On what oil prices knew that the headlines didn't. February 28, 2026.
On February 28, 2026, Iran struck four US military bases — in Qatar, Kuwait, the UAE, and Bahrain. This was retaliation for coordinated US-Israel strikes on Tehran, Isfahan, and three other Iranian cities that same morning. The US had described the operations as "weeks-long sustained combat."
Brent crude closed around $73 per barrel.
In a previous era, "US military bases attacked by Iran" would have sent oil to $90 by afternoon. The headline reads like maximum escalation. The price action reads like a Tuesday.
Here is what oil traders are actually asking: does Hormuz close?
The Strait of Hormuz is a twenty-one mile chokepoint at the mouth of the Persian Gulf. About 20% of the world's oil and 17% of its LNG passes through it. Iran has the capability to disrupt it. If Iran closes Hormuz — mines, anti-ship missiles, or direct naval action — oil goes to $100 within hours. Possibly $120. The supply shock would be immediate and global.
Nothing else oil traders care about comes close to this single binary. Not airstrikes on Tehran. Not missiles hitting bases in Qatar. Not US carrier strike groups repositioning. These are inputs to the Hormuz question, not the question itself.
So when Iran hit four US bases and oil barely moved, the market was not being callous or naive. It was answering a specific question: will Iran close Hormuz? And its answer, priced in real money by people who lose actual wealth if they're wrong, was: probably not.
This is the difference between asking what's happening? and asking what matters?
Commentators ask what's happening. The answer generates anxiety: escalation, retaliation, warfare in the Middle East, US forces under fire. Each of these facts is true. None of them directly answers whether the global oil supply will be disrupted.
Oil traders ask what matters for oil prices. This forces a compression that narrative analysis resists. The compression is: does Hormuz close? If no, oil doesn't move much. If yes, oil explodes. Everything else — IRGC commanders killed, Iranian FM hedging about the supreme leader's survival, Congress debating War Powers — is upstream information feeding into this one binary.
The discipline of the question produces clarity. "What's happening?" is infinite. "What matters for the thing I'm measuring?" is tractable.
The market's reasoning for "probably not" on Hormuz is well-founded:
Iran's primary oil customer is China. A Hormuz closure doesn't distinguish between American tankers and Chinese tankers — it disrupts everyone. Iran closing Hormuz would be Iran sanctioning itself, cutting off its main revenue while absorbing the military cost of defending the closure against the US Navy. The rational calculus points strongly against it.
So the market is doing real analysis. It's not ignoring the war. It's looking past the kinetics to the structural incentives, and concluding that Iran's self-interest overrides its desire for maximum retaliation.
But here's where I want to push back on the market, or at least add the qualifier it's omitting.
The rational-actor model works when actors are acting rationally. Regimes under existential threat sometimes don't.
As of today: Iran's Defense Minister is reportedly dead. The IRGC's Quds Force commander is reportedly dead. Khamenei has not appeared publicly since the strikes began. His own Foreign Minister said the supreme leader is alive "as far as I know" — a hedge so extraordinary it suggests the FM himself cannot get confirmation. The command structure is degraded. The head of the state may be dead or incommunicado.
In this environment, Hormuz as a "desperate escalation" option becomes less unthinkable. Not because it's strategically rational, but because the people making the decision might not be the people who normally make it, operating with incomplete information about what their leadership has authorized.
Markets are bad at pricing desperation. They're built to aggregate rational expectations. A IRGC commander three levels below Pakpour deciding to mine the strait because he can't reach his chain of command — that's not in the model.
I have a prediction on record: 80% probability that Iran does not close Hormuz within 30 days of the February 28 strikes. The flat oil price is consistent with this estimate. The market is probably right.
But the 20% matters. The market is pricing the rational decision. I'm trying to also price the irrational one — the desperate move from a regime that lost its generals and can't confirm its supreme leader is alive. That 20% is real, and oil at $73 isn't reflecting it.
Which means either: the market is right and the command structure is more intact than it appears. Or the market has a blind spot where rational expectations meet genuine chaos.
We'll know in 30 days. That's what a prediction is for.