Greenfield express toll lane projects are still relatively new in the public-private partnership (P3) concession world. Besides the original SR 91 Express Lanes that opened in Orange County, California in December 1995, just six express toll lane projects have been financed as toll concessions thus far: three in Texas, two in Virginia, and one in Colorado. Two freeway reconstructions that include the addition of express toll lanes have been financed as availability payment concessions in Florida. All the other express toll lane projects have been done by state departments of transportation as conversions of existing high occupancy vehicle (HOV) lanes into high occupancy toll (HOT) lanes. In those conversions, the costs are a small fraction of what’s involved in adding new lanes to urban expressways.
The P3 community has been eagerly awaiting Virginia’s third express toll lanes project-I-66 outside the Beltway. Estimated to cost between $2 billion and $3 billion, the project would convert the existing HOV lane and add a new express toll lane each way, for a distance of 25 miles. As of last fall three or four teams were being assembled, and expectations were high that the process would begin soon after the first of the year.
But now there are rumblings of concern over whether the Virginia Department of Transportation (VDOT) actually intends to offer the I-66 project as a toll concession. At the March Commonwealth Transportation Board meeting, VDOT Secretary Aubrey Layne said, “we generally thought that [I-66] would lend itself to a P3, but that decision has not been made yet.” He added that the project would require a “large public sector” contribution and that “we believe there are available funds we can raise to do that through other federal programs.” Layne also pointed out that under the new legislation, in order to do a project as a P3, “the Secretary has to certify that a P3 is the best route.” And that compared with previous P3 projects, “We are in a little different position; the board has monies, and initiating tolling is not the risk that was in previous deals.” At the April meeting, Layne responded to a question about the I-66 project, saying only that deciding whether to do it as a P3 “would be done via the process specified in the new law.”
The suggestion that the revenue from express toll lanes is low-risk strikes me as bizarre. Over the years I have attended many presentations on toll project traffic and revenue projections, many of them by Ed Regan, Senior Vice President of CDM Smith, some of them dealing with managed lanes. He points out that compared with revenue projections on conventional toll road projects, the revenue on a managed lane is very sensitive to small changes in demand. For example, a 10% increase in corridor traffic might mean a 30% increase in revenue, but this cuts both ways-a 10% traffic decrease in corridor traffic due to a recession could mean 30% less revenue. In addition, revenue in the early years of a managed lane is typically low. Compared with traditional toll roads, managed lanes revenue is somewhat less predictable and less stable, especially in the early years. But longer-term, the revenue growth is likely to be greater.
Rating agencies consider managed lanes a good fit for P3 concessions, because compared with typical public sector all-debt financing, the equity investment in a concession provides a cushion during the early years. If toll revenue is below forecast in a $1 billion 100% debt-financed project, debt service must still be paid on debt that provided the $1 billion up-front. But if 25% of the $1 billion was equity and only 75% was debt, then the debt service payments can still be made, even if initial toll revenue is well below forecast.
In April 2012, Fitch Ratings and Standard & Poor’s each released reports on managed lane projects, in both cases focusing on P3 projects in which the new capacity is paid for via toll-based finance. Fitch states that “It is [our] expectation that ML revenue will behave like a derivative, meaning as GPL [general purpose lane] volume grows, ML revenue will grow at faster rates. Likewise, when the amount of GPL traffic declines, ML traffic and revenue will drop more.” Fitch goes on to say, “Given this volatility, higher liquidity levels throughout the life of the debt are critical to help support cash flow during periods of economic weakness. All else being equal, an ML project rated ‘BBB’ needs to have more financial flexibility, either in the form of structured liquidity or a highly flexible debt structure, than a typical toll road given the potential volatility in annual cash flow.” And S&P points out that even a mature project like the 91 Express Lanes in California has suffered outright decreases in toll revenue during recessions.
As Virginia DOT (and other state DOTs) consider whether to do a managed lanes mega-project in-house or via a P3 toll concession, it’s easy to imagine public officials with visions of surplus toll revenues in the later years of the project asking themselves why they should let a private sector company have all that revenue in the out-years. There certainly is long-term potential for such revenues, but deciding whether it’s in the public interest to do the project as a P3 concession must balance possible future revenues against the very real possibility of early-year losses due to the inherent volatility of managed lanes traffic and revenue.
A good example is the I-495 Express Lanes on the Beltway in northern Virginia. The first two years of traffic and revenue are still below projections, so concession company Transurban restructured the project’s finances last year, increasing its equity investment by 80%. Had VDOT done the project, would it have been prepared to invest an additional $280 million in its second year? That kind of traffic and revenue risk transfer is part of the value proposition for doing this kind of project as a P3 toll concession.
As for the potential surplus revenue in the out-years, the obvious answer is to build into the concession agreement a pre-defined revenue-sharing provision with the state DOT. That’s especially appropriate if the state DOT makes an initial investment in the project, as VDOT did on the Beltway project (whose concession agreement does include revenue-sharing).
And of course, that risk transfer is in addition to the company taking on the risks of construction cost over-runs, late completion, and all operations and maintenance, as well as the ability to raise the capital investment for the project now, when it’s needed, rather than having to defer the project to a future decade.
Robert Poole is director of transportation policy at Reason Foundation. This article originally appeared in Public Works Financing.