RATINGS:Fitch-ICBA suggests divided industry

Fitch-ICBA, the New York bond rating agency says that there is a permanent bifurcation of the toll road bond market. Established systems of toll facilities can expect to be rated in the range A to AA, whereas most standalone and startup toll facilities will be rated BB- to BBB. They see a continuing demand for new toll road financings because of what they call a “seemingly unbridgeable gap” between highway needs and the ability to finance them with tax monies that toll projects can often help to fill.

Fitch has rated $25b of US toll facility debt. It reckons the primary market for new bonds for toll projects will remain in the range $2b to $4b/year compared to 1999’s $2.5b. It expects (1) the interurban turnpikes to use bonds to finance rebuilds and in some case expansion (2) urban toll roads to extend their existing systems and add interchanges (3) refinancings (4) TIFIA-related projects (5) ‘public-private partnerships’ and (6) start-up facilities seeking refinancing at better rates.

The report “Ramp-Up at the Plaza: a New Outlook for US Toll Facility Debt” (see www.fitchibca.com) says that the re-ratings this year included $5.1b of debt upgraded including the Maine Turnpike (A+ to AA-), New Jersey Turnpike (A- to A), Harris County Toll Road (BBB+ to A), Ohio Turnpike revenue bonds (AA- to AA). Downgrades affecting $1.8b of debt included the San Joaquin Hills (BBB to BBB-) and the Santa Rosa Bay Bridge (also BBB to BBB-).

State looters

Fitch says that the ever present possibility of state governments siphoning off surplus toll revenues or leveraging them for other borrowings prevents state owned turnpikes from achieving the AAA rating. This is achieved by several overseas toll agencies less susceptible than the US ones to such political predation.

Toll bridges tend to have very inelastic demand because of the lack of substitutes, so they can make very large profits, but these are often used to offset losses on transit. The NYC Triborough Bridge and Tunnel Authority is the industry’s largest debtor and dominates the total debt of US bridge crossing agencies, Fitch notes, but much of this debt was applied to non-toll projects of its loss-making sister MTA transit agencies.

Categorizations

The report has some interesting categorizations. Financial margin per lane mile shows enormous differences. Florida, Ohio and Pennsylvania are all high with profit in the range $40k to $50k/lane-mi, while many others are much smaller and NJ is down as making losses of $25k/lane-mi. NJ Turnpike’s debt level is also the highest per lane-mi: $2.1m/lane-mi compared to most below $1m/lane-mi. Any toll authority that has recently raised money for major enhancement of its system tends to have high debt levels. Fitch says that profit margin per lane-mi indicates a strong ability to finance on a pay-as-you-go basis or to retire debt. That is associated with higher ratings. Pikes able to increase their toll rates by a decent amount tend to get upgrades in their ratings.

It is unlikely that any state owned turnpikes will ever reach AAA: “The key reason is susceptibility to political interventions. All of the turnpikes are creatures of their respective state and a national trend of diminishing autonomy for this group of authorities creates an uncertainty with respect to their financial plans and indenture requirements.” (p7) Fitch cite the governor’s discussions about abolishing tolls in Illinois as an example of political intervention which must perpetuate “uncertainty with respect to these otherwise high credit quality bonds.” Another practice which increases the risk of toll projects is new trust indentures which allow turnpike surpluses to be transferred to non-toll projects. The earlier practice in toll trust indentures was to pledge that turnpike surpluses would NOT be exported.

Other interventions disturbing to investors have been the splitting of the Massachusetts turnpike, legislative oversight of turnpike financial plans since 1981 in PA and since 1998 in OH, and the requirement that the NJ Tpk must get the approval of the governor and a senior state financial official for issuance of all bonds and any changes in toll rates. Indeed the governor can veto any turnpike board decisions, and the turnpike is required by law to transfer $20m/yr to the state. The bridge and tunnel systems are inherently highly profitable but in state control “the key risk is the temptation of policymakers to overleverage their financial strength.” (p9)

Massachusetts unwillingness to use tolls on the mainline of the Central Artery casts a shadow over the finances of the turnpike there.

Devolution

In general there has been a tendency, Fitch notes, for urban systems to be tranferred away from state and into local government control. The clear exception is Illinois where this purely Chicago-area toll system remains in state control. Fitch could have added GA-400, and the Powhite Pwy Ext and Dulles Toll Road in VA. The major examples of devolution of tolling to local authorities are MDX, THCEA, and OOCEA in Florida, NTTA and HCTRA in Texas, the Denver area’s E-470, and California.

Florida DOT’s positive support for local toll authorities contrasts sharply with other DOTs hostility to any independent toll roads and is a factor in the high rating of MDX (A), THCEA (A-) and OOCEA (A).

Fitch says that standalone toll facilities are the weakest since they are most susceptible to project risk. They are heavily reliant on future development to generate traffic, Fitch notes. True certainly of recent toll projects in S.Cal, E-470, Dulles Gway, CA-125S, PGB in Dallas etc. These are all fringe area toll roads.

COMMENT: It is possible to contemplate toll projects catering to existing development and traffic that would be more in the nature of congestion-relievers rather than the fringe area projects traditionally given to tolling in the US. 407-ETR in Toronto, Melbourne CityLink, Dublin’s new toll facilities, and the Birmingham North Relief Road (UK) are such projects outside the US. In the US very solid standalone projects based on better serving existing traffic would be: a new Beltway Wilson Bridge VA-MD, an Oakland Airport to San Francisco Airport midbay bridge-tunnel, a truck toll tunnel under the Hudson River from the NJ Turnpike to Manhattan and Kennedy airport, a Dulles Airport/I-270 at Gaithersburg MD toll road with a new Potomac River crossing, city of San Francisco toll tunnelways (TRnl#48 Apr 00, p1), I-80 in PA (OH-NJ), the northern leg of the Mon-Fayette and the Airport Toll road in Pittsburgh, a rebuilt I-5 right through Los Angeles, new ‘U-way’ toll links in Chicago (TRnl#45, Jan 00 p22), Miami’s planned Central Parkway (TRnl#41 July 99 p1) etc.

Defaults

Fitch says the toll sector has a substantial history of defaults and mentions: Pasco County FL road and bridge bonds, McKinley Bridge bonds St Louis, city of Venice IL (?), Veterans Memorial bridge East St Louis MO, Chicago Skyway (1959-1976 late interest on 1955 and 1957 bonds, courts ordered toll increases 19 74 to 1993), Chesapeake Bay Bridge-Tunnel (1970-1985 late with interest on 1960 bonds, never fully paid), Texas Turnpike Auth 1978 bonds on Houston Ship Channel bridge, West Virginia Turnpike (1958-1979 eventually fully paid), TRIP II Dulles Greenway (1996-1998 no debt service to initial insurance company lenders now 90% paid off).

Most of these defaults occured because of unforseen economic setbacks. The Houston shipping channel bridge opened at a time when the city was utterly dependent on oil, and oil was in heavy surplus and prices had plummeted, devastating the local economy and causing stagnation in traffic levels.

The Dulles Greenway crashed in its early days in part because of a long recession in real estate and a drop away of development in the Gway corridor. Fringe area development is extraordinarily difficult to forecast.

Chicago’s Skyway failed to get traffic because of the collapse in business at steel plants of far northwest Indiana which it served. WV Turnpike was built on the assumption that free interstate connections would be made either end, but they were heavily delayed, so for a long time it was an isolated highway in the mountains. The Chesapeake Bay Bridge-Tunnel at the mouth of Chesapeake Bay was financed on optimistic theories about a new coastal route between the Hampton Roads area and Delaware taking traffic from I-64/I-95, when it is almost entirely a beach vacationers facility.

Fitch says it expects a default rate of 3% to 18% in BB- and BBB debt over the next 10 years. And it projects standalone toll facilities to grow dramatically as a proportion of total toll debt service – from the present 10% to as much as 50% by 2030.

Fitch’s terminology is a bit odd in places. All the categories are intra-state except the crossings to Mexico and Canada. Maybe ‘Inter-urban’ or ‘Statewide’ would distinguish the large turnpikes from the smaller ones. Georgia is miscategorized. It’s a pure standalone Atlanta urban area toll road, by no stretch of the imagination statewide. Rhode Island’s bridges are rural not urban. SJTA’s Atlantic City Exwy we’d argue is really a standalone toll facility not a state turnpike. Anyway, realworld entities rarely fall into neat categories.

Quite a number of important toll agencies are not covered by Fitch – NYS Thruway, PANYNJ, DRPA, NYSBA, MdTA, NJHA, Indiana, Delaware Turnpike, DRBA, DRJBC, NH, Caltrans, to name a few. Biggest bond opportunities are overlooked by Fitch: (1) privatization of state and local pikes, and (2) toll franchises on the interstates!