, September 2015, 16-17.
HM Treasury (2012). A new approach to public private partnerships. Retrieved 1/30/2016 from: https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/205112/pf2_infrastructure_new_approach_to_public_private_parnerships_051212.pdf.
HOCHTIEF, 10/4/2012. HOCHTIEF completes sale of Chilean toll highway. Retrieved 10/19/2015 from: http://www.hochtief-solutions.com/htsol_en/47.jhtml?pid=9211.
International Institute for Sustainable Development (IISD) and IMG Rebel (2015). Sharing Risk and Revenues from PPPs: Perspectives from current practice in the road sector. IISD.
Irwin, T. (2003). Public Money for Private Infrastructure. Deciding When to Offer Guarantees, Output- Based Subsidies, and Other Fiscal Support. Washington, D.C: The International Bank for Reconstruction and Development / The World Bank.
Irwin, T. (2007). Government Guarantees. Allocating and Valuing Risk in Privately Financed Infrastructure Projects. Washington, D.C.: The International Bank for Reconstruction and Development/The World Bank.
Park, T., Kim, B., and Kim, H. (2013). "Real Option Approach to Sharing Privatization Risk in Underground Infrastructures." Journal of Construction Engineering Management, 10.1061/(ASCE)CO.1943-7862.0000636, 685-693.
Kim, Kang-Soo (2013). Valuation of the Minimum Revenue Guarantee in the Urban Railway PPP Project. Korean Development Institute. Retrieved 11/28/2015 from: http://www.kdi.re.kr/kdi_eng/database/pop_summary.jsp?pub_no=13482&type=3.
Kokkaew, N., N. Chiara (2011). A Modeling Government Revenue Guarantees In Privately Built Transportation Projects: A Risk-Adjusted Approach. Transport, Vol. 28, No.2, March 2011, 186-192.
KPMG (2015). Public Private Partnerships. Emerging global trends and the implications for future infrastructure development in Australia. Retrieved on 05/15/2016 from: https://assets.kpmg/content/dam/kpmg/pdf/2015/06/public-private-partnerships-june-2015.pdf
Levinson, D. and K. Krizek (2015). The End of Traffic and the Future of Transport. Kindle Edition. EPUB.
Macquarie (2015). Minimum revenue guarantee summary. Retrieved 10/05/2015 from: https://www.macquarie.com/mgl/mkif/en/mkif-assets/minimum-revenue-guarantee-summary.
Moody's Investor Service (2012). Rating Action: Moody's assigns initial Ba3 rating to Aeropuertos Dominicanos Siglo XXI $550 million senior notes. October 29, 2012.
Park, S. (2014). Improving the Bankability of PPP with Proper Risk Sharing: The Korean Experience. PIMAC. June, 2014.
Parsons (2015) Autoroute 25 Infrastructure, Public-Private Partnership. Retrieved 10/08/2015 from: https://www.parsons.com/project/a25-ppp/.
PricewaterhouseCoopers (PWC) (2011). Funding Infrastructure: Time for a new approach? Retrieved 05/15/2016 from: https://www.pwc.com/gx/en/psrc/pdf/time-for-a-new-approach.pdf.
Quiggin, J. (2005). Public-Private Partnerships: Options for Improved Risk Allocation. Policy Forum: Financing Public Infrastructure. The Australian Economic Review, vol. 38, no. 4, pp. 445-50.
Rus, G.D. and G. Nombela (2000). Least Present Value of Net Revenue: a new auction-mechanism for highway concessions. MPRA Paper No. 12203, posted December 18, 2008.
Rus, G.D. and G. Nombela (2003). Flexible-Term Contracts for Road Franchising. Working Paper, University of Las Palmas (Spain).
Sanchez-Solino, A. and J. M. Vasallo (2006). Minimum Income Guarantee in Transportation Infrastructure Concessions in Chile. Transportation Research Record: Journal of the Transportation Research Board, No. 1960, 15-22.
Sanchez-Solino, A. and J. M. Vasallo (2007). Subordinated Public Participation Loans for Financing Toll Highway Concessions in Spain. Transportation Research Record: Journal of the Transportation Research Board, No. 1996, 22-28.
Shan, L., M. Garvin, and R. Kumar (2010). Collar options to manage revenue risks in real toll public-private partnership transportation projects. Construction Management and Economics, 28: 10, 1057 - 1069.
Stratfor Global Intelligence (2015). Brazilian Growth and Inflation. Retrieved on 10/07/2015 from: https://worldview.stratfor.com/article/brazilian-growth-and-inflation.
Tollroads, 2014. $1.4 Billion Refinancing Improves Long-Term Debt Structure for the 73 Toll Road. Retrieved on 11/22/2015 from: https://thetollroads.com/news/newsroom/newsroom/pressreleases/archive/201410/pr-20141022-14-billion-refinancing-improves-long-t.php.
U.S. Department of Transportation, Federal Highway Administration, Office of Highway Policy Information (USDOT FHWA OHPI) (2015). Traffic Volume Trends. Retrieved 10/16/2015 from: https://www.fhwa.dot.gov/policyinformation/travel_monitoring/tvt.cfm.
U.S. Department of Transportation, Federal Highway Administration, TIFIA Program (USDOT TIFIA) (2015). Projects Financed by TIFIA. Retrieved 10/24/15 from: https://www.transportation.gov/tifia/projects-financed.
U.S. Department of Transportation, Federal Transit Administration (USDOT FTA) (2009). Foreign Public Private Partnership (PPP) Case Study Analysis Report: Canada Line, London Underground, TransMilenio, and Southern Cross Station.
United States Department of Transportation (2009). Public Policy Considerations in Public-Private Partnership (PPP) Arrangements.
Valsangkar, P. (2008). Revenue Sharing Models in a Public Private Partnership (PPP) Context. Computer Society of India.
Vassallo, J. M. (2006). Traffic Risk Mitigation in Highway Concession Projects: The Experience of Chile. Journal of Transport Economics and Policy, Volume 4, Part 3, September 20-06, 359-281.
Vassallo, J. M. (2010). Flexible-Term Highway Concessions How Can They Work Better? Transportation Research Record: Journal of the Transportation Research Board, No. 2187, 1-8.
Vecchi, V., E. Amadio, E. Borgonovo, N. Cusumano, S. Gatti, (2015). Can public manager competencies offset the bidding opportunism generated by public guarantees issued to increase the bankability of PPP projects? Evidences from a multi agent simulation. Bocconi University.
World Bank Group (2015). GDP per capita (current US$). Retrieved 11/22/2015 from: http://data.worldbank.org/indicator/NY.GDP.PCAP.CD.
World Bank Group and Gobierno de España (2012). Best Practices in Public-Private Partnerships Financing in Latin America: The Role of Guarantees. Washington, D.C.: International Bank for Reconstruction and Development.
3 This case primarily relies on published evaluations, in particular of Engel et al (2000, 2002, 2003, 2014) and Vassallo (2006).
4 This case was developed with published materials, an interview with representatives of ANTT on August 17, 2015 organized by the FHWA, and with Marcos Siqueira Moraes, Former Head of the PPP Unit of the State Government of Minas Gerais, and founding partner of the consulting company Radar PPP on 09/15/2015.
5 This case study drew upon both published materials and an interview with Soojin Park, the head of the Policy Team for South Korea's Public and Private Infrastructure Investment Management Center (PIMAC).
