Energy / natural gas Bullish ▲ — and the storage layer is the tightest link

AI / data-center power demand meets trapped, cheap North-American gas; multi-year FCF + buyback story for the E&Ps. Rule: ~$1B/day of sustaining-capital underinvestment makes today's oil prices structural. Muir: turned bullish at the Venezuela bearish-rush capitulation — oil can't go much lower without killing US energy (Hamm not funding new wells); energy is "one of the few places you can hide in because no one owns it," and oil stocks would outperform even in a downturn. Hay (Jun 11, his highest-conviction call): US gas at a ~90% discount to global LNG ($16–20/MMBTU Platts/TTF vs gas "still trading in the threes") is "way too big an arbitrage" — 37 GW of data centers under construction (+146 firm-committed), 20% of global LNG offline for an extended period, EU storage run down, US export terminals building, and positioning still bearish; for the next few years "gas is the only viable solution" to the power crunch. Hay (May 29 POW): his pure-play expression is Expand Energy (EXE) — the largest US gas producer at ~4.2× EV/EBITDA / ~10x P/E with a ~15% FCF yield, deleveraging fast and adding LNG (Delfin) optionality; target $130-132. Larson (Jun 16): "5 years into a 15-year bull cycle for energy" — decades of underinvestment meet a supply-shock world (depleted strategic reserves), and private equity is "kicking the tires" on Canadian energy assets like never before ("never seen more private equity" in Calgary); plays it via the producers (CNQ — "ride Murray Edwards"), oil-field services (Trican) and long-reserve-life names (Athabasca, ~90-yr reserve life at 6–7× cash flow). Rule (Jun 18): energy demand "takes care of itself" — AI may solve half the energy problem, but ~1B people still lack primary electricity (solved over the next 20 yrs); 45 years and ~$10T of alternative-energy investment moved fossil fuels' share only from 83% to 81%, so the projected 2050 consumption "we just can't do" — supply-constrained for decades. Salzman (Barron's, Jun 22): a concrete monetization of the trapped-gas thesis — Chevron (CVX) signs a 20-year deal to sell 2.67 GW (>1M homes' worth) to Microsoft (MSFT) from new off-grid natural-gas plants in West Texas, where a pipeline shortage has pushed local gas to negative prices; long-term offtake converts that stranded gas into steady ~mid-teens-return cash flow (turbines mostly from GE Vernova + a Caterpillar unit; co-developed with Engine No. 1). Power expected by 2028; final investment decision & permits still pending. Hay (Jun 22): reiterates energy equities are "extremely undemanding" on both P/E and Price/Sales — with WTI ~$50 in 2011 dollars (inflation-adjusted) against crisis-low global inventories — and sees a multi-year "valuation elevation" toward growth multiples; go-long names are APA (broke its long-term downtrend), FANG and new name HAL (an oil-service proxy expected to break out) and gas-levered RRC, staying more bullish on natural gas than oil long term. Robotti (RWH046, 2024-JUN-22; SumZero, 2025-JUL-24): the structural case is North America's cheap, non-exportable natural gas — a multi-decade cost edge for energy-intensive industry (steel, cement, ammonia/fertilizer) over the rest of the developed world, independent of IRA carrots or tariff sticks. On oil he flags the "first truly global energy crisis": ~100M bbl/day demand vs only ~2-3M spare, and Saudi spending +60% just to add ~1M bbl/day signals a mature field "on a treadmill" — a tightness unpriced by the market. His book is offshore picks-and-shovels (Tidewater, TechnipFMC, Subsea 7) valued off replacement cost, where supply is being destroyed (no one builds a 30-yr asset into terminal-decline uncertainty). Rule (Jul 5): oil & gas is near-term soft (demand destruction + a Mideast-peace premium unwind) but structurally strong into ~2029–30 — Exxon (XOM) the disciplined big-cap way in, plus a reintroduced small-cap frontier offshore-exploration theme. Paulo Macro (2026-JAN-18, back-fill): the long-term US-gas bull (tight winter balances + LNG buildout + summer datacenter gas-turbine demand) played via Antero (AR — the largest position, lightly hedged), with RRC/EQT/CRK/CNX the screen; near-term, a Sudden Stratospheric Warming set up February cold blasts. Paulo Macro (2026-FEB-02, back-fill): the SSW delivered — gas ripped $3→$7+ in a week, storage neutralized to ~100bcf below the 5-yr average, and a second SSW (a full polar-vortex split ~Feb 8) promised a "Top 5 cold February" with late-injection "drag the trough" potential; watching for a gas re-entry. Rule (Jul 8): still long oil stocks, not selling — the Iran war priced a shortage the world "missed by a week or two" but never suffered; the real one begins 2029–30, structural not war-driven, from ~$1B/day of deferred sustaining capex now made worse by Gulf war damage (Iran can't repair Kharg Island; Qatar/Saudi/Kuwait/UAE facilities need rebuilding). All-in cost of a barrel (incl. cost of capital + taxation) exceeds $60 vs a ~$55 price — the industry "makes it for 60, sells it for 55… 102 million times a day," so it must eventually earn its cost of capital "or your car won't start." Doesn't feel good about oil 2026–27, maybe 2028; feels "real good" 2029–30 — and the discipline is to buy the lower-yield reinvestors, not the popular high-distribution names "cannibalizing themselves" (think 5 years, not 5 weeks). Murti (Jul 11): LNG is the recognized growth market whose capex is already diversifying (the US Gulf Coast build-out, possible Canadian expansion), so supply self-heals; Qatar's Ras Laffan (~17% of exports) is a concentrated but manageable geopolitical risk. Hay (Jul 20): the thermal-coal sell-off on Strait-of-Hormuz-reopening hopes is "misguided" — coal competes directly with LNG, and with LNG once again "trapped in the Persian Gulf" the need for coal as the substitute supply is "becoming increasingly apparent," reinforced by "extremely bullish" supply news out of Indonesia (the top thermal-coal exporter); reiterates his two coal names as urgent adds (Yancoal YACAF −23%, New Hope NHC ~flat — "time is of the essence"), full story in the coming Friday POW!. Smith (Chronometer, Jul 21): the granular well-level version of the US-gas-deficit call — after modeling nearly every US gas well/pipeline/processing asset, the ~20 BCF/d of achievable new supply is already spoken for by the permitted 15→35 BCF/d LNG-export ramp, so AI-compute demand draws working storage below all recorded history by 2029 and the gas price goes "convex and unbounded"; the flat ~$3.50 forward curve is pure complacency (EQT shutting in gas "for later," no rig response, back years illiquid until utilities start hedging 2028 — then "a knife fight to secure natural gas physical"). Best-producer longs Expand Energy (EXE, ~4× EBITDA / ~70% of core Haynesville) and Range (RRC). Full note under the new "US natural-gas deficit / AI-power (2028+)" theme. Hay (Jul 23, Haymaker Daily): Europe's gas in storage is near the lowest of the last 15 years (per a John Kemp chart) — the post-2022 record stockpiles were bailed out by warm winters now being mistaken for permanence — just as escalating Mideast hostilities threaten energy shipments including LNG; with U.S. LNG exports "close to maxed out" and Asia "in dire need of frozen gas" competing for the same cargoes, an average winter — "especially a severe one" — "could prove extremely problematic for the Continent." European authorities "may soon regret" their blasé attitude toward recurring energy needs. McCrea (BMO, Trevor Rose Jul 23): Canadian gas carries "probably some of the most negative sentiment I've seen in gas in a long time" — the contrarian setup: Canada's pure-play gas economics beat the US across BMO's ~200-play database (F&D ~25–50¢/mcf, ~25¢ operating — "almost free," the arbitrage works even if global LNG prices fall), and ~8.2 bcf/d of proposed West-Coast LNG (~40% of WCSB supply; LNG Canada Ph-2 + Ksi Lisims closest to FID, word expected by year-end) would end Canada's end-of-the-pipeline discount by ~2029–31 — the catalyst that brings 10-year-horizon investors back to the gas names. Hay (Jul 24 POW!, the coal leg in full): coal as the LNG substitute — global coal usage +~65% this century; thermal coal is "about as inexpensive as it has ever been relative to LNG" (per Trader Ferg) while LNG prices are "once again ripping" — LNG competes directly with coal for power generation in China, India and Japan (the first two ~3bn people combined). Supply tightens as Indonesia, the world's largest thermal-coal exporter, materially cuts exports, into unusually hot weather in Europe/China/India and drawn-down emergency reserves. Advocates dollar-cost-averaging into Yancoal (YACAF) and New Hope (NHPEF/NHC), explicitly prizing the lonely stance "similar to our extremely against-the-grain aggressive buy on oil at the end of June." Hay (Jul 26): US gas at ~$3/MMBtu vs a world price "pushing over 20" is "just ridiculously cheap" (Currie "incredibly bullish on gas as I am"); Europe's storage has "fallen off the cliff" to at least 15-year lows — the Continent will "pay a severe price this winter." He likes the commodity and "US natural gas producers which are also very cheap." Gas is the data-center bridge fuel via combined-cycle plants, but turbines are the bottleneck — "you just can't get them." 2026-AUG-04/06 — Rick Rule (Stansberry Investor Hour + Rule Classroom Plus): $2.65/MMBtu splits the industry — below cost for a primary gas producer (Hugoton, gassy Eagle Ford), effectively free for a Permian oil producer ("any price north of a nickel makes money"), so "for the US natural gas business as a whole, $2.65 is fine compensation." The oversupply flips as export and domestic-use infrastructure completes: "we're producing too much gas now, but we're building out the infrastructure to export it at the same time… and that imbalance changes." Underinvestment shortfall by 2029–30 upgraded from could to "will" absent a depression. Names: EQT (least-oversupplied Marcellus, "Pennsylvania to Massachusetts"), Devon (interfingered Coterra leases → three-mile laterals); Cheniere is "coining money" only while Qatari gas is inaccessible — "I don't think that lasts." 2026-AUG-11 — Hay (Haymaker Daily): the sentiment-vs-physics case in one page — natural gas is "America's #1 source of electricity generation (at 43% of the total, it is nearly double all renewables combined)" and the only thing keeping a data center from being an "inert shell" now that "wind and solar are unequal to the task of providing reliable baseload power at scale" (the US has "10 times as many" installed data centers as China, before Meta/Microsoft/Amazon/Oracle/Google "invest another $5.8 trillion by 2030"), yet per two John Kemp charts bullishness on the blue fuel is close to its decade low outside Covid, with the real price in the 18th percentile back to 2010 — gas priced "as little more than a scarcely used bridge to nowhere." The apathy has suppressed EXE, RRC and EOG; second injury is the Permian — now the country's second-biggest and fastest-growing gas region — flaring gas that "has often traded at negative prices" for want of pipe. The fix: ~15 Bcf/d of Permian takeaway under construction (~12½% of US marketed output), "a notable earnings kicker for companies with significant Permian gas production," EOG the named beneficiary — a basis trade that needs no Henry Hub rally. 2026-JUN-16 (McDonald, historical): Institutions cutting tech to add gas/coal that directly feeds AI infrastructure; "these Canadian natural gas names are screaming buys, companies like Tourmaline. They have that trapped gas" — move the data center to stranded gas that "would never even be sold because it's too far away from things." 2026-08-19 — Haymaker (David Hay): the sector stance restated with its bias declared — "a decidedly pro-energy outlook — particularly with oil, natural gas, uranium and, reluctantly, coal" in a "long-disdained sector" — alongside a deliberate search for disconfirming data. The equity confirmation: "a further confirmation of the long-term uptrend is the extremely bullish nature of the breakout to an all-time high by the leading energy producer ETF, XLE." The honest half, in the same breath: it is "quite extended on a near-term basis, also indicating a correction might be close at hand" — two independent gauges (the $10 WTI backwardation and the extended XLE chart) flagging the same near-term risk while the multi-year stance is unchanged. Jeff Currie (The Trevor Rose Podcast, 2026-AUG-20) is constructive on US gas on a multi-year view: "with gas, there's going to be a period we're going to be short." Rig counts low, returns poor, no drilling — while the US is being asked to backfill missing gas everywhere, not only Russia but Qatar's Ras Laffan, ~20% of the world's LNG market, out. "Lots of room for investment… it's going to create a tight environment over the next several years." Long-run he's comfortable — "it's just easier to get the molecules out of the ground in gas than it is with oil." John Polomny (AIA Weekly Report, 2026-AUG-29) files Europe's gas premium as manufactured scarcity and the US LNG build as the arbitrage that harvests it: "Europeans now pay eight times more for natural gas than Americans… this is not a geological problem. This is a political choice." One Nord Stream line "is still intact and can be turned on… in a couple months" after refurbishment and nobody is trying; Europe also forbids itself hydrocarbons it has, "not because they're geologically inaccessible, but because it's a political choice." The consequence for US producers and exporters: "the United States benefits because there's an arbitrage… this is why they're building all these LNG facilities along the coast" — Corpus Christi, Brownsville, Sabine Pass, "all trying to put more export trains online" — with gas "three bucks a mcf" domestically selling into Europe "for 20 or 30." 2026-AUG-25 (Adam Rozencwajg, Peak Prosperity, 2026-AUG-25): tight globally, and the US supply that looks resilient is an ageing-field artifact. "All the LNG coming from Qatar has been disrupted, so the LNG gas market is very very tight" — which also strangles ammonia/urea, made by natural-gas conversion. On the US side G&R admits a modelling miss: the Permian produces gas and oil from the same wellbore, "just like carbonation in a can of soda," and as reservoir pressure falls the gas "whooshes" out preferentially. Drilling growth kept the average well young; now the field matures, the stream gassifies, and rising gas/NGL output alongside falling crude is "a sign that that field is getting older… a little bit of a swan song" — the "gas burp." "We didn't have that modeled properly a couple years ago"; it explains why US dry and wet gas "has been able to hang in there better than we would have expected," but it is "this kind of last gasp of gas." Everywhere else US gas "has been quite weak." 2026-AUG-27 — Tony Marino (Tenaz, Rose Bros Podcast): the producer-side counterweight on European gas. TTF spot ~€55/MWh (~C$25/MMBtu against AECO "sub $2" — "at least 10 times as high in Europe"), and the CEO whose company is ~90% levered to it will not extrapolate: the forward curve "has validity for the initial part of the period… two or three years out it's probably not a good predictor at all," and his actual lean is down — "it's actually likely that they would be lower than that over any appreciable forward period." Positioned accordingly: ~55% of 2026 TTF hedged, ~45% of 2027, under 10% of 2028, nothing beyond, struck in the low-€30s/MWh at the time of the NAM Offshore and GEMS acquisitions. The 2028 strip ~€30/MWh sits at roughly half the €50 threshold on his own deal's price contingency — a clean market-implied read on how little the strip believes today's price persists. 2026-AUG-28 — Avi Salzman (Barron's): Goldman expects LNG prices to remain high — LNG "in short supply around the world" and Europe needing to stockpile more ahead of the winter heating season.

Wiederhold (Monetary Matters, 2026-AUG-26) — A structural point from index construction rather than a price view: natural gas is the chronic contango offender across the commodity complex, and contango bleeds a negative roll yield on any long futures position. Bloomberg's Enhanced Roll Yield index (BERY) tilts weights toward backwardation and away from contango, and as a result natural gas carries about half its BCOM weight in BERY — a standing warning that the natural-gas spot story and the natural-gas total-return story are different trades unless the curve cooperates. He also names natural gas 2022 as the reference example of an extreme single-commodity price appreciation, but does not nominate it as this cycle's candidate — he points instead at refined products, the softs (cocoa on El Niño-driven West African drought) and the major industrial metals.

Pernas (Monetary Matters, 2026-AUG-07): gas is the bridge fuel of the AI buildout by default — America's pipeline density is what makes on-site generation the path of least resistance while interconnection queues run 18–24 months, a dynamic Dean Pernas gives 3–5 years. Where no pipeline reaches a site, demand routes to the "virtual pipeline" — liquefying gas, moving it in cryogenic trucks, storing and vaporising it on site — a niche consolidated over two decades to a handful of providers, now signing data-centre supply contracts (he cites one at a couple hundred million dollars over two years) alongside launch-vehicle LNG demand from SpaceX at ~20× current cadence within five years. (Pernas, Monetary Matters, 2026-AUG-07.) 2026-SEP-02 (Prinsights, syndicated from The Contrarian Capitalist): Europe as the parallel (and largely self-induced) constraint — EU natural-gas storage stood near 63% full in late August 2026, well below the seasonal norm heading into the withdrawal season (GIE AGSI; 2026 estimate via Asymmetric Research / ZeroHedge), with Dutch TTF futures rising in sympathy on the monthly chart. Layered with AI data-centre load and broader electrification, the whole energy system tightens — which raises, not lowers, the value of diesel as backup power and for remote operations. 2026-SEP-10 (Toby McKenna, CEO Rockpoint Gas Storage): a self-described 30-year bear turns "secular bull" — the shale price lever that capped every gas spread is structurally broken. Producers now target liquids-rich and oil-associated gas, so the dry-gas price barely enters the drilling decision, and expensive horizontals "can't turn on and off at will" without damaging the well; the result is production sustained below variable cost on the dry leg for extended stretches (AECO ~$1 through all of Q3 2025; California below variable for much of summer 2026). Low prices are therefore bullish for the storage layer, not bearish: gas that stays in the ground can be transacted "over and over again with no risk." Physical evidence of how vulnerable the market has become — Chicago printed $70 last winter, $1,000 in Oklahoma/Texas, $60 in California, in markets with abundant pipe and every economic signal to attract molecules. 2026-SEP-10 (Salzman, Barron's): European gas hit its highest level in nearly four years on Wednesday as the Iran War disrupted global supplies, and Europe is entering the high-demand winter season with its lowest gas in storage since 2009 — the reason Salzman writes that "there are signs that prices won't be coming down anytime soon." The named beneficiaries are the U.S. LNG export chain (Cheniere, Venture Global, NextDecade) plus Shell and Equinor. Low storage into the draw season removes Europe's ability to refuse the marginal cargo, which is the condition under which the Henry Hub–TTF spread widens. (Article body paywalled and not captured — this clause rests on the published headline, standfirst and Barron's own bullet summary.) McGlone (2026-SEP-10) dissents: the US January gas contract at $3.80/MMBtu — lowest since end-2021, vs a ~$9 monthly peak in 2022 — "led the way down after 2022" and "is leading the way down, telling you where energy prices are going"; it has not bounced despite hedge-fund shorts, and the gap to heating oil is "just a shocker." Davolos (Sep 12): analyze gas separately from oil. Gas is critical for data-center power, and the US LNG build-out plus deep basin discounts carry optionality. He expects the Marcellus/Haynesville discount to Henry Hub to narrow. He expresses it through the WhiteHawk (WHK) gas-royalty IPO; oil is constructive in the $60s, not underwritten at $100. 2026-SEP-14 (Josef Schachter): gas/NGL equities 'very very cheap relative to historic', the place for new money. AECO C$1.40/mcf and US ~$2.80 vs Europe ~$25 and Asia ~$23/MMBtu (Bangladesh paid $28.50 spot); the marginal buyer sets the price. US LNG exports 19 Bcf/d heading to 30 by decade-end; Canada 2 Bcf/d, possibly 6–7 by 2031–32. Winter demand ~1.5 Mb/d above shoulder season. Paul Sankey (2026-SEP-15): Europe cuts ~15 Mt/yr of Russian LNG on Jan 1 2027, adding diesel tightness; LNG carriers (~$250M, ~5x a VLCC) proved drone-vulnerable. Arjun Murti (Super-Spiked EP227, 2026-SEP-12): his self-graded 2026 miss was not naming natural gas a through-theme — a "super fuel" across Power Surge, energy-rich, AI/energy convergence and Geopolitical Super Vol; "the peak global gas demand call has been as bad as the peak oil demand call. It's even, frankly, more ridiculous." Shale producers have been "one of the least exciting ways to play natural gas" — the opportunity is midstream/downstream (LNG, pipelines). Chris D'Agnes (Hamlin Capital, 2026-AUG-10): energy is still the cheapest sector at ~13x earnings, priced for oil back to $50-60 after the Iran war - 'I'm not sure it's going to be that easy': strategic reserves, inventories and oil on the water are drawn down, countries want their own reserves rather than US shale, and data centers need natural gas. Hamlin's equity-income strategy is overweight energy (SLB, Chevron for the Middle East and Venezuela). Energy is currently negatively correlated to the S&P. Jay Singh (2026-SEP-20): a new natural-gas long, Expand Energy (EXE), bought at the year's lows on warm-winter fears. It trades at a 15% 2026 FCF yield with about $3bn net debt (from $10bn+ pre-2019); two-year model bear $80 / base $129 / bull $152 on $70 oil and $3.34-3.75 Henry Hub. His case: US LNG exports growing 10-15% a year, Europe's structural gas deficit with Qatar offline, and AI data-centre power demand. 'It could be a good time over the next six months to accumulate.' 2026-SEP-21 Hay (Haymaker Portfolio Update): EQT rated a Strong BUY at a 12.4x P/E (~40% below the S&P, "silly-cheap") despite being ~23% below entry. U.S. gas at ~$3/MMBtu vs >$20 in Asia/Europe; ~50 GW of gas-fired plants under construction or planned make it "hard to believe U.S. gas will stay around $3." Japan is moving ahead with ~18 GW of U.S. gas-power commitments (incl. a 9.2 GW Portsmouth, Ohio plant, the largest U.S. gas plant ever), mostly in or near Appalachian supply; gas turbines are in extremely short supply.