Why Texas toll roads stay tolled after construction

Construction-cost recovery, a road’s original bond, the wider system’s debt, public charging authority and a private concession’s remaining term answer different questions. A mature road can pass one test and remain subject to another. Applying each test on its own prevents an accounting fact from becoming a legal conclusion—or a legal power from becoming a claim that the current policy is wise.

Texas has no single statewide toll-dollar ledger. NTTA closes its year on December 31; Harris County, TxDOT, CTRMA, and Fort Bend use different periods and structures. Accrual expenses do not necessarily describe that year’s cash payments, and pledged revenue may include investment and non-toll income.

The normalized chart therefore compares only the Central Texas Turnpike System and Grand Parkway. Both rows use the same TxDOT FY2025 report, operating-revenue denominator, and category schedule.

For CTTS, each $100 of FY2025 operating revenue supported $30.17 of operating expense, $19.29 of repairs and maintenance, $47.49 of debt service, and $3.05 of required reserve. For Grand Parkway, the corresponding figures were $15.03, $4.82, $44.59, and $35.56.

The common basis permits a comparison of the two reported allocations. Judging which system provides better service at lower lifecycle cost requires more evidence. A required reserve still claims cash today, even though it may fund later work rather than current expense. Removing it from the comparison would make a financing commitment look like an unexplained profit.

The FY2023–FY2025 extension follows NTTA’s own-road System, CTRMA’s combined activities, Harris County’s toll fund and TxDOT’s Central Texas Turnpike System. These are financial boundaries, not all the roads on a TollTag or EZ TAG bill. SH 130 segments 1–4 belong to CTTS; the privately financed segments 5–6 have their own concession.

NTTA’s System toll revenue rose from $1.131bn to $1.253bn. Its FY2025 report-wide total was $1.300bn because the Non-System tolling-service fund adds another $47.1m. Assigning every road on the bill to the collector would count the wrong enterprise.

HCTRA’s FY2025 cash-flow statement closes the funding loop. Operating activities generated $593.4m. Construction used $315.3m, bond principal and interest $242.0m, and county transfers $398.6m. Net commercial-paper borrowing added $92.34m; other capital financing used $7.04m and net investing used $31.78m.

The result was a $308.99m reduction in cash and cash equivalents, from $1.3047bn to $995.75m. All four cash-flow sections reconcile. The cash includes restricted balances, so the ending total is not a measure of unrestricted spending capacity. Audited cash flows, PDF p24 ↗

It also shows why “principal paid” and “debt reduced” differ. HCTRA retired $96.685m of bond principal while commercial paper rose $92.340m. Those combined principal balances fell just $4.345m. The larger $33.393m decline in financing carrying amount also includes bond-premium amortization.

Across FY2023–FY2025, Harris County’s toll-road fund transferred $961.3 million to other county funds, including $398.6 million in FY2025. Note 10 identifies transportation facilities outside the toll system as the purpose. Toll users are therefore helping finance a wider county network. FY2025 audit, Note 10.

I would judge those transfers by the commitments toll users should finance and the cash the system can release after meeting its own obligations. Retiring the original construction bond would settle only part of that decision.

The comparison uses each entity’s fiscal year and accounting basis. Blank fields identify categories the reviewed records cannot support on a comparable basis; they do not mean zero.

HCTRA’s reported toll revenue rose $153.1m between FY2024 and FY2025. Its road components contributed $52.3m; interoperability contributed $50.7m; administration contributed $50.1m. The three movements reconcile exactly. The road subtotal grew 6.0%, compared with 17.5% for the headline figure.

CTTS has another accounting discontinuity. Its FY2025 $311.477m net toll revenue includes roughly $272.3m gross revenue plus a $39.2m decrease in the doubtful-account allowance. Its 5.3% net increase is therefore not a traffic-growth estimate.

A construction-recovery claim first needs a declared cost base: right-of-way, capitalized interest, grants, financing fees, electronic collection and later widening or rehabilitation. Gross tolls cannot simply be added until they exceed one construction invoice. The historical return must account for what it cost to keep the road operating while that revenue was collected.

There is also a reason a fully funded road might continue charging a price. FHWA’s congestion-pricing primer describes charges that shift some trips away from the busiest periods and help preserve a reliable travel option. That is an operating purpose, distinct from recovering construction cost. It supplies no presumption that a particular Texas toll is optimal; the corridor’s traffic response, alternatives and users still need to be assessed. Traffic-management rationale.

The converse deserves equal weight. A road can need future maintenance without needing the existing toll schedule to finance it. Taxes, a different user charge or a changed capital program might produce another allocation of cost and benefit. Those alternatives require a funded comparison. The decision must address historical cost recovery, present contractual restrictions, and the merits of the next funding policy.

A road can have no remaining standalone bond while its revenue still supports a pooled system. Refunding bonds replace prior debt to change rates, maturities, or creditor terms; they do not represent another construction project.

NTTA’s FY2025 System reported $1.253bn of net toll revenue and $8.854bn of financing debt carrying amount, including current portions and unamortized premium. Principal and accreted interest before premium were $8.411bn. The HB803 dashboard assigns $675.7m to debt service and makes its projected 2048 payoff conditional on no further system financing.

These are Trust-wide figures. Pledged revenue funds system operations and required debt payments before later uses. Coverage tests require more revenue than bare debt service. The Trust Agreement also governs additional borrowing, reserve releases, and subordinate obligations.

The December 2025 monthly report shows $1.3575 billion of gross Trust revenue, $252.7 million of net operating expense, and $666.9 million of preliminary cash debt service. The arithmetic remainder is $437.9 million.

The $437.9 million is a calculated remainder, with no named destination in this schedule. Treating it as unrestricted cash, profit, a reserve deposit, or a capital allocation would exceed the record. Later uses remain subject to the Trust.

NTTA received about $630.5 million of refunding-bond proceeds in FY2025. These change creditor claims, maturities, and interest cost without creating another construction project or an automatic toll-removal date.

A pooled pledge can make a new link financeable by using the reliable traffic of older roads. It can also keep a mature road’s users paying for projects elsewhere in the network. A toll-removal proposal should identify the governing release and coverage restrictions, the remaining lifecycle work and the replacement revenue. Invoking “system debt” without tracing those commitments is as incomplete as invoking the age of the road without them.

HCTRA transferred $369.3m in FY2023, $193.4m in FY2024 and $398.6m in FY2025: $961.3m in three years. Note 10 identifies the FY2025 total as surplus-revenue transfers, including the annual mobility transfer, for roads and related facilities outside the toll system. It is neither a private shareholder withdrawal nor solely one annual mobility payment.

The county still has system debt, construction and maintenance commitments. Its revenue-bond covenant requires at least 1.25 times coverage under the governing calculation. A rate-reduction proposal must therefore identify which commitments can change and what replaces any lost revenue.

The strongest case for the county transfer is network investment: toll revenue can fund connections that improve access to the tolled system and substitute for tax-funded transportation work. OMB’s August 2023 record states that rationale and describes repair and liquidity priorities. It describes the adopted framework; completed benefits and later compliance need their own evidence. County policy record.

The distributional question remains. A toll user finances part of a wider network, while a traveler elsewhere may benefit without paying this charge. That can be defensible if the connections and funding substitution justify it. The audit alone cannot decide how broadly the costs should be shared. It gives the public a specific transfer to examine.

TxDOT’s FY2025 report places excess revenue from its debt-financed systems in project reserve accounts required by bond documents. On that schedule, CTTS and Grand Parkway had no surplus available for an unrelated transfer.

Rates, coverage and reserves make a promise to lenders that precedes a later rate debate. That constraint can support borrowing. It also narrows future choices. To assess the price of flexibility, identify the covenant, its remaining life and the cost of releasing or replacing the pledge. A reserve balance is neither a free-spending account nor evidence that every dollar must stay committed forever.

Capital spending also continues after opening. TxDOT’s FY2026-FY2030 plans include rehabilitation, widening, interchanges, technology, and safety work.

CTRMA’s FY2025 audit reports $276.4 million of toll revenue, about $2.448 billion of debt principal, and 2.28 times debt-service coverage. Its June 2025 statements add detail, but their categories cannot be reconciled with the audit without assumptions. CTRMA therefore stays outside the normalized chart.

Fort Bend County has two toll authorities. FBCTRA reports $68.54 million of FY2025 toll revenue and $526.91 million of bond principal. Its $43.0 million total expense includes noncash and financing items and is not treated as cash O&M.

FBGPTRA has its own FY2025 report and financing structure. Combining the authorities, or substituting countywide figures, would obscure those differences.

North Tarrant Express, LBJ Express, and NTE 35W occupy public corridors under long-term private concessions. Their project companies finance and operate dynamically priced lanes, service debt, and can distribute cash once contractual and financing tests are met.

Ferrovial’s 2026 Factbook reported, at the 100 percent project-company level for the three Texas concessions, $935 million of FY2025 revenue, $776 million of adjusted EBITDA, and $553 million of project-company distributions. The same filing reported $318 million as Ferrovial’s ownership-adjusted share of distributions.

The project-company totals and Ferrovial’s share use different denominators. Neither measures payments to TxDOT.

The filing also reported $46.7 million of FY2025 payments to TxDOT: $26.4 million of NTE 35W revenue sharing, $8.1 million from NTE, a $6.6 million NTE 35W refinancing-gain payment, and $5.6 million of LBJ Wishbone revenue sharing.

These concessions can distribute cash to private investors after the senior claims have been met. That residual is the return their equity was committed to earn, and it can be lost when traffic, costs or the entry price disappoint. The public question is what risk and service the contract purchased in exchange. High distributions leave return adequacy and public value unresolved.

The annual totals are not a reconstructed source-of-funds bridge for those distributions. Do not divide dividends by current-year toll receipts and call the result the share of each driver’s payment sent to investors: earlier cash, reserves and financing movements require reconciliation. Nor is adjusted EBITDA a distribution. The sponsor filing reports those measures; a complete cash waterfall needs the project accounts and financing restrictions.

The 2016 comprehensive development agreement gave Blueridge Transportation Group a long-term package of financing, construction, operation, maintenance, and toll-revenue rights. TxDOT held public title.

In 2024, the Texas Transportation Commission used the agreement’s termination-for-convenience mechanism. The stated payment was $1,731,730,721, less permitted adjustments. Control transitioned to the public structure on October 8, 2024.

Texas later reduced rates and continued tolled managed lanes. Ending the private contract did not end public tolling authority.

The buyout paid to terminate a package of future contractual rights. The federal financing record says the proceeds repaid outstanding debt first, with the remainder available to shareholders. Calling the gross payment an unpaid construction balance would erase both the debt priority and the value of the remaining concession. Federal repayment record.

Public control now permits a different rate choice while bringing the future operating and financing case under public responsibility. Lower tolls can benefit users and reduce receipts at the same time. The comparison needs the buyout financing, maintenance and capital obligations, expected traffic response and the value of the changed service. Replacing a private claimant with a public one changes the bargain; it does not remove its costs.

Transportation Code §284.008 provides a default transition for certain county toll projects. Subject to its exception, a project joins the state highway system and is maintained without tolls once project-revenue bonds and interest are paid, or sufficient money is held in trust to pay them through maturity.

Under subsection (d), a county may request a Texas Transportation Commission order preventing that transition. A February 25, 2016 agenda records Harris County’s request. It does not establish the final order’s full scope, so this article assigns the exception to no named road without the order.

Other structures use different provisions. Section 366.407 addresses surplus revenue for regional tollway authorities. Section 370.174 addresses regional mobility authorities. Section 228.006 addresses surplus revenue from TxDOT toll projects. Bond documents and contracts may further constrain each statute.

The mandatory county-law transition and a discretionary proposal under another structure have different governing conditions. For the latter, the following questions identify commitments and funding to resolve; they are not six additional statutory conditions inserted into §284.008. The governing law and actual project documents determine which apply.

Authority depends on the structure. The NTTA board sets rates subject to Trust covenants. Harris County government and bond documents govern HCTRA policy. CTRMA’s board acts under Chapter 370. CTTS, Grand Parkway, and SH 288 fall under the Texas Transportation Commission and related public financing entities.

A concessionaire can price within its contract’s formula, caps, performance standards, and public controls. Political promises and press releases do not replace those covenants.

A proposed reduction needs a forward corridor budget: resulting trips and revenue, operating work, capital payments, debt coverage and the alternative source of funds. Holding traffic fixed is one sensitivity. Pricing can also change route choice, congestion and reliability. The financial model should therefore be paired with the effect on users and adjacent routes, with neither a pure social gain nor a pure funding loss assumed.

The October 21, 2026 extension reconstructs FY2023–FY2025 accounts for four public reporting scopes from audited reports. It adds 231 page-linked observations, 53 calculation expressions, a 31-row road/collection crosswalk and 111 component-year revenue records. The earlier Grand Parkway, Fort Bend and SH 288 material uses its cited evidence and original cutoff.

The revised CTRMA FY2025 report contains differences worth preserving: its basic statement places the $10m CAMPO contribution outside operating expense; the direct cash-flow total differs from its reconciliation by $125,190; and HB803 debt service differs from the cash-obligation/interest sum by $41.737m. MoPac allocations and $1 component differences also remain visible. No disputed amount is silently harmonized.

HCTRA’s FY2023 bond-note total includes $11.684m accrued interest. The comparable financing measure removes it before adding commercial paper. NTTA’s System/report totals and CTRMA’s System/MoPac scopes overlap; they must not be added together.

The remaining acquisition questions are specific: SH 130 annual accounts, settlement ledgers, current funding/equity schedules and speed-payment election; HCTRA’s project-specific §284.008(d) orders and transfer-recipient ledger; CTRMA’s reconciliation; and the earlier FBGPTRA page-level gap.

The phrase “paid off” compresses several legitimate questions into one. Drivers can reasonably ask when the promised financing burden ends. Lenders can enforce claims the public system has already granted. Governments can choose to fund a wider network or preserve a reliable peak-period route. None of those interests settles the others.

My reading of these accounts is that toll persistence is increasingly a decision about the next use of a mature revenue stream. Some commitments are already binding; others deserve a fresh public justification. I’d demand a dated explanation of both: what remains pledged, what future work is proposed, and what would change under a lower or zero toll.

That makes the policy accountable. A financial history can explain why the original bond’s retirement leaves the toll in place. The continuing charge earns its justification through the service and obligations it finances now, with the costs of alternatives visible.

The public package contains the analysis, page-marked source text, calculation scripts and data. The primary-source ledger below links the agency reports and contracts.

The financial sources below link the basic statements and notes. The download also identifies management discussion, HB803 schedules, original URLs, and file hashes.