Forest for the Trees (FFTT) · former Cleveland Research / Midwest Research analyst · author ("The Mr. X Interviews") — running synthesis of his interviews, with per-appearance breakdowns and a stock index. Top-down macro: US fiscal/Treasury plumbing, the dollar, oil, gold & the move toward a multipolar settlement system.
Structural scarcity money cannot fix, now stated as a preference: 'I like them. I like silver. I like copper. Copper quietly is what, like almost seven bucks?' The grid is the driver — US electricity generation was flat 2004-2024 while China's went to 2x — and the build cannot be financed into existence: 'you cannot build a grid with dollar swap lines.' The AI/reshoring build-out needs ~50 mega mines in 20 years and the deposits don't exist.
Owned but deliberately subordinate — 20-25% against 75-80% bullion. The gold/oil ratio (a miner-profitability proxy) has run 6x→60x and fiscal dominance pushes it higher, but 'if gold's going back into the system there are risks of nationalization' and 'it's a lot easier to grab gold in the ground'; he'd rather not be 'wrong for the right reason.' Sep-13 folds them into one sized bucket — 'probably 40% gold and gold miners' — with no miner-specific argument.
Core position and the destination of every argument — 'all roads lead to gold.' The trade wins on both branches of the 4.7-4.8% line in the sand: 'if it goes over 4.8 and goes into a debt spiral, you want to own gold. And if they inject liquidity to stop it at 4.8, you want to own gold.' Gold is now a bigger share of FX reserves than treasuries, central-bank buying returned to record highs in Q2, and his own one-month fix starts with the ESF bidding gold before a §2.10 revaluation. Held as bullion in private vaults, 75-80% of the metals book. Sep-07 adds the pre-positioning argument: when trillion-dollar balance sheets finally agree, 'they're going to go to hit the sell button and it's not going to work' — markets shut two to three weeks and 'when they reopen, you will own what you own at the new allocation.' The 1998 Ukrainian bank holiday is his evidence: the gold and silver holders 'were fine. Nothing changed for them.' Sep-13 (with Darius Dale) adds the regime tell: gold up ~1.5% on a day the 10-year sold off 5bp — 'when you have 120% debt to GDP… when rates go up gold is a buy not a sell' — and the ninth inning is upsized buybacks with no pretenses and 'gold moving $100, $200, $300 days.' Stated allocation: ~40% gold and gold miners.
First Trust NASDAQ Clean Edge Smart Grid Infrastructure ETF
Second component guide (with PAVE) to US electrical-infrastructure equities — grid/power-equipment names that are 'generation agnostic' (gas/coal/nuke/hydro), riding years of open-field order backlogs. Sep-13 (theme only, no ETF named): US generation flat 2004-2024 while China went from 30% of US capacity to 2.5x — 'very early days… a lot of open field running.'
Liked outright — 'I like Bitcoin long term' — with a sovereign-scale caveat attached in the same breath: Bessent 'has been talking about controlling the pipelines and the on and off ramps,' and while purists don't need ramps, 'at sovereign levels that's a bit much.' Gold moves without permission or banking pipelines; Bitcoin captures the same debasement but is the second choice for the buyer that matters. Sep-07 reframes it as pre-positioning rather than price: 'there isn't going to be a shift, an orderly shift or even a one month shift of trillion dollar balance sheets into gold and Bitcoin. They'll shut the markets and then they'll reopen them two weeks later and Bitcoin will be where it is.' Sep-13 sizes it: 'probably six 7% Bitcoin, five six% Bitcoin' of the book; Dale adds that the dollar debases ~35% a year against Bitcoin, faster if Paradigm D is pulled forward.
Industrials / steel & iron ore (theme — no single ticker named)
Positioning, stated flat: 'it tells me I want to own industrials.' Steel and iron ore go in the same bucket — 'you cannot build a grid with dollar swap lines.' The cycle is ahead, not behind: private manufacturing construction is down 18% year-over-year even with AI running, so today's inflation is what you get before the build starts — 'what do you think inflation is going to be when construction spending is actually up?'
Japanese industrial equities (theme — no single ticker named)
Reshoring theme — the US can't rebuild without Japan (China off-limits); Japan's industrials have lagged the AI-headline stuff and 'also do very well.' Still early days, a multi-year setup.
US electrical-infrastructure play — 20 yrs of near-zero added power capacity now colliding with AI/reshoring demand; named as a component guide to the theme (alongside his private metal-fabricator PE deal seeing multi-year backlogs). Sep-13 sizes the theme at ~15% of the book (no ETF named): 'not as sanguine on AI specific. I am as sanguine on the buildout. I prefer to play it via electrical infrastructure equities.'
Positive — his first actual view on silver after the Aug-14 title-only mention. Asked about non-gold metals and minerals: 'I like them. I like silver. I like copper.' Sits in both buckets at once: a monetary metal riding the same debasement as gold, and an industrial input to the grid/reshoring build. No target and no allocation given — a stated preference, not a developed thesis. Sep-07 keeps it there but adds both halves of the freeze argument: the 1998 Ukrainians holding 'gold and silver' came through intact, while 1980's Hunt-brothers COMEX episode — 'the buy button stopped working' — is his template for a market simply switched off when everyone arrives at once.
Long in dollars, short in gold — stated as a rule: 'shorting American stocks in dollar terms is not a good idea. Shorting them in gold has been a great idea.' S&P total return is −30% against gold since the Fed began hiking in early 2022 and −50% since 2000, and he expects gold to keep outperforming for two to five years. The nominal side goes vertical under his own fix — 'Dow probably goes from 50,000 to 100,000' — alongside 10-15% inflation. Sep-13 names it the 'Argentinization' of US stocks — 'S&P up in dollar terms but down in gold terms,' down 10-15% in gold since Powell's Q4-2018 pivot while up ~200% in dollars — and holds blended large caps as the balance of the allocation.
Positive — ~15% of the book (Sep-13), held for two reasons. Optionality in 'a highly political market' where narratives whipsaw gold (Dan Oliver's Reichsmark chart: the trend won, but levered longs 'lost all your money four or five different times in five years'; 'sell gold, Warsh is a hawk' is today's version). And carry: 'I look at my cash position as earning a yield on my gold' — T-bills at ~3.5% make a 'positive one and a half% carry across my cash and bullion.'
Referenced through Larry Fink, and by the host — 'are you doubting Larry Fink's ability to raise money? I'm not.' Gromen's rejoinder is about the source rather than the firm: 'he's probably over in Saudi Arabia getting money. And I would say, from who? What money do the Middle East have to invest now?' — the UAE having just needed emergency swap lines. Sep-07 makes it a historical exhibit rather than a company view: via a Jim Rickards book, 'Treasury's got a direct line into BlackRock… in a crisis, Treasury can pick up the phone, make one call and lock down 5 trillion of capital. That's it. No sales. And the rest of the market would follow' — evidence the exit closes administratively, not economically.
Referenced, not a pick — BYD used to illustrate Chinese 'cheaper and better' competitiveness (rode them in the UK, 'a good quality product'); the export offer the US won't let in.
Referenced, not a pick — the third major named at the White House mining press conference. Sits inside the copper-scarcity argument rather than carrying its own view: even permitted and financed, a large new mine needs a thousand people on site for 12-18 months, and the deposits the build-out needs don't exist.
Referenced by the host, not a pick — Google's reported '800 billion dollars in forward purchase commitments' offered as evidence that Q4 results will confirm what the off-balance-sheet agreements already show. Gromen concedes the near-term point ('the bubble hasn't popped yet') while insisting every prior capex boom ended in a bust.
Referenced, not a pick — named beside BYD and solar as the goods base that makes the yuan worth accepting, the first leg of China's yuan-invoicing / gold-settlement architecture.
Named in passing, no thesis — the obvious expression of the war's one clean winner after the host notes refined products stayed 'stubbornly high' while crude fell, so 'the refining margins have been really really high.' Gromen's answer is the ticker itself; no position is implied.
Referenced, not a pick — listed with Rio Tinto and Freeport at the White House mining event. No company view; the discussion is 40 years of financialization leaving 'not the bench depth… not the bench at all' in US mining.
Disclosed as a conservative allocation slice — AAA-rated life-insurance equity earning ~6% federal- and state-tax-free; a stable cash-substitute holding, not a market call.
Referenced, not a pick — cited as evidence the AI bubble is still being fed credit rather than as a company call: 'you got NVIDIA, Jensen bragging about creating a compute security derivative… No judgment. I know what that is. That's making it easier to get more credit to them. So the bubble's not over yet.' Sep-13 mention is Darius Dale's, not Gromen's: with 85-90% of federal receipts coming from workers AI displaces, redistributive taxes after 2028 go where the money is — 'they're gonna go Nvidia.'
Referenced, not a pick — named by the host as a major invited to the White House hard-rock-mining press conference. Gromen's reply is about the policy, not the company: $100 million on mining education 'over an undefined period' against $37 billion spent in Iran in four months.
Semiconductors / AI equities (sector — no single ticker named)
Take profits, don't short — 'I wouldn't short them here.' New credit is still arriving (SEC loosening AI securitization rules, PE syndication, a compute derivative), and manias end when the credit stops, not when valuations get silly, so 'the bubble's not over yet.' But his own history work is the timing rule: two to three years into every prior capex boom you did better selling most of it and buying gold, which won over the full cycle every time. Sep-13 restates it against Darius Dale's outright AI bull case (capital deepening at a record 22.3%, a bubble into end-2027/mid-2028): AI is the sixth and largest US capex boom, 'I tend to be more cautious about the equities there,' and AI borrowing to eliminate labor competes with Bessent's employment-based tax base.
Negative — the interventions keep failing. Bessent sold ~11 billion euro of reserves to buy yen, trailed on Instagram so hedge funds would front-run him, and 'it bounced back pretty quickly… already retraced over half of that full intervention.' The war did the damage: 'the yen gets killed, energy costs go up on the yen, so now the JGB markets sell.' The constraint is symmetric — too strong a yen unwinds the yen carry trade, too strong a dollar the dollar carry trade — so the only play is recurring liquidity to hold a band.
Negative, inferred from the trade that didn't happen. Insurers swapped long treasury duration for SOFR-linked private loans; then the 10-year hit 4.7 and they still didn't buy. 'If they could sell it at a decent mark and buy a 10-year Treasury bond at 4.7, 4.75, they would have — and they didn't. That gives you all you need to know about the actual liquidity and solvency.' He ties the UAE's scramble for swap lines to being 'up to their chin in private credit' when Hormuz shut — and reads it straight through to more liquidity: 'buy gold.' Sep-07 adds Lyn Alden's precision: the redemption gate is a liquidity feature written into the contract (closer to full-reserve banking than a bank run, since the lenders are pensions/insurers/family offices lending savings, not payroll), while solvency is a separate question — 'on the margins we do see solvency issues… it's still unclear how big some of those solvency areas could be.' Sep-13 puts a number on the standoff (citing Nick Neoth's Substack): of a $10trn life-insurance industry, $1.54trn is affiliated reinsurance against ~$647bn of total reserves — 'if the marks are bad enough, they're out of reserves,' and they sell Treasuries and mortgage-backs to fill the hole. Expected fix: regulatory relief, 'just QE through the life insurance industry.'
Negative in real terms and deliberately not nominally — nominal-default odds are 'zero,' and 'I don't know that I'm really nominally bearish on the long bond here.' The damage comes through the numeraire: in gold terms TLT is down 90-95% since 2014, when central banks stopped buying treasuries on net, with 'another 90 to 95% to go against gold' and the next leg 'all gold.' Vol stays elevated: hedge funds now own 8.5% of the treasury market via the levered basis trade. Sep-07 adds Lyn Alden's buyer-by-buyer census from the other direction — foreigners short of the run rate as a share of issuance, a balance-sheet hawk at the Fed, banks needing further SLR relief, insurers and pensions unable to lever — 'so I do think that they're getting squeezed,' a 'pretty orderly degradation of the global bond market'; Gromen's version is 'a nonlinearity facing him at the long end.' Sep-13 (with Darius Dale): 'they're going to lose the long end no matter what they do' — a hike is 'a pay raise to 65 million boomers' that widens the deficit and lifts the dollar into foreign selling, a cut into 8% nominal growth lifts it too. He accepts Dale's five-model 10-year fair value of 5.87% and argues the path is convex ('48, 52, 58, 62 happen fairly quickly') because life insurers can't buy at any yield.
In one line: the US is cornered into a single choice — sacrifice the dollar (inflation) or the bond market (higher rates) — and as of 2026-AUG-19 the choice has been made in public: Bessent doubled Treasury buybacks in the 10-to-30-year sector, which Gromen reads not as a panic but as the next rung on an eighteen-month ladder — "essentially a version of operation twist… another soft form of yield curve control," the same vector as the UAE/Japan FIMA swap lines, the yen intervention and the stablecoin push. The arithmetic that forces it: entitlements + interest + veterans' benefits = 105% of receipts, growing 7.5% against receipts at 4%, with $1.4trn of net borrowing over the next two quarters — so it was "mathematically impossible" for Warsh to be a hawk, and the Iran war (10-year 3.94% → 4.74%) was "the straw that broke the camel's back." The whole book reduces to one level: over ~4.8% on the 10-year you own gold, and if they inject liquidity to stop it at 4.8 you own gold. The template is 1946–51 — debt/GDP 110% → 55% in five years, real rates −3%, bondholders losing half to two-thirds — and his own one-month version is explicit: ESF bids gold, settle China deficits in gold, revalue the certificates under Fed accounting manual §2.10 (≈$5trn into the TGA), buy back the long end, stuff the rest into 60bp stablecoin T-bills, at the cost of 10–15% inflation and a Dow at 100,000. Positioning: long stocks in dollars, short stocks in gold; gold bullion 75–80% / miners 20–25%; own industrials, silver and copper; stay away from the long end (TLT −90–95% vs gold since 2014 with "another 90 to 95% to go… mostly via gold"). The duration he wants is "0% yielding, infinite face value, infinite duration, and finite issuance." 2026-SEP-07 sharpens two things and adds a third. The rate call is now stated as an impossibility, not a probability — true interest expense (gross interest + Social Security + Medicare + Medicaid + VA) is 105% of receipts growing 7–12% against receipts at 4%, one hike takes it to 107% growing 8–9 and a second to 110% growing 10, so "Warsh isn't going to hike rates. He's not. He can't," and the "we owe it in our own currency" defence dies because entitlements are owed in a hard currency: "we didn't owe my dad a payment for Medicare. We owed him a knee." The long end is a nonlinearity, not a slope — every patient buyer is gated at once (hedge funds only while vol is low, foreign central banks net-absent 12–13 years, insurers and pensions unable to mark down private credit). And the new leg is the pre-positioning argument: you will not get out. When trillion-dollar balance sheets finally agree, "they're going to go to hit the sell button and it's not going to work" — 1980 COMEX silver, Treasury's one-call line into BlackRock, a two-to-three-week closure — and "when they reopen, you will own what you own at the new allocation." Alongside him Lyn Alden reaches the same structure by a different road and lands softer on the near term (base case zero-to-one hike, "if we get the one, it'd be kind of symbolic"), which is the useful cross-check on this hub's central call. 2026-SEP-13 (Thoughtful Money, with Darius Dale of 42 Macro) puts him in the "sixth, seventh, or even eighth inning" of a bond-market crisis — the gauge being true interest expense at 105% of receipts per the Q3 TBAC report, and the tell being gold rising with yields while buybacks climb from $6bn toward 8 and 10 — and, unusually, states the allocation: ~15% T-bills, ~40% gold and gold miners, ~15% electrical-infrastructure equities, ~5–7% Bitcoin, the balance in blended large caps ("S&P up in dollar terms but down in gold terms"). Dale's five models independently put the 10-year's fair value at 5.87% and both expect explicit yield curve control by end-2027 to end-2028; they split on whether the Fed should hike (Dale yes; Gromen: a hike is "a pay raise to 65 million boomers" and "they're going to lose the long end no matter what they do").
The propaganda tell — and why it was predictable. "Anytime I see something happen and I get basically 30 versions of the same thing said about it, that's usually your first clue someone's attempting to propagandize you" — in this case "Warsh is a hawk." The arithmetic that kills it: ~100% of receipts are already interest, entitlements and veterans' benefits, so "you can't have a strong dollar that doesn't blow up the fiscal math increasingly quickly." The template is DOGE: universal belief, a six-week gold sell-off, then "how stupid was that." Aug-14: the propaganda is "in the process of being thrown in the trash."Aug-20, with the receipts: entitlements + interest + VA are 105% of receipts and growing 7.5% against receipts at 4% — "in the same way that the math suggested there was no way that Elon could DOGE a trillion, there's no way Warsh could be a hawk. It is mathematically impossible." The dovish breadcrumbs were in plain sight all along — trimmed-mean inflation, and a first press conference floating that "there doesn't have to be a zero at the end of the inflation thing."
The fix is QE with extra steps — and nobody works the derivatives. Cut the front end, fund it with stablecoin-backed T-bills, re-regulate banks into the long end and backstop them with swap lines: "that's just QE — when you cosign a loan for your kid, it ain't your kid taking out the loan." Second derivative: inflation. Third: hot prints, upward pressure on the long end, pressure to raise rates. Release valve: "it comes out in the currency. It's really good for gold. Should eventually really be good for Bitcoin. It's good for stocks."
Nobody is short dollars — so dollar strength gets paid for in Treasuries. $13–14T of dollar debt against $60T gross / ~$20–25T net dollar assets including $9.5T of Treasuries. A too-strong dollar forces holders to sell Treasuries — either to buy dollar-priced commodities or to defend a currency — and Japan hit both at once. The response is always the same: dollar liquidity. Warsh pre-announced it ("a fair price for assets in a crisis"), and the defended 10-year level has been walked up from 4.4 to 4.6–4.7 (from 3.94% on Feb 28) — "that's not a sign of strength."
Synchronised debasement is invisible in the crosses. Within five days of the NATO meeting the US, UK, Germany, Korea and Japan all announced defense borrowing — "defense stimmies," circular because US dollar liquidity is what keeps their bond markets from forcing Treasury sales. The genius: "all their currencies debase against gold and against stocks and against inflation but not against each other" — higher inflation said not to be inflation, a weaker dollar that doesn't look weak on screens, and anyone naming $6–7,000 gold "will be ostracized."
The variant perception: in the risk-off, long yields go UP. When equities break you get a drop in long Western yields for "5 days, 10 days, maybe 3 weeks" — then they rise faster as stocks fall. The mechanism is the creditor swap: a Fed white paper (Oct-2025) puts 37% of net note/bond issuance since 2022 with Cayman-Islands hedge funds running the levered basis trade, so an equity-vol spike forces them flat and turns the market's biggest buyer into its seller. The loop only ends when policymakers inject dollar liquidity — 2020 QE, Yellen's 2022-23 dollar weakening / front-end shift / RRP drawdown, the 2024 buybacks Bessent criticised then doubled. "An emerging-market debt crisis with the American flag pasted on top." That injection, not the panic, is the buy signal.
Treasury-market functioning is the Fed's real #1 mandate. Powell's own euphemism is "the Fed's shadow third mandate"; at 120% debt/GDP and 6% deficits it is the first one. Consensus expects Warsh to subordinate it to price stability — "there's not a chance. The only question is how long." Policy-reaction levels on the 10-year: 4.4% used to trigger a back-down, 4.65-4.7 is tolerated now, 4.6-4.9 is the problem zone; Bessent's "three arrows" are dead. Aug-20 makes it explicit and operational: "over 4.8 on the 10-year, bad things" — and the response is now visible rather than inferred, with the Aug-19 doubling of 10y–30y buybacks. He is bearish in real terms only: nominal-default odds are "zero," so the loss arrives through the numéraire, not the price. Each back-down erodes credibility — and the military-backs-the-Treasury protection racket erodes with it, as a navy is stood off by cheap missiles and drones.
Contagion is already priced abroad. UK, Japan, Germany and France are all flashing red; Japan and the UK are now the #1 and #2 foreign creditors of the US, and 10-year gilts and 10-year USTs "just lockstep — they're tied at the hip." "Once one of them breaks, they're all going to break in very short order." The lone outlier is China, whose 10-year has gone from the highest of the group to below Japan's.
Equities: complacent now, rational in dollars, falling in gold — and structurally load-bearing. Near-term "extremely complacent": debt-financed, cash-negative big tech can't have anything go wrong into rising rates. Structurally rational (the Venezuela framing — resilience is the currency), yet the S&P total return priced in gold is −40% since Jan-2000, −8% since Q4-18, −21% since Jan-22. And a sustained −20% blows out the deficit through cap-gains/stock-comp receipts, so stocks back the Treasury market and vice versa — zugzwang. Don't own long-term bonds: "bonds get crushed by either devaluation or war."
Gold is being installed as the settlement asset — deliberately. The dollar is down ~70% vs gold in three years (1,800→5,400); the pullback to the low-4,000s is blow-off-top digestion, and China buys every lower price bigger — 173 tons last month ≈ $23B against a $105B trade surplus, ~a quarter of the surplus settled in gold. China wants gold, not the yuan, to replace the Treasury bond: yuan-invoiced commodity buying plus offshore yuan clearing banks in every major gold hub (London, Switzerland, Dubai, Singapore, Hong Kong, Shanghai) make the closed capital account two-way through gold. "Gold is a 0% yielding bond of finite issuance, infinite face value; a Treasury is a 4% yielding bond of infinite issuance, finite face value." At $16,000 gold China's trade balance flattens — and if gold becomes the backstop the risk-free rate drops to 1-2% real, which is very bullish equities. Aug-14: the bottom is in — "I do think it'll get back to 5,000" this year.
Own the bullion, weight the miners down. He owns both, "probably an 80/20 split, maybe a 75/25 split, bullion to miners," physical, in private vaults, almost all US with a little in Switzerland. The reason is legislative, not operational: "if gold's going back into the system there are risks of nationalization of assets… it's a lot easier to grab gold in the ground than to go door to door asking people to take you to their private vault." He calls the failure he's insuring against being "wrong for the right reason."
Wartime footing is a slogan until you price it — and copper can't be bought at all. 1940 meant a 25%-of-GDP deficit (~$8trn today), a Fed balance sheet up 10× funded at 3/8%, 30–50% inflation, capital controls ("that's your big reset right there — it's forced") and a 90% top tax rate. So reshoring in 5–10 years is "a freaking pipe dream — 10 to 15 best case, probably more like 20." And the physical constraint is absolute: ~50 mega copper mines needed in 20 years and the deposits don't exist — "the dollar has hyperinflated against the major copper mines number five through number 50… there is no amount of dollars that can get you them." China bought that option cheap through Belt-and-Road "opaque lending" (Carmen Reinhart's term) while the US spent '02–2020 in the Middle East (the Longer Telegram window).
Yen intervention is the recurring trigger — watch the reaction to the second one. They intervened at 163–164 on USD/JPY; it is already back to 159.25 with 160 the reported next trigger, because "nothing's changed with the underlying" (Japan structurally short dollar oil, oil back over 80, a war-driven current-account deficit). "I'm going to be much less interested in the event and much more interested in the reaction of the markets… everyone will be like I got fooled again" — equities rip, gold really rips, possibly Bitcoin. And on "growing out of it": possible only "with some form of yield-curve control and letting the currency take the hit — growth in dollar terms, austerity in gold terms."
The twin tripwire — they were finished two years ago; everything since is time-buying. Two carry trades sit on opposite sides of the same pair. Too strong a yen unwinds the yen carry trade and forces global selling of stocks and bonds; too strong a dollar forces the world's net-long dollar holders ($22trn net / $65trn gross, incl. $9.5trn of Treasuries, against $13–14trn of offshore dollar borrowing) to sell Treasuries. "That's where I knew they're done." The only playable path is brake-and-spur liquidity to hold USD/JPY in a band — which is why the yen intervention (≈€11bn of reserves sold, trailed on Instagram so hedge funds would front-run it) retraced more than half within days. The August-2024 version of this call — cut rates, then inject liquidity, good for gold, Bitcoin and industrials — he now scores "check, check, check."
The one-month fix, and the option already sitting on the books. Asked what he'd do: ESF bids gold aggressively → announce all China trade deficits settle in gold → instruct Warsh to revalue the gold certificates (Fed financial accounting manual §2.10: held at $42.22, revaluable at the Treasury Secretary's sole discretion; 261m oz × ~$4,000 ≈ $1trn each, so $20,000 gold books ≈$5trn straight into the TGA) → buy back everything past ~five years for cash → Clarity Act, stablecoin T-bills at 60bp. "That's like the MMT platinum coin trick except it's actually on the books." Cost: 10–15% inflation for a couple of years, the midterms, and "Treasury holders get killed" — banks and boomers, "the richest generation in history… consuming 80% of the budget." Paradoxically it lowers yields on the other side: gold high enough doesn't destroy the Treasury market, it collateralizes it (Judy Shelton's groundwork; 2–3% long rates). The durable version pegs gold to oil, never to a currency — 500–1,000 barrels an ounce by US–China–Arab agreement ($60 oil ⇒ $30,000 gold), which ends exorbitant privilege and caps Chinese mercantilism at once. And it is stated policy in embryo: Hamiltonian economics — neutral reserve asset, tariffs, self-sufficiency — from Bessent's New York Economic Club speech + same-day WSJ op-ed, Greer at Davos, Vance in 2023.
AI is the boom that eats the tax base — sell it into gold at year two or three. AI bids for capital against the Treasury, and to be worth the trillion in debt, trillion in lease commitments and trillion in semiconductor commitments it must destroy white-collar employment — healthcare administration is the largest employer in 39 states and half the tax base comes from jobs. Either way receipts fall: "a snake eating its own tail." Every one of the five prior capex booms back to the 1840 canals ended in a bust, and the historical rule is specific: two to three years in, you did better selling most of the bubble and buying gold — gold won over the full cycle every single time (sell the Dow for gold in May 1929 and you're still +15% a century on). But not a short: new credit is still arriving (SEC loosening AI securitization, PE syndication, a compute derivative), and manias end when the credit stops. The circular reference nobody names: the UAE money funding US AI came from Bessent's swap lines — "he's just creating the money."
The trades beyond gold.US electrical-infrastructure equities — US generation was flat 2004→2024, so "the US didn't really grow on a real basis for 20 years," and AI plus reshoring now reverse it; PAVE/GRID as the holdings screen (no financial relationship, recommended to clients for years) plus his private metal-fabricator PE deal. Japanese industrial equities by process of elimination — reshoring must be done well and non-inflationary, and of Germany / China / Japan only Japan is available. Bitcoin (does well under global YCC/debasement — Aug-20 upgraded to "I like Bitcoin long term," with the caveat that Bessent's talk of controlling the on/off ramps is "a bit much" at sovereign level, where gold needs no one's permission) and a nuanced semis hold — an add on any US-AI wobble, but China's "cheaper and better" AI moment caps the long-run multiple. Aug-20 adds the physical leg outright — "I want to own industrials," plus silver, copper ("almost seven bucks, and no one's talking about it"), iron ore and steel, because "you cannot build a grid with dollar swap lines." The build hasn't started: US electricity generation was flat 2004→2024 while China went to 2× the US grid, and private manufacturing construction is −18% y/y — so today's inflation is what you get before the boom. And a new negative: private credit, read off the non-bid — insurers didn't buy the 10-year at 4.7 because they can't mark their loan books, which removes a natural Treasury buyer and guarantees another liquidity injection.
You won't get out — so the allocation has to exist first.(Sep-07, BTC Sessions, with Lyn Alden.) The exit closes administratively, not economically. Precedents he stacks: COMEX silver 1980, where "the buy button stopped working" on the Hunt brothers; the Jim Rickards account of Treasury's direct line into BlackRock — "one call and lock down 5 trillion of capital… no sales, and the rest of the market would follow," in place "20 years ago nearly"; and the gate already demonstrated at small scale in private credit ("we want three billion" / "you can't have it"). So "whatever your allocation is to everything — bonds, stocks, gold, Bitcoin — that's going to be your allocation. You're not going to be able to move"; markets shut two-to-three weeks and reopen with stocks gapped higher, Treasuries destroyed in relative terms, debt/GDP 120% → 20% — paid for by the holders. The 1998 Ukrainian bank holiday is the personal evidence: five cars' worth of savings bought a month of groceries, while the people holding gold and silver "were fine. Nothing changed for them." The framing he uses for the whole position is "a Weimar gold reparations problem" — explicitly not a hyperinflation call, but obligations owed in a hard currency that inflation-adjusts, larger than receipts, today. No date: "could it be next week? Sure. Could it be 20 years? Sure."
The Lyn Alden cross-check — same structure, softer near-term call, different instruments to watch.(Sep-07.) She reaches the squeeze by census rather than arithmetic: foreigners buying in dollars but "not nearly enough" as a share of issuance, a self-described balance-sheet hawk at the Fed, banks with capacity only behind further SLR relief, and insurers/pensions as "fairly honest balance sheets" that must sell to buy — while stuck in private credit, where she separates the liquidity gate (contractual, "closer to full reserve banking") from the solvency question ("still unclear how big"). Her verdict is deliberately calibrated: "a pretty orderly degradation of the global bond market," no MOVE-index stress yet. Her endgame tell is not a crash but an embarrassment — "the central bank has to come in and start buying bonds and has trouble explaining why" (2019 repo; the BoE's cancelled QT speech in the 2022 gilt crisis) — and the soft version is already live as Treasury operation twist "until the midterms." Where she departs from Gromen: zero-to-one hike is the base case and a hike would be "symbolic," because under fiscal dominance hikes aren't a working tool (Congress doesn't respond to 5% vs 4%; money-market boomers get a raise). And she redirects attention outside the Fed's band entirely: crack spreads, not the crude price — oil never hit $150 but record refining margins have "diesel priced as though oil itself is over 100," which matters more to the inflation path than 25bp. Also hers: Japan's GPIF repatriation "nuclear option", and the observation that fiscal dominance has gone mainstream, so what holds the market together now is perception — the cascade risk is a belief cascade.
The Darius Dale cross-check — same timing on the bond crisis, opposite advice on the Fed, and the allocation in numbers.(Sep-13, Thoughtful Money.) Dale's paradigm framework (A fiscal dominance → B cut → C grow / run it hot → D default via debasement → E political realignment and war) puts the US in C and "third or fourth inning" toward E; Gromen's true interest expense at 105% of receipts puts the bond-market crisis in the "sixth, seventh, eighth," and once the question is framed that way they agree — Dale: that crisis arrives "by the end of next year, certainly not by the end of 2028," with explicit YCC. Dale's five-model 10-year fair value is 5.87%; Gromen supplies the reason the path is convex: life insurers' $1.54trn of affiliated reinsurance against ~$647bn of reserves means "there's no level at which they're a buyer," and the expected regulatory relief is "just QE through the life insurance industry." They split on the Fed — Dale says hike to keep Paradigm C alive (higher rates are income support for $11trn of household cash); Gromen says a hike accelerates the deficit and "they're going to lose the long end no matter what they do." Both reject stablecoins as a fix: Gromen traces the other side of the balance sheet (a eurodollar run into T-bill stablecoins → dollar shortage → foreign selling of $22trn of US assets → non-withheld receipts collapse → a bigger deficit within 9–12 months), and adds that the world wants China's goods, offered with a yuan exchangeable for gold. On AI Dale is the bull (capital deepening at a record 22.3%, a stock/gold/Bitcoin bubble into end-2027/mid-2028, then redistributive taxes after 2028 — "they're gonna go Nvidia"); Gromen owns the build-out via electrical infrastructure rather than AI equities because gold beat the boom sector over every one of the five prior US capex booms. The allocation, stated flat: ~15% T-bills (optionality in a political market plus ~1.5% blended carry on the gold), ~40% gold and miners, ~15% electrical infrastructure, ~5–7% Bitcoin, balance in blended large caps — the "Argentinization" of US stocks. Closing advice: "keep your Overton window wide open. Keep your leverage low."
Appearances
One dated page per interview — each has its stock table, talking points and an "In plain English" section. Newest first.
Date
Title / analysis page
Show
Video
Text
Actionable insights
2026-SEP-13
Which Inning Are We In?(joint discussion with Darius Dale, 42 Macro — Dale's views attributed)