← Analysis page  ·  Luke Gromen hub  ·  Research hub

Actionable insights — These Are The Catalysts That Could Send Gold & Silver To New Highs

The repeatable analysis behind the calls: not what he owns, but how he finds and times it — written so the process can be rerun later on different events.
2026-AUG-14 · Goldfinger Capital (Robert Sinn) · Luke Gromen (Forest for the Trees / FFTT) · ▶ Watch · full analysis · transcript
How to read this page: each insight is a method — the tell that put him onto a read, the steps that turn it into a position or a rule, and the signal to watch when re-running it. The boxed line shows how it played out in this interview. Timestamps deep-link into the video.

7:08 1. The uniformity tell — 30 versions of the same take means you're being propagandised

The repeatable method
  1. When a new narrative lands, don't first ask whether it's true. Count how many independent-looking sources are saying the same sentence, in the same words, in the same week.
  2. High uniformity is the signal. Genuine analysis produces dispersion; a placed narrative produces near-identical copies. "That's usually your first clue that someone's attempting to propagandize you."
  3. Ask cui bono — which faction needs this belief held right now, and what does it buy them (time, a stronger currency, a calmer bond auction)?
  4. Then test the narrative against an arithmetic constraint that doesn't care about opinion. If the two conflict, the narrative is the thing that has to give — and the timing of its collapse is your trade.
Here: "It's Warsh is a hawk. Warsh is going to be a hawk" — repeated everywhere. The arithmetic that contradicts it: ~100% of receipts already consumed by interest, entitlements and veterans' benefits (1:08), so "you can't have a strong dollar that doesn't blow up the fiscal math increasingly quickly." Outcome: after the yen interventions, "the propaganda that Warsh is a hawk is in the process of being thrown in the trash," and gold rose 14% in five days (11:26). He offers the DOGE episode as the template for how these unwind — universal belief, a six-week gold sell-off, then "everyone now is like, oh, DOGE, how stupid was that" (26:21).
Watch for

3:25 2. Walk a policy fix out to its second and third derivatives before pricing it

The repeatable method
  1. Write out the proposed fix as its authors describe it, in full mechanical detail — every leg, including the backstops nobody mentions.
  2. Grant that it works. Most analysis stops here, and so does most mispricing.
  3. Now ask what the fix produces (the second derivative): what has to be created, borrowed or printed for each leg to function, and what that does to prices.
  4. Then ask what those consequences do to the fix itself (the third derivative). If the output feeds back into the input, you have a loop, not a solution — and the loop's release valve is where the money is.
  5. Identify which variable is allowed to absorb the strain. If yields can't and spending can't, it is the currency — so hold what the currency is measured against.
Here: the fix is cut the front end, fund it with stablecoin-backed T-bills, re-regulate banks into long duration, and provide swap lines so a rate rise doesn't break them. Granted, it works. Second derivative: "your money-financing stuff at the front end — that's inflationary… you're basically doing QE at the long end just through the banks — that's inflationary." Third: "you're going to start having hot prints, and now you're going to have more upward pressure on the long end… now there's 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." (GLD, IBIT, SPY.)
Watch for

8:46 3. The nobody-is-short-dollars flow test — net the stocks before forecasting the currency

The repeatable method
  1. Before taking a view on a reserve currency, net the world's position in it: gross liabilities (dollar-denominated debt) against gross and net assets (deposits, securities, official reserves).
  2. If the world is net long, a strong dollar is not a squeeze — it is an incentive to sell the assets. Work out which asset is the most liquid and the most likely to be sold: Treasuries.
  3. Enumerate the two conditions that force selling: needing dollars to buy dollar-priced commodities, and needing dollars to defend a currency. Watch for a country in both at once.
  4. Then predict the policy response rather than the price: rising yields from that selling will be met with dollar liquidity, because the alternative breaks the fiscal math. Position for the liquidity, not the selling.
Here: "They've borrowed 13 to 14 trillion in dollar-denominated debt… and they also have $60 trillion in dollar-denominated assets on a gross basis and probably 20, 25 trillion in net assets including $9.5 trillion in Treasuries. Nobody's short dollars." Japan hit both conditions at once — oil back over 80 with a current-account deficit, plus a yen to defend — so it sold Treasuries; and Bessent "did exactly what Powell did, exactly what Yellen did" (9:58). The trade that followed was gold, not the dollar.
Watch for

14:21 4. Track the defended yield level, and treat every upward revision as weakness

The repeatable method
  1. Anchor to the yield that prevailed when the current stress began — the pre-event baseline, not a long-run average.
  2. Record the level at which policymakers have observably backed off: the walk-back, the tweet, the abandoned strike, the issuance change. That is the defended level.
  3. When the defended level is raised, do not read it as increased tolerance or strength. Read it as an admission that the old level could no longer be held without visible inflation.
  4. Estimate the ceiling from what happens above it: at some yield the response stops being jawboning and becomes buying. Name that yield in advance and treat crossing it as the trigger for soft yield-curve control.
  5. Trade the sequence, not the level: approach to the defended level implies a policy back-down; a forced upward revision implies the release valve (currency, gold) is next.
Here: "10-year Treasury yield was at 3.94% when we bombed Iran on February 28th. And now it's 4.7." Through April, "every time it hit 4.4, Trump backed off"; then "to their credit they backed up 4.6, 4.7 — but that's not a sign of strength. That's we can't defend that level without inflation picking up, we need to defend a higher level." The named ceiling: "practically speaking the 10-year never would have gone to six. It would have gone to five and then they would have started buying 10-years one way or another" (13:34) — which is Warsh's own pre-announcement of "a fair price for assets in a crisis."
Watch for

24:54 5. Trade the reaction to the second intervention, not the first

The repeatable method
  1. When an authority intervenes to hold a level, check whether anything in the underlying flow changed. If not, the level will be retested — so the first intervention is information, not an event.
  2. Measure the retracement speed. The faster the market gives the intervention back, the shorter the credibility that intervention bought.
  3. Identify the belief the intervention was defending — usually a belief about the people in charge ("these ones are different"). That belief, not the price level, is the position being held.
  4. Wait for the second intervention. Then watch the market's reaction, because that is when the belief breaks and the repricing happens all at once.
  5. Be positioned before it in what benefits from the capitulation — the debasement assets — rather than trying to trade the intervention itself.
Here: intervention at 163–164 on USD/JPY (FXY); "we're at 159.25, so we're almost back to 160," and "the yen's already retraced over half of the strengthening move" (20:19). Underlying unchanged: Japan short dollar oil with oil over 80. The belief being defended: "this aura around Bessent, this aura around Warsh — they're Wall Street's guys, they're Druckenmiller's guys, they're Soros's guys. They're going to have to do the same stuff because the math doesn't math." The trade: "I'm going to be much more interested to see the reaction of the markets… everyone will be like I got fooled again. You could see equities rip, gold really rip, and I think it could be good for Bitcoin."
Watch for

22:19 6. When everyone debases together, stop measuring in FX crosses

The repeatable method
  1. When several major sovereigns announce the same fiscal expansion within days of each other, treat it as coordination, not coincidence — and as a single policy, not five.
  2. Recognise that simultaneous debasement cancels out in the relative measure. FX crosses and DXY will show nothing, because every denominator is falling too.
  3. Switch the numéraire to something that cannot be issued by any of them — gold — and re-plot everything against it. That is where the effect is visible.
  4. Expect the official statistics to deny it, because the FX screens support the denial. Treat the gap between the gold-denominated series and the official inflation series as the size of the trade.
  5. Position in the numéraire and in the assets that inflate nominally, and expect the fixed-income leg to absorb the loss.
Here: "within like a five-day period we had the US, the UK, Germany, Korea and Japan all come out and say we're going to borrow more money and spend it on the military. That's basically just defense stimmies." The mechanism: "if the US, UK, Germany, Japan and Korea all do this at the same time, then all their currencies against each other all debase against gold and against stocks and against inflation — but not against each other. You'll get higher inflation that is said to not be higher inflation, and a weaker dollar that will look on our screens like it's not a weaker dollar." And the social enforcement: "anybody in the establishment who mentions the fact that gold is 6,000 or 7,000 bucks will be ostracized."
Watch for

50:51 7. Split the position by legislative risk, not by expected return — bullion versus miners

The repeatable method
  1. Once a thesis implies an asset is being pulled back into the monetary system, add a risk column that valuation models don't have: the risk that a government simply takes it.
  2. Rank each expression of the thesis by how seizable it is. Fixed, licensed, immovable assets inside a jurisdiction (mines, pipelines, refineries) rank worst; portable, private, dispersed holdings rank best.
  3. Size accordingly — don't exclude the higher-return expression, weight it down. Keep the majority in the least-seizable form.
  4. State the failure mode you are insuring against in one sentence, so the sizing can be revisited when it changes: "I don't want to be wrong for the right reason."
  5. Store the safe leg with jurisdictional diversification, so the insurance isn't concentrated either.
Here: "I would prefer to own gold bullion. To the miners, I own both — it's probably an 80/20 split, maybe a 75/25 split, bullion to miners," in physical form in private vaults at different locations, "almost all in the US, a little bit in Switzerland." The risk being priced: "if gold's going back into the system, there are risks of nationalization of assets. We've already seen that — the United States is threatening to do it or has done it." And the asymmetry that decides the weighting: "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" (GLD over GDX).
Watch for

37:32 8. Price the slogan — run the wartime-footing arithmetic before believing a build-out

The repeatable method
  1. When a policy slogan invokes a historical precedent ("wartime footing like 1940"), go get that period's actual numbers rather than its mood: deficit as a share of GDP, central-bank balance-sheet growth, the funding rate, the inflation rate.
  2. Scale them to today's economy to get the required run-rate. That number is usually absurd, and the absurdity is the finding.
  3. Then price the consequences for each asset class in turn — the long end, mortgages, bank collateral, anything discounted off the 10-year.
  4. Look up the parts of the precedent nobody quotes: in this case capital controls and confiscatory tax rates. Ask whether the advocate would accept those too. If not, the slogan is not a plan.
  5. Convert the finding into a timetable, and use the timetable — not the announcement — as your investment horizon.
Here: 1940 was a 25%-of-GDP deficit (about 4× today's 6%, i.e. $8 trillion a year), the Fed's balance sheet grew 10× in three years, and it was funded at 3/8 of a percent. "So you're going to have to have the Fed finance $8 trillion a year at 3/8 of a percent in the T-bill market. What do you think inflation's going to do? 30, 40, 50% for the next three, four, five years. Now which part of the long end of the curve do you want to own?" (TLT.) The unquoted parts: capital controls — "that's your big reset right there, it's forced" — and a 90%+ top marginal rate. The timetable that falls out: reshoring in 5–10 years is "a freaking pipe dream. Maybe 10 to 15 best case. Probably more like 20" (39:53).
Watch for

45:06 9. Find the quantity that cannot be bought — then read its price as already hyperinflated

The repeatable method
  1. For any announced build-out, convert the plan into physical units — mines, transformers, megawatts, trained people — rather than dollars.
  2. Count how many of those units exist, are permitted, or are even discovered. Add the build lead-time (people on site, years to first production).
  3. If the required count exceeds what physically exists, stop modelling the price and note the conclusion: the currency has already hyperinflated against that unit, whatever the spot quote says.
  4. Invest in the constrained unit or its nearest tradeable proxy, and treat the announced timetable for the build-out as fiction.
  5. Check who spent the previous cycle acquiring those units cheaply — that tells you who set the price and who has the option.
Here: ~50 mega copper mines needed in 20 years, and "there aren't actually 50 deposits available right now"; a single large mine needs a thousand people on site for 12–18 months before it operates (41:25). Gromen's conclusion: "if something's impossible, you know what the dollar value of 50 major copper mines is? It's a fugazi. The dollar has hyperinflated against the major copper mines number five through number 50 that we need. There is no amount of dollars that can get you them, because they don't exist" (Copper). Who bought the option: China, via what Carmen Reinhart called opaque Belt-and-Road lending, "securing these supplies dirt cheap" while the US spent '02–2020 in the Middle East (44:42). And the human constraint is real: "40 years of moving away from manufacturing and mining to financialization — there's not the bench at all" against $100 million of mining education versus $37 billion spent in Iran in four months (37:04).
Watch for

17:51 10. Score the deterrent as part of the bond bid — production rates as a macro input

The repeatable method
  1. Accept the premise explicitly rather than as metaphor: part of what makes reluctant foreign holders buy and hold Treasuries is the security guarantee.
  2. Score the guarantee mechanically. In an attrition contest, compare the two sides' production rates — offensive weapons per month against interceptors per month — not inventories or headline capability.
  3. If the cheaper side's production rate exceeds the expensive side's, the deterrent has an expiry date, and it is calculable the way the Allies calculated German factory output in 1943.
  4. Translate the expiry into the financial variable: a deterrent that can't deter removes a non-economic source of demand for the sovereign's paper.
  5. Look for corroboration in leaks and personnel behaviour — senior officers publicly counselling restraint is the visible output of the same arithmetic.
Here: the IRGC's claim that "we can produce these offensive weapons faster than the enemy can produce defensive weapons," which Gromen says he heard from a well-placed US source three months earlier. The end state briefed to the President: "we're heading toward a point where we can't defend them at all, Mr President — and once we hit that point, you will have completely destroyed the credibility of the US defense umbrella around the world" (17:31). The corroboration: leaked stories of Dan Caine and Pentagon generals advising against escalation (15:24). The financial translation: a senior officer's line that "part of the military's job is to be the muscle to threaten people into buying Treasuries" — so "guess what that implies for yields, for the dollar, for a lot of things."
Watch for

47:59 11. Stress-test the "we'll grow out of it" claim on its own terms

The repeatable method
  1. Take the optimistic growth claim at face value and ask what else must be simultaneously true for it to solve the debt problem.
  2. Look for the internal contradiction — usually a productivity boom that must also create no unemployment, because the leverage in the system can't survive a rise in joblessness.
  3. Then apply the bondholder test: at the claimed nominal growth rate, who willingly holds the sovereign's paper at the current yield? If nobody does, yields must rise — which cancels the growth.
  4. Conclude with the only configuration that closes: cap the yields (yield-curve control) and let the currency absorb the difference.
  5. Restate the outcome in a non-printable numéraire so the result is legible: growth in dollar terms, austerity in gold terms.
Here: against 8% GDP talk — "you'd need an AI productivity boom, but it cannot drive any layoffs in the short run; if unemployment goes up 5%, the entire levered system comes unwound. So we need a productivity boom where nobody loses their job. That's a contradiction." The bondholder test: "who's going to hold bonds at four if you're growing eight nominal?" The only closure: "they can grow out of it. They just have to do some form of yield-curve control and let the currency take the hit. You'll have nominal growth, but in gold terms you won't have any growth — austerity in gold terms, growth in dollar terms" (SPY up in dollars, down in gold; TLT the funding source).
Watch for

Methods distilled from the public YouTube video (transcript in transcript.txt) for personal study. Not investment advice. © Goldfinger Capital / Luke Gromen / Forest for the Trees (FFTT) for source material.