Luke Gromen — As The Conflict Turns

2:19 Setup — the Iran conflict, revisited

• Last time on, the Iran crisis was just starting; he and Erik were ridiculed for suggesting it could run into July. It has — and, recorded July 21, Hormuz is "closed again."
• The duration call worked; the surprise is where prices ended up.

3:27 The mea culpa — right on duration, wrong on price

• High conviction that the war and the Hormuz closure would last far longer than consensus: correct.
• Wrong on the reaction function: had Hormuz stayed shut, he'd have expected lower equities, higher inflation and higher oil. None of that fully showed up.

4:11 China absorbed the shock — 3–4M bbl/d of demand destruction

• In H1 2026 China shifted 1.4M bbl/d of oil demand to EVs and cut total oil demand 3–4M bbl/d (part of it drawing down its SPR) — an astonishing move that capped the crude spike.
• Consensus had China as the biggest loser; instead exports were up 27% y/y in May and corporate profits up ~19% YTD. "They weathered it much better than anyone would have thought."

6:23 Who blinked first — the Treasury market, not China

• This was always a pain contest. Consensus: China hit first. His view: the Treasury market first — the MOVE index hit near-dysfunctional levels (~120) on March 27, then bond and equity vol peaked and fell within 12 hours of each other. "Odd" — smells of intervention by Treasury/Fed.
• China, by contrast, showed it could take the pain "well beyond" the point the Treasury market would break.

7:56 China's export playbook — solar, EVs, batteries + yuan swap lines

• China showed potential customers a de-dollarization kit: buy our solar panels / EVs / battery arrays (which China dominates), cut your oil demand, and cut your demand for dollars (since oil is dollar-priced).
• Yuan swap lines already exist with basically every country except the US — "pay us in yuan or however you like." It bought China real geopolitical goodwill across the global south.

10:00 Where does energy go — the $50-or-$200 fork

• Bull case: "we're just getting started" in Iran, all uphill from here. Bear case: China just pulled off a miracle containing prices and could keep draining SPRs to manipulate energy lower into the elections.
• He can argue either side with conviction — which is exactly why he won't trade oil directly.

10:50 China's optimal strategy — keep it simmering

• If he were China: keep oil low enough to avoid a crisis, high enough to keep inflation rising, sell everyone on every side the weapons and components they need, and "let the bond markets do what they're going to do."
• The only bond market not straining on the upside is China's. A prolonged, inflationary conflict breaks Western bonds while China gains.

12:33 China's 20-year currency goal — gold replaces the Treasury

• Since 2009 China has said it wants a non-credit-based currency with a neutral settlement asset: "we don't want the yuan to replace the dollar — we want gold to replace the treasury bond as the neutral reserve asset."
• Internationalize the yuan (buy oil/gas in yuan), net-settle in gold. The war advances that goal without China having to force it.

13:51 Contagion — bonds break, then "save the bond market or the dollar?"

• Enough inflation creates a bond-market crisis in the US, Japan, Europe, UK — "once one goes, they're all going to go."
• That returns to the recurring 2020-onward question: does Warsh / does Bessent save the bond market or the dollar? You can't do both.

14:11 Hamiltonian economics — the whole administration is on board

• Bessent gave a NY speech (and an op-ed) saying Trump's plan is Hamiltonian; Greer said the same at Davos in January; Vance and Trump echo it. Definition: high tariffs + a neutral reserve asset that floats in all currencies (i.e. gold settlement).
• Trump's "1870–1913, tax foreigners to pay for our people" is a version of it. "If I'm China, this is going great" — the US is voluntarily walking toward China's preferred system.

16:32 The next move — rare earths and the helium-ban signpost

• Continued rare-earth constriction is exactly what you'd expect if China wants the conflict to last. But the standout signpost: China's helium export ban — despite helium prices being "on the floor" and the US/Qatar being the biggest producers.
• Why ban an abundant, cheap export? Because China expects the US to weaponize helium (critical for semiconductor production) against it — i.e. China is bracing for a much longer fight.

18:46 Chinese AI at the frontier — "cheaper AND better"

• Chinese AI is starting to disintermediate US frontier models. The Rust-Belt movie he's watched his whole career: "they're cheaper but never better" → "about as good" → "cheaper and better." "How about fool me six or seven times?"
• China sits back, lets the US stay at war, and quietly builds AI, power and energy dominance.

21:40 The endgame — yield-curve control

• The outcome forces the West into "some form of implicit or explicit yield curve control." War is always inflationary — "there's never been a deflationary war ever in history."
• Longer war → longer supply-chain interruptions → sell bonds.

22:25 Defense stimmies — creditors turn into borrowers

• In three weeks the US, UK, Germany, Korea and Japan all landed on the same idea: borrow to build a defense base. The structural creditors of the last 50 years (Japan, Germany, Korea) are turning borrowers.
• Where 2021 had "consumer stimmies," this is "defense stimmies" — inflationary because none of it can be sourced from China, the world's biggest industrial base. "It all says sell bonds, sell bonds, sell bonds" → rates up until something triggers YCC.

24:10 Past the trigger — currencies fall together; gold up, Bitcoin does well

• Once one country goes to YCC, all follow; currencies "all weaken together against the yuan and against gold," so none looks obviously weaker against the others.
• Local equity markets go "bonkers on the upside" in nominal terms but fall in gold terms. "I think Bitcoin would do really well in that case." The open questions: when, and does Warsh play tough-guy and hike first (the catalyst for rates running higher).

27:51 Gold is already the #1 US export — 8 of the last 10 months

• Over the last 10 months, gold has been the US's number-one export in 8 of them — bigger than pharma, jet engines or oil (oil/pharma took the other two).
• They have a strategic North Star (Hamiltonian, gold-settled) but "tactically they're just messing it up as they go" — the Iran war is a disaster for actually getting there, and Trump "doesn't want to hear" bad news.

31:16 Oil is "too hard" — so buy gold instead

• Too many swing factors (SPR rundowns, Red Sea / Houthis, China's discretion) make oil a puzzle he won't trade: "for me, it's too hard. Just buy gold."
• Gold does well with less volatility and "at least keeps up with oil" — and he thinks gold goes "way higher relative to oil" over time. Oil still stays at levels "good for US and global producers."

31:59 The regime flip — war-on now means gold-up

• For five months the pattern was war-hot = gold-down, war-off = gold-up. "I think that's about to reverse — it may have already." This week: war on, rates up, oil up, gold up — different from the prior five months.
• Analog: the Iraq war, gold ~$300–350 in 2003 → ~$1,000 by 2008, and that was from a US surplus. The longer this goes, "the better it is for gold."

33:26 CIPS record + gold settlement — the alternative system is live

• CIPS (China's cross-border payment system) hit an all-time record in May — ~$2T (≈14T yuan). The longer the war runs, the more the world "makes other arrangements."
• The only other arrangement is yuan trade net-settled in gold — which diverts world trade into yuan and into gold settlement: "really, really good for gold."

34:29 US electrical infrastructure — PAVE, GRID + a private PE deal

• "Still really like US electrical-infrastructure equities" — power bottlenecks after ~20 years of near-zero added capacity, while China brute-forces power after 20 years of subsidy.
• One of his biggest positions is a private-equity metal-fabricator (generation-agnostic — gas, coal, nuke, hydro) seeing 3–5 years of "open field running" backlog. Public-company proxies: "look at the components of the PAVE ETF, the GRID ETF."

36:14 Japan — you can't reshore without it

• "The US is not going to be able to reshore without Japan" — to build things now you need Japan or China to build them for you, and China won't. Korea/Germany help only at the margins.
• Japan's industrials have lagged the AI headline stuff; "I think Japanese industrial equities also do very well." Still "pretty early days."

37:05 Semis, part 1 — too much too fast, and "AI is real estate"

• Gun to his head, the semi rally is "too much too fast." Same Rust-Belt arc is arriving in AI: cheaper-but-worse → about-as-good → cheaper-and-better.
• Flags a must-read: a "Groundbreaker" substack (early July) reframing AI hyperscalers as real-estate companies — the problem isn't demand falling but the second derivative of growth slowing so you can't refinance. Published weeks before the Kimi performance shock hit these names.

40:42 Semis, part 2 — down near-term, capped long-term

• Weaponizing semis against China (2022, Biden) just guaranteed China would build a hedge — and this week Z.AI unveiled a ~1GW data center running entirely on Chinese chips, no NVIDIA.
• Near-term: AI chips get dragged down with any US-AI wobble (an add opportunity). Long-term (within ~5 years): the multiple you'll pay for semis has to compress once China goes "cheaper and better" — "ultimately they probably end up higher," but with a lower multiple.

43:07 "When the Chinese come, we get an airport" — BYD and the world's view

• The world's choice: America disrupting your oil/fertilizer/dollar flows, vs China offering yuan swap lines, cheap solar, cheap EVs and battery arrays. In Asia, solar now generates more electricity than gas.
• He rode BYDs in London/Scotland — expected junk, found "a good quality product." America's "squeeze you with the dollar" strategy "is not working"; the US needs to pivot to building itself out.

45:45 FFTT — bottleneck investing

• FFTT (fftt-llc.com) aggregates public data to identify developing economic bottlenecks. Over 30+ years: "sectors set to benefit from economic bottlenecks earn excess returns; those hurt by it don't." Products for both institutional and mass-market investors.

46:22 Trading desk (hosts) — Ceresna's GLD "trade of the week"

• Patrick Ceresna's trade, not Gromen's: own GLD ~$376 wrapped in a collar to Sep-18 — buy the $370→$350 put spread for downside, finance it by capping upside at $415 (~+10%). Net cost ~$1.75/share. Built off Gromen's rates-up/oil-up/gold-up observation.
• COT color: gold specs near the bottom of their one-year range (room to run to $415) but sticky through the 30% / six-month correction — "nobody left."

54:09 Trading desk (hosts) — crude squeeze, copper lesson

• Crude +35% in three weeks (back to a ~$90 handle, Brent near $100) yet vol only ~65% (vs 120% in March) and large specs sold into the rally — "this move is running on fundamentals." Fair-value debate: a higher $90–100 range.
• Copper lesson: an extreme crowded reading (100 COT score in May) is "a condition, not a signal" — copper resolved its crowding through time, not a crash, because the physical market kept buying.