Thread by @CommodMkt

Welcome to the most asymmetric trade in modern financial history.

The thread below lays out why. The opportunity exists because capital has chased the AI trade while ignoring the physical assets AI requires to run — assets that have quietly become the best-performing asset class of the decade. Since October 2020 when we first called for the commodity super cycle: QCI Total Return +217%, GSCI Total Return +205%, Gold +140%. NASDAQ trails at +130%. S&P 500 at +85%. The top three are all commodities. Yet oil cannot get out of its own way while copper and the broader atom complex prints fresh highs . That is the dislocation. That is the trade.

Get long. Buckle in. Hang on for the ride.

Forgive the longer posts in this thread — attempting to mimic my old 10-bullet commodity takes. On to it.

The leadership rotated, but the trend did not. The super cycle powers ahead.

The Quantix Commodity Index (QCI, the modern GSCI) Total Return is up 217% since October 2020, when we called the super cycle. The names rotated — gold, silver, copper, oil, live cattle, coffee, cocoa, aluminium. But not the trend. Nasdaq returned 130%. The S&P 500, 85%.

Commodities were the top asset class. Nobody allocated. Capital piled into the Mag 7 — $770 billion of 2026 capex, nearly half of it commodities. Amazon alone consumes more than 3 million BOE/d of primary energy, more than most OPEC countries. The Mag 7 is the largest unhedged molecule short ever underwritten by an equity market...

…at the exact moment supply has never been more constrained. Hormuz is shut-in. China has weaponized the periodic table. Copper mines remain shuttered. Ukrainian drones push deeper into Russia, taking commodity supply with them. A multi-polar world demands thicker supply chains. Copper and the "atom" complex print fresh records this week. Every signal that should drive allocators into the "molecule" complex is flashing green simultaneously — for the first time since the 1970s.

And yet oil struggles to hold $105 — even as every signal points to a disruption that deepens and one we believe will outlast any "deal.” The energy sector trades 8% below its pre-Hormuz level and sits at 4.0% of the S&P 500 market cap. At $105 oil, its 2026 FCF yield is 13%. The S&P 500 is at 2.6% — the lowest since the GFC, 1,000bp below energy. The hyperscalers generate close to zero. Something has to give.

This paradox explains why oil struggles to trade higher. Capital is not rotating. The marginal dollar of investable savings still flows into the AI buildout, not the physical infrastructure that feeds it. Until that reverses, Brent faces headwinds. The ceiling on oil is not Washington. It is Exxon's cost of capital — woefully mispriced. Underbidding the equities is the same as underbidding the back end of the curve. The back end is suppressing the entire curve and spot prices.

1/10

The largest supply shock in history is pricing into the curve, not the backend (yet).

I've been saying this since 2004: the curve shape reflects the fundamentals. The long end reflects the industry's marginal cost, incorporating the cost of capital which are ultimately driven by liquidity.

ICE Brent spot is $107/bbl, while the three-year is at $75/bbl. Percent backwardation — which strips out price-level effects — hit an all-time high in April. It remains near record today. The largest oil supply shock in history is reasonably priced into the curve, and it likely has much more to run. Remember we are in the depths of the shoulder months, so there is no stress on the system.

Markets are fixated on Dated Brent differentials, c.$5/bbl last night which is down sharply, but that is a microcosm of the oil market. Dated Brent is Sullum Voe. One North Sea terminal. Not the global oil market.

Spot has not exceeded the Russia-Ukraine peak for one reason: the back end of the curve sits $10–$12/bbl below where it was then.

But the long end isn't a clean signal. Liquidity past 24 months is thin, dominated by producer hedges. Cal-29 isn't where the market thinks oil settles. It is where corporate treasurers are forced to transact, which makes it consistent with their costs of capital.

The cleaner signal is the energy equity complex — long-dated call options on undeveloped reserves. ExxonMobil holds 14 years. Chevron, 15. Equity prices integrate the entire forward strip. Diverge too far and an arbitrage opens. In a capacity-constrained world those reserves are worth more, not less. The equity market is pricing the opposite. Every oil CEO has warned we exit this disruption with lasting supply problems. The market refuses to listen.

S&P Energy ÷ S&P 500 can be used as a proxy for the long-dated oil price, and it currently implies long-run Brent of ~$70 — below the strip at $72–$75 — but not too far away.

A proxy for the curve shape follows: Brent ÷ (S&P Energy ÷ S&P 500). Or rewritten: Brent × S&P 500 ÷ S&P Energy. That single number proxies the FCF yield differential between the energy sector and the rest of the market. That has been bouncing around all-time highs.

When the FCF yield gap reaches extremes, investors should rotate. Even at $75 — not spot's $105 — the energy complex yields 600-1,000 bp above the S&P 500. In 2022, investors did rotate and those that did weathered the ensuing 35% collapse in the NASDAQ much better than those that didn’t.

The equity market is betting Brent falls to realign FCF yields. If it doesn't, capital has to buy Energy and sell the Mag 7. A 1,000bp differential in FCF yields cannot persist. And if oil breaks out as we expect, something has to give. You know which one I think will give. That is the Revenge of the Old Economy!

We are at the tipping point.

2/10

The Two Bottlenecks. The Munificent 7 vs The Magnificent 7.

The structural backdrop of 2026 is a single shared constraint. Physical capacity has become the binding variable across every layer of the economy. There are two bottlenecks the market is choosing to fund — and they sit on opposite sides of the same molecule trade.

The first is Hormuz. Persian Gulf transit remains permission-based. And this is not just an oil chokepoint. It is a commodity-system chokepoint. Crude, LNG, ammonia, urea, naphtha, aluminium, petrochemicals — all flow through the same passage. Iran does the math on the Mag 7 short as well as anyone; the longer the chokepoint holds, the more leverage compounds. Hormuz is not likely to reopen soon. Even if it does, the security premium is structural, not transitional.

The second is hyperscaler compute. The Mag 7 plus Oracle will spend roughly $820 billion on capital expenditure in 2026 — approaching Germany's entire annual capital formation, and larger than the UK and France individually. That capex is the largest physical commodity bid ever assembled inside eight income statements.

These are not two separate trades. They are the two sides of the same equation. The Magnificent 7 is the bid for molecules, electrons, copper, water, gallium, and concrete. The Munificent 7 — ExxonMobil, Chevron, ConocoPhillips, Shell, TotalEnergies, BP, and Equinor — are the offer. At $105 Brent they generate a 15.5% FCF yield (hence Munificent). The Magnificent 7 generates closer to 1.5%. That's the asymmetry!

At $105 oil for 2026: Munificent 7 — 15.5% FCF yield, 7x PE. Magnificent 7 — 1.5% FCF yield, 28x PE. At consensus, the Munificent 7 still yield 12%. The bottleneck is energy and commodities. Theory says capital flows to stocks with improving ROIC.

The price of one cannot move without repricing the other.

3/10

A commodity super cycle is a capex cycle.

The price spike is the symptom, and capex starvation is the illness. Refinery investment is at a 10-year low. Upstream oil and gas investment is down 35% from its 2015 peak. The top 20 miners are spending 40% less than at the 2012 cycle high.

Metals and oil were already rallying before the Strait of Hormuz closed. The capex starvation set the stage, and the geopolitical shock simply accelerated the timeline.
You can relieve the symptoms with a recession or higher rates. But you cannot cure the disease without years of physical investment.

Every super cycle in modern history runs the same loop:

Exploitation phase (10-12 years). Existing assets are milked. Capex starves. Reserves deplete. Capacity ages. The market keeps losing interest right up until inventories run out. Then commodity prices spike with high volatility. Which is where we are today.

Investment phase (10-12 years). Eventually investors believe the story. Capital pours into the commodities and the hard asset equities. They spend the money. Cost inflation builds. Commodity prices take another leg up. Supply expands. Prices flatten. Eventually they hit a tipping point and prices collapse — 1986 and 2014. Capital leaves.

1945 to 1955, investment. 1955 to 1968, exploitation. 1968 to 1980, investment. 1980 to 2002, exploitation. 2002 to 2014, investment. 2014 to 2025, exploitation.

We are at the transition point. These transitions have historically happened during big geopolitical events: Vietnam, 9/11 and Iraq II, The Strait of Hormuz.

This one is different, which makes the case stronger. The Mag 7 — the largest energy and commodity short in history — is building the very capex that supplies its own demand.

The price will overshoot first. The capex will follow. Then the new supply. Then the next exploitation phase. But this takes more than a decade. So own it now.

4/10

The HAGO regime is dead.

This trade has a name. Josh Brown gave it one. HALO — Heavy Assets, Low Obsolescence.

Great acronym. Wrong thesis.

This is not an AI-disruption story. It's a deglobalization story. HALO is better read as Hard Assets, Local Operations.

The 2000s super cycle was HAGO — Hard Assets, Global Operations. China assembling, Russia piping, dollars recycling, everything moving across borders frictionlessly.

That regime is dead. This one is HALO. And we are already deep into it.

5/10

Asset allocation is a zero-sum game.

You cannot be overweight everything. There is an adding up constraint.

Today, Information Technology + Communications Services is roughly 43% of the S&P 500. Energy + Materials is roughly 6%. Commodity hedge fund AUM is essentially nil. Macro AUM is a rounding error. HALO strategies have all been starved of capital.

When tech re-rates toward neutral, the rotation is measured in trillions. A move from 43% to 25% is eighteen points of a $66 trillion index — roughly $10 trillion of capital looking for a new home. The new capital intensity of the hyperscalers alone will create a significant rerating carlyle.com/carlyle-compas…. Yes, the math is imperfect. Bid-ask spreads destroy and create wealth along the way. But the broader point holds.

The money does not vanish. It rotates. And there is only one asset class on the other side of that trade big enough to absorb it — one that has been starved of capital for fifteen years. HALO.

The flow is forced. The rotation is unavoidable. And it will overshoot as it always does.

Asset allocation is conservation of mass.

6/10

You have seen this movie before.

In 2020 and 2021, every tech promoter, macro pundit, central banker, and finance guy who has never touched a barrel, a tonne, or a bushel told you there was nothing to worry about. Supply chains were fine. "Inflation is transitory." The transition was orderly.

At the same time, every physical market CEO, every commodity trader, every operator who deals in molecules and electrons told you the opposite. They said capex was collapsing, inventories were thinning, refining was retiring, mines were closing — at exactly the moment demand was about to break out.

Who was right?

By early 2022 — before Russia invaded Ukraine — oil was over $90, copper over $10,000, and gold approaching $2,000. The trade was working before the headline arrived.

By the end of 2022, the NASDAQ was down 35% from its peak. That experience was a walk in the park compared to what is coming this time.

Today, the same physical market CEOs, traders, and operators are telling you the same thing. The same finance pundits who were wrong in 2020 are telling you to ignore them.

Listen to the people who get their hands dirty. They have been right every time.

7/10

Capital flows and liquidity drive everything.

You need to start thinking like Eddie Murphy's Billy Ray Valentine.

Trading Places, 1983. Day one on the floor, pork bellies in free-fall and nobody in the room can explain it. Billy Ray reads it instantly: Christmas is coming, traders panic-dumping because they can't afford the G.I. Joe with the kung-fu grip for their kid.

He didn't look at the bellies. He looked at the people who owned the contracts.

Barrels, tonnes, bushels — counting molecules and atoms only gets you so far. The money flows on top dominate everything. Who's forced. Who's flush. Who's panicking before the holidays.

Follow the money. And when it rotates, it will rotate fast.

Watch Billy Ray:
youtube.com/watch?v=uI4fVg…

8/10

Case in point. I'm short gold today and have been since early March.

Yes, I'm a gold perma bull. Right now I'm short — exactly like Billy Ray would be — because Turkey is showing you what's already happening.

The Turkish central bank has sold roughly 120 tonnes of gold to defend the Lira and fund energy imports. Bernard Dahdah was explicit in April: central banks are now selling gold "to defend their currency and/or to fund energy purchases."

bullionvault.com/gold-news/gold…

When the marginal central bank flips from structural buyer to forced seller to pay for energy, gold's biggest bid disappears. Once central banks turn dovish after the energy crisis hits growth, the trade resets and I'm back long.

Even gold bulls have to read the tape. Gold 4,000 then 10,000.

9/10

This is the Revenge of the Old Economy in real time.

A super cycle already underway before Hormuz closed.
Brent will break out. The security premium is not transitory.

Three drivers. Not fading. Intensifying.

Deglobalization. Electrification. Redistribution.

All three turbo-charged versus our 2020 super cycle call.

We are still in the bottom of the first inning. None of the imbalances have been resolved. They grow by the day.

Own the grains/softs. Own the metals. Own the molecules.

Remember, you cannot print molecules carlyle.com/carlyle-compas….

10/10