Williams Companies and the Toll Road at a Full Price
What the business is
A good toll road can remain a good toll road even when too many investors arrive at the same destination.
Williams Companies is a natural-gas midstream operator built around interstate pipelines and gas-processing infrastructure. It handles roughly one-third of US natural gas, about 33 Bcf/d. Its central asset is Transco, a corridor running from the Gulf Coast to New York that transports about 15% of the gas consumed in the country.
The economics are less dependent on guessing commodity prices than the word “gas” might suggest. Most revenue comes from long-term, fee-based contracts and cost-of-service rates approved by FERC. Williams is principally selling transportation capacity rather than taking a direct wager on the price of the molecule passing through its pipes.
The moat rests on efficient scale and hard-to-replicate intangible assets: regulatory approvals and established rights-of-way through densely populated corridors. Building a competing interstate route faces cost, permitting and local-opposition barriers. FERC-regulated rates also allow a return on invested capital and can pass inflation into rates, although this is bounded pricing power rather than an invitation to charge whatever the traffic will bear.
The internal thesis considers the moat to be widening because permitting conditions for new competing pipelines are deteriorating, while expansions along existing corridors retain a cost advantage over new routes. That claim is separate from recent operating growth: better results do not by themselves prove a stronger competitive position. The moat case depends specifically on scarce corridors, approvals and contracted capacity.
Williams is now reaching beyond conventional pipeline expansion into data-centre power demand. The programme includes the 682 MW Neo project and a $5.34 billion power joint venture with Blackstone. This could deepen the role of its existing gas network, but it also changes the capital-allocation question. A regulated toll road is fairly understandable; a toll road embarking on a very large construction programme deserves a second clipboard.
What the numbers say
The required valuation panel is straightforward:
- Market cap (price-adjusted, 2026-10-01 close): $85.00 billion
- Owner-earnings yield (price-adjusted): 0.08%
- Share price (2026-10-01 close): $69.22
- 10-year Treasury: 5.24%
On those screening figures, the 0.08% owner-earnings yield sits 5.16 percentage points below the 5.24% Treasury. This is not a close call disguised by decimal places.
There is an important data-quality wrinkle. The annual screening calculation reports $72 million of owner earnings, but the thesis says that figure deducts growth capital expenditure in full. The source material does not identify the accounting period behind that annual figure or provide a complete line-by-line reconciliation, so it is not fully reproducible from the supplied record.
The thesis instead normalises earning power from $2.79 billion of net income, about $2.2 billion of depreciation and amortisation, and approximately $0.9 billion of maintenance capital expenditure. It adopts owner earnings of roughly $4.0–$4.5 billion, cross-checked against $5.86 billion of 2025 AFFO and $5.9 billion of operating cash flow. That is economically more sensible for a pipeline business, but it does not create a margin of safety at the current price under the supplied valuation range.
That range is $48–$68 billion. The price-adjusted market capitalisation stands 77% above the low end and 25% above the high end. Even the optimistic case is below the market’s current appraisal.
Capital allocation has improved since the failed Energy Transfer transaction and the 2016 dividend reduction from $0.64 to $0.20 per quarter. The later record includes a roughly 4.0 times leverage target, 5% dividend CAGR, a conditionally opportunistic $1.5 billion repurchase programme announced in 2021, and 2025 AFFO dividend coverage of 2.40 times.
The concern is the next chapter. Planned growth capital expenditure increased from $2.6–$2.9 billion in 2025 to $7.0–$7.6 billion in 2026 under CEO Chad Zamarin, who took over in July 2025. The Blackstone venture shares some of the burden, but the central test remains whether projects are contracted before construction and whether leverage stays near its stated target.
What Wall Street believes vs what I see
The available headlines lean constructive. They frame Williams as a beneficiary of an AI-related power buildout, a formidable competitor in the Northeast and a vehicle for a new natural-gas growth cycle. Another headline emphasises superior dividend growth. This is not a formal live consensus, but the public mood represented here plainly prefers the growth story.
I see a strong asset accompanied by ambitious spending and a valuation that already asks for considerable success. Data-centre electricity demand may increase gas demand, and existing corridors may capture part of it. Neither proposition establishes that every new dollar of capital will earn an attractive return for shareholders.
The owner-earnings arithmetic is also less flattering than the headlines. Under the prescribed screen, a 0.08% owner-earnings yield against a 5.24% Treasury fails even the modest requirement that the business yield more than the bond. Under the thesis’s normalised earning-power approach, the intrinsic-value range still tops out at $68 billion, below the price-adjusted market capitalisation of $85.00 billion.
The useful monitoring list is therefore practical rather than thematic: the proportion of growth projects sold under contract before construction, quarterly adherence to the 4.0 times leverage target, AFFO dividend coverage relative to 2.0 times, the number and credit quality of data-centre power contracts, and the price at which repurchases resume. Narratives travel quickly. Contracted cash flows tend to use the pipeline.
Verdict
Verdict: pass.
The price-adjusted owner-earnings yield of 0.08% is below the 5.24% Treasury, so it does not satisfy the rubric’s watch threshold. The $85.00 billion market capitalisation also stands 77% above the $48 billion low end of intrinsic value and 25% above the $68 billion high end, leaving no qualifying margin of safety. I would change the verdict if the owner-earnings yield moved above the 5.24% Treasury, or if price and demonstrated earning power produced a genuine discount to the intrinsic-value range.
This analysis is AI-generated, educational, and not investment advice. Figures may contain errors or be delayed. Disclaimer