Actionable insights — Underestimating the Strait of Hormuz crisis

1. The SPR / crude-draw leading-indicator dashboard

1. Treat headline oil price as a lagging, manipulable number; track the physical buffer instead. The buffer is finite, so when it empties, price must rise to ration demand.
2. Watch a basket, not one series: weekly EIA/API crude draws (looking for an acceleration — "big-ass crude draws"), the crack spread, and product inventories (diesel, jet fuel, gasoline) versus their 5-year averages. Multiple gauges below average and falling together = the buffer is depleting.
3. Note what's opaque and adjust: China's >1B-barrel reserve is unreported, so infer from its import behavior rather than expecting a number.
4. Accept you can't time the date ("we know there's land to the west; we don't know when we hit it") — size and stage accordingly rather than waiting for confirmation that arrives only after the move.

• An acceleration in weekly draws; product inventories breaking below 5-year ranges; SPR/exports masking the depletion (US "exporting" SPR barrels); a sudden conciliatory pivot timed to the buffer running low.

2. Own the molecule toll-collector, not the producer or the winner

1. When a theme (AI, electrification) has an obvious but crowded "winner," refuse to pick it. Instead ask: what physical input must everyone in the theme buy, regardless of who wins?
2. Move one or two derivatives down the chain — the distribution/transportation and byproduct layers — where you collect a toll on the input rather than betting on the end-product's economics or margins.
3. Prefer the under-owned layer to the over-owned one (energy ~3% of the S&P vs tech ~40%): the asymmetry favors what's neglected.

• A theme whose winner is consensus and richly priced; an indispensable physical input with constrained supply; a distribution/byproduct layer trading at a fraction of the end-product's multiple.

3. The critical-minerals "follow-the-government-money" screen

1. Start from the official scarcity map — the USGS / US-government critical-minerals list (tungsten, antimony, rare earths, nickel, copper, uranium…) — rather than from a chart.
2. Separate the geology from the cycle: the minerals exist, but a decade-plus of underinvestment means new supply is in poorer jurisdictions, needs more technical expertise, and costs more and takes longer — so the incentive price has to rise.
3. Follow the policy capital: governments that neglected this for 20 years "will spend trillions trying to fix it." Position where that spending must flow, and back proven mine-builders with track records of creating shareholder wealth in hard jurisdictions.
4. Recognize the demand was already on an s-curve (the global south/east) before AI — AI just pulls it forward; don't make the thesis depend on AI.

• Names on the official critical list with structural deficits; government funding/onshoring programs; operators with proven jurisdictional/technical track records; a thesis that holds even without the AI demand kicker.

4. The crowding / "is it luck or skill" test for a hot trade

1. For a beloved trade, measure its share of the index (concentration) before judging the company — a sector at 40% of the S&P is a crowding warning regardless of fundamentals.
2. Pose the one-decision test: forced to buy-and-hold for 10 years, would you rather own the single most-loved name at today's price, or the entire under-owned sector on the other side? The answer reveals the asymmetry.
3. Apply the luck-vs-skill filter to recent outperformance: beating the index by owning one hot name is "a $100 bill in the gutter" — ask whether it's repeatable over 5–10 years before crediting it as process.
4. Use a personally-lived analog (the 1998–2000 internet bubble) to recognize the same FOMO pattern in a "commodity" priced as a secular winner.

• A sector at an extreme share of the index; a "commodity"-type business priced like a secular compounder; outperformance concentrated in one or two names; bubble-era language ("arms race," "the universe is limitless").

5. The asset-light-to-heavy-industry drift test (and the financing tell)

1. For a franchise prized as high-margin and asset-light, check whether it has quietly started building physical infrastructure (power plants, substations, turbines, reactors) — capex that belongs to a low-margin business, not a software one.
2. If so, re-underwrite it at low-margin economics and ask "where are the profits?" — capex without a profit stream is the bubble signature.
3. Read the funding ladder for desperation: a normally cash-rich company reaching for debt (a bond auction) or new equity to sustain the buildout means the spend has outrun internal cash flow.
4. Note who is on the other side of the financing — if even disciplined value money is buying the paper, the whole market is "on one side of the canoe."

• Mega-cap capex/revenue climbing; commentary about "building" physical plant; the shift cash flow → debt → equity to keep funding it; disciplined investors financing the buildout.

6. The jurisdictional-contrarian watch-list method

1. Keep standing watch lists of out-of-favor, ignored jurisdictions (frontier countries, beaten-down sectors) — not to buy now, but to be ready.
2. For each, write the specific catalyst that would make it investable: a privatization/IPO that brings transparency, a government change, a tax-regime rollback. Don't act until the catalyst fires.
3. Prefer the liquid, transparency-forcing vehicle: a partial public listing of state-owned enterprises (the Franklin Templeton/Romania template) puts "sunlight on the situation"; a London-listed national fund is the easy way in.
4. Anchor on durable compounding (a young, under-indebted economy growing 6–7%/yr) so you're early, not just contrarian — "stay ahead of the crowd; they'll never be the first ones in."

• A partial-privatization/IPO that forces disclosure; a pending election/government change; a punitive tax regime poised to roll back; durable 6–7% growth and low debt as the underlying engine.