In March of this year, the Texas Department of Transportation (DOT) signed a precedent-setting deal. Its very first long-term toll concession agreement (or Comprehensive Development Agreement [CDA], as they are known in Texas) showed that the private sector could finance more than twice as much as a conventional public-toll-agency financial model.
The conventional toll-agency model could provide only $600 million of the $1.3 billion cost of SH 130 (Segment 5 and 6) between Austin and San Antonio. But under a reasonable set of CDA terms (50 years, inflation-adjusted toll rates, modest competition limits, and normal buyback provisions), a long-term CDA would raise the entire $1.3 billion. Not only would the private sector pay for the entire project, it would share some of the revenues with the government during the 50-year lease deal.
The entire U.S. transportation community was awed by this project. It received a “deal of the year” award from the American Road & Transportation Builders Association. The implications went far beyond just SH 130: If the private sector could take a marginal toll corridor with modest traffic forecasts and fully fund it, when traditional public-agency toll finance could not, this implied that CDAs could fund a great many of the big-ticket expressway projects needed to accomplish Texas's aggressive and much-needed congestion-reduction goals.
That was bad news for those who want to stop the expansion of toll roads and public-private partnerships. And for some reason, it was also seen as a threat by the two well-run public-sector toll agencies in Texas— North Texas Tollway Authority (NTTA) in Dallas/Ft. Worth and Harris County Toll Road Authority (HCTRA) in Houston. Never mind that there are plenty of new toll road projects to go around, some of which meet the normal, conservative funding criteria for public-sector toll agencies. Some consultants and others who make their living financing public toll roads began to peddle the idea that anything the private sector could do via CDAs, the public toll agencies could do better.
The first confrontation has come with a second new toll road: SH 121 in the Dallas suburbs. In contrast with SH 130, this one is much easier to finance. First, the sum that must be raised to build it is less, since part of it has already been funded and portions already built. Plus, unlike the rural area between Austin and San Antonio, the SH 121 corridor is a very high-growth suburban area with a strong need for the road. Anyone could make this one work as toll road. Hence, it was obvious that if the same basic CDA model as in SH 130 was used, there could be a very large up-front payment, as well as potentially large revenue sharing. So Texas DOT held a competition, several teams competed, and a winning deal was announced. Besides fully funding the toll road, the CDA with Cintra would provide the public sector with a $2.1 billion up-front payment, guaranteed annual lease payments worth (present value) $700 million, and significant potential revenue sharing during the 50-year term.
But this deal was being finalized just as the anti-toll, anti-CDA forces in the legislature were developing a two-year moratorium on CDA projects. In fact, close examination of the moratorium bill's provisions shows that it would so greatly increase the risks and reduce the potential profits as to effectively kill CDAs in Texas. That's certainly what some of the bill's supporters intended. But many others went along, hoping the moratorium would simply permit fine-tuning of the details.
And that's where the public-toll-agency campaign comes in. If reasonable people can be convinced, during the remaining few weeks of the legislative session, that public toll agencies can do everything the private sector can to via CDAs, then even if the moratorium actually kills off CDAs and drives the private sector toll companies away from Texas, the toll agencies will be there to do all the projects themselves.
But there's one huge problem with this assertion: it's wrong.
The first thing to remember is that SH 121 is a very special case. Most other needed toll road projects will be far more costly to build, with less-robust traffic (i.e., more like SH 130 than like SH 121). So public agencies will not be able to fund all of these needed roads and the CDA model will be needed to build them. If that model is thrown away due to a false belief that public agencies can meet all the needs, Texas will be left with a huge funding shortfall.
It's not possible to fully analyze NTTA's SH 121 “proposal,” since all they have shared with the public thus far are glittering generalities and grandiose claims, not the underlying assumptions on which those claims rest. For now, the best we can do is to review the generally similar deal NTTA sketched out in a letter to Sen. John Carona on March 12, 2007 (where they did disclose many assumptions).
In order to raise the $2.1 billion up-front payment to compete with Cintra, NTTA proposed mortgaging their entire toll road system. That might be do-able—but just this one time. As a general solution for funding numerous needed toll projects in the Metroplex over the next 25 years, it's a dud. NTTA will be out of assets and too far in debt to fund other projects. By contrast, an approach based on CDAs would finance each toll road based on its own revenue stream, meaning that approach could be used again and again. Moreover, once the bond rating agencies get a chance to look at an actual, detailed proposal from NTTA, the odds are very high they will downgrade the agency's bond rating, due to a much higher level of risk (see below). Moody's has already suggested as much.
A major flaw in NTTA's March proposal is that it was based on a far more “aggressive” projection of population, land-use, and traffic than is typical for public-sector toll agencies. Public-sector agencies almost always rely on more conservative forecasts because they need to get an investment-grade rating on their bonds, to sell them to conservative tax-exempt bond-buyers. NTTA's riskier projection is one of the factors likely to lead to a lower bond rating (which means a higher cost of debt service, for that huge new level of NTTA debt).
An equally large flaw in NTTA's proposal is the assumption that it will be able to do what a company operating a 50-year CDA could do: raise toll rates at up to the consumer price index (CPI) every year for 50 years. That is a completely non-credible assumption—and the rating agencies and bond buyers will see it as such. Political interference in toll-setting is a problem that has plagued public-sector toll agencies ever since the Pennsylvania Turnpike opened to traffic in 1940. No one has ever come up with a mechanism to bind future elected bodies from interfering with future toll increases. But a long-term CDA with CPI caps is legally enforceable. That's why investors will gladly fund CDA projects.
And that leads to the third fatal flaw in the purported NTTA proposal: the discount rate used to estimate the present value of potential future revenues. Any financial analyst knows that when you try to value uncertain future flows of revenue, you need to choose a discount rate that properly reflects the degree of uncertainty. A highly likely flow of future revenues might be discounted at somewhat more than the expected long-term rate of inflation—say 5 percent. But a much less certain flow of future revenues needs a larger discount rate—7, 8, or 9 percent perhaps. A recent analysis of NTTA's March proposal by KPMG and Goldman Sachs concluded that an 8.5 percent discount rate was appropriate, given the high level of risk embedded in its traffic projections and its assumed annual toll increases over 50 years. By contrast, the CDA annual lease payments are guaranteed, so a 5 percent discount rate was judged appropriate. Comparing the two proposals, this analysis concluded that the CDA deal for SH 121 produced greater dollar value for the public sector than NTTA's deal.
Thus, Dallas-area decision-makers would be unwise to assume that the NTTA proposal would give them as much or more as the now-threatened CDA for SH 121. And looking beyond that one project, they should think very hard about the importance of keeping CDAs available for the whole raft of toll projects needed between now and 2030.
And that brings us back to the moratorium. As approved and sent to Gov. Rick Perry in HB 1892, this bill is very clearly a CDA-killer, not just a moratorium. The arbitrary 40-year limit on CDA terms would dramatically reduce any toll company's revenue over the life of the deal. That means significantly less would be available for up-front payments, annual lease payments, and revenue sharing.
And the buy-back provision alone could eliminate up-front payments. It basically tells toll road companies: “Go ahead and finance a billion-dollar project, based on 50 years worth of toll revenue, but if we change our minds 10 years into it, we won't pay you the market value of the roadway you've created. No, we'll just reimburse your construction costs plus some return on that.” Any firm that had made a major up-front payment for such a project would, in effect, have that payment expropriated. This is banana-republic stuff, unworthy of the great state of Texas. So there will be no up-front payments if this kind of buy-back provision is mandated by law.
There are also serious legal questions, already raised by the Federal Highway Administration, about subverting the SH 121 CDA procurement process. You don't change the rules after a procurement is done, and one party has won fair and square. That's what accepting an after-the-fact NTTA proposal would mean. Besides subjecting the relevant agencies to potential loss of federal funding and to litigation risks from losing competitors, this kind of decision would send a message to the private sector, loud and clear: there's high political risk in Texas. And that would send the toll road companies and untold billions in private capital to other fast-growing states that know how to play by the rules of fair competition.
Timing is everything as this drama plays out. If the NTTA proposal creates the illusion that Texas doesn't need CDAs, then the moratorium may seem low-risk, and may survive the Governor's probable veto. If that course is followed, over the next few years Texas will discover that public toll agencies can actually fund only a small number of projects. But the private sector will have walked away—to Florida, Georgia, Utah, Virginia and the other fast-growing states with sound public-private partnership laws and comparably large transportation funding shortfalls. That will make the opponents of tolling in Texas happy. But it will leave millions of motorists stuck in traffic congestion, with no relief in sight.