Barron's staff writer (energy, commodities & markets) — running archive of the securities & macro views surfaced in his articles, with per-article breakdowns and a stock index.
Cameco — owns 49% of Westinghouse and is the only public window into its numbers: its share of 2025 Westinghouse revenue was C$3.46B ($2.47B), up 20% y/y, implying ~$5B group revenue. A Westinghouse IPO would put a market price on a stake currently buried inside Cameco's financials. (Quoted in-article on its Toronto line, CCO.) Aug 26: a third leg added — Westinghouse "has also been testing its own microreactor," plans to deploy some at Army bases, and was one of only four firms to reach criticality this year under the DOE program. So Cameco now carries the fuel, the incumbent AP1000 licensor heading for a listing, and a free option on the small format developed by the one organisation with real licensing experience — the capability none of the venture-funded entrants has (no microreactor holds a commercial operating licence).
Cipher Mining — first of the five Bitcoin miners Morgan Stanley's Stephen Byrd "expects to benefit amid the backlash" (Aug 24; Barron's wording "Cipher Digital"). The thesis is not crypto but grid access as a scarce asset: with governors from New York to Texas pausing approvals for new data centers seeking to hook into the grid, the miners "already have access to the electric grid at their existing sites, so they're less exposed to the political backlash" — and every cancelled or delayed rival project "just makes these companies' sites more valuable." Warehouses are being converted to AI data centers and rented out as power or as shell, income "steadier than Bitcoin." The exposure that remains is the expansion pipeline, not the installed base.
Equinor — Norway's state-controlled producer and, since Russian pipeline gas was cut, Europe's largest pipeline gas supplier; named in the standfirst as a beneficiary of the price surge. The simplest position of the five: large volumes sold into Europe at European prices from fields already producing through pipelines already built — a higher price is close to straight-through profit. Least conditional name in the article, and the least upside if the disruption reverses, since it is pure price with no growth story attached. Row uses the NYSE ADR; primary listing Oslo. Barron's body paywalled and not captured.
Galaxy Digital — third of Byrd's five (Aug 24), grouped with the miners because for this argument the only thing that matters is whether a company already controls connected megawatts and can sell that power to data centers. Byrd calls the group "a safe harbor in the selloff." Caveat that applies across the five: they differ in how much value sits in the installed site versus an unapproved pipeline, and the call protects only the former.
GeoPark — the one Latin American beneficiary Demichelis singles out from the Venezuela deal (Aug 31): a small Colombian producer whose asset is the neighbourhood — regional operating experience becomes scarce and valuable if Venezuela genuinely reopens to private capital. It is not a party to the agreement, and the market paid a lot for the mention: +13% on Monday, six times Valero's move, on the thinnest argument in the article. The textbook shape of a policy headline expressed through the highest-beta available instrument — fastest to price it in, fastest to give it back if the deal stalls on funding, legality or term.
Halliburton — the second U.S. oil-services name (with SLB) flagged as a winner from Washington's control of 17 Venezuelan oilfields (Aug 31): existing "operations in Venezuela" plus American domicile, which Jefferies' Alejandro Anibal Demichelis thinks "might get preferential treatment to operate on U.S.-controlled land." More concentrated in the drilling-and-completion work a field-rehabilitation programme actually consists of, so more upside per dollar of Venezuelan capex and less cushion if it never starts. Services revenue is the last link in the chain to be paid and the first cancelled — and the chain's first link (who funds the "tens of billions"?) is still empty, with the contract term unresolved (25 years or 100).
Hut 8 Corp. — second of Byrd's five (Aug 24). Same mechanism as Cipher: an already-interconnected site portfolio being converted from mining to AI hosting, revalued upward by a permitting freeze that blocks everyone else's greenfield. The analogy is a warehouse with a rail spur in a year when no new spurs are permitted — the asset is unchanged, its scarcity is not.
Kinder Morgan — building pipelines to Gulf-Coast LNG terminals; projects US natural-gas demand could jump +28 Bcf/d (>25% above 2024) by 2030 — though it was the 2015 dividend-cut cautionary tale.
Cheniere Energy — the named LNG-export champion of the Corpus Christi boom, quoted by the port's CEO: "Every ship that leaves Cheniere Energy powers a million homes in Europe for a month. Had it not been for the supply coming out of here, the lights go out in Europe." A geopolitical framing, not a rating — and a water-hungry facility in a rationing region. Sep 10: named again as a beneficiary of European gas hitting a near-four-year high (Iran War supply disruption + Europe entering winter on its lowest storage since 2009) — though most Cheniere capacity is sold on fixed-fee contracts, so a spike lifts the uncontracted edge and the case for the next expansion more than current earnings. Barron's body paywalled and not captured.
MARA Holdings — fourth of Byrd's five (Aug 24) and the clearest illustration of the mechanism: as miners shifted to the data-center model "their stocks have traded like AI names," so they sold off with AI power on the state approval pauses. Byrd's claim is that this is inherited beta, not inherited exposure — the moratoria act on new interconnection requests, not on electricity already flowing to a built site — and that "investors are misunderstanding the risks."
Marathon Petroleum — headlined alongside Valero as one of the winning refiner stocks of the record diesel squeeze (Aug 18), also at an all-time high. Wholesale diesel +109% YTD and pump diesel $5.47 (+48% y/y) against a WTI-to-diesel spread of $101.86 — the first ever triple-digit crack — with BofA's Michael Widmer expecting the market to stay "tight, volatile, and expensive well into next year" as harvest-season demand accelerates into low inventories. Aug 21: the gasoline-side confirmation — refiners "like Valero and Marathon Petroleum are making very high margins on the fuel they sell" while Washington's only lever against a $4.11 pump price ($1 above a year ago) is an EPA blend waiver worth 10–30¢ for a few weeks; the policy toolkit is still aimed at the consumer price, not the refining margin. Aug 31: named with PBF as a refiner that "also process[es] heavy crude" and "would also benefit if there's more of the heavy stuff on the market" after the U.S. took control of 17 Venezuelan oilfields — a second, feedstock-side margin tailwind stacked on the record product crack. Note the weaker form of the claim: it depends on the global heavy-light differential widening, not on Venezuelan barrels specifically (MPC is not in TPH's Venezuelan-crude top three).
Matador Resources — oil-levered Permian producer named by KeyBanc's Tim Rezvan (Sep 18) as a beneficiary of a tightening physical market: spot Brent $33 over futures, the SPR near its floor, and China importing again. An analyst call relayed by Salzman, not a Barron's pick; it reverses if the Saudi pipeline is repaired early or the Iran war ends.
NextDecade — the developer of the group, third of the three American LNG exporters named as benefiting from the European price surge. What a developer gains from a supply shock is not cash but the ability to sign long-term offtake contracts (buyers commit when they are frightened), which is what unlocks project financing — so the benefit is a step closer to existing, not a bigger number this year. The most speculative link in the chain and the most exposed to the crunch resolving before commitments are made. Barron's body paywalled and not captured.
Terra Innovatum — one of only three publicly traded microreactor names, at a $625M cap against private marks of $1.9B–$6B for peers (Aug 26), an inversion of the usual liquidity discount. Its distinguishing fact is the specialist's capital: "Segra has invested in Valar Atomics and Terra Innovatum" — and Segra's Arthur Hyde supplies every bear point in the article (24/7 armed security that does not scale down, prohibitive permitting and site-selection costs). What it lacks: not among the Army's five selections, and not named in the criticality cohort — the theme's tailwind plus an informed backer, without the contract or the technical milestone.
Oklo — named as one of the novel small/micro-reactor developers whose success Standard Nuclear's TRISO-fuel demand depends on; part of the "broader nuclear ecosystem" that must scale for the fuel-supply-chain thesis to pay off. Sep 17: an upstart "making steady progress at inking commercial deals for new reactors," and +12% the day Holtec withdrew its $10B IPO — one less large issuer competing for nuclear-dedicated capital. Relief, not new evidence: its reactors "won't be turned [on] for years" and face "significant regulatory hurdles."
PBF Energy — named with Marathon as a refiner that "also process[es] heavy crude" and "would also benefit if there's more of the heavy stuff on the market" after the U.S. took control of 17 Venezuelan oilfields (Aug 31). A pure refiner with no upstream, midstream or retail cushion, so it is the highest-torque expression of the article's mechanism: profit is almost entirely the spread between cheap heavy crude and the fuel made from it. The claim is broader and weaker than Valero's — it rides the global heavy-light differential rather than Venezuelan barrels specifically (PBF is not in TPH's Venezuelan-crude top three) — and the same lack of diversification that magnifies the gain leaves nothing to offset a compressed spread.
PPL Corp. — the named beneficiary of Pennsylvania's new data-center rules (Aug 19). Gov. Shapiro's executive order forces data centers to bring their own power and win community buy-in, and strips them from the grid fast track — which is exactly the product of PPL's Invitium Energy JV with Blackstone, already holding 5 GW of reserved gas turbines to build dedicated plants in-state. BTIG's Alex Kania: PPL "has a way to benefit"; CEO Vincent Sorgi expects Invitium to sign a developer by year-end — the dated, checkable catalyst. Its transmission arm is two-sided (fewer data-center wires, but tech companies forced to fund more of the build-out).
Phillips 66 — the third U.S. refiner named among "the biggest beneficiaries" of the record crack (Aug 18) and likewise at an all-time high. The margin is driven by product-side supply withdrawals stacked deeper than the crude-side ones: Hormuz shut (~20% of world oil), 2.8m bbl of Russian refining capacity droned offline (BofA), Russian export restrictions into next year and Chinese export cuts, leaving the US as exporter of last resort at record weekly diesel exports. Aug 31: ranked third — behind Valero and Chevron — among refiners of Venezuelan crude by TPH, so a direct if smaller beneficiary of a wider heavy-barrel discount on the Gulf Coast as Washington takes control of 17 Venezuelan oilfields. A ranking mention rather than an argued case; "refiners' stocks were up on Monday."
Riot Platforms — last of Byrd's five (Aug 24), and where the article's one caveat bites hardest: "while some new Bitcoin-to-AI projects in states like Texas may see delays, most Bitcoin miners are still in a strong position to convert their warehouses for AI use." Underwrite it in two halves — energised megawatts are the protected, appreciating asset; announced megawatts face the same Texas audit and interconnection queue as any hyperscaler's greenfield.
Seadrill — the second leg of Goehring & Rozencwajg's $100-oil trade (Aug 26): offshore oil services that "have struggled in recent years as demand fell. Some went bankrupt and have restructured" — a permanently shrunken fleet with cleaned-up balance sheets meeting returning demand as "more companies have been open to offshore drilling as they search for ways to grow production." Goehring says companies like Seadrill "should benefit." High-torque and pure — day rates and utilisation move non-linearly once the market tightens — with the balance-sheet history as the standing risk.
Shell — named in the standfirst (not the bullet summary, so a secondary mention) as a beneficiary of European gas at a near-four-year high. As the world's largest LNG trader as well as a producer, it earns on dislocation rather than price level: a war-driven disruption plus record-low European storage is textbook conditions for whoever controls the ships and contracts to move gas between a desperate region and a well-supplied one — and it requires no increase in Shell's own production. Row uses the NYSE ADR; primary listing London. Barron's body paywalled and not captured.
SLB (Schlumberger) — named with Seadrill as an offshore-services beneficiary if operators shift capital back offshore to replace flattening U.S. shale growth (Aug 26). The quality-and-liquidity leg of the trade rather than the high-torque one — and note the internal tension with the managers' own stated rule (they avoid diversified companies when they want exposure to a single variable): SLB's breadth across product lines, geographies and onshore/offshore makes it steadier and, for the same reason, less levered to an offshore capex turn. Aug 31: a second services catalyst — named first (with Halliburton) as an oil-services company "with operations in Venezuela" that could "win more business" if drilling picks up under the U.S. takeover of 17 oilfields. The differentiator Demichelis adds is jurisdictional, not technical: "because they're American companies, they might get preferential treatment to operate on U.S.-controlled land" — a moat made of politics, revocable by an election on either end, and gated behind "tens of billions" of investment nobody has yet volunteered.
SM Energy — oil-levered U.S. producer on KeyBanc's (Rezvan) list of names that can keep rising "even after a very strong year" as supply buffers run out (Sep 18). An analyst call relayed by Salzman, with no valuation given.
Talos Energy — oil-levered Gulf of Mexico producer on KeyBanc's (Rezvan) list of beneficiaries of the Saudi-pipeline escalation and China's return to imports (Sep 18). An analyst call relayed by Salzman, with no valuation given.
Valar Atomics — privately-held microreactor developer, the most richly funded of the group: raised $1B this month at a $6B valuation and reached criticality this year under the DOE program (Aug 26). It is a Segra position, and Hyde's case answers his own bear argument: rather than fight the fixed 24/7 security and permitting overhead that makes a lone remote microreactor uneconomic, dilute it — "Valar can deploy at least 30 small or midsize reactors at one site." Gas-cooled like Radiant, so it also sells industrial process heat, "expand[ing] the markets it can serve." Against that: private, no commercial licence, not an Army selection, and a $6B mark that already assumes much goes right.
Venture Global — listed first among the names European gas's near-four-year high is "creating opportunities" for, and named in Barron's bullet summary among the American LNG exporters benefiting. The higher-torque way to own the move: a larger share of its cargoes has historically gone to the spot market rather than into long fixed-fee contracts, so it captures a European price spike directly — and has no contracted floor when storage rebuilds. Barron's body paywalled and not captured; stance is "named as a beneficiary," not an argued case.
Valero Energy — named first among "the biggest beneficiaries" of the record diesel crack (Aug 18): the WTI-to-diesel spread hit $101.86, the first triple-digit print ever, U.S. refiners are earning triple last year's per-barrel profits (OPIS) and the stock has made an all-time high. A pure-play on the refining margin rather than on crude, with Melius' James West seeing "structurally higher refining margins over the next two years" on a flat-to-shrinking global fleet through 2027. Earlier (Aug 7) it was the named refiner in the Corpus Christi water-rate fight: among the large users challenging the city's doubling of industrial water rates before state regulators while posting blockbuster earnings ("We're here worried about charging them a little bit more, and they're still fighting"). Margins are excellent; the local licence to operate is what's being spent down. Aug 21: named again as "making very high margins on the fuel they sell amid elevated prices at the pump" — and the administration's response to $4.11 gasoline is an early switch to cheaper, higher-butane winter-grade fuel (10–30¢/gal for a few weeks), a relaxed constraint rather than a windfall tax, with California and New York declining to waive their own stricter rules. Aug 31: the second margin lever arrives on the feedstock side — TPH ranks Valero the top refiner of Venezuelan crude, and with Washington taking control of 17 Venezuelan oilfields, "the more Venezuelan crude they can purchase, the wider those margins are likely to get" (heavy, sour barrels trade at a discount and Gulf Coast plants are built for them). The safest part of the thesis needs no deal at all — U.S. imports from Venezuela already ran 137k bbl/d (Jan) → 544k (May) → ~700k (mid-July). Stock +2% on the day: an incremental improvement to an already-excellent margin environment, not a transformation.
Williams Cos. — natural-gas-focused pipeline operator with a data-center power side business; the article's best performer (+10% YTD) and squarely in Sanghani's preferred "natural-gas over oil/liquids" bucket.
X-Energy — listed advanced-reactor developer (earlier in this archive as private X-energy, also a TRISO fuel-pellet maker competing with Standard Nuclear). Sep 17: an upstart "making steady progress at inking commercial deals for new reactors," +12% the day Holtec pulled its $900M / $10B IPO — the flow benefit of a rival listing disappearing. Positive with the article's caveat: reactors years from switch-on, "significant regulatory hurdles" ahead.
Saudi Aramco — the Saudi producer whose Hormuz workaround was hit (Sep 11): projectiles struck the pumping station of the East-West pipeline to the Red Sea, which diverts 7 mb/d around the strait and feeds up to 5 mb/d of exports, and the kingdom shut it down. Two-sided — the bypass is proven vulnerable and the damage undisclosed, while Brent jumped to $109.23 and the article says prices "are almost certain to stay high." Named only in the photo caption; Neutral pending the size and duration of the outage.
Aalo Atomics — privately-held developer that has raised "over $300 million" and reached criticality this year under the DOE program, but was not among the Army's five selections (Aug 26). The cleanest test in the article of what the technical milestone is worth alone: criticality proves the reactor starts; it does not produce a customer, and the Army order was the only customer available. Watch whether it converts DOE validation into a commercial contract without a military anchor — precisely the question the piece closes on ("the industry… will have to convince other buyers too").
Antares Nuclear — privately-held developer, one of the five companies the U.S. Army selected for its microreactor program and one of four firms to reach criticality this year under the DOE program; raised $370M last month at a $2.1B valuation (Aug 26). Two of the article's three de-risking markers, which is more than most — but private, pre-revenue, no commercial operating licence, and carrying the sector's unresolved question: a military anchor order validates that the machine works, never that it is economic off a base that already pays for round-the-clock security.
Bank of America — anchor tenant of the half-built 30-story Dallas tower that will also house the Texas Stock Exchange, the physical symbol of "Y'all Street"; evidence of the financial-hub migration whose arithmetic is California's 10.8% bank income tax versus Texas' none (a 0.75% margin tax on gross profit).
Brookfield Corp. — Westinghouse's majority owner alongside Cameco's 49% (Barron's names the Brookfield side loosely as "Brookfield Asset Management"). A Westinghouse listing is the textbook monetization/mark-up event for the stake, though the read-through is diluted inside a much larger diversified parent.
BWX Technologies — the listed valuation anchor for the Westinghouse IPO: $2.35B of 2025 revenue at a $15B valuation (~6.4× sales), on a revenue base slightly below Westinghouse's implied ~$5B. Aug 26: promoted from yardstick to participant — BWXT Advanced Technologies is one of the five companies the U.S. Army selected to build microreactors at bases, inside a program of up to $2.2B for at least 20 reactors with first switch-on as soon as September 2028. The distinction that matters versus the venture-funded entrants: the order lands on an existing naval-nuclear manufacturing business with revenue and clearances, so it is an added contract rather than the event that must justify a valuation already paid. Sep 17: back to yardstick — "a nuclear company that investors consider comparable to Holtec," off 19% this year, part of why Holtec pulled its IPO; rose only 0.5% on the day versus +12% for the pure-play upstarts, i.e. priced on orders, not theme flows.
Blackstone — PPL's partner in Invitium Energy, the JV that has reserved 5 GW of gas turbines to build dedicated data-center power plants in Pennsylvania (Aug 19), and therefore on the winning side of Shapiro's "bring your own power" order. No view expressed on Blackstone itself and the exposure is one JV inside a vast alternative-asset manager — a signal of where private capital is positioning rather than a way to own the outcome.
Chevron — signed a 20-year deal to sell 2.67 GW from new off-grid West Texas gas plants to Microsoft's data centers; monetizes stranded (negatively-priced) regional gas as steady ~mid-teens-return contracted cash flow, decoupling it from oil's boom-and-bust and planting it in the AI-power theme. Aug 31: a second, unrelated exposure — TPH ranks Chevron the second-largest refiner of Venezuelan crude behind Valero, so its Gulf Coast refining arm benefits from a wider heavy-barrel discount if the U.S. takeover of 17 Venezuelan oilfields eventually adds supply. A one-line ranking mention; the article leaves unremarked that Chevron is the U.S. major with the longest-running Venezuelan operating history. Sep 11: a third, Middle East option — "considering adding oil-drilling sites in Iraq," Chevron "may help build pipelines" in the Gulf's Hormuz-bypass build-out; doubly conditional, and undercut by the same piece's point that the bypasses are now targets too (Saudi Arabia's East-West line was hit from Iraq).
Engine No. 1 — the (private) investment firm co-developing Chevron's West Texas gas-to-data-center power project; the two have discussed it publicly for over a year.
ERock — Houston maker of on-site natural-gas "rockblock" gensets giving data centers speed-to-power around multi-year grid-interconnection queues (customers Meta/Microsoft/Foxconn, $1.3B backlog); a direct AI-power play, but lofty at ~32× 2025 sales (vs Caterpillar's 6×) on a $59M loss.
Energy Transfer — building natural-gas pipelines to three Oracle data centers (two in Texas), but serves both oil & gas and is down 15% YTD; the mixed oil/liquids exposure Sanghani favors less than pure-gas names.
Exelon — named with PPL as a Pennsylvania transmission builder caught both ways by Shapiro's order (Aug 19): the data-center slowdown "could reduce how many wires they build… but it could also help them by forcing tech companies to pay for more of the buildout." Unlike PPL, no new-generation vehicle is named to capture the "bring your own power" mandate — it gets the ambiguity without the offset.
Deep Fission — the smallest of the three public microreactor names at "around $400 million" (Aug 26), appearing as part of the count of what is already listed rather than as an Army selection or a criticality-cohort member. Worth watching for a link the article does not draw: its underground siting concept is at least a plausible answer to the piece's single strongest objection — Hyde's point that guarding a small reactor at a remote surface site "becomes almost impossible to have that economically make sense."
GE Vernova — named as the supplier of most of the gas turbines for the Chevron–Microsoft West Texas plants; a direct equipment beneficiary of the off-grid AI-power build-out. Sep 17: its nuclear JV GE Vernova Hitachi named as one of the "seasoned operators" among dozens of reactor developers — the incumbent competition that makes Holtec's SMR push its "riskiest gambit."
Goldman Sachs Group — the fullest illustration of the Dallas build-out: 4,500+ local staff, up from 900 in 2017, with a much bigger campus under construction; its Dallas head, exiled there in 2016, now calls Michael Lewis' "Equities in Dallas" insult "a relic of the early '80s." Capital-committed migration evidence, no stance on the stock.
Foxconn (Hon Hai Precision) — named as an ERock customer; the Taiwanese contract manufacturer, increasingly a data-center builder, buying on-site power.
Holtec Nuclear (private; planned IPO ticker HNUC) — New Jersey nuclear services, equipment and reactor developer: decommissioning old plants, spent-fuel casks, small modular reactors, and the Palisades (Michigan) restart. Sep 17: canceled its IPO (~$900M at a $10B valuation, below a year-ago level) on "market conditions" — plus terms giving CEO Kris Singh majority voting control via Class B shares with no way to oust him or vice chair Martha Singh, which "gave some shareholders pause." Neutral: steady legacy lines, but the reactor arm is "its riskiest gambit" in a crowded field; check the share structure if it refiles.
Intercontinental Exchange (NYSE) — named in passing as having "ramped up" its Texas presence, the third of the venues behind Dallas' boast of being "the only city in the world with three major exchanges"; no view expressed and the Texas venues remain under 1% of US equity volume.
JPMorgan Chase — the scale marker for the headcount shift: more employees in Texas than in New York, home to its new $3B-plus headquarters; part of the 100,000+ financial jobs the Dallas area added in a decade (now nearly 400,000).
Morgan Stanley — the wealth-management pull of Texas' former wildcatters turned "Old Money barons": considering a $1.3B Dallas office complex near the Dallas Fed. Still under consideration, not committed — the lowest rung on the commitment ladder among the migrating banks.
Nasdaq — launched Nasdaq Texas in March as a defensive flag-plant (trades still run on East Coast wires; it "allows companies to establish a listings presence in Texas while maintaining all of the benefits of their Nasdaq listing"). The contested part is the 15-trading-day Nasdaq-100 fast track created around SpaceX's listing — Nasdaq says the change predated it and keeps the index representative.
VanEck Uranium & Nuclear ETF — the sector gauge: down 13% this year (Sep 17), Salzman's evidence that "the euphoria about nuclear that built up in the past two years has worn off" and the "market conditions" behind Holtec's withdrawn IPO.
Nano Nuclear Energy — the largest of the three already-public microreactor names at a $1B cap, in a group that "attracted some interest but their market caps remain modest" (Aug 26). Run the article's own three de-risking tests and it passes none: not one of the five companies the Army selected for its $2.2B / 20-reactor program, not among the four firms that reached criticality this year under the DOE program, and not a Segra position. Not a judgment on the company — a statement that nothing in this news de-risks it.
Radiant Industries — privately-held novel-reactor developer named as part of the small/micro-reactor ecosystem Standard Nuclear's fuel business relies on. Aug 26: the largest single Army award — up to 15 microreactors at Fort Benning for up to $750M — with the company saying it hopes the project will "prove that its technology works and open up commercial opportunities." >$500M raised (Andreessen Horowitz, Chevron Technology Ventures) at a $1.9B Pitchbook mark; high-temperature helium gas-cooled design "meant to have a lower risk of meltdowns." Neutral: private, not in the criticality cohort, and Salzman's own brake applies hardest here — the Army contracts "are not going to justify multibillion-dollar valuations for all these firms," because a base already pays the 24/7 security bill a commercial site would not.
Charles Schwab — the completed-migration example: moved its headquarters from San Francisco to the Dallas area in 2021, an outright HQ relocation rather than a satellite office, in the same tax arbitrage drawing the banks.
SpaceX — the showcase of the Texas governance model: its bylaws force disputes into the Texas Business Court, which its own filings say "may discourage lawsuits against us and our directors, officers…", and it dual-listed on Nasdaq's New York and Texas exchanges — around which Nasdaq created a 15-trading-day fast track into the Nasdaq-100 that critics say hands index investors "inordinate risks."
Standard Nuclear — Tennessee maker of poppyseed-sized TRISO ceramic uranium fuel pellets for the small/micro reactors being built for data centers; packages others' mined/enriched uranium, is already producing/selling and a DOE nuclear-acceleration partner ("more demand than we have capacity"), but its $150M/$15 IPO downsized and broke price ($13.50 open → $12.26 close) on a $7.7M Q1 loss as new-nuclear sentiment soured.
Texas Stock Exchange (TXSE, private) — opened as a trading venue in July, still procuring its own listings; will share the new 30-story Dallas tower with Bank of America's regional HQ. Pitch is subtraction (no "wasteful fees and onerous rules," no board-diversity disclosure, a constitutional ban on securities-transaction taxes) — but the Texas venues together are under 1% of US equity volume.
Tortoise North American Pipeline Fund — sector ETF up just 4% YTD vs the S&P's 16%, the article's proxy for midstream's underperformance despite the record capex boom.
Tesla — the trigger for the Delaware-to-Texas incorporation wave: after a Delaware judge denied Musk's $56B pay package in 2024, he moved Tesla's legal domicile to Texas, "whose rules are considered friendlier to corporations." Cited as the catalyst for the shareholder-rights "race to the bottom," not as a stock view.
Westinghouse Electric (private, pre-IPO) — owner of the AP1000, "the most advanced U.S. design on the market," and "one of the few nuclear developers that actually makes money"; plans to go public at potentially tens of billions while pre-revenue reactor upstarts already trade in the billions. Post-bankruptcy it is asset-light (others build; it licenses the design and sells services, currently earning off overseas reactors), targets construction starts on 10 US reactors by 2030, and carries a US–Japan $80B financing deal in limbo plus a potential government stake/warrants of up to 20%. Aug 26: the incumbent hedges into the small format — it "has also been testing its own microreactor," plans to deploy some at Army bases, and reached criticality this year alongside Valar, Antares and Aalo. The competitive point for the startups: none of them holds a commercial operating licence, and licensing experience is the one capability Westinghouse has more of than anyone. Ownable only via Cameco (49%) and Brookfield.
CoinShares Bitcoin Mining and Digital Power ETF — cited only as the sector scoreboard in the Aug 24 piece: down 16% over the past month as states paused approvals for new data centers looking to hook into the grid. It is the measurement of the dislocation Byrd calls a misunderstanding, and the crudest basket expression of his call — it will not distinguish miners whose value is a protected installed site from those whose value sits in an unapproved pipeline.
X-energy — privately-held advanced-reactor developer in the small/micro-reactor ecosystem; notably it "is also making fuel pellets," making it a direct competitor to Standard Nuclear in the TRISO-fuel niche.
ExxonMobil — best quarter since 2022 (Q2 revenue $116.02B, adjusted EPS $3.52) yet the stock fell 2.1%: a four-cent miss on "elevated" H1 refining maintenance against a +30% YTD run. Record diesel output and $17.2B of quarterly FCF (vs ~$16B expected) confirm Woods' call that the world's fuel deficit persists — "a very robust refining market with very high margins" — but 750k bpd of Middle East production is at risk if Hormuz stays shut through Q3, and BofA cut it to Neutral on fading war momentum ($158 target). Aug 18: named as an integrated "also profiting" from the first-ever triple-digit WTI-to-diesel spread ($101.86) — Woods' call corroborated, but with the upstream Hormuz shut-in as the internal hedge that dilutes the crack windfall versus the pure refiners. Aug 26: named as the deliberate anti-example in the Goehring & Rozencwajg $100-oil interview — they "wouldn't buy Exxon Mobil to play rising oil prices, because it's too diversified in other areas such as chemicals." Not a view on the business (which is earning record cash flow off exactly that diversification) but a rule about instrument selection: when the thesis is one variable, buy the security with the highest sensitivity to it and the fewest offsetting segments. Their expression instead: CNQ/SU (oil sands) and SDRL/SLB (offshore services).
Constellation Energy — named first among the losers of the data-center political backlash (Aug 24): governors slowing or stopping new construction "from New York to Texas" has "hurt stocks of companies that produce electricity for data centers, like power plant owners Constellation Energy and NRG." Structurally the same loss as Talen/Vistra/PSEG five days earlier — the plant keeps running and keeps selling into the market; what disappears is the option on a specialised, premium data-center contract that requires the project to be approved in the first place.
Canadian Natural Resources — the first-named "big bet on Canadian oil producers" through which Goehring & Rozencwajg express their call for crude over $100 for much of 2027 (Aug 26). Selection criterion is reserve life, not price beta: "the Canadian oil sands will be able to sustain production longer than U.S. shale wells, which deplete quickly" — the right asset when the thesis is that shale growth turns negative within months and OPEC "has lost its biggest source of competition," repeating 2002–08 when oil rose four- to five-fold. The managers compound 14% annualised since 2015 and 21% over five years, but are up only 6.4% through July 31 this year, having missed the refinery rally. Aug 31 — the same attribute, opposite sign: named in the lede of the Venezuela piece as a producer of "heavier crude that competes against Venezuelan resources," with Canadian producers "among the most vulnerable" if the heavy market "gets flooded with new supplies." Reserve life and grade are different attributes: long life protects the volume thesis and does nothing for the differential. Owning CNQ for the $100-oil call also means being short the heavy-light spread — plus TPH's second-order risk that a US-sponsored Venezuelan revival erodes Canada's tariff-exemption "card" "even before physical production grows."
Cenovus Energy — third of the four Canadian heavy producers named as losers from the Venezuela deal (Aug 31): "most Canadian crude is on the heavy side, including the country's enormous reserves in Alberta's tar sands," so a heavy market "flooded with new supplies" would "weigh on profits." Its partial structural defence is integration — it owns refining capacity, so some of the margin a Gulf Coast refiner captures from cheaper heavy crude is recaptured in-house; a producer-refiner sits on both sides of the spread. The larger, slower risk is TPH's: Canadian energy keeps its U.S. tariff exemption because it is an "important feedstock for U.S. refiners," and a credible substitute erodes that card "even before physical production grows."
Imperial Oil — fourth of the Canadian heavy producers exposed to new Venezuelan supply (Aug 31), and the one where the article's own caveat cuts most in its favour: TPH's Jeoffrey Lambujon notes "Midwest refineries remain dependent on Canadian crude" — pipeline-fed inland plants that seaborne Venezuelan cargoes landing on the Gulf Coast cannot easily reach. So read the loser side as graded, not uniform: producers competing for waterborne Gulf Coast demand are most exposed, those selling into captive inland markets least. The risk no Canadian producer escapes is strategic — a credible alternative supplier weakens Canada's tariff-exemption leverage regardless of which refinery buys which barrel.
J.B. Hunt Transport Services — the named casualty of record diesel (Sep 16): warned late Tuesday that soaring diesel and other costs will cut Q3 profit 5–10% versus Q2, CFO Brad Delco citing "some of the most radical and abnormal swings in fuel prices," and the stock fell 13%. The mechanism is the fuel-surcharge lag: with diesel +86¢ in a month to a record $6.31 (GasBuddy sees $7 in several states), the trucker pays before it can bill. Negative while diesel keeps accelerating; a flattening price — or an export ban that holds more diesel at home — is what lets the surcharge catch up.
NRG Energy — the second power-plant owner named as hurt by the approval freeze (Aug 24). Held next to the miners it makes the article's structural point clean: two companies can both be "AI power" and sit on opposite sides of one rule. The one that needs a new project approved to monetise its megawatts loses; the one already behind an existing interconnection gains, because the freeze is a supply cut on its competition.
Public Service Enterprise Group — third of the existing-plant owners named in Zimbardo's negative read on Shapiro's order (Aug 19). Same mechanism as Talen and Vistra: a lost option on a premium data-center contract rather than a hit to current earnings. Keep it in proportion — one state, by executive order, reversible at the next election — but the direction of travel (NY moratorium, TX audit, PA order) narrows the "sell my existing megawatts to AI" story in several places at once.
Suncor Energy — the second Canadian producer in Goehring & Rozencwajg's oil-sands bet (Aug 26), chosen on the same axis: long-lived, slow-decline barrels against a fast-depleting U.S. shale base whose growth Goehring expects to turn negative "in the coming months." The method generalises — when a forecast is about a commodity over five years, company selection collapses to which producer's output survives long enough to sell into it. Fast-declining assets deliver a good quarter; slow-declining assets deliver the thesis. Aug 31 — the reversal, five days later: named first among the losers from the U.S. takeover of 17 Venezuelan oilfields, because oil-sands barrels are "heavier crude that competes against Venezuelan resources" and new heavy supply widens the discount Suncor is paid. Same characteristic, opposite sign — the grade that makes the reserves long-lived is the grade that competes with the new barrel. The exposure is to the differential, not the benchmark.
Talen Energy — named first among owners of existing Pennsylvania plants for whom Jefferies' Paul Zimbardo thinks Shapiro's executive order "could be a negative" (Aug 19). The co-location trade — selling output from an already-built plant to a data center under a specialized premium contract — is what the "build or bring your own power" requirement closes, and the order applies "whether or not they connect to the larger electric grid," catching the behind-the-meter workaround. What is lost is an option on a contracted premium, not current earnings — but that option is much of what the market has been paying for.
Vistra Energy — second of the three existing-plant owners Zimbardo flags as hurt by Pennsylvania's new rules (Aug 19): the hoped-for specialized data-center supply contracts off plants already standing are the channel the "bring your own power" mandate shuts, with the order's scope clause explicitly reaching projects that never touch the public grid. Plants keep selling into the market; the contracted uplift is what goes.
In one line: Avi Salzman is a Barron's writer, not a money manager — this archive captures the named securities and cross-cutting energy/markets views his reporting surfaces (often via the analysts he quotes), tracked the same way as the commentators in the hub. The beat is energy supply meets AI-era power demand: midstream/pipelines (leaning on Westwood's Parag Sanghani and Tortoise's Rob Thummel — a record build-out, lagging stocks, "favor natural gas over oil/liquids"), on-site and off-grid generation for data centers, the new-nuclear supply chain and its IPO wave, and — since the Iran war — oil-major earnings, refining margins and the Hormuz supply shock. As of Aug 2026 the beat widens from the molecules to the jurisdiction they sit in: Texas as a hypercapitalist experiment — the HQ/incorporation/listing migration, the shareholder-rights "race to the bottom" it has started, and water as the physical limit on the AI build-out. By mid-August the energy thread resolves into its cleanest expression yet: the war trade has moved out of the barrel and into the margin — a first-ever triple-digit WTI-to-diesel crack, refiners at all-time highs, and a global refining fleet flat-to-shrinking through 2027 — while Washington's only answer to a $4.11 pump price is an early winter-blend waiver worth 10–30¢ for a few weeks, i.e. the policy lever is still pointed at the consumer price, not the refining margin. Alongside it a second front opens: state-level pushback against data centers (a New York moratorium, a Texas audit, Pennsylvania's "bring your own power" order) that delays rather than derails the AI-power trade — and redistributes it from owners of existing plants to whoever can build the newly-mandatory dedicated generation. By the end of August that redistribution has a general form: because the moratoria act on new interconnection requests and not on operating assets, an already-connected site is the asset the backlash marks up — which is why Morgan Stanley's Stephen Byrd calls the Bitcoin miners a safe harbour and the power-plant owners the casualties. The energy thread meanwhile splits in two: the refining-margin trade the archive has been tracking, and a contrarian crude call — Goehring & Rozencwajg reading diesel tank bottoms and a U.S. shale rollover to $100+ oil through 2027, expressed deliberately in long-life Canadian barrels and restructured offshore services rather than a diversified major. And the nuclear thread reaches its demand side: the U.S. Army becomes the microreactor industry's first real customer. By Aug 28 the war trade has a clean physical explanation: the Hormuz blockade has been routed around rather than lifted — regional Gulf exports back to ~80% of prewar even with strait traffic at half — which is why crude sits in a "not-so-bad equilibrium" near $89 and why the whole geopolitical hedge has migrated from the barrel to the constrained processed product (diesel, LNG). The archive's single most durable read of 2026: in this cycle the scarcity is capacity, not crude.
Reporter, not advisor. Stances in this archive reflect how each name is framed in the article (and the sourced analysts' lean), not a personal price target. Treat the views as attributed to the quoted experts.
Pipelines / midstream (Dec 2, 2025). A record pipeline-building boom — ~$53B of growth capex (past the 2019 peak) on LNG exports + data-center power demand — but the stocks lag (Tortoise fund +4% vs S&P +16%) on weak oil and 2015-bust memory. The screen: favor natural-gas pipelines (Williams, Kinder Morgan) over oil/liquids transporters (Energy Transfer, Oneok); bigger cash-flow cushions than 2015 de-risk the dividends.
AI power — on-site generation (Jun 10, 2026). ERock (EROC) IPOs on the NYSE: on-site natural-gas "rockblock" gensets that give data centers speed-to-power around multi-year grid-interconnection queues (customers Meta/Microsoft/Foxconn, $1.3B backlog; competitors Caterpillar, Generac, Cummins). A direct AI-power play — but down 5% on debut and lofty at ~32× 2025 sales (vs Caterpillar's 6×) on a $59M loss. Framed Neutral: real demand vs a price that needs the ramp to show up.
AI power — oil major as power seller (Jun 22, 2026). Chevron (CVX) signs a 20-year deal to sell 2.67 GW to Microsoft (MSFT) from new off-grid West Texas gas plants (turbines from GE Vernova + a Caterpillar unit; co-developed with Engine No. 1). The investment case Salzman frames positively: it monetizes stranded, negatively-priced regional gas as steady ~mid-teens-return contracted cash flow, decoupling Chevron from oil's boom-and-bust and giving it an AI-power foothold — yet the stock barely moved ("it arguably should"). Gating items: no final investment decision yet, permits pending, power not until 2028.
Nuclear fuel supply chain (Jul 16, 2026). Standard Nuclear (STDN) IPOs: a Tennessee maker of poppyseed-sized TRISO ceramic uranium fuel pellets for the small/micro reactors being built for data centers — packaging others' mined/enriched uranium into fuel. Salzman frames it as a potential "central player in the new nuclear supply chain": already producing/selling, a DOE nuclear-acceleration partner, CEO says demand exceeds capacity. But the deal downsized and broke price ($150M at $15 → $13.50 open → $12.26 close) on a $7.7M Q1 loss, as investors soured on new-nuclear stocks. Its fate rides on the reactor ecosystem (Oklo, Radiant, X-energy — the last also a fuel-pellet competitor). Framed Neutral: real strategic niche vs a cooling trade and an unproven ramp.
Nuclear scale-up — the profitable incumbent lists (Jul 31, 2026).Westinghouse, owner of the AP1000 ("the most advanced U.S. design on the market"), says it plans to go public — potentially at tens of billions, because it is "one of the few nuclear developers that actually makes money," versus upstarts valued in the billions "despite having no approved designs or significant revenue." Owned by Brookfield (BN) and Cameco (CCJ, 49%), whose filings are the only window on the numbers: a C$3.46B ($2.47B) revenue share in 2025, +20% y/y → ~$5B group revenue, against BWXT's $2.35B revenue at a $15B value (~6.4× sales). Post-bankruptcy it is asset-light — others build, it licenses the AP1000 and sells services, currently off overseas reactors — targeting construction starts on 10 U.S. reactors by 2030. Overhangs: the U.S.–Japan $80B financing deal is in limbo, leaving a potential government stake (20% of distributions above $17.5B, or warrants up to 20% of the stock) unresolved. Framed Neutral: a real earner arriving into a theme whose last IPO (STDN, two weeks earlier) broke price.
Oil majors — the war windfall and its expiry (Jul 31, 2026).ExxonMobil (XOM)'s Q2 is its best since 2022 — revenue $116.02B, adjusted EPS $3.52 (from $1.64), $17.2B of free cash flow (more than the prior three quarters combined) — and the stock still falls 2.1%: a four-cent miss on "elevated" H1 refining maintenance, into a +30% YTD run. The operating read is strong and structural — record diesel output into a war-driven shortage, and CEO Darren Woods saying the fuel deficit persists ("a very robust refining market with very high margins"). The offsets are geopolitical and positional: Exxon itself quantifies a 750,000 bpd y/y Middle East production hit if Hormuz stays closed through Q3 (~20% of volume there shut in, Qatar unclear), and BofA's Salisbury cut it to Neutral on the war moving toward resolution while raising her target to $158. Also flagged: XOM trades at a premium to CVX on cash flow despite similar prospects. Framed Neutral — the next tell is how the cash gets deployed (buybacks/dividends favored over M&A; Guyana advancing).
Texas hypercapitalism — the jurisdiction trade and its two limits (Aug 7, 2026). The reported survey that widens the beat from assets to venue. The pull is quantified: 57 Fortune 500 HQs (past California this year), most jobs added of any state three years running, the world's eighth-largest economy, more gas plants under construction than the next seven states combined plus leading renewables, and a Corpus Christi port shipping ~2.5m bbl/d of crude (half of US exports, up tenfold in a decade) — the migration evidence running from Schwab (SCHW)'s outright HQ move and Goldman (GS)'s 900→4,500+ Dallas staff to JPMorgan (JPM) having more Texas than New York employees, Morgan Stanley (MS)'s mooted $1.3B complex and the Bank of America (BAC) tower that will house the Texas Stock Exchange. The arbitrage is explicit — California taxes bank income at 10.8%, Texas at zero (0.75% margin tax). Limit 1 — governance. Texas has extended the pitch to legal domicile and listing: after Delaware voided Musk's $56B package Tesla (TSLA) reincorporated there; companies may block sub-3% holders from proposals and certain suits, proxy advisors are curbed, and SpaceX (SPCX)'s bylaws route disputes to the Texas Business Court, which its filings concede "may discourage lawsuits." Nevada and Delaware have already followed — "a race to the bottom" (AFR's Natalia Renta) — while the venue war stays cosmetic (three Dallas exchanges, under 1% of US volume; Nasdaq (NDAQ) Texas trades still run on East Coast wires) except for one real lever: the 15-trading-day Nasdaq-100 fast track created around SpaceX's listing. Limit 2 — water, not power. Corpus Christi is running dry (Choke Canyon 8% full): residents cut usage 19% while large users cut 3%, industry takes ~60% of city water, ExxonMobil (XOM) is the biggest user at ~13m gal/day and Valero (VLO) fights a doubled industrial rate on blockbuster earnings. Statewide, a severe drought means a ~20% supply shortfall by 2030, with data centers going <1% → 3.7% (2030) → 9.1% (2040) of state water use — and Gov. Abbott has halted new data-center approvals pending a water audit. Reportage, so all names Neutral; the transferable read is that the constraint on the Texas AI build-out is water, and the cost of the tax/regulatory arbitrage is paid in shareholder recourse.
Refining margins — the war trade moves from the barrel to the crack (Aug 18, 2026). Crude has chopped on Iran headlines; the fuels made from crude have gone one way. Wholesale diesel is +109% YTD, pump diesel $5.47 (+7% in a month, +48% y/y), and the WTI-to-diesel spread hit $101.86 — the first triple-digit crack on record, with U.S. refiners earning triple last year's per-barrel profits (OPIS). The named beneficiaries — Valero (VLO), Marathon Petroleum (MPC) and Phillips 66 (PSX) — are all at all-time highs; integrateds with refining arms like ExxonMobil (XOM) "are also profiting," though Exxon's upstream carries the offsetting Hormuz shut-in. The cause is that the product market is more disrupted than the crude market: Hormuz removes ~20% of world oil, Ukrainian drones took 2.8m bbl of Russian refining capacity offline as of July (BofA), Russia has restricted diesel/gasoline exports until next year, and China has cut fuel exports — leaving the U.S. as exporter of last resort at record weekly diesel exports that still can't supply the world. With fuel reserves already low and harvest season accelerating, BofA's Michael Widmer sees diesel "tight, volatile, and expensive well into next year." The structural leg is Melius' James West: more refineries closing than opening, EV substitution too slow to balance the market, the global fleet flat-to-shrinking through 2027 and the next openings only in 2028–2030 — "structurally higher refining margins over the next two years." Framed Positive on the refiners; the single named exit trigger is a full resolution of both the Iran and Russia conflicts.
The state pushback against data centers — delay, not derail, and a transfer of value (Aug 19, 2026). The AI-power theme meets its regulatory counterweight, and it is bipartisan: a one-year moratorium in New York, an industry "audit" in Texas pausing grid connections, and Pennsylvania Gov. Josh Shapiro's executive order forcing data centers to bring their own power sources and secure community buy-in before the DEP will consider an application — applying "whether or not they connect to the larger electric grid," and stripping every data-center project out of the state's grid "fast track." Johns Hopkins' Abe Silverman: "We are a split country, except on this one issue." Shapiro is not hostile (no ban, no moratorium; he praised two Amazon projects) and BTIG's Alex Kania sees no cancellations — the state has 100+ proposals, surplus generation and abundant gas — so the effect is timing, plus a redistribution. Loser: owners of existing in-state plants — Talen (TLN), Vistra (VST), PSEG (PEG) — whose hoped-for specialized co-location contracts the own-generation mandate closes (Jefferies' Paul Zimbardo); what dies is the option on a contracted premium, not the megawatt. Two-sided: the transmission builders PPL and Exelon (EXC) — fewer data-center wires, but tech companies forced to fund more of the build-out. Winner: PPL, via the Invitium Energy JV with Blackstone (BX), which has already reserved 5 GW of gas turbines — the scarce complement a "bring your own power" rule makes compulsory — with CEO Vincent Sorgi expecting a first developer deal by year-end. The transferable rule is the closing line: "power companies that can help them catch up to the new political reality should profit."
Gasoline politics — the lever is the blend, not the margin (Aug 21, 2026). The EPA lets stations sell cheaper winter-grade gasoline from Sept. 1 instead of mid-September — E10 with more volatile, cheaper components like butane — which it says could add "hundreds of thousands of extra barrels of gas per day" and cut 10–30¢ a gallon (AAA) against an average of $4.11, about $1 above a year ago. The trade-off is more smog (the summer volatility rule exists for ozone), and the reach is limited: the waiver is federal but California and New York haven't waived their own stricter rules, and the underlying cause is that "oil and fuel supplies have been strained by the wars in Iran and Ukraine." The equity clause is the tell — refiners "like Valero (VLO) and Marathon Petroleum (MPC) are making very high margins on the fuel they sell amid elevated prices at the pump" — and so is the choice of instrument: a cheaper, higher-volume blend relaxes a refiner constraint. When the political response to expensive fuel is still a specification waiver rather than an export ban, price cap or windfall levy, the margin is not yet the target. Watch the tool, not the pump price.
Grid access as the scarce asset — the backlash marks up the incumbent (Aug 24, 2026). The Aug 19 redistribution argument generalised, by Morgan Stanley's Stephen Byrd. Governors "from New York to Texas" have slowed or stopped new data-center construction, hurting the power-plant owners who needed those projects approved — Constellation (CEG) and NRG. But the moratoria act on new interconnection requests, not on electricity already flowing to a built site, so the companies already behind an existing connection are revalued upward: "most of the miners already have access to the electric grid at their existing sites… being connected to the grid is worth more money now than ever before," and every cancellation or delay "just makes these companies' sites more valuable." The unlikely beneficiaries are the Bitcoin miners — Cipher (CIFR), Hut 8 (HUT), Galaxy Digital (GLXY), MARA and Riot (RIOT) — who queued for those connections years ago and have spent two years converting warehouses into AI data centers, renting out the power or the shell for returns "steadier than Bitcoin." Having rerated into the AI-power theme, they sold off with it (WGMI −16% in a month) on a risk Byrd says they do not carry: inherited beta, not inherited exposure. The one kept caveat splits the exposure cleanly — "some new Bitcoin-to-AI projects in states like Texas may see delays," so the installed footprint is protected and appreciating while the pipeline faces the same queue as anyone's greenfield.
The contrarian crude call — tank bottoms, shale rollover, $100 oil in 2027 (Aug 26, 2026).Leigh Goehring & Adam Rozencwajg (14% annualised since 2015, 21% over five years — but +6.4% through July 31 this year, having missed the refinery rally this archive documented) argue the war's non-spike is misleading. Six months in, Hormuz is "mostly closed" yet Brent trades just under $90, banks see $78 next year and the curve $77 — while "the relative calm in the broader oil market is masking much deeper stress." The measurement is physical, not price: diesel near record highs, refineries in Russia, China and the Middle East making less of it, stockpiles "near tank bottom," and buyers "panicked throughout the world." The non-obvious consequence — the bullish event is the resolution: when the strait reopens, countries import more crude to refill inventories, demand stacked on consumption, so "the damage is already done." Rozencwajg's analogue is February 2020: alarmists look wrong for two months "and then it hit all at once." Supply cannot answer — U.S. shale growth turns negative "in the coming months" for lack of investment and "no other oil project in the world can make up for diminishing U.S. output," so "OPEC has lost its biggest source of competition… repeat exactly what happened between 2002 and 2008, where oil prices basically went up four- to five-fold" — crude over $100 for much of 2027. The expression is the transferable part: equities not futures (a multi-year view cannot survive roll and margin), and deliberately concentrated — they "wouldn't buy ExxonMobil (XOM)… because it's too diversified." Vehicles: Canadian Natural (CNQ) and Suncor (SU) for oil-sands reserve life against fast-depleting shale, and Seadrill (SDRL) / SLB for offshore services off a fleet the last downturn destroyed. Closing discipline: "Just because it hasn't happened yet, doesn't mean it won't."
Microreactors get their first customer — the U.S. Army (Aug 26, 2026). The nuclear thread (TRISO fuel in July, the Westinghouse listing at month-end) reaches demand. The Army selected five companies to build microreactors at bases — up to $2.2B for at least 20 reactors, first switch-on "as soon as September 2028" — machines that "fit on the back of a truck" and power 500–1,000 homes versus the 1m-home behemoths. Selected: privately held Antares, General Atomics EMS and Radiant (up to 15 units at Fort Benning for $750M), plus BWXT; and Westinghouse — owned by Cameco (CCJ) and Brookfield Renewable — "plans to deploy some too." The buyer's objective, per Army Secretary Dan Driscoll, is "energy resilience… without relying on potentially vulnerable external grids" — resilience, not cost, which is why the least price-sensitive customer is first. Three usable screens come out of it. (1) Walk-away cost: big reactors take a decade and "over $10 billion each," while "even if a microreactor project fails, the cost of walking away will be in the millions, not the billions" — Segra's Arthur Hyde: "at a minimum, this reactor is not making me bankrupt." (2) Criticality as the binary de-risking milestone — Valar, Antares, Westinghouse and Aalo reached it this year under a DOE program, "equivalent to getting a car to successfully start when you turn the key" (no electricity yet). (3) Security-cost economics as the commercial screen: 24/7 armed guarding does not scale down, so a lone remote unit "becomes almost impossible to have that economically make sense" — bases already pay that bill, commercial sites do not, and Third Way calls the whole approach inefficient. Hyde's constructive exception is aggregation: Segra owns Valar Atomics and Terra Innovatum (NKLR), and he thinks Valar can put "at least 30" reactors on one site, with gas-cooled process heat widening the market. Note the valuation inversion — private marks (Valar $6B, Antares $2.1B, Radiant $1.9B) sit above the public trio (NNE $1B, NKLR $625M, FISN ~$400M), and the criticality cohort is entirely private. Salzman's brake: "the Army contracts are nice, but they're not going to justify multibillion-dollar valuations for all these firms." The open question that ends the piece is the whole thesis: "for the industry to succeed, it will have to convince other buyers too."
Hormuz — the blockade got routed around, not lifted (Aug 28, 2026). The war's oil story resolves into a flow-accounting problem, and it is a macro-only piece — no securities named. Capital Economics' Kieran Tompkins: Persian Gulf outflows, "using the strait and alternate exit routes," are back to about 80% of prewar levels, leaving a "not-so-bad equilibrium" that explains Brent near $89 despite "little progress" in U.S.–Iran talks. Goldman's Daan Struyven puts crude-plus-products at 15–16 mb/d (still 7–8 mb/d under prewar) while strait traffic is only 8–10 mb/d, roughly half prewar — the divergence is the whole insight: track the region's exports, not the chokepoint. The gap is closed by adaptation, not supply — night "dark crossings" with transponders off "aided by the U.S. military," ship-to-ship transfers onto Asia-bound tankers, and routes around the strait or out via the Red Sea: "producers and shippers are adapting." Two forward reads follow. (1) The duration signal is in freight, not crude: "elevated shipping rates indicate that oil companies are preparing for the strait to remain at least partially closed into next year." (2) The instrument changes: because crude is in a "strange semi-equilibrium" — enough moving to prevent a spike, not enough to let prices fall — Goldman tells clients to "turn to other products to hedge against geopolitical disruption," recommending diesel futures (refineries in Russia, the Middle East and China running at reduced capacity keep diesel richer than crude) and expecting LNG to stay high on short global supply plus Europe's winter stockpiling deadline. This is the third independent desk in eleven days — with Aug 18's record triple-digit diesel crack and Aug 26's "tank bottom" stockpiles — locating the bottleneck in refining capacity rather than barrels. Consumer read unchanged: rebounding supply "could keep gasoline prices from jumping," but "consumers shouldn't expect much near-term relief at the pump."
Venezuela — a winners/losers chain that runs entirely through crude grade (Aug 31, 2026). Trump says the U.S. has "inked a partnership with a private company" for 65 billion barrels of Venezuelan reserves — more than America's own 46 billion barrels of proven reserves — "at no cost to the American Taxpayer," paid per the WSJ in warrants that cost nothing up front (the White House would not discuss terms). President Delcy Rodriguez says it covers 17 oilfields and could add 1.5m bbl/d. Salzman's brakes are stacked three deep: "none of that will happen with a snap of the fingers" — it needs "tens of billions of dollars worth of private investment" from a payer nobody has named; the plan must be shown "legal, enforceable and financially sound"; and the term is unresolved (early reports said 100 years, Rodriguez says 25) — "for producers putting billions of dollars at risk, the length of the contract will matter." Jefferies' Alejandro Anibal Demichelis: "legal and execution risks remain high, with any major production uplift likely years away," modelling ~1.1m bbl/d today to 1.5m by end-2027 — i.e. the official increment is roughly the analyst's total. What is already happening needs no deal at all: half of Venezuela's oil goes to the U.S. and imports ran 137k bbl/d (Jan) → 544k (May) → ~700k (mid-July). The transmission channel is grade, not price: heavy, sour barrels "trade at a discount," Gulf Coast plants are built for them and "the more Venezuelan crude they can purchase, the wider those margins are likely to get." Winners — Valero (VLO), the top refiner of Venezuelan crude per TPH, then Chevron (CVX) and Phillips 66 (PSX), with heavy-capable PBF and Marathon (MPC); services SLB and Halliburton (HAL), which have Venezuelan operations and, "because they're American companies," may get "preferential treatment to operate on U.S.-controlled land"; and one LatAm wildcard, GeoPark (GPRK). Losers — Suncor (SU), Cenovus (CVE), Imperial Oil (IMO) and Canadian Natural (CNQ), whose Alberta barrels are the competing heavy grade. Two readings carry beyond the trade. (1) The archive contradicts itself productively: five days earlier CNQ and SU were the chosen vehicles for $100 oil because oil sands are long-lived heavy barrels. Reserve life and grade are different attributes — long life protects the volume thesis and does nothing for the differential. (2) TPH's Jeoffrey Lambujon supplies the second-order point: a U.S.-sponsored Venezuelan revival "supports the alternative to Canadian heavy barrels… which may erode the value of that card over time for Canada even before physical production grows" — Canada's tariff exemption exists because its oil is an "important feedstock for U.S. refiners," so a credible substitute damages the bargaining position years before it ships a barrel. The bound is in his own parenthesis: "Midwest refineries remain dependent on Canadian crude." Finally, the positioning tell: VLO +2% (ranked #1, benefit already accruing) versus GPRK +13% (mentioned once, every gate still shut).
The bypass becomes the target — Saudi East-West pipeline hit (Sep 11, 2026). The Aug 28 "routed around, not lifted" equilibrium meets the event that breaks it. Projectiles struck the pumping station of Saudi Arabia's East-West pipeline to the Red Sea — "the biggest pipeline in the Middle East," diverting 7 mb/d around Hormuz with up to 5 mb/d exported from the Red Sea terminal — and the kingdom shut it; the Saudi foreign ministry believes the strikes came from Iraq (Iran-linked militias), with the Houthis also hitting Red Sea infrastructure. Damage undisclosed. Brent +8.7% on the week, then +4.4% to $109.23 — ~$20 above the Aug 28 level. Rystad's Janiv Shah: "supply is becoming a prized commodity." Salzman's call: "fewer and fewer places to hide from Iranian violence," so prices are "almost certain to stay high." The structural read: the region is building more bypasses (UAE capacity, an Iraq–Syria–Mediterranean route, Chevron (CVX) possibly helping) and Bessent calls the strait "worthless" in two years — "but if the pipelines are vulnerable, too, Iran could retain its power over the market." Names: Saudi Aramco (2222.SR) and CVX, both Neutral.
Record diesel reaches the income statement — and the export-ban lever (Sep 16, 2026). The product squeeze the archive has tracked since Aug 18 sets a record at the pump: $6.31/gal (AAA), +37¢ in a week and +86¢ in a month versus $3.70 a year ago, with GasBuddy's Patrick De Haan seeing $7 in several states and NEADA expecting heating-oil bills of ~$2,520 (from $1,749). The first named casualty is J.B. Hunt (JBHT): diesel and other costs cut Q3 profit 5–10% vs Q2, stock −13% — Negative. The cause is unchanged (Iran war flows plus Ukrainian strikes on Russian refineries and Russian export curbs); the U.S. is exporting 1.61m b/d (vs 1.25m average) with refineries near full capacity and distillate stocks ~13% below last year. What changes is the policy rung: from Aug 21's blend waiver to a first-ever diesel export ban — SoFi's Liz Thomas calls the odds before the midterms "high," Senate Majority Leader John Thune is open to it — which, unlike a waiver, aims at the export-margin channel; producers warn it would eventually curb supply and raise prices.
The nuclear IPO wave stalls — Holtec withdraws (Sep 17, 2026). The archive's nuclear-listing thread (STDN breaking issue Jul 16, Westinghouse's plan Jul 31) gets its first outright withdrawal: Holtec Nuclear (planned ticker HNUC) canceled a ~$900M raise at a $10B valuation on "market conditions" — "the euphoria about nuclear that built up in the past two years has worn off," with NLR −13% and comparable BWXT −19% YTD — plus its own terms: CEO Kris Singh's majority voting control through Class B shares, with no way to oust him or vice chair Martha Singh, "gave some shareholders pause." The business (decommissioning, waste casks, SMRs, the Palisades restart) is real; reactor development is "its riskiest gambit" against dozens of rivals incl. GE Vernova Hitachi. The flow read: pure-play upstarts X-Energy (XE) and Oklo (OKLO) both +12% on Holtec's exit (BWXT +0.5%), though their reactors are years away with "significant regulatory hurdles." Salzman's close: not "a sign it's time for investors to abandon the industry."
Articles
One dated page per article — each has its full stock table, talking points, and the saved text. Newest first.