Market Commentary
When the Physical World Catches the Financial One: The Iran War, the Hormuz Cascade, and the End of the Free Ride
The Thesis Nobody Wants to Hear
I have argued for years that the value chain of artificial intelligence does not terminate in software. It runs through compute, through energy, through materials, through the physical infrastructure that makes the digital economy possible. Code value traces back to compute cost, which traces back to energy cost, which traces back to materials cost. At every step, what looks like a technology problem is actually an infrastructure problem. What looks like a software margin is actually an energy margin in disguise.
This argument has been unpopular. During the years when capital flowed into AI at rates that would have embarrassed the South Sea Company, pointing out that data centres consume electricity, that electricity requires fuel, and that fuel transits chokepoints was treated as a category error — the complaint of someone who did not understand that software scales infinitely. The market priced AI as though it existed in a vacuum. Fourteen days into the Iran war, with Hormuz at 97 per cent closure, Brent above $100, and the US Energy Secretary admitting the Navy cannot escort tankers, the vacuum has been punctured.
This is not a note about oil prices. It is about what happens when a financial system built on the assumption of infinite, cheap, uninterrupted energy collides with a physical world that is none of those things. The collision was always coming. The war determined the timing.
The Coiled Spring
On 3 March, as US and Israeli bombs were falling on Tehran, Dr Michael Burry published what he called a "Foundations" article on his Substack. Burry — the former hedge fund manager whose correct bet against the US housing market before the 2008 financial crisis was documented in Michael Lewis's The Big Short — has closed his fund and now writes exclusively for subscribers. His track record is defined by a willingness to do the work that arrives at conclusions the market does not want to hear, and then to hold those conclusions through years of being told he is wrong. He has earned the right to be taken seriously when he says the market is mispriced.
His argument is structural, not cyclical. It rests on three pillars that have never converged before. The first is the passive investing transformation: over 60 per cent of US equity fund assets are now in index funds that buy the entire market regardless of valuation, destroying the price discovery mechanism that is supposed to be the market's social purpose. The second is demographic: Baby Boomers whose 401(k) contributions created the passive investing boom are now reaching the age of Required Minimum Distributions, which Burry calculates at $250 billion per year today, peaking above $1 trillion in the early 2030s. By 2028, for the first time in the defined contribution system's history, redemptions will exceed contributions. The engine that inflated US equities for a generation reverses polarity.
The third pillar is the disappearance of the corporate buyback bid. The Magnificent Seven technology companies have collectively slashed buybacks by 74 per cent year-on-year in Q4, to $12.6 billion. Oracle has borrowed $25 billion, Meta $30 billion, Alphabet $15 billion or more. These are capital expenditure loans to build AI infrastructure. Pivotal Research projects Alphabet's free cash flow will fall 90 per cent in 2026. Morgan Stanley and Bank of America both see Amazon at negative free cash flow. Burry's Shiller cyclically adjusted PE analysis shows the S&P 500 has not touched its long-term mean in 34 years — a record by a wide margin — and his log return decomposition demonstrates the current run is overwhelmingly a multiple-expansion story, not an earnings-growth story. When the multiples compress, there is no earnings floor to catch the fall.
Burry timed his publication deliberately, noting he released it "given all that is going on in global geopolitics." He is not predicting the war causes a crash. He is observing that the market is a coiled spring, and the war is a potential trigger. The spring was coiled long before the first missile struck Tehran.
The Hormuz Cascade
Here is where Burry's financial architecture analysis and our energy-infrastructure thesis connect — and where the data has moved beyond anything the consensus was modelling even a week ago.
Goldman Sachs's Oil Tracker of 12 March reports that estimated flows through the Strait of Hormuz have collapsed to 0.6 mb/d — a 97 per cent decline from normal levels. The net hit to Persian Gulf exports stands at 16.2 mb/d, which Goldman describes as sixteen times larger than the peak April 2022 hit to Russian oil production. Only 22 tankers have crossed the Strait since 1 March, with S&P Global data showing most operating with AIS transponders switched off. The IEA estimates at least 10 mb/d of crude and condensate production losses. Refinery outages across the Middle East have reached 2.0 mb/d, including a precautionary shutdown of the UAE's Ruwais refinery. Brent has breached $100. The call options implied volatility skew — the cost of hedging against higher oil — has reached an all-time high in Goldman's data.
Crude prices dropped briefly as Trump suggested the war could end "very soon." Hours later, US intelligence confirmed Iran had deployed naval mines in the Strait. Oil surged more than $10 per barrel in fifty minutes. This was the moment the disruption decoupled from the war's timeline. Missiles and drones go away when the fighting stops. Mines do not.
They sit in the water until physically cleared by specialised vessels — a process requiring weeks to months. CENTCOM destroyed 16 mine-laying vessels on 10 March, but US intelligence assesses Iran retains 80–90 per cent of its smaller boats and can deploy hundreds more. The IRGC has layered mines with explosive boats and coastal missile batteries to create what amounts to a persistent denial zone.
The cascade effects compound daily. Iraq has cut production from 4.3 million bpd to 1.3 million as storage filled and southern export terminals became unreachable — Iraq has no bypass route. Kuwait has cut 1.5 million bpd. Goldman estimates Gulf producers have approximately 25 days of storage before forced shutdowns cascade through the system. Shutting wells in mature fields is not costlessly reversible; some require months of workover to restart. Every additional day of closure creates production losses that extend long beyond the disruption itself.
And then there is the grade problem. Gulf producers export medium to medium-light crude (28–34° API) that Asian and European refineries were built to process. US shale is significantly lighter (39–42° API). Goldman's own data shows the WCS–WTI heavy-versus-light quality differential has narrowed sharply as the market prices scarcity of medium grades. Substitution is not one-for-one. This is simultaneously a volume shortage and a type-of-oil shortage. Qatar's Ras Laffan LNG facility — handling roughly 20 per cent of global LNG supply — has declared force majeure on contracts to Asian buyers. Saudi Arabia's Ras Tanura refinery is offline with no restart timeline.
The G7 has announced the largest emergency reserve release in history: 400 million barrels over 120 days. Goldman assumes actual releases of 213 mb at 2.4 mb/d, dampening the hit by roughly half. Half, against a 16.2 mb/d shortfall. And OECD strategic reserves are already over 200 million barrels below their early-2022 levels before this draw even begins. Polymarket now prices only a 19 per cent probability the conflict ends in March, with the base case at April to mid-May.
Hormuz remains impaired for a minimum of three to six months after any ceasefire, with full normalisation six to twelve months out. The price floor is set not by the war's duration but by the physical time required to clear mines, rebuild infrastructure, and restart shut-in wells.
The Three Bids That Disappear Simultaneously
Now connect the threads. Burry documents two structural bids that have supported US equity valuations for a generation: passive fund inflows and corporate buybacks. Both are reversing. The Iran war introduces a third.
Gulf sovereign wealth funds hold enormous positions across both public and private markets. ADIA, PIF, QIA, and KIA collectively manage trillions in assets — and they are among the world's largest limited partners in private equity and venture capital. These funds are financed by hydrocarbon revenue. If Gulf oil exports drop 80 per cent for three to six months — which the Goldman data suggests is the current trajectory — the fiscal arithmetic changes fundamentally. Vision 2030, defence procurement, social spending, and domestic subsidy programmes do not pause because export revenue has.
The contagion runs in both directions. If sovereign funds begin liquidating public equity positions, that is a third structural bid disappearing on top of the passive reversal and the buyback collapse. But the private markets are equally exposed. When the largest LPs in the world shift from deploying capital to recalling it — or simply stop committing to new funds — the denominator effect alone forces markdowns across private equity and venture portfolios. The valuation mirage that the private markets sector has worked to maintain for two years breaks under exactly this kind of liquidity stress. And forced PE exits compress public market comparables, which in turn pressure the next round of private marks. The reflexivity is underappreciated.
The IMF research Burry cites adds a further dimension: the traditional negative correlation between stocks and bonds has broken down since 2019. Oil-driven inflation pushes rates up, pushing bonds down, while simultaneously compressing equity multiples. Stocks and bonds fall together. The 60/40 portfolio — the foundation of institutional asset allocation — offers no protection.
The AI Paradox
Here is where the market's central narrative fractures. The companies borrowing tens of billions to build AI infrastructure are building physical things: data centres, GPU clusters, cooling systems, power connections. These physical things consume electricity at industrial scale. That electricity, globally, traces back to fuel — and a meaningful share of that fuel traces back through the Strait of Hormuz.
The market prices AI as a software margin business. It is an energy margin business. When the energy costs 40 per cent more for twelve or more months, the economics of every AI capital expenditure decision made in the last two years need to be revisited.
The companies that borrowed $70 billion or more to build infrastructure are now servicing that debt in a higher energy cost environment while generating less free cash flow than their lenders modelled. The gap will show up in earnings, in credit spreads, and eventually in equity multiples — the same multiples Burry has documented are at historic extremes with no earnings floor beneath them.
Meanwhile, the defence production bottleneck compounds the problem. The White House convened CEOs of seven major defence contractors last week to discuss quadrupling production of missile systems and interceptors. You do not quadruple production for a three-week conflict. That meeting tells you the US defence establishment is planning for a multi-year demand surge — rearmament across NATO, Gulf procurement, and Indo-Pacific readiness simultaneously. Every Patriot and THAAD interceptor expended over the Gulf is one that cannot be delivered to another customer for years.
What China Sees That the Market Does Not
Goldman's vessel-level data confirms what we identified in our earlier analysis: no Asian-flagged tankers have been attacked. Of the 22 vessels that have crossed the Strait since 1 March, the majority are Iranian-flagged, sanctioned, or operating under false flags — almost all bound for China. Beijing is receiving Gulf crude through a corridor that everyone else is shut out of.
Chinese data centres running on discounted Iranian crude face structurally lower energy costs than their American counterparts running on Permian shale or Appalachian gas at war-premium prices. If this differential persists for six to twelve months, it compounds into a meaningful advantage in compute cost — the single largest variable cost in AI model training and inference. The AI race is, at bottom, an energy race. And right now, China is running it on cheaper fuel.
For principals with access to Chinese-intermediated supply chains, this creates potential arbitrage — with corresponding sanctions, legal, and reputational risks that must be assessed case by case. For those without such access, it represents a structural competitive disadvantage that the market has not begun to price.
Tactical Positioning
The Goldman data validates and strengthens our earlier assessment: any model assuming Gulf flows normalise within 60 days of a ceasefire is mispriced. The mine clearance timeline alone extends beyond that window. For entities with contracts denominated in specific Gulf crude grades, assess basis risk immediately — hedging Brent does not cover the medium-grade scarcity that Goldman's quality differential data now confirms.
Three bids degrading simultaneously — passive inflows, buybacks, sovereign fund holdings — mean concentration in market-cap-weighted US index funds is concentration in the assets most exposed to the unwind. Physical infrastructure, energy, materials, and defence sit on the right side of this trade.
Canadian oil sands, Norwegian North Sea, Brazilian pre-salt, and West African producers track oil prices and are the only available supply if Gulf impairment persists. Maritime security and mine countermeasure contractors face sustained demand. European cleantech and critical infrastructure technology become more valuable under sustained $100+ oil.
Every GCC state has been struck. Goldman's data shows refinery outages at 2.0 mb/d and rising. US defence primes are sitting on order backlogs extending years; Gulf states diversifying toward non-US suppliers create a parallel procurement architecture. Advisory mandates bridging European defence and infrastructure technology to Gulf sovereign buyers offer high-margin positioning with long tail.
Track PIF, ADIA, QIA, and KIA portfolio disclosures and secondary market activity. Watch PE secondary markets for forced exits — Gulf LPs reducing commitments will create a repricing wave across private markets that precedes and compounds the public equity adjustment.
The Longer View
Wars end. This one will end. But the structural dislocations it has exposed will not reverse when the shooting stops. The market's two-decade assumption that energy is cheap, abundant, and uninterruptible has been falsified. The passive investing architecture that inflated US equities beyond any historical valuation precedent is approaching its demographic reversal point. The AI capital expenditure boom has converted the market's most valuable companies from cash generators to leveraged borrowers in an environment where the cost of the energy they consume has spiked, the buybacks that supported their share prices have collapsed, and the sovereign funds that anchored both public and private allocations may be heading for the exit.
Burry calls this a coiled spring. I would add that the spring is coiled around a physical foundation — energy, materials, infrastructure — that the financial markets have spent twenty years pretending does not exist. The Iran war has made it impossible to pretend any longer.
The Hormuz cascade — mines, storage saturation, well shutdowns, grade mismatches, LNG force majeure, and a 16.2 mb/d shortfall that strategic reserves cannot close — has ensured the disruption outlasts the conflict itself by a factor of three to six.
The principals who understood that AI was always an infrastructure problem, that energy was always the binding constraint, and that the Gulf's security architecture was fracturing before the first bomb fell — those principals are not scrambling to reprice their models this week. They repriced months ago. The question now is not whether the repricing happens. It is whether you are positioned on the right side of it when it does.
Market Commentary