Transport Report: Toll roads- the way ahead
1999 was a watershed year for the development of toll roads with
the successful close of four of the biggest road deals ever
constructed on a non-recourse basis. These projects demonstrate the
enormous appetite that the private sector, and the capital markets
in particular, have for traffic and construction risk.
But if the size of these deals astounds, it is the detail that provides signposts for road finance in the second millennium.
What three of the deals of 1999 most importantly show is the recent trend to part privatisation, part project finance on a concession basis to develop new roads. This is particularly so in the case of Highway 407 ? of the C$3.8 billion of project cost, C$3.1 billion is the cash sum paid to acquire the existing facility. Only C$900 million will be invested in new infrastructure.
A similar structure was used on the N3 in South Africa, a much smaller road deal, which was also financed in 1999. R1.35 billion was spent acquiring the 420 km Johannesburg ? Durban toll road in exchange for a concession which includes the obligation at a future date to upgrade the De Beers Pass, a key section of the road.
But nowhere is this emphasis on part privatisation, part project finance better demonstrated than in the whole roads programme of Portugal. The table below shows how significant a role existing stretches of road play in its programme of 16 road concessions (of which both the Northern Concession and the SCUT Beira Interior are part). In the table the first two numerical columns show the existing stretches transferred with the concessions while the third shows the new kilometerage.
The nature of the role of the existing stretches can differ. For instance, the Grande Porto is primarily about new construction (46.4 km out of a project total of 62.7 km) whereas the IP5 involves widening an existing stretch of road (175 km out of a project total of 179.5 km).
Such an approach is surely privatisation by another name. It demonstrates the growing fusion between privatisation and concessioning, or project finance, and it addresses one of the fundamental problems of road construction by the private sector; its inability to make the numbers add up.
Most roads, and indeed most transport projects, involve a large amount of capital cost. There is no such thing as a standard cost per kilometer for the construction of a road. The cost depends upon ground conditions, the number structures used, the local cost of materials and labour, the length of the annual construction season and many more variables. However, a rule of thumb of $5 million/kilometer provides a good guide. It is very rare for a road to be able to charge sufficient at the toll booths to be able to pay operating costs, service debt and make a return to equity at such levels of cost. Unlike a power or a water concession, where the output tariff is typically paid by a host government or one of its agents, the toll on a road is normally paid by the users. In a power or water project the output tariff can be set at a level which is sufficient to service the project as a whole. A toll cannot be set in such a way. If it is set too high it will drive away users who cannot or will not pay it. This diversion effect can only be offset by subsidy or by the use of shadow tolls (paid by the government not the user ? see below). The traditional approaches to this issue have included the provision of grants, guarantees, shadow tolls or favourable tax treatment by host governments.
The use of existing stretches of road in concessions to build new or extend stretches of motorway helps address this problem. In the first place it provides the project with a cashflow not associated with the cost of construction.
In the case of the West and Northern concessions the bidders had to offer a price for taking over the existing road. (the A8 on the West and the A7 on the Northern). In these projects the existing cashflow from the existing roads is available from the first day of the concession. This reduces funding requirements during construction which benefits the concession.
In the case of the seven shadow toll roads, the concessionaires will be paid a fixed fee at a relatively low level during construction. This will be converted into a shadow toll payment based upon vehicle flows on both existing, widened and new stretches of road after five or six years, depending on the particular project.
The other benefit of including existing stretches of road in concessions is that it makes forecasting traffic a little more predictable. Traffic forecasting is a notoriously difficult business. Autostrade used to say that an accuracy factor of +/? 10% is good. There have been some notable disasters ? traffic on the Dulles Greenway (financed in September 1992) turned out to be around only 30% of estimated levels. Borrowing large amounts of money on forecasts which may only be reliable to +/? 10% is not an easy process. By including existing motorways in the concessions the Portuguese state is providing concessionaires with a traffic flow which is readily verifiable ? it can be counted and does not have to be estimated (except for growth predictions, which tend to be related to GDP) in quite the same way as traffic on a new road has to be forecasted.
The Information Memorandum for the syndication of the West Concession, in 1999, drew attention to this fact.
The map of the Northern concession illustrates quite well how the existing and the new can be blended together to form an overall project which ensures a large amount of investment and construction and yet benefits from the predictability and availability of cashflows accruing to the existing stretch of road, from Famalicao to Guimaraes, which was transferred to the concession.
How the Road Tolls
One of the four deals, the SCUT Beira Interior, is a shadow toll road. If there is anything new in road financing it is the shadow toll.
Under a real toll mechanism the user of the road pays a toll. Therefore, the user may be deterred from travelling by the cost of the toll. The road will need to incorporate toll plazas at which tolls will be collected or where traffic will be measured through toll barriers, either physically or electronically, and the user will be debited by some form of accounts system. The toll plazas add cost, both capital and operational, to the project and the project bears patronage risk, both in the sense of traffic volume generally and the diversion effect of the toll charge.
Under a shadow tolling system the user of the road does not pay the toll. Instead the host agency, in this case JAE (the Portuguese Highways Agency), pays the concessionaire tolls in relation to the number of vehicles travelling on the road. It does this on the basis of a banding structure which is related to the vehicle kilometerage (?vkm?) travelled on the road. Bidders for the shadow toll projects must define their own banding structure, within the broad framework of three bands ? viz:
PTE 0
Band 3
PTE Z per vkm
Number of
vehicle Band 2
kilometers PTE Y per vkm
Band 1
PTE X per vkm
Time
PTE = Portuguese Escudos
Any traffic above the third band receives no toll, thereby capping JAE's toll payments and the concessionaire's potential returns. The tolls may be indexed, but only growing by 90% of the Consumer Price Index increase in Portugal each year. The precedents in the UK for financing road projects in this way suggest that bidders use each band to cover different elements of the project's cost profile.
Band 1 has typically been used to cover fixed operating and maintenance costs and senior debt service.
Band 2 covers variable operating and maintenance costs and subordinated debt service. Band 3 tends to be used to pay dividends and for quasi-equity debt service.
Shadow tolls have now been successfully used in the UK, Finland, Madrid and Portugal. They have been considered in India and at one stage it was proposed to structure the Brisbane City Valley By-pass in the same way.
The concept of payment by the host government in proportion to some performance measure is gradually being extended. For example the A130 in the United Kingdom includes as an element of the payment stream a fee for availability ? the concessionaire is paid to keep the road open. This puts greater emphasis on the transfer of operations and, in particular, maintenance risk.
In principle, provided it can be measured objectively, any yardstick can be used to evaluate the performance of a concessionaire and be the basis for a payment mechanism. It is not beyond the bounds of imagination to envisage payment mechanisms related to such measures as the reduction in accident levels, journey time reliability or skid resistance. In Poland, the Highways Agency is considering proposals whereby it pays to a concessionaire the financially necessary revenue stream, based upon shadow tolls and availability, and charges the road user the economically sustainable toll.
?Gentlemen prefer bonds?
The Mellon family's famous dictum might just as easily be applied to toll road financing because of the particular nature of the cashflow profile of a toll road. One member of the class of 1999, the Highway 407, refinanced its senior bridge facility with the issue of C$1.1 billion of bonds and a private placement of over C$600 million of real return bonds. An interesting comparison again comes from the N3, which issued index linked bonds on the basis of an A- rating by Standard and Poors. CPI linked debt also featured in the Cross ? Israel Highway; $850 million of an index linked Shekel facility was raised. In principle, index linked facilities should suit most toll roads, whose tolls are invariably related in some way to local inflation adjustments. 1999 also saw the Polish government agree to provide a guarantee to secure the issuance of bonds to finance part of the A2 highway concession.
Bonds have a number of advantages over bank debt. The main one is that they can be structured over considerable maturities. This is critical to the cashflow profile of any road, where, typically, toll revenues in the early years of a project are low but grow considerably over time. Stretching out the maturity over a longer period can help overcome this early pinch point. Similarly, extensive grace periods can be achieved, which address the same problem. By tapping the bond markets roads are broadening their funding base, thus ensuring greater competition from the financial markets. In contrast, a major disadvantage of bond financing is the rigidity of the structure.
The exotic featured quite highly in three of the four big deals. The 407, the Cross-Israel Highway and the Northern Concession all used subordinated debt in their financial structures. In addition, the Cross-Israel Highway also issued $250 million of loan notes in a private placement. It is also worth noting that the Cross-Israel Highway achieved an extraordinary maturity of 28 years for some of its senior debt. Bond financing and some unusual debt structures are sure to feature ever more significantly in road financing.
?An Easy Toll? ? Conclusion
1999 was undoubtedly a successful year for road financing. The deals done, on several continents, show the way ahead for toll road financing. The basis of much of it will surely be the part privatisation, part concession model. Only in this way will the age old problem of generating enough funds at the toll booths to pay back the cost of investment be solved.
There will be considerable focus on payment mechanisms with the payment mechanism of the 1990s ? the shadow toll ? being developed into other performance measures. And, with such a requirement for projects and funds, the traditional bank debt markets will have to be complemented by more exotic financing instruments that both increase funding capacity and address the traditional cashflow profile of a road. Thus will the ?easy toll? continue to be the way ahead.
Nigel Purse is a Partner in the Project Finance and Privatisation Investment Banking department of PricewaterhouseCoopers, where he is responsible for road financing. He led the teams which financed the Northern Concession and the SCUT Costa de Prata in Portugal and he is currently advising the Highways Agencies in Poland, India and South Africa on toll road concessions.
But if the size of these deals astounds, it is the detail that provides signposts for road finance in the second millennium.
What three of the deals of 1999 most importantly show is the recent trend to part privatisation, part project finance on a concession basis to develop new roads. This is particularly so in the case of Highway 407 ? of the C$3.8 billion of project cost, C$3.1 billion is the cash sum paid to acquire the existing facility. Only C$900 million will be invested in new infrastructure.
A similar structure was used on the N3 in South Africa, a much smaller road deal, which was also financed in 1999. R1.35 billion was spent acquiring the 420 km Johannesburg ? Durban toll road in exchange for a concession which includes the obligation at a future date to upgrade the De Beers Pass, a key section of the road.
But nowhere is this emphasis on part privatisation, part project finance better demonstrated than in the whole roads programme of Portugal. The table below shows how significant a role existing stretches of road play in its programme of 16 road concessions (of which both the Northern Concession and the SCUT Beira Interior are part). In the table the first two numerical columns show the existing stretches transferred with the concessions while the third shows the new kilometerage.
The nature of the role of the existing stretches can differ. For instance, the Grande Porto is primarily about new construction (46.4 km out of a project total of 62.7 km) whereas the IP5 involves widening an existing stretch of road (175 km out of a project total of 179.5 km).
Such an approach is surely privatisation by another name. It demonstrates the growing fusion between privatisation and concessioning, or project finance, and it addresses one of the fundamental problems of road construction by the private sector; its inability to make the numbers add up.
Most roads, and indeed most transport projects, involve a large amount of capital cost. There is no such thing as a standard cost per kilometer for the construction of a road. The cost depends upon ground conditions, the number structures used, the local cost of materials and labour, the length of the annual construction season and many more variables. However, a rule of thumb of $5 million/kilometer provides a good guide. It is very rare for a road to be able to charge sufficient at the toll booths to be able to pay operating costs, service debt and make a return to equity at such levels of cost. Unlike a power or a water concession, where the output tariff is typically paid by a host government or one of its agents, the toll on a road is normally paid by the users. In a power or water project the output tariff can be set at a level which is sufficient to service the project as a whole. A toll cannot be set in such a way. If it is set too high it will drive away users who cannot or will not pay it. This diversion effect can only be offset by subsidy or by the use of shadow tolls (paid by the government not the user ? see below). The traditional approaches to this issue have included the provision of grants, guarantees, shadow tolls or favourable tax treatment by host governments.
The use of existing stretches of road in concessions to build new or extend stretches of motorway helps address this problem. In the first place it provides the project with a cashflow not associated with the cost of construction.
In the case of the West and Northern concessions the bidders had to offer a price for taking over the existing road. (the A8 on the West and the A7 on the Northern). In these projects the existing cashflow from the existing roads is available from the first day of the concession. This reduces funding requirements during construction which benefits the concession.
In the case of the seven shadow toll roads, the concessionaires will be paid a fixed fee at a relatively low level during construction. This will be converted into a shadow toll payment based upon vehicle flows on both existing, widened and new stretches of road after five or six years, depending on the particular project.
The other benefit of including existing stretches of road in concessions is that it makes forecasting traffic a little more predictable. Traffic forecasting is a notoriously difficult business. Autostrade used to say that an accuracy factor of +/? 10% is good. There have been some notable disasters ? traffic on the Dulles Greenway (financed in September 1992) turned out to be around only 30% of estimated levels. Borrowing large amounts of money on forecasts which may only be reliable to +/? 10% is not an easy process. By including existing motorways in the concessions the Portuguese state is providing concessionaires with a traffic flow which is readily verifiable ? it can be counted and does not have to be estimated (except for growth predictions, which tend to be related to GDP) in quite the same way as traffic on a new road has to be forecasted.
The Information Memorandum for the syndication of the West Concession, in 1999, drew attention to this fact.
The map of the Northern concession illustrates quite well how the existing and the new can be blended together to form an overall project which ensures a large amount of investment and construction and yet benefits from the predictability and availability of cashflows accruing to the existing stretch of road, from Famalicao to Guimaraes, which was transferred to the concession.
How the Road Tolls
One of the four deals, the SCUT Beira Interior, is a shadow toll road. If there is anything new in road financing it is the shadow toll.
Under a real toll mechanism the user of the road pays a toll. Therefore, the user may be deterred from travelling by the cost of the toll. The road will need to incorporate toll plazas at which tolls will be collected or where traffic will be measured through toll barriers, either physically or electronically, and the user will be debited by some form of accounts system. The toll plazas add cost, both capital and operational, to the project and the project bears patronage risk, both in the sense of traffic volume generally and the diversion effect of the toll charge.
Under a shadow tolling system the user of the road does not pay the toll. Instead the host agency, in this case JAE (the Portuguese Highways Agency), pays the concessionaire tolls in relation to the number of vehicles travelling on the road. It does this on the basis of a banding structure which is related to the vehicle kilometerage (?vkm?) travelled on the road. Bidders for the shadow toll projects must define their own banding structure, within the broad framework of three bands ? viz:
PTE 0
Band 3
PTE Z per vkm
Number of
vehicle Band 2
kilometers PTE Y per vkm
Band 1
PTE X per vkm
Time
PTE = Portuguese Escudos
Any traffic above the third band receives no toll, thereby capping JAE's toll payments and the concessionaire's potential returns. The tolls may be indexed, but only growing by 90% of the Consumer Price Index increase in Portugal each year. The precedents in the UK for financing road projects in this way suggest that bidders use each band to cover different elements of the project's cost profile.
Band 1 has typically been used to cover fixed operating and maintenance costs and senior debt service.
Band 2 covers variable operating and maintenance costs and subordinated debt service. Band 3 tends to be used to pay dividends and for quasi-equity debt service.
Shadow tolls have now been successfully used in the UK, Finland, Madrid and Portugal. They have been considered in India and at one stage it was proposed to structure the Brisbane City Valley By-pass in the same way.
The concept of payment by the host government in proportion to some performance measure is gradually being extended. For example the A130 in the United Kingdom includes as an element of the payment stream a fee for availability ? the concessionaire is paid to keep the road open. This puts greater emphasis on the transfer of operations and, in particular, maintenance risk.
In principle, provided it can be measured objectively, any yardstick can be used to evaluate the performance of a concessionaire and be the basis for a payment mechanism. It is not beyond the bounds of imagination to envisage payment mechanisms related to such measures as the reduction in accident levels, journey time reliability or skid resistance. In Poland, the Highways Agency is considering proposals whereby it pays to a concessionaire the financially necessary revenue stream, based upon shadow tolls and availability, and charges the road user the economically sustainable toll.
?Gentlemen prefer bonds?
The Mellon family's famous dictum might just as easily be applied to toll road financing because of the particular nature of the cashflow profile of a toll road. One member of the class of 1999, the Highway 407, refinanced its senior bridge facility with the issue of C$1.1 billion of bonds and a private placement of over C$600 million of real return bonds. An interesting comparison again comes from the N3, which issued index linked bonds on the basis of an A- rating by Standard and Poors. CPI linked debt also featured in the Cross ? Israel Highway; $850 million of an index linked Shekel facility was raised. In principle, index linked facilities should suit most toll roads, whose tolls are invariably related in some way to local inflation adjustments. 1999 also saw the Polish government agree to provide a guarantee to secure the issuance of bonds to finance part of the A2 highway concession.
Bonds have a number of advantages over bank debt. The main one is that they can be structured over considerable maturities. This is critical to the cashflow profile of any road, where, typically, toll revenues in the early years of a project are low but grow considerably over time. Stretching out the maturity over a longer period can help overcome this early pinch point. Similarly, extensive grace periods can be achieved, which address the same problem. By tapping the bond markets roads are broadening their funding base, thus ensuring greater competition from the financial markets. In contrast, a major disadvantage of bond financing is the rigidity of the structure.
The exotic featured quite highly in three of the four big deals. The 407, the Cross-Israel Highway and the Northern Concession all used subordinated debt in their financial structures. In addition, the Cross-Israel Highway also issued $250 million of loan notes in a private placement. It is also worth noting that the Cross-Israel Highway achieved an extraordinary maturity of 28 years for some of its senior debt. Bond financing and some unusual debt structures are sure to feature ever more significantly in road financing.
?An Easy Toll? ? Conclusion
1999 was undoubtedly a successful year for road financing. The deals done, on several continents, show the way ahead for toll road financing. The basis of much of it will surely be the part privatisation, part concession model. Only in this way will the age old problem of generating enough funds at the toll booths to pay back the cost of investment be solved.
There will be considerable focus on payment mechanisms with the payment mechanism of the 1990s ? the shadow toll ? being developed into other performance measures. And, with such a requirement for projects and funds, the traditional bank debt markets will have to be complemented by more exotic financing instruments that both increase funding capacity and address the traditional cashflow profile of a road. Thus will the ?easy toll? continue to be the way ahead.
Nigel Purse is a Partner in the Project Finance and Privatisation Investment Banking department of PricewaterhouseCoopers, where he is responsible for road financing. He led the teams which financed the Northern Concession and the SCUT Costa de Prata in Portugal and he is currently advising the Highways Agencies in Poland, India and South Africa on toll road concessions.
Request a Demo
Interested in IJGlobal? Request a demo to discuss a trial with a member of our team. Talk to the team to explore the value of our asset and transaction databases, our market-leading news, league tables and much more.