Hired pork doctors inflate toll road revenue forecasts
Roads to riches
Paved with bad projections
By Chuck Plunkett
Denver Post Staff Writer
As Colorado, other states and federal officials increasingly look to toll roads to spur growth or clear clogged highways, a review of 23 new turnpikes nationwide shows that a clear majority are failing to meet revenue projections to justify their costs.
Even with adjustments for the break-in period in the opening years, 86 percent of new toll roads in eight states failed to meet expectations in their first full year.
By year three, 75 percent - 15 of the 20 that have been open that long - remained poor performers.
Despite that history of flawed forecasts, Colorado officials are promoting tolling as a way to build new roads or express lanes in an era in which other funding sources for roads are shrinking.
Three companies nationally do most of the revenue projections relied upon to sell bonds to cover road construction costs. Their representatives offer several possible explanations for consistent overestimates of road popularity.
Two scenarios never mentioned are troubling to securities experts:
Cases where the consultants doing the revenue and traffic forecast either had an interest in seeing the road get built or were later awarded additional work on that road.
Cases where the road's revenue projections were used as a negotiating tool to secure favorable financing terms rather than as an impartial scientific study.
Both situations apply at metro Denver's Northwest Parkway.
The $416 million, 11-mile parkway from Broomfield to E-470 has attracted just half the cars forecast since it opened in 2003.
Its director, Aurora City Councilman Steve Hogan, said that before seeking outside investors in the road, he didn't believe the optimistic forecasts for its profit potential. But, he said, he treated those estimates as a tool to persuade bond experts to give the debt a favorable rating, not as a solid predictor.
"My personal opinion was that the numbers were probably a little high," said Hogan, who thought the projections for the critical, early years could be as much as 25 percent above the mark. He expected bond raters to trim the revenue estimates and base their ratings for investors on more realistic projections, but they did not.
In the modern world of financing toll roads, those are the rules of the game, he says. Others agree.
"Big numbers win big prizes," said Robert Bain, a London- based analyst for bond-rating agency Standard & Poor's who has conducted international studies of toll roads. "Quite often, people shop around until they find the people who provide the numbers."
But if that's the case, it's a game that former enforcement officers with the Securities and Exchange Commission say should not be played. Inflated numbers expose investors to more risk and lower returns than they should be getting.
"To believe that they're 25 percent off - I would be extremely concerned about that," said Carr Conway, a former SEC enforcement officer, who spoke after current SEC officials declined to comment. "That's just flat wrong. ... It's not a negotiation. It's not anything like that."
Though consultants say they abhor mistaken studies, a review by The Denver Post of tolling projects in eight states finds there is no incentive for the estimates to be accurate. Even when wrong, the bonds are simply refinanced and the consultants are paid again for their work on new studies to support the new bonds.
New investors, now earning a higher return, pay off the old investors and hope that enough development follows the new road to make good on the payments. Even in a worse case, such as a Texas toll road that defaulted on $75 million in bonds, taxpayers or insurers can be counted on to bail out the lenders. And the same consultants are hired again and again.
Additionally at the Northwest Parkway, a consultant on the traffic study was promised a lucrative job - paying $350 an hour - if the road was built. Another consultant on the study was promised a contract to serve as oversight engineer if the deal went through. That contract proved to be worth $5.7 million.
Similar situations have occurred elsewhere.
In South Carolina, the company hired to prepare projections for Greenville's Southern Connector also was promised a pair of contracts worth millions of dollars if the bonds sold.
In Florida, too, the state's tolling officials hired their traffic consultant to do additional work for three of the roads funded by its studies.
One of those five roads will open this year, so its performance can't yet be measured against the projections.
In each of the other four cases, actual use of the roads ranged from 34.5 percent to 67.5 percent of their estimated traffic in their first year of operation. The projections remained dramatically off in the third year, where toll collections were from 35.3 percent to 78.4 percent of the forecast amount.
Consultants said their revenue projections were never influenced by the prospect or promise of future work. Instead, they said, estimating use is a tricky business, subject to the vagaries of development, the economy, weather and even terrorist acts.
Ed Regan, who directs forecasting at South Carolina engineering firm Wilbur Smith Associates, said bias is not an issue for his company. As proof, he pointed out that not every road idea gets to market.
Five times in the past 21 years, his company was paid for comprehensive studies of proposed roads and concluded that a reasonable tolling scheme would not work.
But Wilbur Smith's president and chief operating officer acknowledges that traffic consultants are under pressure from project sponsors to supply numbers that sell bonds.
"Some of them will come directly to us and say, 'Your numbers are too low,"' said Hollis A. Walker Jr. "It is not a rare situation to have clients try to influence the numbers, and we refuse. We make our best estimates, and then we stick by them."
Still, critics bristle at a process they say lacks the oversight of traditional road-building.
"It's a no-lose situation for the state," said George Price, who opposed Greenville's Southern Connector. "If the bonds sink, the bondholders are left holding the bag, but you will have a road.
"No matter what happens, you will get a road. And that's how they're selling them across the United States."
When planning a new toll road, backers usually turn to one of just three companies to conduct the traffic projections: Wilbur Smith, URS Corp. and Vollmer Associates, which did the projections for the Northwest Parkway.
Although each company worked on a project reviewed by The Post that surpassed expectations (the only three that did), collectively the three companies missed first-year projections on 19 of the 22 operating roads.
Several tolling officials said they had little choice in picking a traffic consultant because Wall Street analysts expect experienced firms. But the analysts say they have wondered about the errors.
"The track record for startup toll roads has been spotty," said Scott Trommer, a senior director at Fitch Ratings.
So much so that investors increasingly demand greater returns on the bonds, forcing toll authorities to borrow more money to use as a hedge against roads that don't perform as expected. The extra money is set aside to cover payments in early years should the toll collections not be sufficient.
But traffic and revenue consultants say many toll roads do finally come around and produce, once the new houses and businesses they make possible move in.
Colorado's E-470 is a good example, Hogan says, adding that he expects the Northwest Parkway to emulate the road he led for several years before taking over the Northwest Parkway.
But E-470 is still lagging well behind its original projected toll revenues. Vollmer predicted in 1995 that E-470 would collect $97.9 million in 2005. The road collected $77.8 million, a miss of more than 20 percent. But the miss is bigger than it seems. Vollmer's 1995 predictions were based on an E-470 only 34 miles long. The toll road opened to its present 47 miles in the first days of 2003.
Consultants at the Northwest Parkway blame missed forecasts on the terrorist attacks of Sept. 11, 2001, which occurred after the bonds were sold that summer, and caused a drop in the expected trips between Denver's northern suburbs and Denver International Airport.
The collapse of the high- technology industry also hurt employment in the corridor, they say, and plans for a nearby housing development are just now getting underway.
But traffic at DIA is higher than it has ever been, and the tech-industry collapse had started well before Sept. 11. Yet the road has failed to come close to its projections.
The projections Vollmer gave to investors predicted that in its first full year, 2004, the Northwest Parkway would collect $6.3 million before expenses. Instead, the new road collected $4.2 million.
Vollmer said that toll plazas would collect $10.4 million in 2005. Instead, they collected $5.6 million.
Hogan predicts the authority will take in between $6.3 million and $8.5 million in 2006. Vollmer predicted $17.3 million.
Vollmer officials declined repeated requests to comment.
Analysts say that at its current rate of collections, the authority could be in danger of default when it is required to begin repaying debt in 2008. Hogan says that the money it saved in construction costs and some reserve funds could carry the authority until 2009.
"A reasonable basis"
In 1994, the SEC first warned that anyone selling municipal bonds should make sure they were giving potential investors the full picture.
"Municipal dealers must have a reasonable basis for recommending the purchase of securities," the SEC said.
Former SEC enforcement attorney David Zisser said the federal agency could easily become concerned about the failure of a tolling authority to disclose doubts about their revenue forecasts.
Hogan's admission that he doubted the numbers struck rating-agencies analysts as surprising. Analysts said they expected good-faith, rigorous estimates.
"It's interesting to actually hear an open acknowledgment of it, because when we sit here, we wonder if that's going on," said Tom Paolicelli, a senior analyst at Moody's, one of three agencies that gave the parkway strong initial ratings. "We wonder if there is an inflation because they expect us to cut (the revenue estimate) down."
Hogan, who earns in excess of $166,000 annually as the authority's director and whose signature is on the official statement, said his doubts wouldn't have mattered to investors.
"If I were buying the bonds, ... I would say, 'Well, here's an executive director whose job it is to be cautious, whose job it is to be careful ... but who has never issued bonds on his own."'
His explanation is echoed by others in the industry who say that the bonds are purchased by sophisticated investors who understand what they are getting into. And the official statements warn that the revenue projections could be in error and that the bonds are for the consideration of experienced investors.
"These are not mom-and-pop people," said Pamela Bailey- Campbell, a consultant Vollmer hired to help it prepare the traffic study.
But Colorado mutual-fund manager Chris Johns, who buys Colorado tax-exempt bonds for Kirkpatrick Pettis Smith Polian, says had he known of Hogan's doubts, he would have returned to the bargaining table.
Hogan's concern that the numbers might have been 25 percent too high means "more risk," Johns said. "Which means we would require a higher return."
Startup costs covered
One of the major differences between a traditional, taxpayer- constructed road and the public- private partnerships that create toll authorities such as the Northwest Parkway is that the contractors hired to build the toll road are required to cover many of the startup costs.
Because of this arrangement, the firm chosen to build the Northwest Parkway was required to cover the costs of hiring various consultants, including the contract with Vollmer to conduct the traffic and revenue study.
The firm, now known as Washington Group, also paid most of a $1.1 million fee engineering firm Carter & Burgess charged for an environmental-impact study.
The result is that Washington Group knew upfront that it would lose the money it spent if the bonds didn't sell and the road wasn't built.
If the bonds did sell, however, the construction company would win an immediate fee of $7.75 million, and the company would move forward on a project worth $191.6 million.
Washington Group partnered with another contractor in 2001. Its cut plus the $7.75 million would be worth $83.7 million in gross revenues.
Hogan says Washington Group's work was kept separate from Vollmer's work to prevent any undue influence on the studies that gauged the project's feasibility. But he acknowledged that the contractor, with so much to gain from the sale of the bonds, had key meetings with the consultants as the critical projections were being crafted.
A year before the bond sale, and with the blessing of Northwest Parkway officials, Washington Group representatives went to Vollmer's New York offices to get the preliminary picture on whether revenue generated by the road would be enough to persuade bondholders to invest. They returned and informed Hogan that the projections looked good.
The bonds sold. The Northwest Parkway Public Highway Authority was suddenly flush with cash. But officials discovered soon after it opened that it might not be able to repay that money. Now, because traffic has fallen so far short of expectations, some of that cash is being spent again in preparation for refinancing the bonds through sale to new investors who probably will have to accept more risk.
The authority paid $500,000 to Vollmer last year to recalibrate its traffic model in a failed start at refinancing, and another $42,465 to Bailey-Campbell's consulting firm for work on the refinancing. The authority abandoned the effort when it learned the interest costs would be too high.
The canceled bond sale cost the authority an additional $324,071 in payments to rating agencies and disclosure attorneys.
For its next attempt, the authority has agreed to pay Vollmer another $625,000 to update its traffic and revenue study. Bailey-Campbell, the rating agencies and attorneys will all be paid again.
The result could be that the authority doesn't default on its loans but that users of the road could pay tolls 10 more years beyond the expected bond payoff date of 2041, Hogan says.
The Northwest Parkway, like some others, paid to insure its bonds. It paid about $16 million to two insurance companies from the money it borrowed from investors.
So even if the authority fails to refinance and can't make payments, bondholders should get their money.
Johns, of Kirkpatrick Pettis, which bought $2 million of the Northwest Parkway bonds, said he knows what the insurance meant to him: Without it, his company would not own the bonds.
The policies, Johns says, give an extra advantage to his clients, whom he describes as mostly working people adding to their retirement savings and young couples saving for their children's college educations.
But Johns predicts that the certainty the insurance provides isn't so certain for future bond issues. The insurers may not wish to maintain or increase their coverage, he said.
"Would you insure the new bonds?" Johns said. "Here's what will happen: The existing bondholders aren't going to lose anything. ... When the new ones are issued, they will find a new group of investors who have to analyze the risk to determine whether to invest.
"It's going to be a kind of 'eyes wide open."'
Johns' point is all too clear to Hogan as the Northwest Parkway tries again to sell new bonds.
Hogan says he's thankful the toll road managed to save $20 million in construction costs it set aside as a rainy-day fund.
"If we hadn't done that, we would have been in trouble," he said. "Big, big trouble. There's no two ways around that."
Staff writer Jeffrey Leib and staff researcher Barbara Hudson contributed to this report.
Staff writer Chuck Plunkett can be reached at 303-820-1333 or email@example.com.
© 2006 Denver Post: