Williams Turns a Record Quarter Into a Financing Puzzle

Record Q3 EBITDA and a 16% internal growth target collide with $27.9B net debt, negative free cash flow, and a power backlog still waiting on signatures

WMB · Earnings Call · 2026-11-03

The quarter itself was not the problem. Williams delivered $2.04 billion of Q3 adjusted EBITDA, up 6%, a record for a third quarter, and adjusted EPS and AFFO both rose 10%. Management now expects 2026 EBITDA in the upper half of the $8.3 billion-$8.5 billion range. Chad Zamarin called it “another strong quarter for Williams” — Chad J. Zamarin, President and Chief Executive Officer · 2026-11-03 and ticked off the Momentum Midstream close, four expansion projects in service, and the final Socrates phase. The stock, however, is not celebrating: WMB sits about 8.7% below its May peak and has returned -3.5% over the last 90 trading days, a flat tape against a very busy project narrative. The tension is capital. Capex was $1.8B in the latest quarter, up 89% year over year, while free cash flow was -$475M and effective net cash was -$27.9B. Quarterly capex of $1.8B, up 89% year over year is the physical expression of the growth story, but free cash flow of -$475M shows the near-term cost. Effective net cash of -$27.9B and interest coverage of 3.2x explain why the second half of the call was less about record EBITDA and more about how to pay for the next wave.

Power Innovation: a real backlog, but the announcement clock keeps resetting

Power Innovation is no longer a new theme for Williams. It has led the company's keyword list for several quarters: power project, speed to power, and invest capital are now core vocabulary. What is new is the language around scale and tenor. Chad said the next projects should broaden the customer base and extend contract length, and Julien Dumoulin-Smith pressed on whether future announcements would show larger phases. Chad's answer: “you will see projects from us that ... crystallize, commercialize, a larger portion of the backlog and show larger scale phases” — Chad J. Zamarin, President and Chief Executive Officer · 2026-11-03. That matters because Williams has spent multiple calls saying the 6-gigawatt opportunity set was an order of magnitude, not a contract. In May, Chad cautioned: “I wouldn't try to focus on precision with the 6-gigawatt” — Chad Zamarin, President and CEO · 2026-05-05. The market has heard that for a while. The fresh disclosure is the equipment commitment. Spiro Dounis flagged $5.9 billion of long-lead equipment purchases for projects not yet sanctioned. John Porter's response was unusually candid: “I do not think there is a real strong connection between those 2” — John D. Porter, Chief Financial Officer · 2026-11-03 — the $5.9B and the 6GW-plus backlog. That is not a red flag by itself; Williams says it has backstop agreements and cancellation fees rather than absolute commitments. But it does mean the balance sheet is carrying option value on projects that have not yet been signed. The company still expects one to two more power projects by year-end. It said the same in August: “We expect to commercialize additional projects between now and the end of the year” — Chad Zamarin, President and CEO · 2026-08-04. Repeating the target is fine; repeating it two quarters running is the reason to watch the calendar rather than the headline.

The 16% target and the 4x leash

The most consequential new item was buried in a question about compensation. The 10-Q appears to point to a 10%-16% EBITDA CAGR through 2030, with the low end congruent with the 11%+ target management raised last quarter. John confirmed the structure includes a hard leash: “there is no payout if our balance sheet is over 4x” — John D. Porter, Chief Financial Officer · 2026-11-03. That is a strong alignment statement, but it also frames the problem. Williams expects year-end leverage around 3.75x, leaving roughly $2 billion of capacity versus its internal 4x ceiling. The next power projects, the SESI II expansion, and the Momentum integration all compete for that capacity.

On leverage, excluding the impacts of any additional power innovation projects that we announce prior to year end, we would expect year end leverage to be around 3.75x, reflecting the Momentum acquisition on a full year run rate basis. And as we previously discussed, 3.75x provides another $2 billion of incremental capacity versus our internal 4x leverage ceiling to cover the initial spend on any additional power projects we announced before year end.
John D. Porter, Chief Financial Officer · 2026-11-03

This is not a new concern. In August, John called leverage “really just a '26 and '27 issue” — John Porter, Chief Financial Officer · 2026-08-04, with 2028 earnings growth resetting capacity. In May, he described leverage as “temporarily move modestly above our long-term target range” — John Porter, Chief Financial Officer · 2026-05-05. The difference now is that the 16% upside case makes the capital appetite bigger while the 4x gate makes the funding mix less forgiving. The Blackstone JV at 6.35% was a proof point; John noted on this call that market pricing has moved with Treasuries, so a repeat is an option, not a fixed solution.

Storage, LNG, and the permitting overhang

Away from power, Williams is leaning into gas storage and Mountain West. The 9.4 Bcf Clay Basin expansion — Clay Basin — is small in dollars but strategically large: it is the Rockies' largest storage asset and a gateway between production and western markets. Management expects more storage announcements in coming quarters. LNG demand remains the long-cycle pull: LNG demand and the Haynesville position are why the Momentum deal matters, and why Line 200 and the Woodside capacity keep coming up. The offset is NESE. The Third Circuit remanded the New Jersey certificate, and General Counsel Lane Wilson said “there is nothing in that remand that is really jeopardized the project” — Judge Terence Lane Wilson, General Counsel · 2026-11-03. Chad added that it should not affect financial forecasting but could delay in-service for next winter and hurt affordability in New York. The call used NESE to argue for permitting reform, with a Senate cloture vote scheduled for November 9. That is a political catalyst the stock does not price quarter to quarter, but it sits underneath the entire Transco footprint expansion story. The market's own tape is a contrast. Global 90-day decliners include AI data center and data center capacity names, down sharply, even as Williams keeps selling speed to power. That divergence is the opportunity and the risk: either Williams is monetizing a real bottleneck while the AI trade cools around it, or the market is starting to discount the pace of those projects before the contracts are signed. The record quarter earns management the benefit of the doubt. The next two announcements, the 2027 financing plan, and the 4x gate will decide whether the stock re-rates from here.