Here is the one idea. J.P. Morgan can no longer model the Iran war. Gold does not have to.
Natasha Kaneva, the bank’s head of commodities, told clients in the latest Oil Markets Weekly that this is the first time since the conflict started that her team has no baseline view. “We simply don’t know how to model the endgame.” Tyler Durden ran the note Thursday and called it the most honest report since the war began. The honesty is the market event. A desk that exists to forecast barrels has put down the pencil.
Gold is what people buy when the pencil goes down. Not because Kaneva issued a gold price forecast. She did not. Because an unpriceable tail is exactly the job bullion claims. Gold mining stocks inherit that bid only after diesel, yields, and dilution take their cut. That is the whole chain.
What the Bank Thought It Knew
At the start of the war, Kaneva’s team assumed Washington had economic red lines it would not cross. Oil at $100. Gasoline near $5. Headline inflation at 4%. A 5-handle on the 10-year Treasury. Those caps were supposed to force a deal to reopen the Strait by June.
Six months later the lines are behind us. Oil is above $100. The 10-year has the 5-handle. Gasoline at $4.37 is a record for the season. Diesel is at an all-time high of $6.31 a gallon heading into winter, the peak demand window. Inventories sit at cycle lows. The exit strategy is less clear, not more.
J.P. Morgan’s own rule of thumb is blunt. Every 1 million barrels a day of disruption adds about $4 to the oil price. Brent near $106 against a $90 “fair” value implies the market is already pricing another 4 million barrels a day of risk on top of the 10 million already disrupted. That $16 premium is not a forecast of peace. It is a surcharge for not knowing.
New pressure points keep arriving. Houthis toward Bab al-Mandeb. A hit on Saudi Arabia’s East-West line. Ukrainian drones on Russia’s Slaviansk, Taneco, and Syzran refineries — Taneco more than 1,200 kilometres from the border. Russia answering on Ukrainian cities. Trump is due to see Xi in Washington on Sept. 24. Absent a breakthrough there, the idea that this disruption is “temporary” gets harder to keep.
How the Market Actually Cleared
Commodities have to balance. Surplus becomes inventory. Shortage comes out of inventory. Kaneva’s April “Illusion of Plenty” note expected a fast draw — about 1.6 billion barrels — toward operational floor levels by September. Prices were supposed to live near $100 even after the Strait reopened.
The world did something else. Governments and consumers hoarded. They kept barrels for a longer war. Global draws of crude and products came to 555 million barrels. That is about one-third of what the bank expected. The United States accounted for 219 million. China 147. Europe 77. Japan 69. South Korea added 12 million. The rainy-day tank was not emptied. Demand was.
Since March, demand has run about 4.4 million barrels a day below last year. That is one and a half times the relief that came from stocks — a 2.9 million barrel-a-day demand hit. Brent has averaged only $94 since the war began. Falling stocks lift price. Falling demand sits on price. The war was absorbed less by the tank farm than by the closed factory and the cancelled trip.
Kaneva says fear that inventories are about to hit zero is premature. For now. The next dry powder is uneven. U.S. commercial stocks are already near the bottom of their recent range. The Strategic Petroleum Reserve can add only about 30 million more. Japan, South Korea, and Europe might free about 85 million without much strain. With the United States, call it 120 million from the OECD. China could add about 120 million by year-end at a 1 million barrel-a-day pace, plus a sliver from other non-OECD holders. That pile can pull the system toward J.P. Morgan’s 7.65 billion barrel “stress” line. Beyond that, demand has to fall harder, or governments have to step in again.
If Middle East flows stay where they are, the bank would lift its own fourth-quarter and December forecasts by $7 to $8. Contained. For now. Those two words are doing a lot of work.
Why Gold Cares That the Model Died
Gold does not need Kaneva to pick an endgame. That is the point.
A baseline view is a story about how the war ends and how oil mean-reverts. No baseline is a story about duration risk that cannot be hedged with a strip of futures. Duration risk of that kind leaks into inflation, into the 10-year, and into the dollar. This week the Federal Reserve hiked anyway. Diesel printed a record into winter. Those two facts can live together because the shock is supply, not a wage spiral the Fed can fine-tune. Kaneva said the quiet part. The red lines were crossed. Policy did not produce a map.
That is bullish for the gold price in the slow way, not the day-trader way. Official buyers already treat gold as the reserve that cannot be frozen. Households treat it as the asset that does not require a Strait forecast. When a house like J.P. Morgan says the endgame is unmodelable, the case for a non-model asset gets simpler. You do not have to believe in $10,000 gold. You have to believe that uncertainty with a $6.31 diesel tag and a 5% 10-year is not a week of noise.
The opposite gold case is also in her numbers. Demand destruction of 4.4 million barrels a day is another name for a slower world. A slower world can knock industrial metals and risk assets together. Gold can fall in that wash for a month, as it did after Wednesday’s hike, and still be the thing that holds when the model stays broken into winter. A bounce on Thursday after the Bank of England held does not settle the argument. The argument is the missing baseline.
What It Does to Gold Stocks
Gold mining stocks are not gold. They are gold minus diesel, minus yields, minus the next raise.
A record diesel price is a cost line. Open pits and remote Canadian camps burn the refined barrel, not the Brent speech. If Kaneva’s “contained for now” fails and product stays tight into winter, margins leak even if bullion holds. If product eases because demand keeps dying, the metal can sag with growth and the miner still does not get a clean tape.
A 5-handle 10-year is a discount rate. Junior gold stocks live on that rate. Producers with cash and no debt can wait. Explorers cannot. Rick Rule’s point this month still applies. The stocks can be cheap to the metal and still be a bad purchase if the company cannot survive the fuel and the finance.
Higher gold from unmodelable war risk helps the names that already pour ounces. It does not automatically rescue a PEA that assumed cheap energy and a friendly window. Screen the cost curve and the cash before you treat Kaneva’s confession as a sector-wide green light. This publication does not pick the tickers. The one idea is enough. When the oil desk drops its baseline, the hedge that never needed one gets a bid. The equity that still needs diesel has to earn it.
Disclaimer
Based on J.P. Morgan’s Oil Markets Weekly as excerpted by ZeroHedge on 17 September 2026 and attributed to Natasha Kaneva. Oil, diesel, Treasury, gold, and mining-share prices move. J.P. Morgan’s inventory and demand figures are the bank’s. This is not investment advice and not a recommendation to buy or sell gold or gold mining stocks.