The Funding Question Answered, Then Spent: A 6.35% Partner, a $5.5B Bolt-On, and a Growth Target Two Points Higher

Results vs. Consensus

Williams reported after the close on Monday, Aug. 3, and held its call the following morning at 9:30 a.m. ET. The release carried three separate pieces of news, and the smallest of them was the quarter. Second-quarter results were unremarkable and slightly soft. The Blackstone joint venture, signed in July and disclosed here in full, resolves the single largest open question in the investment case. The Momentum Midstream acquisition, announced in the same release, opens a new one.

The adjusted EPS panels disagreed. One desk carried $0.50 into the print and called the result in line; two carried $0.52 and recorded a 3.8% miss, with one noting that its $0.52 had already been cut 0.7% over the preceding 30 days. The honest characterization is that Williams landed at the low end of a narrow band, matching the most recently revised estimate and falling about 4% short of the broader average. The revenue spread is wider and is a modelling artefact rather than a disagreement about the business: reported total revenues contain a mark-to-market derivative line that swung from a $359M loss in the first quarter to a $94M gain in this one, and panels that model revenue gross of that mark land near $3.08B while those that model it net land near $2.83B. No published consensus for total adjusted EBITDA exists for this quarter, because no estimate was published for the marketing segment, so the $1,948M figure above is the sum of the four segment estimates that were published and is labelled as such wherever it appears.

Income Statement, Year Over Year

Sequentially, vs. 1Q 2026

Of the $333M sequential decline in adjusted EBITDA, $228M is the marketing segment giving back the winter. That was the central claim of our first-quarter note, and it is now settled fact rather than forecast: Gas & NGL Marketing Services earned $227M in the March quarter and lost $1M in the June quarter. The three infrastructure segments together gave back $86M sequentially, which is ordinary seasonality in transmission and a genuine soft patch in West.

Revenue

The first-quarter note argued that the revenue line at Williams is a poor instrument for measuring the business, and this quarter proves the point from the other direction. Three months ago a $359M derivative loss inside the revenue subtotal produced a reported 9% shortfall against the Street on a quarter that was operationally fine. This quarter a $94M derivative gain sits in the same place and helps the optics on a quarter that was operationally softer. Neither figure is revenue in any economic sense, and management removes both from adjusted EBITDA. Judged on fee streams, growth slowed: service revenues of $2,152M grew 5.4% against 10.1% in the first quarter, a deceleration that is fully explained by the Transco rate settlement lapping into the base. Product sales of $762M grew 16.0% on higher NGL production in West and the Northeast, and pass most of that through to the cost line.

Margins and operating leverage

Operating income rose 25.1% on revenue up 9.8%, and unlike the first quarter the leverage is mostly real. Total costs and expenses rose only 1.9%. Depreciation fell to $592M from $605M and product costs rose 7.4% against product sales up 16.0%, so gross commodity margins widened. The $12M gain on asset sales sitting inside the costs subtotal is immaterial this quarter, against $182M in the first quarter. What is not clean is the operating expense line. Operating and maintenance expenses rose 4.4% to $597M and general and administrative rose 7.1% to $180M, the latter faster than adjusted EBITDA at 6.2%, and the concentration is in West, where segment operating and administrative costs rose 10.7% year over year and 11.4% sequentially. That is where the West segment's shortfall against published estimates originates, since every published volume statistic for the segment came in ahead.

EPS and below the line

Interest expense of $371M rose 6.0% against adjusted EBITDA up 6.2%, so the relationship that we flagged as inverting in 2027 has not yet turned. It is close. First-half interest expense of $747M annualizes to $1,494M, and management re-cut the full-year guidance assumption to $1,535M from $1,485M, which implies $394M per quarter in the second half. Our first-quarter model called for $1,530M to $1,570M against a company assumption of $1,485M; the company moved inside our range. The more consequential below-the-line item is noncontrolling interests, which the guidance bridge raised to $220M from $180M. That is the first visible cost of the joint venture, and it is a line that grows every time a power project enters service, because consolidated adjusted EBITDA carries 100% of those projects while Williams owns 51%.

Segment Performance

The composition inverted from the first quarter in a way that is more reassuring than the headline suggests. Three months ago the marketing segment supplied 69% of the variance against published estimates and the infrastructure segments were in line. This quarter the marketing segment contributed nothing, and the misses came from the other two infrastructure segments while the Northeast beat. Growth of 6.2% with the weather-levered line at zero is a cleaner result than growth of 13.3% with the weather-levered line at a record, even though the second number is bigger.

Transmission, Power & Gulf

The largest segment grew 6.2% to $959M and missed the published estimate by 2.5%. The growth decomposition matters more than the number. Regulated interstate transportation, storage and other revenues rose 2.8% to $917M, a sharp slowdown from the first quarter because the Transco rate case settlement that lifted 2025 comparisons has now lapped. The growth came from elsewhere: gathering, processing, storage and transportation revenues rose 16.5% to $254M, and management sized the Gulf businesses at 23% growth and natural gas storage at 23% growth. Operating and administrative costs rose 4.9% to $300M. Sequentially the segment fell 5.0%, which is normal seasonality in transmission plus a $20M decline in other fee revenues.

"Our overall financial performance continues to be led by our Transmission and Gulf businesses, which improved $56 million or about 6%. Growth in this segment was led by our Gulf businesses, which grew 23%, reflecting the combined effects of our recent Gulf expansion projects."
— John Porter, Chief Financial Officer

Assessment: this is the segment carrying 91% of the quarter's capital expenditure and growing at 6%. That is the correct shape for an infrastructure builder three years into a program whose in-service dates cluster in 2027 through 2029, but it does mean the reported growth rate of the largest segment will stay unexciting until Southeast Supply Enhancement and the power projects land. The estimate miss is not a quality problem; it is the rate-case tailwind annualizing on schedule.

Northeast G&P

The best relative quarter in the portfolio. Adjusted EBITDA rose 7.8% to $540M and beat the published estimate by 4.2%, the only segment to do so materially. The driver is a mix shift into rich gas that the volume table makes visible. Consolidated gathering volumes were 4.16 Bcf/d against 4.15 a year ago, essentially flat, while plant inlet volumes rose 7.9% to 2.04 Bcf/d and NGL production rose 20.3% to 166 Mbbls/d. Proportional EBITDA from equity-method investments rose 9.1% to $168M on Blue Racer Midstream and Bradford within Appalachia Midstream. Operating and administrative costs fell 3.5% to $109M.

Assessment: our first-quarter note flagged an unsized 8.7% year-over-year decline in consolidated Northeast gathering volumes as the one operating datum nobody on that call addressed. It has stabilized. Volumes are flat year over year, sequentially up 3.7% from 4.01 Bcf/d, and the segment is earning more on the same gas because the barrel content is richer. That is a better outcome than a volume recovery, because it does not require the dry-gas rig count to move.

West

The weakest segment relative to expectations. Adjusted EBITDA rose 5.3% to $359M and missed the published estimate by 7.8%, the largest variance in the portfolio, and fell 12.4% sequentially from $410M. Volumes were not the problem: gathering volumes of 6.03 Bcf/d rose 1.5% and beat the published estimate of 6 Bcf/d, NGL production rose 15.7% to 118 Mbbls/d, and NGL equity sales of 14 Mbbls/d nearly doubled a published estimate of 7.30. The problem is on the cost line and in minimum volume commitments. Net gathering, processing, transportation, storage and fractionation revenues rose 6.6% to $454M, while operating and administrative costs rose 10.7% to $166M. The company sized the Eagle Ford drag at a $9M revenue decline in the quarter and $19M in the half, attributed to contractual step-downs in minimum volume commitments and lower gathering volumes.

Assessment: a $16M year-over-year increase in segment operating costs against an $18M increase in segment EBITDA is the least attractive relationship in the release, and neither the release commentary nor the call addressed it. The Eagle Ford step-downs are contractual and known, and they will keep grinding through 2026. What is new is that the segment is now spending very little capital, at $56M in the quarter against $274M a year ago, while carrying a rising cost base. West is being harvested to fund the Gulf Coast and power programs, which is the right allocation decision and a poor short-term optic.

Gas & NGL Marketing Services (Sequent)

Adjusted EBITDA of negative $1M against negative $15M a year ago. Commodity margins were $12M against negative $16M, operating costs were $18M, and proportional EBITDA from equity-method investments contributed $10M. The segment's modified EBITDA of $123M looks very different because it includes a $120M unrealized derivative gain, which management removes to reach the adjusted figure. Natural gas sales volumes fell 10.5% to 5.52 Bcf/d and NGL sales volumes rose 8.8% to 185 Mbbls/d.

Assessment: exactly what the seasonal pattern predicted, and the reason this note argued three months ago that none of the first quarter's marketing upside should be carried forward. Sequent earned $227M in the March quarter and $226M across the first half; it earned $53M across the second half of 2025. The segment is a weather option that pays in winter and costs nothing in summer, and the correct treatment is to model the first quarter generously and the rest of the year at roughly zero. That the total company still grew 6.2% with this line contributing nothing is the more important fact.

Other

Adjusted EBITDA fell 17.9% to $64M and beat a small published estimate by 12.9%. Net realized product sales fell 14.4% to $125M on the January divestiture of the South Mansfield upstream interests, and operating costs fell 3.9% to $73M. The segment's modified EBITDA of $98M includes the $12M gain on an upstream asset sale and a $22M unrealized derivative gain, both removed from the adjusted figure.

Assessment: the shrinkage is deliberate and disclosed. Upstream is being sold down, and the segment's contribution will keep falling as the divestitures annualize through the second half. Model this line down toward $290M for the full year against $369M in 2025.

Operating Statistics

Two lines are worth isolating. Northwest Pipeline transportation volumes fell 16.7% while its firm reserved capacity rose 8.1% to 4.0 MMdth, a divergence partly explained by the Naughton coal-to-gas conversion placed in service in April, which added 98 Mdth per day of contracted capacity. Contracted capacity is what Williams gets paid for on a regulated pipeline, so the volume decline is close to economically irrelevant. Second, the rich-gas indicators in the gathering segments moved up sharply: NGL production rose 20.3% in the Northeast and 15.7% in West, and West NGL equity sales nearly doubled. Barrel content, not gas volume, drove the gathering segments this quarter.

Key Topics & Management Commentary

Overall Management Tone: the more confident of the two calls we have reviewed, and confident about the balance sheet rather than about the operations, which is a reversal. Three months ago management deflected every financing question at the level of principle; this quarter it led with the structure, the cost of capital and the remaining headroom, and pushed the growth target up two points. Where the tone thinned was on the acquisition. Management would not size Momentum's synergies, its growth rate or its historical earnings, and answered the one detailed question about the joint venture capital bridge in a single sentence.

1. The Power Innovation Joint Venture: What Williams Actually Sold

In July, Williams sold a 49% noncontrolling interest in its five committed power projects, Socrates, Apollo, Aquila, Socrates the Younger and Neo, in exchange for $5.34B of committed capital led by funds managed by Blackstone Credit & Insurance, in partnership with Apollo and with insurance vehicles and accounts managed by KKR. The initial contribution of approximately $3.75B landed in July, with the balance due through early 2027. Williams retains 51%, operatorship and commercial control.

"The joint venture provides $5.34 billion of committed capital, including $4.4 billion for 49% of the expected total growth capital expenditures plus $900 million of additional consideration to Williams. Importantly, that capital comes at an attractive capped 6.35% cost of equity, which is a very efficient way to fund these near-term Power Innovation projects without diluting the value of the platform we are building."
— John Porter, Chief Financial Officer

The economics are better than the ownership split implies, and the reason sits in the distribution waterfall rather than in the headline. Cash distributions "will generally align with ownership percentages," but distributions to the investor above a target return "serve to reduce its investment balance," and Williams holds a buyout right between years 7 and 14 struck at the investor's outstanding balance. A partner whose return is capped at 6.35%, whose capital amortizes with every distribution above that cap, and who can be bought out at whatever remains, is structurally closer to an amortizing preferred security than to a 49% common partner. Because Williams retains control, the accounting treats the whole thing as an equity transaction, increasing capital in excess of par value and noncontrolling interests rather than debt.

"Specifically, if you look at the ratio of the total cash flow Williams will see from these 5 projects, to the total invested capital, that ratio improves about 56% with the joint venture. And again, that is only over the primary term of the contracts and doesn't include any of the upside we expect to develop both within the primary term and well beyond."
— John Porter, Chief Financial Officer

Assessment: this is the single most important development of the quarter and it resolves the bear case we opened the coverage with. Three months ago the risk we named was that funding the buildout would quietly transfer a share of the very EBITDA growth the multiple was capitalizing. That did not happen. Williams sold a capped, self-liquidating claim rather than a permanent equity slice, took $900M of cash consideration on top, and kept a repurchase option. The offsetting caution is one of presentation rather than substance: consolidated adjusted EBITDA, and therefore the 11% growth target quoted on it, now includes 100% of assets Williams owns 51% of. Per-share arithmetic has to run through the noncontrolling-interest line, which the guidance bridge already raised by $40M for a partial year.

2. Socrates Phase One Is in Service

The first behind-the-meter project reached commercial operation in late July, roughly on the timeline management set out at the first quarter. Phase two remains guided to the fourth quarter, and the combined Socrates North and South facilities represent 556 MW of expected capacity in New Albany, Ohio under a 10-year, primarily fixed-price power purchase agreement with a customer extension option.

"Last week, we achieved in-service for Phase 1 of Socrates, delivering a utility scale 200 megawatts of power to our customer in under 18 months since commercialization."
— Chad Zamarin, President and Chief Executive Officer

The operating detail was more useful than the milestone. Commissioning included load testing designed specifically to validate that the plant could follow a data-center load profile, which is the technical question that separates a behind-the-meter power plant from a merchant peaker.

"And the commissioning has gone extremely well. We did a lot of load testing prior to actual start-up to facilitate to make sure that we actually could see the AI load following actually the work the way it was intended to. So those tests went really smoothly."
— Larry Larsen, Chief Operating Officer

Assessment: the proof point the coverage needed. Every valuation argument for the power franchise until now rested on a build multiple management asserted and a contract structure investors could not inspect. Williams has now delivered a utility-scale plant on schedule and within budget, demonstrated AI load-following in commissioning, and reported the remaining four projects on schedule and on budget. The claim that this is repeatable industrial execution rather than a one-off is materially more credible than it was in May.

3. The Momentum Midstream Acquisition

Williams agreed to acquire 100% of Momentum Midstream for total consideration of up to $5.5B, comprising approximately $3.5B of cash and debt and roughly $2B of Williams equity. The assets add more than 4,000 miles of pipe and over one million dedicated acres across four Haynesville gathering areas with 6 Bcf/d of combined capacity, plus three take-or-pay pipelines capable of moving 4.05 Bcf/d. The stated multiple is approximately 8.5x projected 2027 EBITDA on a consolidated basis, and management put it at roughly 9x net of the noncontrolling interest in the NG3 pipeline joint venture. The transaction is subject to Hart-Scott-Rodino clearance and is expected to close later this year.

"The combined Williams and Momentum assets will form the backbone that connects our country's fastest-growing supply basin with our fastest-growing demand corridor."
— Chad Zamarin, President and Chief Executive Officer

Working backwards from the multiple, $5.5B at 8.5x implies roughly $645M of consolidated 2027 EBITDA, and roughly $610M net of the noncontrolling interest. Against Williams' own trading multiple of 14.4x on 2026 guidance, the acquisition is bought at close to a six-turn discount, which is the mechanical source of the accretion management claims to both AFFO per share and earnings per share.

Assessment: the strategic logic is the strongest part of the case, because Williams already owns the Transco corridor these volumes have to reach and the Gillis interconnect they have to pass through. The financial disclosure is the weakest. Investors were given a forward multiple on a forward year and nothing else. No historical Momentum EBITDA, no growth rate, no synergy figure, and two direct analyst attempts to obtain any of the three were declined. An 8.5x entry multiple on assets in the right basin is defensible on its face; it is not verifiable from anything in this release.

4. Delta Access and Shelby Connector

Two projects were announced alongside the acquisition. Delta Access is a $1.5B transmission project along the Transco corridor with initial capacity of 2.25 Bcf/d, expandable to 3.5 Bcf/d, serving LNG and power customers in Louisiana, with in-service guided to the first quarter of 2029. It is conditional on the Momentum acquisition closing. The Shelby Trough Connector expands the Louisiana Energy Gateway system into the Shelby Trough with 750 MMcf/d of initial capacity, expandable to 1.5 Bcf/d, guided to the second quarter of 2028.

Both were commercialized by Momentum and are described by management as the reason the acquisition multiple compresses over time rather than as separate projects with their own returns.

"So to be clear, the multiple we've been speaking to does not include -- that's a current run rate multiple. So that does not include any consideration of the future growth that you would see from the Shelby Connector, Delta Access, any additional growth within the platform."
— Chad Zamarin, President and Chief Executive Officer

Assessment: Delta Access is the more consequential of the two, because a fully contracted 2.25 Bcf/d transmission line along the Transco corridor is exactly the kind of project the February analyst-day backlog was supposed to produce and had not yet produced organically. Note the sequencing carefully: the project is conditional on the acquisition, which means the $5.5B purchase price buys both the existing platform and the option on a $1.5B expansion whose returns management says sit inside its normal build-multiple range. That is a fairer characterization of what is being bought than the 8.5x headline alone.

5. The Growth Target Moves to 11% Plus

The 2025 to 2030 adjusted EBITDA growth target was raised to 11%-plus from 10%-plus. The more informative number is the contracted-book figure underneath it, which management has now moved twice: approximately 8% at the February analyst day, approximately 9% in May, and 11% now.

"And now after layering in the Momentum transaction as well as the other projects we've announced today, we feel confident in moving our target up to 11% plus."
— John Porter, Chief Financial Officer

Pressed on whether 11% embeds conservatism, management said it does, and specified what is excluded.

"we're really talking about a number here that continues to be centered on our existing contracted book of business. And so we're excluding the commercialization of any additional power or pipes projects, which you're aware of the extensive backlog we've got in both the power side and the pipes projects."
— John Porter, Chief Financial Officer

The arithmetic is worth spelling out. An 11% compound rate on 2025 adjusted EBITDA of $7,750M reaches $13.06B by 2030, against $12.48B at 10% and $11.39B at 8%. The 2026 guidance midpoint of $8.40B represents 8.4% growth in year one, so the remaining four years must compound at roughly 11.7% to reach the 2030 figure. Management is explicit that the shape is back-loaded, with the earnings step-up arriving in 2028.

Assessment: our first-quarter note said that a quarter in which the contracted CAGR did not step up toward the 10%-plus target would be a signal. It stepped up two full points and cleared the target outright. This is the second consecutive quarter of upward revision, it excludes the entire uncommercialized backlog, and it is the strongest single argument for owning the shares. The back-loading is real but it is disclosed, contracted and tied to specific in-service dates rather than to a growth assumption.

6. Leverage Resets from 4.1x to 3.75x

Quarter-end debt to adjusted EBITDA was 3.67x against 3.80x a year ago, and against a guided full-year midpoint of approximately 4.1x three months ago. Management now guides year-end 2026 leverage to approximately 3.9x with three months of Momentum included, and to approximately 3.75x on a full-year Momentum run rate, against a stated 3.5x to 4.0x target range and an internal 4.0x ceiling.

"3.75x leverage opens up in excess of another $2 billion of incremental capacity versus our internal 4x leverage ceiling. And that's without considering bringing in any partners on future power innovation opportunities, which will remain an attractive and relatively easy thing to do."
— John Porter, Chief Financial Officer

"Most importantly, though, as we previously discussed, the balance sheet leverage tightness is primarily an issue for '26 and '27 before the historic earnings growth we expect in '28 and beyond."
— John Porter, Chief Financial Officer

One mechanical caveat belongs on the reported 3.67x. The company's definition nets debt against cash and, for 2026, against $777M of cash purchases of reimbursable long-lead power equipment. Without that adjustment the ratio is 3.76x, so the adjustment is worth roughly 0.09x, and it has grown from $439M at the first quarter to $777M now.

Assessment: a 0.4-turn improvement in the guided midpoint in three months on a company that was heading the other way is the clearest evidence that the July transaction was not a financing of convenience. The equipment adjustment is a real presentational flatter and it is disclosed in the same footnote where the ratio appears, which is the right place for it. The more useful number for a forward view is the $2B of stated headroom, because it defines how much additional power commercialization can be absorbed before another partner has to be brought in.

7. How Much of the Guidance Raise Is Operational

Full-year adjusted EBITDA guidance moved to $8.3B to $8.5B from $8.05B to $8.35B, a $200M increase at the midpoint. Management separated the two contributions qualitatively but not numerically.

"For full year '26 adjusted EBITDA, our existing businesses continue tracking toward the upper half of the guidance framework we discussed earlier in the year. On top of that, the accretive Momentum acquisition adds incremental EBITDA, taking the full year outlook to $8.3 billion to $8.5 billion."
— John Porter, Chief Financial Officer

That statement is sufficient to do the split. The upper half of the old range is $8.20B to $8.35B, midpoint $8.275B. The new range is $8.3B to $8.5B, midpoint $8.40B. The difference between the two midpoints is $125M, which is the assumed contribution from roughly one quarter of Momentum ownership. The remaining $75M of the $200M headline raise is the base business moving from the midpoint of the old range to the midpoint of its upper half. Roughly 62% of the raise is an acquisition that has not yet closed.

Assessment: not a criticism of the raise, which is real, but a correction to how it will be read. A $75M base-business improvement on an $8.2B guide is a 0.9% upward revision, which is a good quarter and not a step-change. The $125M is contingent on Hart-Scott-Rodino clearance and on closing in time to contribute a full quarter. Model the base and the acquisition separately, because they carry different probabilities.

8. Transco's Organic Queue

Four organic commercializations landed in the quarter alongside the acquisition. Customer agreements were signed on Transco's Leidy Access and Garden Connector, two expansions serving residential, commercial and power demand in Pennsylvania and New Jersey. Power Express was upsized again, to 800 MMcf/d from the 750 MMcf/d disclosed in May. An extension of Line 200, the 3.1 Bcf/d line being built from Gillis to serve the Woodside LNG terminal, adds a seven-mile lateral into Lake Charles to serve incremental power demand.

"We also further upsized our Transco Power Express project, which now represents an 800 million cubic feet per day expansion of Transco to serve load growth, power demand and data center growth in Virginia."
— Chad Zamarin, President and Chief Executive Officer

On the existing construction program, Southeast Supply Enhancement remains on track for partial in-service at the start of 2027 with full service targeted for the third quarter, and the Northeast Supply Enhancement offshore build is guided mostly to 2027.

"We've got Southeast Supply Enhancement that is under construction. We still believe we'll have some early in-service for the pipeline segment of that, that could start beginning of the year in '27 and then full in-service still targeting for third quarter."
— Larry Larsen, Chief Operating Officer

Assessment: the Transco franchise is doing exactly what the bull case requires, converting demand growth on an existing right of way into contracted expansions at rates set by a settled rate case. Four separate commercializations in one quarter without a single new pipeline corridor is the definition of a scarce asset. The one caution is that the segment's reported growth is now 6% while its capital intensity is at a record, so this pillar will look better in the model than in the income statement for the next two years.

9. The Northeast Stabilized

Consolidated Northeast gathering volumes were 4.16 Bcf/d against 4.15 a year ago, ending the decline that produced an 8.7% year-over-year drop last quarter. The earnings improvement came from mix rather than throughput, with plant inlet volumes up 7.9% and NGL production up 20.3%. Management describes its own forecast for the region as deliberately cautious.

"I think as John mentioned in his comments, as we look at our outlook, we've been somewhat conservative on our growth for the Northeast, but we're seeing a lot of demand in and around each -- the region that's going to help support pricing and activity."
— Larry Larsen, Chief Operating Officer

Assessment: the segment that looked most fragile three months ago produced the only meaningful beat this quarter. More importantly, it did so without a volume recovery, which means the improvement does not depend on Appalachian dry-gas rigs returning. Over $2B of annual EBITDA modelled conservatively is a real option on the region, and it is one the market is unlikely to be paying for.

10. The West Segment Missed on Costs and Minimum Volume Commitments

West adjusted EBITDA of $359M grew 5.3% and fell 7.8% short of the published estimate, the largest variance in the portfolio. Volumes and NGL statistics all beat. Segment operating and administrative costs rose 10.7% year over year to $166M, and Eagle Ford revenues fell $9M in the quarter and $19M in the half on contractual minimum-volume-commitment step-downs and lower gathering volumes. Neither the release commentary on the segment nor any question on the call addressed the cost line.

Assessment: the least satisfying disclosure in the release. A segment whose volumes beat, whose barrel content improved sharply and whose EBITDA still missed by 7.8% is a cost story, and the cost story is not told anywhere. The Eagle Ford step-downs are contractual, disclosed and will continue. The $16M increase in segment operating costs is neither.

11. The Marketing Segment Gave Everything Back

Gas & NGL Marketing Services adjusted EBITDA was negative $1M, against $227M in the first quarter and negative $15M a year ago. Management framed the outcome as ordinary seasonality and treated it as such in the guidance discussion, explicitly declining to count on a repeat of first-quarter conditions.

"things like Sequent, obviously, occasionally can have pretty fantastic early winter results but not something we count on when we do these guidance updates."
— John Porter, Chief Financial Officer

Assessment: the risk we flagged in the initiation has now been tested and resolved rather than deferred, and the resolution is favourable. The concern was never that Sequent would lose money in summer; it was that the first quarter's headline growth of 13.3% would be read as a run rate. It was not a run rate, the market now has the evidence, and the company still grew 6.2% year over year with the line at zero. Treat this segment as a first-quarter option worth roughly $230M to $280M a year and nothing in the remaining nine months.

12. The Pace of Power Commercialization

Management guided to additional project commercializations before year end, and was unusually explicit that it is deliberately slowing the announcement cadence to match delivery capacity.

"I will say, I know everyone is excited about and looking forward to the next announcement, but the team is also doing a really good job of pacing the commercialization of projects so that we can maintain this steady growth throughout the end of the decade and beyond."
— Chad Zamarin, President and Chief Executive Officer

The binding constraint is equipment rather than demand or capital. Turbines are locked up against the existing backlog, and the pacing question is the rest of the plant.

"And it's not only just the turbines. We've highlighted that we've locked up the turbines to be able to support our backlog, but it's the rest of the balance of plant."
— Larry Larsen, Chief Operating Officer

Assessment: a company telling investors it could announce more and is choosing not to is either genuine discipline or a way of managing expectations it cannot meet. The evidence this quarter favours the first reading, because Socrates landed on time and within budget and the other four projects are reported on schedule. Equipment lead times are a real constraint and an honest one to name. Watch whether the promised year-end commercializations arrive; that is the test of whether pacing is a choice.

Guidance & Outlook

The second-quarter release publishes a midpoint column only, where the first-quarter release published low, mid and high, so every per-share and AFFO comparison above is midpoint to midpoint. Two lines in the bridge deserve attention beyond the headline raise. Interest expense was re-cut upward for the first time this year, to $1,535M from the $1,485M that had been carried unchanged since February, which is the acknowledgment we argued was overdue when first-quarter interest alone annualized above the full-year assumption. And the noncontrolling-interest line rose $40M on a partial year of the joint venture, which is the first quantified per-share cost of the July transaction.

Implied second-half ramp: first-half adjusted EBITDA of $4,175M grew 10.0%. The $8.40B midpoint implies $4,225M in the second half against $3,953M a year ago, or 6.9% growth. The $8.5B top end implies $4,325M and 9.4%; the $8.3B floor implies $4,125M and 4.4%. Stripping out the assumed Momentum contribution, the base business is guided to roughly $4,100M in the second half, or 3.7% growth against a first half that grew 10.0%. Part of that deceleration is arithmetic, because the first quarter carried the marketing segment's entire annual contribution. The rest is the Other segment shrinking on divestitures and a deliberately cautious posture on the second half.

"Yes, it's still early August, and there's still quite a few things to play out for the year. So those are some of the reasons why we try to stay, I'd say, fairly conservative."
— John Porter, Chief Financial Officer

Street at: the published full-year consensus for adjusted EPS is $2.35, which is precisely the new guidance midpoint, so the Street is parked on the guide rather than taking a view around it. The published third-quarter estimate is $0.56 on revenue of $3.19B. No published full-year consensus for adjusted EBITDA exists, so the guidance range remains the honest anchor.

Guidance style: conservative in the range, confident in the framing, and for the first time this year willing to re-cut its own assumptions. Management named hurricane season, weak summer gas prices and rig activity as the reasons for holding a $200M range with five months left, and declined to count on a strong early winter at Sequent. That is the same posture as the first quarter with better follow-through: the bridge that was carried forward line for line in May has now been updated where it was wrong.

Analyst Q&A Highlights

Conservatism Embedded in the New Growth Target

The opening question of the call went straight at whether the two-point CAGR increase was itself understated, on the reasoning that the acquisition and the pipeline project announced with it should on their own add roughly 200 basis points to a target that started at 10%. Management neither confirmed nor denied the specific math, but conceded conservatism in three places: the figure is contracted-book only, it excludes the entire uncommercialized power and pipeline backlog, and the Northeast is modelled cautiously.

Q: "So the EBITDA CAGR here was increased to 11% from 10% 5-year EBITDA CAGR. I guess if we just simply layer in EBITDA from Momentum and Delta Express (sic) [ Delta Access, ] I mean, it seems like on our math, those projects alone would add 200 basis points to the CAGR, take it up to 12%. So I'm just -- is that 11% target incorporating a degree of conservatism? Or are there other kind of headwinds, puts and takes to consider in the forecast?"
— Praneeth Satish, Wells Fargo

A: "Yes, I mean, like I said in my comments, we do feel well positioned to exceed 11%. And so like I said in February, plus is plus."
— John Porter, Chief Financial Officer

Assessment: the most valuable exchange on the call. A management team that has raised its contracted-book estimate twice in six months, and that responds to a suggestion of a further 100 basis points by saying "plus is plus," is signalling that 11% is a floor. The question also usefully establishes that the Street is capable of doing the layering itself, which means the target is unlikely to be the binding constraint on how the stock is valued.

The Basis of the Acquisition Multiple and Unquantified Synergies

A follow-up pressed on two distinct points: whether the 8.5x was struck on consolidated or net EBITDA given the noncontrolling interest in one of the acquired pipelines, and whether any operating or cost synergies could be sized. The multiple question was answered precisely. The synergy question was declined.

Q: "So I guess beyond the organic projects that you've identified, Delta Access, are there specific operating or cost synergies that you expect to get from Momentum? And if so, can you help quantify those? And then just as a point of clarification on the deal, the 8.5x acquisition multiple, is that multiple calculated based on Momentum's consolidated EBITDA? Or is it based on Williams' net share after reflecting -- 35% interest?"
— Praneeth Satish, Wells Fargo

A: "First off, I'd say we're not going to quantify yet what those synergies will be. But I think if you look at the footprint and the overlap between the 2 companies, there will absolutely be operational synergies."
— Chad Zamarin, President and Chief Executive Officer

Assessment: the clarification is more useful than the deflection. Management confirmed the 8.5x sits on consolidated EBITDA and put the net-of-minority figure at roughly 9x, which is the number that belongs in a model. Declining to size synergies at announcement is standard practice and not itself a red flag; declining to disclose any historical earnings figure for the target in the same breath is less standard, and it means the entry multiple cannot be checked against anything.

Exclusivity Across Hyperscaler Counterparties

A recurring investor concern about the behind-the-meter franchise is whether serving one hyperscaler forecloses the others. The question was put twice, once in general terms and once as a direct confirmation, and both times management said the constraint does not exist.

Q: "And just to confirm, I guess, with -- doing business with 1 major hyperscaler, you don't think precludes your commercial negotiations with signing up another major hyperscaler."
— Jeremy Tonet, JPMorgan Securities LLC

A: "No. Look, we're a big company. I mean, we want to provide energy infrastructure solutions for every business in America."
— Chad Zamarin, President and Chief Executive Officer

Assessment: management also stated that the backlog "is supported by multiple customers beyond just our first primary customer," which is the first time the customer set has been described as plural. That matters more than the exclusivity answer, because the concentration risk in this franchise has never been whether Williams may sign a second hyperscaler but whether it has. The counterparty on every committed project remains undisclosed, so the claim cannot be verified, but the framing has moved.

Commissioning Evidence from the First Power Project

The most operationally detailed exchange of the call. The question asked directly for proof of concept on a business model the investor base has no reference points for, and specifically whether the plant is ramping as expected and whether the learnings pull forward any subsequent in-service dates.

Q: "Can you just maybe give us a sense of how that start-up process went and how it's going so far? And I ask in the context of all this being somewhat novel to us and the investor base and really trying to see the proof of concept here. So curious, is it operating, ramping as expected?"
— Spiro Dounis, Citi

A: "And as of this week, we're delivering first power to the facility and expect to see that ramp up over the course of the month. And so far, so good. So excited to see that ramp up to full capacity in the near term."
— Larry Larsen, Chief Operating Officer

Assessment: management explicitly declined to pull forward any in-service date on the strength of the learnings, saying the remaining projects are trending on schedule and on budget rather than ahead. That is the right answer and the credible one. The useful disclosure was the load-testing detail, which addresses the one technical question that separates this asset class from merchant generation.

Reconciling the Joint Venture Capital to the Program Cost

A housekeeping question that exposed the only genuine gap in the quarter's disclosure. The joint venture's $4.4B for 49% of expected growth capital implies a project program of roughly $9.0B, against the $9.6B figure management has used for the power program. The bridge was answered in a single sentence and never sized.

Q: "I think that implies something closer to $9 billion. How do I reconcile that with the $9.6 billion? I think, on Slide 30, I know certain items like capitalized interest rate is excluded, but that seemed like a pretty big delta for just capitalized interest. If you could just help us bridge that, that would be helpful."
— Ameet Thakkar, BMO Capital Markets

A: "Yes. I think capitalized interest is the biggest component of that. So a noncash from the standpoint of the partnership."
— John Porter, Chief Financial Officer

Assessment: the questioner was right that the gap is large. Roughly $620M of capitalized interest on a $9.0B program is about 6.9% of project cost, which is plausible for multi-year construction at current rates but is not obviously "the biggest component" without being shown the other components. The substantive point is that the partner is funding 49% of cash capital, not 49% of capitalized cost, so Williams carries the financing charge on the whole program while sharing the asset.

Whether the Partnership Structure Repeats

Given how favourable the terms look, the natural question is whether the same vehicle can be scaled or whether each future project needs its own negotiation. Management's answer was that the diligence infrastructure is reusable but the pricing is not.

Q: "And then more broadly for incremental funding options, can we think about the existing JV with Blackstone being expanded, so more kind of assets being brought into it? Could the next ones be a different structure? Maybe just walk us through some of the options."
— John Mackay, Goldman Sachs

A: "I think in general, though, it's probable in my mind that each one of these deals could perhaps be a unique separate partnership just because I think each partnership sort of has to price the opportunity in a somewhat unique manner."
— John Porter, Chief Financial Officer

Assessment: the honest reading is that the 6.35% cap was priced against a specific set of five projects with signed power purchase agreements and a partly de-risked lead asset, and a future vehicle would be priced against whatever is in it. Management's claim that a repeat process could be run "in a much, much faster time frame" is credible on process and says nothing about price. The reusable asset here is the counterparty depth, not the terms.

Mechanics of the Acquisition Share Issuance

Roughly $2B of Williams equity forms part of the acquisition consideration, and the first question after the strategic discussion was about the overhang that creates.

Q: "Maybe a first easy one here, if I can. How are you thinking about the lockup here on the shares here being issued as part of the transaction?"
— Julien Dumoulin-Smith, Jefferies

A: "I mean instead of a traditional lockup, we'll be releasing the shares over a 180-day period. And then once those shares are released, we'll have a trading restriction that basically will limit their trading to a small percentage of our average daily trading volume. So we really don't see any negative pressure on the shares as a result of this transaction."
— Robert Wingo, Executive Vice President, Corporate Strategic Development

Assessment: a volume-linked trading restriction is a better instrument than a cliff lockup, because it removes the single dated event that a conventional structure creates and replaces it with a rate limit. Roughly $2B of stock is about 2.3% of the guided share count, and the guidance bridge already carries 8M additional diluted shares, so the dilution is visible in the numbers rather than pending. This answer coincided with the stock recovering from its opening gap.

Whether Large Transmission Growth Still Comes Organically

A pointed observation that the two largest recent Transco expansions each arrived attached to a transaction, and the implied challenge that the organic backlog may not be producing on its own.

Q: "But as I reflect on kind of this Transco expansion and then the one you announced last year, they were both a result of some M&A. And so it leads to the natural question of do you see kind of large-scale organic Transco opportunities that are still available within your backlog? Or do you need these kind of outside deals to unlock some of that attractive growth?"
— Jason Gabelman, TD Cowen

A: "And yes, definitely, we see organic opportunities on Transco. We're continuing to have those discussions. I mentioned earlier. It doesn't require M&A transaction to help facilitate those. The deals that we've done with SESE and Power Express, those are all organic opportunities."
— Larry Larsen, Chief Operating Officer

Assessment: the answer is factually correct and the question is still the right one. Southeast Supply Enhancement and Power Express were organic, and so were Leidy Access and Garden Connector this quarter. But management also volunteered that most of the larger projects under discussion sit in a 2030-plus time frame and are held up on regulatory certainty and customer timing. The organic queue exists; it is slower than the acquired one, which is precisely why the acquired one is attractive.

What Moves the Guide to the Top of the Range

The last question of the call asked what separates $8.5B from $8.3B. The answer was a list of things that could go wrong rather than things that could go right.

Q: "I just quickly wanted to understand the guidance was raised for 2026, which is very positive. What could drive you towards the top end of that $8.5 billion guidance versus the midpoint or the lower end, if you could help us with that?"
— Manav Gupta, UBS

A: "But I think some of the things that could be impactful would be what kind of hurricane season we have, what happens to prices here as we move into winter and overall levels of rig activity, things like Sequent, obviously, occasionally can have pretty fantastic early winter results but not something we count on when we do these guidance updates."
— John Porter, Chief Financial Officer

Assessment: three of the four swing factors named are outside management's control and two of them are weather. That is an accurate description of what actually moves the last $200M of this company's annual EBITDA, and it is consistent with the argument that the marketing segment is an option rather than a business line. It also confirms that the guidance range does not embed a strong early winter, which is where upside to the top end would come from.

What They're NOT Saying

1. There is still no power-segment financial disclosure, and now there is an operating power plant. The Transmission, Power & Gulf revenue disaggregation runs regulated interstate transportation and storage, gathering and processing, commodity consideration, other fee revenues and product sales. There is no power caption. Socrates entered service in late July, so the first power revenue is a third-quarter event, and as of this release there is no disclosed line to carry it. This was the first item on our list a quarter ago and nothing has changed.
2. No historical Momentum earnings figure was disclosed anywhere. Investors were given a forward multiple on a forward year and an assurance that it compresses. There is no 2025 or trailing EBITDA, no revenue, no contract-duration profile and no customer concentration. Two analysts asked for a growth rate or a synergy number and both were declined.
3. The base-business share of the guidance raise was never stated. Management said the existing businesses track toward the upper half and the acquisition adds on top, which is enough to derive roughly $75M of base improvement and roughly $125M of acquisition contribution. It is not enough to know whether the base is at the bottom or the top of that upper half, and the release publishes a midpoint column only.
4. Cogentrix has disappeared from the narrative. On the first-quarter call management said it expected to divest the investment later this year. The divestiture is not mentioned in this release, it was not raised on the call, no analyst asked, and the investment remains on the balance sheet at $289M with its carrying value reflecting the favourable fair-value impact of an announced agreement to sell a significant portion of the underlying power plant assets.
5. The forward path of the noncontrolling-interest line is not given. The 2026 bridge raises it $40M for a partial year. Consolidated adjusted EBITDA carries 100% of five projects Williams owns 51% of, and the 11% growth target is quoted on that consolidated measure. No 2027 or 2028 noncontrolling-interest figure was provided, so the per-share translation of the power franchise cannot be modelled from disclosure alone.
6. The gap between the joint venture's implied program cost and the stated program cost was answered but not sized. Roughly $600M separates the two, capitalized interest was named as the biggest component, and no other component was identified or quantified.
7. The counterparty on all five power projects remains undisclosed, and the geographic concentration is now visible. Four of the five committed projects are in Ohio, representing 2,068 MW of the 2,588 MW committed. Neither the concentration by customer nor the concentration by state was discussed.
8. A second Permian divestiture is in progress and appears nowhere in the release or the call. Williams signed an agreement in June to sell certain West-segment gas gathering assets in the Permian basin, designated them held for sale at June 30, and expects to recognize a gain on closing in the third quarter. No proceeds figure and no earnings impact were given.
9. West segment cost inflation is unaddressed. Segment operating and administrative costs rose 10.7% year over year against segment EBITDA up 5.3%, in the quarter with the portfolio's largest shortfall against published estimates. The release attributes the segment's year-over-year change to Louisiana Energy Gateway, gathering volumes and minimum volume commitments, and says nothing about costs. No question on the call raised it.
10. The AFFO definition changed and the effect on the comparison was not shown. Beginning this quarter, contributions from noncontrolling interests are excluded from AFFO, disclosed in a single footnote. The change lands in the same quarter Williams signed a joint venture that will route billions of partner capital through noncontrolling interests. The change is conservative and correct; the prior-year comparative was not restated, so the reported 10.1% growth is on a shifted basis.

Market Reaction

• Pre-print setup: WMB closed at $70.43 on Aug. 3, entering the print up 17.2% year to date against the S&P 500's 11.0%, down 3.7% over the trailing 30 days and up 16.9% over the trailing 12 months. The 52-week closing range was $56.51 to $79.40, so the stock went into this print 11.3% below its best close of the past year. Three months ago it went in within 1.2% of that year's high.
• Overnight and premarket: shares were weak on the release, with coverage attributing the move to adjusted EPS landing at $0.50 against a broader Street average of $0.52 and to the roughly $2B of Williams stock forming part of the acquisition consideration.
• Next-day session: the Aug. 4 session opened at $69.57, down 1.2%, traded a range of $68.52 to $72.53, and closed at $71.51, up 1.5% or $1.08. The recovery occurred across the 9:30 a.m. ET call.
• Volume: 13.8 million shares against a 30-day average of 7.7 million, or 1.8x normal.
• Relative: the S&P 500 rose 1.8% on the same session, so WMB underperformed the index by roughly 30 basis points on the day it announced a $5.5B acquisition, a $200M guidance raise and a two-point increase in its long-term growth target.

The intraday shape repeats the first quarter exactly: a gap down on the print and a recovery through the call. What produced the gap was legible within minutes. Adjusted EPS printed at the low end of a narrow band, the largest segment missed, and the release disclosed roughly $2B of stock to be issued to a private seller. What produced the recovery was equally legible and arrived in sequence: the guidance raise, the growth-target upgrade, confirmation that the first power project is delivering power, and the answer on the share-release mechanics that replaced a feared cliff lockup with a volume-linked restriction.

What the day again failed to produce was a re-rating. A company that resolved its funding question on better terms than the market was pricing, raised guidance, raised its long-term growth target by two points and put its first power plant into service finished the session up 1.5% against an index up 1.8%. That is now two consecutive quarters in which a materially good print has been treated as confirmation rather than news.

The difference is the entry point. Three months ago that non-reaction was a warning, because the stock was a fraction below its all-time high and the growth was fully in the price. This time the same non-reaction happened after a quarter in which WMB fell 7.5% while the S&P 500 rose roughly 5.5%, roughly thirteen points of underperformance, leaving the shares 11.3% below their 52-week closing high. The market has spent three months de-rating a company whose contracted growth rate went up two points and whose guided leverage went down 0.4 turns. That gap is the investment case.

Street Perspective

Debate: Is the Joint Venture Equity or Leverage in a Costume?

Bull view: the bull case on the Street is that this is the cheapest capital in the midstream sector. A capped 6.35% cost of capital against assets built at a 5x multiple, which is a 20% EBITDA yield on invested capital, is a wide spread, the accounting treatment is equity because control is retained, the partner's balance amortizes with distributions above its target return, and Williams can repurchase the position from 2033 at whatever balance remains. On this reading Williams sold a temporary financing, not a permanent share of the franchise.

Bear view: the bear camp argues that a capped return, an amortizing balance and a buyout struck at that balance describe a debt instrument in every respect except its accounting label, that rating agencies are unlikely to treat it as pure equity, and that the 3.75x leverage figure therefore flatters a balance sheet carrying a further $5.34B of hybrid obligation. On this reading the leverage improvement is presentational.

Our take: the bulls have the better of it on economics and the bears have a fair point on optics. The instrument genuinely is preferred-like, which is why it is cheap, and cheap hybrid capital funding assets at a 5x build multiple creates value regardless of where the agencies classify it. The number to watch is not the headline leverage ratio but the noncontrolling-interest line in the adjusted-income bridge, which is where the partner's claim becomes visible per share. It rose $40M this year and will keep rising as the projects enter service. Investors who model consolidated EBITDA and ignore that line will overstate the franchise.

Debate: Was Momentum Bought Well, or Bought Because the Capital Was There?

Bull view: the constructive read is that 8.5x on 2027 earnings for the leading gathering position in the basin that has to supply the Gulf Coast LNG build-out is a six-turn discount to Williams' own multiple, that the acquisition unlocks a $1.5B fully contracted Transco-corridor expansion the organic backlog had not produced, and that the accretion to earnings and cash flow per share is immediate and mechanical.

Bear view: the skeptical read is that a company which raised $5.34B of partner capital in July announced a $5.5B acquisition in August, that no historical earnings, growth rate or synergy figure for the target was disclosed, that growth capex has now been raised in two consecutive quarters, and that a management team which describes its own growth target as a floor has an incentive to buy the growth that clears it.

Our take: the sequencing is uncomfortable and the strategic logic is nonetheless sound. Williams already owns the Transco corridor these volumes must reach, and owning the gathering in front of it is the textbook adjacent position. The honest concern is disclosure, not price: an entry multiple that cannot be checked against a single historical figure is an assertion. We would underwrite the acquisition at management's stated net multiple of roughly 9x, assign no value to unquantified synergies, and treat Delta Access as a separate project on its own build multiple rather than as multiple compression.

Debate: Is 11% a Target or a Floor?

Bull view: the bull argument is that the contracted book alone has moved from 8% to 9% to 11% in six months, that the figure excludes every uncommercialized power and pipeline project in a backlog management describes as extensive, that the Northeast is deliberately modelled low, and that the first question on the call proposed 12% without being contradicted.

Bear view: the bear argument is that a 2026 guidance midpoint of $8.40B represents only 8.4% growth in the first year of a five-year 11% path, so the remaining four years must compound at roughly 11.7%, and that a target which back-loads that heavily is a promise about 2028 through 2030 rather than a description of the business today.

Our take: both are right and they are not in conflict. The back-loading is real, disclosed and mechanically explained by in-service dates that cluster in 2027 through 2029. The difference between this and an ordinary back-loaded target is that the assets are contracted, under construction and named. The 11% is a floor on the contracted book and a genuine target on the total, and the risk is execution and timing rather than commercial demand. That is the risk this company has spent the last two quarters demonstrating it manages well.

Model Update Needed

Valuation. At the Aug. 4 close of $71.51 on the 1,237M guided diluted share count, equity value is approximately $88.5B. Adding net debt of $30.6B from the June 30 balance sheet, noncontrolling interests of $2.2B and preferred stock produces an enterprise value of roughly $121.3B, or 14.4x the $8.40B guidance midpoint. That compares with 15.3x on the corresponding basis three months ago. The shares trade at 30.4x the $2.35 guided adjusted EPS midpoint against 33.2x last quarter, and at 13.9x the $5.15 AFFO-per-share midpoint against 15.1x, for a 2.9% dividend yield on the $2.10 annualized rate.

The forward framework below runs on 2027 estimates and on a pro-forma capital structure, because the reported June 30 balance sheet predates both the July joint-venture funding and the pending acquisition. Net debt is set at $32.8B, which is what management's approximately 3.9x year-end leverage guide implies on the $8.40B midpoint. Noncontrolling interests are set at $6.0B, comprising the $2.2B reported plus most of the $5.34B of committed joint-venture capital phased in through early 2027, less the undisclosed portion that lands in capital in excess of par value rather than in noncontrolling interests. Book value is the right deduction here rather than a discount or a premium, because the buyout right is struck at the investor's outstanding balance. That assumption is the largest single judgment in the framework, and every $1B of difference is worth $0.79 per share.

Twelve-month target: $82, raised from $80. The multiple moves up to 15.0x from the 14.5x we used a quarter ago, because the reason for the discount has been removed. Three months ago we applied a contraction to reflect peak leverage and an undefined funding plan; leverage now falls rather than rises, the funding is contracted at a capped cost, and the growth rate underneath the multiple is two points higher. That implies 14.7% price appreciation from $71.51 plus the 2.9% dividend yield, for a total return of roughly 18%. The bear case at 13.0x, which assumes the acquisition closes and delivers nothing beyond its entry multiple and the power franchise is capitalized as merchant generation, sits 6.1% below the current price. The asymmetry has moved decisively since the initiation, when the base case sat essentially at spot and the bear case was 20% below it.

Thesis Scorecard Post-Earnings

The pillars below are the ones established at initiation and carried in the standing thesis. Three of the six move this quarter, and one new risk is added.

Overall: the thesis strengthens materially. Every bull pillar is confirmed, and the two risks that carried the Hold rating have both been challenged by events rather than argued away. The company did precisely the two things we said would change our view, and it did them in the same quarter that the shares de-rated. The offsetting new risk is real but second-order: an acquisition-led growth strategy carries integration and disclosure risk, not solvency risk, and it is being pursued from a balance sheet that just got stronger rather than weaker.

Action: buy. The quarter itself was the weakest of the year, and that is the point. Williams is a better business than it was in May, with a funded program, a higher contracted growth rate and an operating power plant, and it is 6.1% cheaper than when we initiated. The two things that would reverse this view are a Momentum integration that requires the multiple to compress on synergies that never get quantified, and a third consecutive quarter of growth capex being raised without a corresponding step-up in the contracted book. Neither is visible today.