The Market That Remembers

Part three of three. The oil market says the war is ending. One market disagrees, with money, and it has the longest memory in the room.

To insure a single supertanker for one trip through the Strait of Hormuz right now costs somewhere between three and eight percent of the value of the ship. For a large crude carrier, that is three to eight million dollars, for one passage. Before the war, the same cover ran about a quarter of one percent. The price of getting a tanker through the strait has risen roughly fifteen to thirty times over, and as of the end of June, with the framework signed and the oil price walking back toward normal, the underwriters were not cutting it.

Hold that beside the story the last two parts have told. The oil price decided the war was over, walked itself back to where it started, and then, on July 7, flinched when two more tankers were hit. The insurance market never flinched, because it never agreed in the first place. Part two explained that these two markets measure different things, intent versus execution, and that the insurers are the ones who actually understand the water. This part is about why they are so slow to change their minds, why this time slow is correct, and why the one country that most needs the war over is quietly the reason it may not be. The answer runs through the third gear of this series, and it is the one the desk uses on itself.

The outside view

There are two ways to forecast anything. You can look hard at the case in front of you, the specific deal, the specific promises, the mood in the room, and reason forward. Forecasters call that the inside view, and it is where confidence comes from and where most bad predictions are born. Or you can ignore the story for a moment and ask a colder question: when things like this have happened before, how did they usually turn out? That is the outside view, the base rate, and it is the single most reliable move in forecasting.

Think of two doctors. One listens to how strong you feel this morning and how sincere you are about the gym. The other has read a thousand charts like yours and knows what usually happens next. When you want the truth rather than the comfort, you want the second doctor. The insurance market is the second doctor. An underwriter does not price the Strait of Hormuz by the mood in the room at Doha. He prices it by remembering what happened the last time, and the time before that.

The memory is not comforting. During the Tanker War of the 1980s, Iran and Iraq attacked more than 400 commercial ships, and war-risk premiums went from under a tenth of a percent of hull value to more than five percent, a fifty-fold jump. When that war finally reached a ceasefire in 1988, the premiums did not follow the diplomats home. Rates took an estimated twelve to eighteen months to come back to normal, and that was with United States warships escorting convoys through the Gulf the entire time. The pattern has held in every Gulf shipping crisis since: premiums spike within days of the first shot and take many months, sometimes years, to climb down after the last one. The insurers are not being slow. They are remembering that finishing has always taken far longer than starting.

Why the number stays high even when the odds fall

There is a second reason the premium is sticky, and it is not stubbornness. It is arithmetic.

An underwriter writing war-risk cover is making a very lopsided bet. If nothing happens, he keeps a premium measured in millions. If the ship strikes a mine, he pays out a loss measured in hundreds of millions, the hull, the cargo, the cleanup, the lives. When one side of a wager is bounded and the other is not, the odds stop being the thing that matters. A one-in-fifty chance of a total loss is not a small risk to the man who has to write the check; it is a catastrophe he has to charge for now, because he only gets to be wrong about it once. So the premium does not track the average expected outcome, the way a bookmaker’s line does. It tracks the tail, the disaster at the edge of the distribution, and the tail is slow to shrink because mines are slow to clear and commanders are slow to stand down.

This is worth saying plainly, because it dismantles the obvious objection. You do not need the insurers to believe the war is likely to reignite. You only need them to believe it is not yet impossible, and that a single failure would be ruinous, and both of those are true. That is enough to keep the number at fifteen times normal while the oil price sits near where it started.

When is it actually over

Ask a simple question and the whole series turns on it: on what day does the war end?

There is no such day. The framework was signed on one date, but the mines were not lifted on that date, and the last Revolutionary Guard commander did not get the message on that date, and the freight desks did not exhale on that date. Peace here is not an event. It is a slope, and a shallow one. Philosophers call this the sorites problem, the paradox of the heap: no single grain of sand turns a scatter into a pile, and no single day turns a war into a peace. You cannot point to the moment. You can only look back, months later, and see that the water got safe somewhere along the way.

This is why the oil market’s clean, bright signal is misleading, not wrong. A price is a single number, and a single number wants a single answer, so it snapped back to $67 as if a switch had flipped. The strait does not have a switch. It has a slope, and the only instrument calibrated to a slope is the one that reprices every week and remembers every prior war. In part one of this series, the Federal Reserve chapter noted a truth about signals: the ones worth trusting are the ones that cost the sender something. The oil price is nearly free to move. The insurance quote costs three to eight percent of a hull to send. One of these markets is putting its money where the risk actually is.

The one who gets poorer when it ends

Here is the sharpest reason the insurers are right to wait, and it is the one this week keeps confirming.

A peace is supposed to pay everyone. This one does not. Iran is the single party at the table whose finances get worse when the shooting stops, and seeing why unlocks the whole thing. Recall the hole from part one: an empire lost before the war even started, a third of its oil exports gone, a central bank counting twelve years to recovery. Oil is most of what Iran has left to sell, and the deal just handed back its right to sell it. But the same market that priced the war away priced Iran’s rescue away with it. Tehran’s barrels are worth whatever the market will pay, and a calm market pays $68, not the $112 it paid when the strait was on fire. Cheap, peaceful oil is a tax on Iran. Expensive, frightened oil is a subsidy. And the one instrument that turns the first into the second is the strait, which Iran holds.

Set the motives aside, because you cannot read them and you do not need to. Maybe the July 7 strike was ordered from the top to lift the price. Maybe it was a hardline commander wrecking a deal he hates, or an answer to the American president’s threat that morning to finish the job. It does not matter to the point, because the incentive does not require anyone to be cynical. It is simply true that Iran’s oil is worth more when the strait is frightening, and simply true that a government watching its currency die will feel the pull of anything that lifts the price of the one thing it has left. The economist Hyman Minsky argued that calm itself breeds the risk-taking that ends it, that stability is the cause of the crash. Here the mechanism is unusually direct: the cheaper the peace makes oil, the more the cornered party is paid to break it. A calm this profitable to shatter does not stay calm. Which is why, when the market called the war over and the strait answered inside a week, the people least surprised were the ones writing the insurance.

The call

So the desk makes its bet, and puts it on the board where it can be graded.

The bet is the series in a sentence: this does not end quickly. War-risk insurance for a large tanker through the Strait of Hormuz stays high, at least ten times its pre-crisis level, through August 21, the day the American oil waiver is set to expire. That cover runs fifteen to thirty times normal today. For us to be wrong, the underwriters would have to cut it by more than half in six weeks. We put that at 80 percent, and the base rate is the whole argument: after the last tanker war, rates took twelve to eighteen months to normalize even under United States Navy escort, and nothing about a strait taking fire this week suggests a faster clock. The metric is public, the date is fixed, and the scoreboard will settle it.

We carry one older bet alongside it, from before this series began: that the energy-stock selloff holds, that XLE does not rebound 10 percent by July 31. After this week that is a closer call than it was, because a fresh war scare is exactly the kind of shock that lifts energy stocks, and July 7 delivered one. We are leaving it on the board as written. A call you keep only while it is comfortable is not a call.

Watch the underwriters, not the diplomats. When the people who have to pay for a ship meeting a mine decide the water is safe, they will cut the rate, and not a week before. Until then, the truest number about this war is not the price of a barrel. It is the several million dollars it costs to send one ship through the water everyone insists is at peace.

---

Sources

• Current Hormuz war-risk premium (3-8% of hull, $3-8M per large-tanker transit, ~0.25% pre-crisis base, roughly 15-30x), as of June 28 2026: Insurance Business; Strauss Center. (Wide spread reflects real dispersion by flag and week.)
• 1988 normalization ~12-18 months under US escort; 1980s Tanker War (400+ ships, <0.1% to >5% of hull): House of Saud; Strauss Center. The exact 12-18mo figure is single-sourced but consistent with the 1980s and 2003 analogs; verify against Lloyd’s List before publish.
• July 7 2026 re-escalation (two tankers hit, LNG carrier at risk, Qatar blames Iran; oil +~2.3-2.4%; talks paused; Trump “one way or the other”): Reuters via WHBL; CBS News live. Attribution as-reported; keep attributed.
• Iran macro (economy contracting, oil the main export): IMF WEO Apr 2026 via IranWire. WTI ~$68 vs war peak ~$112: The Understated desk series.
• Base rates / outside view: Kahneman-Tversky; Tetlock. Asymmetric wager: decision theory (bounded vs unbounded loss). Sorites paradox: classical philosophy. Costly signaling callback: Spence 1973 (The Understated, Fed piece). Minsky moment: Hyman Minsky.