Matt Smith (Chronometer) — The 2028 Natural Gas Crisis No One Sees Coming (Invest Like the Best, Jul 2026)

Ticker	Name	Research	View	What he said	At
EXE	Expand Energy	QT · SA · STK · FA	Positive	"Far and away the biggest winner" — controls ~70% of remaining core Haynesville wells, some of the highest-quality rock in the country; CEO-less after a search, stock down ~6 months, trading ~4× EBITDA / low-to-mid-teens FCF yield on a forward curve "where no one believes" his higher-gas case.	25:27
RRC	Range Resources	QT · SA · STK · FA	Positive	"Highest-quality upstream company in Appalachia" — significant room to grow production and materially grow returns to investors; a mature-portfolio producer positioned for the gas squeeze.	26:01
XIFR	XPLR Infrastructure (fmr. NextEra Energy Partners)	QT · SA · STK · FA	Positive	Utility-scale solar yield-co with a "windfall coming in the latter part of the decade / early 2030s" — marks its PPAs to market at much higher values (as gas lifts power prices) with no incremental capex.	27:16
CWEN	Clearway Energy	QT · SA · STK · FA	Positive	"Another one, similar circumstance" — a second utility-scale solar yield-co set to win from PPA mark-to-market as gas sets a higher marginal power price with zero fuel cost.	27:41
CCJ	Cameco	QT · SA · STK · FA	Positive	One of the two companies "most levered" to the AP1000 large-scale-nuclear fix — owns 49% of Westinghouse (Brookfield 51%), which is "deeply undervalued within Cameco today" and could come public as the US government lines up commitments/early procurement.	32:57
BN	Brookfield Corporation	QT · SA · STK · FA	Positive	The other company "most levered" to the AP1000 build — owns 51% of Westinghouse alongside Cameco's 49%; benefits from the coming large-scale-nuclear cycle.	32:57
BWXT	BWX Technologies	QT · SA · STK · FA	Positive	"A super interesting company" — the primary supplier of nuclear for the US Navy; significantly benefits from the coming nuclear cycle with "lots of dollar content in the AP1000s."	33:23
Westinghouse	Westinghouse Electric (AP1000)	—	Positive	The AP1000 reactor OEM Smith sees as the only durable large-scale-nuclear answer (~2033-34); held privately 49% Cameco / 51% Brookfield and "deeply undervalued within Cameco today"; expected to come public.	32:57
LNG	Cheniere Energy	QT · SA · STK · FA	Neutral	Historical reference — the first US LNG exporter ("starting with Cheniere"), the origin of the export ramp that has taken the US from zero to ~15 BCF/d today, scheduled to reach 35 by 2030.	3:13
EQT	EQT Corp	QT · SA · STK · FA	Neutral	Cited as evidence of the complacency: "nobody is investing in gas — in fact EQT is shutting in natural gas right now because they think it'll be more valuable later." A mature-portfolio Appalachian producer (grouped with Expand) whose low replacement cost he flags.	21:04
GEV	GE Vernova	QT · SA · STK · FA	Neutral	Maker of the "large-scale, most-efficient" combined-cycle gas turbines; every GE Vernova press release "means more gas, not less." Grouped in the gas-turbine boom-bust caution — orders could "slow very meaningfully" beyond 2029-30 as it may stop making sense to build large-scale gas gen.	34:47
CAT	Caterpillar	QT · SA · STK · FA	Negative	Doubling its Solar Turbines capacity by end-2029 "at the exact wrong time" — a replay of the early-2000s gas-plant boom/bust — just as buyers may question deploying those gensets because gas is far more expensive than planned.	35:54
BE	Bloom Energy	QT · SA · STK · FA	Negative	"More poorly positioned than investors appreciate" — its fuel cells won't be able to secure gas at 2 GW+ scale amid the scarcity; Smith treats fuel cells as backup generation only, and deploying them as baseload just adds to the convex gas squeeze.	36:16

0:00 Cold open — the conclusion

By mid-2028 gas in storage breaks materially below anything in history; by 2029 below all known historical evidence. "A knife fight to secure natural gas physical in 28" — the biggest loser is the US consumer.

1:32 The finding — an historic US gas deficit

Gas is >40% of US power generation and is overtaking petroleum as the most important fuel. 26-27 is well supplied; from 28, compute demand + continued LNG exports eat into working storage — the nexus of supply and demand — and the upside price risk is "unbounded and convex," felt as electricity prices in 28-30.

2:53 How we got here — shale → LNG exports

The die was cast long before AI. Shale (from ~2010) turned the US from importer to exporter; starting with Cheniere, exports have reached ~15 BCF/d of nameplate on ~110-112 BCF/d of production (~12-15% of daily supply), and permitted projects scale that to ~35 BCF/d by 2030 — "mostly cast."

4:48 Sources vs uses — ~20 BCF/d of new supply

Having evaluated every producing well + the whole pipeline/processing/gathering system, the US can add ~20 BCF/d of production. That was already matched to the 15→35 LNG ramp before AI compute — so the compute demand is incremental strain on an already-balanced system.

6:06 The P50 base case (and the P30/P0 tail)

With an outside partner, every compute power source is probability-weighted. Base case P50 (has approvals + a PPA + interconnection) = ~5 BCF/d of credible incremental gas demand. Move to P30/P0 and it more than doubles to 12-15 BCF/d by the early 2030s — "there isn't gas for that unless you take it from something else."

7:48 Why not just shut off exports?

Contract law: tens of billions of project-financed LNG contracts, and the US will be ~a third of global gas supply — allies (FTA and increasingly non-FTA, e.g. Japanese utilities) depend on it. It could be curtailed but it's a lever of last resort with big geopolitical cost.

11:59 Is there enough rock? — depleting known inventory

Resource in the ground is well-measured (Appalachia/Marcellus-Utica, Haynesville the swing basin, Permian/Eagle Ford as oil-directed). There is gas — but the existing captured inventory of the companies gets largely developed in 4-5 years, and the maximum deliverability tops out ~128-132 BCF/d (a flow metric, not a stock). Higher prices could unlock legacy uneconomic basins, but there's no infrastructure to accelerate them.

17:14 The infrastructure constraints — processing, gathering, pipelines

Even the ~20 BCF/d needs more processing (2-3 yr builds; not yet in place), far more gathering, and interstate pipelines — of which only one (Mountain Valley) was built in the last decade. Permitting is "a nightmare"; the administration is trying to cut barriers but urgency is high.

21:04 Complacency — flat curve, EQT shutting in

Nobody's investing in gas (EQT is shutting some in for later). Gas sits at $3.50-3.60 with a flat forward curve out to 2030 because everyone believes it's abundant — "gas has lulled everybody to sleep." The rig count shows no one is onto the 29-30 tightness.

22:21 Storage draws below all history

Working storage ~4 TCF total. Meaningful draws begin mid-28; by mid-28 storage breaks below anything in history, by 2029 below all known historical evidence. Past spikes (Russia-Ukraine to $8-10/MCF, polar vortexes) were transitory — this is structural demand vs a known production ceiling that don't match.

24:07 Convex, unbounded prices — the export dilemma

He won't put a price target on it, but at $8-10 you'd start shutting off spot cargos; spot alone can't solve it — you'd have to shut contracted cargos (hard, given allied counterparties). The deficit gets "convex and unbounded."

25:00 Winners — the best gas producers

Expand Energy the top winner (~70% of remaining core Haynesville, ~4× EBITDA, CEO-less, stock washed out). Range the highest-quality Appalachian upstream with room to grow. The forward curve doesn't believe the higher-gas case, and the assets haven't changed.

26:21 The dispatch curve — solar's windfall

Gas sets the marginal power price, so as gas rises so does electricity. Free-fuel solar (90% of the interconnect queue with batteries) gets a margin windfall for no capex — utility-scale yield-cos XPLR (XIFR) and Clearway mark PPAs up; residential solar is one of the only consumer hedges for the 10am-6pm peak and should "grow exponentially" even without tax incentives.

28:49 Nuclear — large-scale AP1000 the only durable fix

No silver bullet and no bridge fuel but solar/wind until nuclear arrives. Large-scale nukes are the only real answer, pointing to Westinghouse AP1000s (built commercially on a known timeline, unlike still-experimental SMRs). It must come on by 2033-34. Levered names: Cameco (49% Westinghouse), Brookfield (51%), and BWXT (US Navy nuclear supplier, dollar content in AP1000s).

33:23 The losers — the US consumer

An "awful trade-off": export gas to allies, feed AI compute, or protect consumer bills. Electricity prices rise (already visible on some forward curves), feeding NIMBYism. The consumer pays the bill through the 2030-2035 window.

34:47 "Think more gas" — turbine boom-bust

Bring-your-own-generation (BYOG) means more gas, not less — every Bloom/GE Vernova press release "means more gas." The turbine/genset makers (the market's biggest winners) are adding capacity again like the early 2000s: Caterpillar doubling Solar Turbines "at the exact wrong time"; Bloom's fuel cells won't get gas at 2 GW+ scale. Large-scale gas-gen orders could slow meaningfully past 2029-30.

37:36 The DRAM analogy — energy 10% → 20-40% of compute cost

A recent guest called it "DRAM two years ago — slowly at first, then all at once." Energy is budgeted at ~10% of hyperscaler cost today; if gas doubles/triples structurally it could be 20-40% of the cost of compute by 2029. Everyone underwrites the flat LCOE forward curve — a very attractive but complacent input.

39:29 Devil's advocate — what would make him wrong

The base case already includes the 7+ BCF/d of Permian pipelines coming 26-30; more Permian gas needs much higher oil for longer (which only worsens the consumer crisis). The real disconfirming risk is a step-function battery breakthrough (sodium etc.) — "a watershed moment I would welcome."

45:17 Czar for a day — build AP1000s, reinvigorate solar

If in government: build 2-4 AP1000s end-to-end to derisk the supply chain (the DOE loan office has ~$260B); incentivize residential solar again before it's needed; and build a 1-2 BCF/d Canada→US Midwest (MISO) pipeline to tap Canada's deep, trapped gas.

51:06 What surprised him — the coming knife fight

Tens of billions have flowed into short-life distributed gas gensets that "may not even run" in 4-5 years, and almost no one — even respected firms — has asked whether there will be enough gas or contracted for it. 28 is illiquid, so the curve hasn't moved; once utilities start hedging/buying 28 gas, "a knife fight to secure natural gas physical" begins — six-plus months out, not two years.

54:33 Closing challenge — lock physical supply

To any CEO with energy as an input: know exactly where your gas comes from, lock up physical supply, and understand your counterparty risk ("imagine being short memory 18 months ago"). Have a backup plan for gas at $10+, not $3.50.

Expand Energy is the largest US natural-gas producer, sitting on roughly 70% of the best remaining wells in the Haynesville — a top-tier gas field. Smith calls it "far and away the biggest winner" of the coming gas shortage.

The setup: the company has no CEO right now (a leadership search dragged the stock down over six months), and it trades at only about 4× its cash earnings (EBITDA) and a low-to-mid-teens free-cash-flow yield — cheap. Crucially, the market prices in today's low gas price staying flat for years. Smith's whole argument is that gas is heading much higher, so a great asset is on sale precisely because "no one believes" that. The wells haven't changed; only the sentiment has.

Range Resources is what Smith calls the highest-quality gas producer in Appalachia (the Marcellus/Utica region). It has plenty of room to grow how much it produces and to return more cash to shareholders.

Like Expand, it's a high-quality, long-life gas company that should benefit directly as gas prices climb into the deficit he forecasts — a straightforward "own the best producers" call.

XPLR Infrastructure (formerly NextEra Energy Partners) owns large solar farms and sells their power under long-term contracts (PPAs — power purchase agreements). Here's the clever part of Smith's thesis: in each regional power market, the price everyone pays is set by the last, most expensive plant needed — usually a gas plant. So when gas gets expensive, the whole market's electricity price jumps.

Solar's fuel (sunlight) is free and its cost doesn't change. So as gas pushes power prices up, a solar owner's profit margin expands with zero extra spending — and when XPLR renews or re-marks its contracts to those higher prices, it captures a "windfall" in the early 2030s without building anything new.

Clearway Energy is a second utility-scale clean-power owner in the same boat as XPLR — "another one, similar circumstance," as Smith puts it.

Same mechanism: it sells solar/renewable power under contracts, its fuel is free, and it re-prices those contracts higher as expensive gas lifts the market price of electricity — margin expansion for no extra capital.

Cameco is one of the world's biggest uranium miners, but Smith's angle here is different: Cameco owns 49% of Westinghouse, the company that makes the AP1000 — a big, proven nuclear reactor. Smith thinks large-scale nuclear (not small experimental reactors) is the only lasting fix for the power crunch, and AP1000s are the horse to bet on.

He argues that Westinghouse stake is "deeply undervalued" inside Cameco today, especially as the US government lines up orders and financing to get reactors built — and that Westinghouse could eventually be taken public, surfacing that value.

Brookfield is a large asset manager and owner of real assets. It's the other big owner of Westinghouse — 51%, to Cameco's 49%.

Smith names it as one of the two companies "most levered" to the AP1000 nuclear build-out, so if large-scale nuclear becomes the fix he expects, Brookfield participates through its Westinghouse ownership.

BWX Technologies is the main company that supplies nuclear reactors and components for the US Navy (submarines and aircraft carriers). That gives it deep, specialized nuclear-manufacturing expertise most companies can't match.

Smith likes it as a picks-and-shovels play on the nuclear revival: as big reactors like the AP1000 get built, there's "lots of dollar content" (i.e. lots of parts and work) that flows to a supplier like BWXT.

Westinghouse makes the AP1000, the large commercial nuclear reactor at the center of Smith's "only durable solution" for the late-decade power shortage. It isn't separately listed — it's owned 49% by Cameco and 51% by Brookfield — so investors get exposure through those two.

His view: it's "deeply undervalued" inside Cameco today, and because SMRs (small modular reactors) are still mostly science experiments, the proven AP1000 — which needs to come online around 2033-34 — is where serious money should go. He expects it may eventually come public.

GE Vernova makes the large, efficient gas turbines ("combined-cycle" plants) that are the workhorses of gas power. Smith's point is double-edged: every new order it announces literally "means more gas" demand — reinforcing his shortage thesis — so its press releases are a signal, not a stock tip.

He also folds it into a caution: gas-turbine makers have been the market's biggest winners and are ramping capacity again, echoing the early-2000s boom that later busted. If it stops making sense to build large-scale gas plants past 2029-30 (because gas is too expensive), those orders could "slow very meaningfully." Hence a neutral, watch-it stance rather than an outright short.

Caterpillar's Solar Turbines unit makes smaller on-site gas generators. Smith says it's doubling that capacity by the end of 2029 — "at the exact wrong time."

His worry: this repeats the early-2000s gas-plant boom that ended in a bust. Just as all this new capacity comes online, buyers may not want it, because the gas to run those generators will be far more expensive than they assumed. So he sees the market underappreciating the risk to that business.

Bloom Energy makes fuel cells — devices that turn natural gas into electricity on-site. The stock has been a hot AI-power play, but Smith thinks it's "more poorly positioned than investors appreciate."

The reason is his whole thesis: at large scale (2 gigawatts and up), those fuel cells simply won't be able to secure the gas they need in a shortage — they'll be competing with everything else that burns gas. So he treats fuel cells as backup power only; deploying them as always-on baseload power would just make the gas squeeze worse.

Matt Smith — The 2028 Natural Gas Crisis No One Sees Coming

1. Stocks & names mentioned

2. Talking points

0:00 Cold open — the conclusion

1:32 The finding — an historic US gas deficit

2:53 How we got here — shale → LNG exports

4:48 Sources vs uses — ~20 BCF/d of new supply

6:06 The P50 base case (and the P30/P0 tail)

7:48 Why not just shut off exports?

11:59 Is there enough rock? — depleting known inventory

17:14 The infrastructure constraints — processing, gathering, pipelines

21:04 Complacency — flat curve, EQT shutting in

22:21 Storage draws below all history

24:07 Convex, unbounded prices — the export dilemma

25:00 Winners — the best gas producers

26:21 The dispatch curve — solar's windfall

28:49 Nuclear — large-scale AP1000 the only durable fix

33:23 The losers — the US consumer

34:47 "Think more gas" — turbine boom-bust

37:36 The DRAM analogy — energy 10% → 20-40% of compute cost

39:29 Devil's advocate — what would make him wrong

45:17 Czar for a day — build AP1000s, reinvigorate solar

51:06 What surprised him — the coming knife fight

54:33 Closing challenge — lock physical supply

3. In plain English

EXE — Expand Energy Positive

RRC — Range Resources Positive

XIFR — XPLR Infrastructure Positive

CWEN — Clearway Energy Positive

CCJ — Cameco Positive

BN — Brookfield Corporation Positive

BWXT — BWX Technologies Positive

Westinghouse — Westinghouse Electric Positive

GEV — GE Vernova Neutral

CAT — Caterpillar Negative

BE — Bloom Energy Negative