Crude from the Middle East is almost back. The fuel a gold mine burns is not. On September 29, 2026, JPMorgan said crude shipments from the region had recovered to 17.5 million barrels a day. That is 98 percent of the pre-war pace. Flows of refined products, the gasoline and the diesel, were 3 million barrels a day. That is 58 percent. Put together, crude plus products were 89 percent of 2025 levels, on a 10-day average. The bank’s Natasha Kaneva called it a remarkable recovery for a region still at war, and an uneven one.
The Market Ear, writing the next morning, asked who that unevenness hits. Not the bar of gold. The bar does not buy diesel. A miner does. Crude can reprice on a headline. A miner only feels the headline when the diesel it actually burns gets cheaper, or when a contract it already signed rolls off. Some contracts will not roll off soon. Agnico Eagle has already bought most of the diesel its Nunavut mines will use through the 2027 sealift, at a price about 30 percent above its 2026 budget after hedges.
This piece has one theme for investors. The ounce has no fuel bill. The mine does. If gold stops rising while that bill stays high, the metal is the cleaner way to hold the gold story, and the leverage that lifted miners on the way up works against them on the way down. If gold keeps rising fast, the fuel bill can still be a footnote. The whole argument is that condition. It is not a signal to sell every miner.
Nothing here is advice to buy or sell gold, a bullion fund, or a mining share. A fuel footnote can be overwhelmed by the gold price. It can also be the thing that eats a margin when the gold price stalls.
Two recoveries, not one
The war between the United States and Iran was in its eighth month when JPMorgan published the note. Shipping risk had not ended. Threats remained in the Strait of Hormuz and at Bab el-Mandeb. What changed was volume. Saudi Arabia led crude back through Hormuz. Those flows were almost back to the late-June highs of nearly 13 million barrels a day. The kingdom’s east-west pipeline, damaged earlier in September, had about half its throughput back. That line carries crude to Red Sea ports. The chart The Market Ear ran, sourced to JPMorgan, draws a normal line at 18 million barrels a day of Middle East crude flows. The stack of routes fell off a cliff early in the year and climbed back toward that line by late September. The eye sees a recovery. The product number says the recovery is crude, not fuel.
A refinery is not a valve on a tanker. Crude can move, and still not become diesel in the place a mine needs it. Product exports at 58 percent of the pre-war pace mean the tight market is the one miners actually pay. A lower crude quote helps a miner only after it passes through a refinery, a ship, a port, a truck, and a contract. Any one of those steps can keep the old price. JPMorgan also said the rise in Hormuz crossings was not a sign that danger had faded. It was a sign the trade had learned to move oil while the danger stayed. A learned route can break again. Investors who treat 98 percent as the end of the fuel story have read the wrong barrel.
There is a second split inside the 98. Other watchers were less sure the water was that full. TotalEnergies’ chief executive had put crude and products getting out closer to 10 million barrels a day. U.S. officials had talked about higher numbers. The honest use of JPMorgan’s figure is as one serious count, not as a law. Even on that count, products lag. The lag is the part that matters for a pit.
Why a mine cannot shop today’s oil price
A city truck can buy diesel this week. A remote gold mine often cannot. Nunavut is the clearest case because the sea ice sets the calendar. Supplies go north on a sealift in a short open-water season. Miss the ship and you do not reorder in November. Agnico Eagle said it had purchased about 70 percent of the diesel its Nunavut operations need for the rest of 2026 and through the 2027 sealift. That is about 130 million litres. The price, after hedges, was about 30 percent above the 2026 budget. The company’s own cost guide for the year had assumed a diesel benchmark of $0.78 a litre, before transport and taxes.
Read that slowly. The company bought security of supply. It also bought a higher bill. Oil can fall this afternoon. Those litres will not be repriced by the fall. The Market Ear’s line is the right one. A miner that already bought expensive diesel does not get the benefit today. Agnico said the purchase cuts its exposure to further diesel swings, and that it would try to finish the remaining Nunavut buys in the third quarter and add hedges if the market allowed. A hedge against a further rise is not a refund on a price already paid. The stronger U.S. dollar and higher by-product prices, especially silver and copper, had helped offset some of the fuel pressure versus the 2026 cost guide. Offset is not erasure.
Diesel is not a rounding error at Agnico, and it is not the whole cost either. The company said diesel is about 10 percent of operating costs. About 7 points of that is fuel burned in trucks and in on-site power. About 3 points is freight. A 10 percent move in the diesel price, on the company’s estimate, would change total cash costs by about $4 an ounce in the second half of 2026 from the direct burn, and about $2 more from the freight. Six dollars an ounce is not a thesis by itself at a gold price above $4,000. It becomes a thesis when gold is no longer adding dollars faster than costs are, and when the expensive litres are already in the tank through 2027. The sealift is a clock. The gold quote is a different clock. Investors who mix them will think the miner is a leveraged bar. It is a bar plus a fuel contract.
Agnico is one company with a northern problem. It is also a clean example of a general rule. Fuel is bought on a schedule. Gold is priced every second. The schedules do not match. A miner in Nevada can turn a tanker faster than a miner in the Arctic. It still cannot turn a hedge, a yearly supply deal, or a local shortage of refined diesel as fast as a futures screen. The 58 percent product figure is the world version of the same delay.
Barrick’s ten-dollar rule
Barrick put the oil math in its own 2026 outlook, not in a metaphor. The guide assumes $70 West Texas Intermediate and $75 Brent. A move of $10 a barrel, up or down, changes attributable all-in sustaining costs and total cash costs by about $12 an ounce. The same $10 swing changes attributable EBITDA by about $56 million, and consolidated EBITDA by about $61 million, on the table in the second-quarter release. The Market Ear reported that Barrick had also tied part of the rise in second-quarter production costs to higher fuel, along with lower grades and higher royalties. Grades and royalties do not care what Hormuz does tomorrow. They sit beside the fuel line and remind you that diesel is one bite, not the whole meal.
The $12 is the number to keep. It is a direct diesel sensitivity, not a promise that every $10 in the headline oil price arrives at the mine on Friday. Barrick’s own table is a planning tool. It assumes the oil move passes into the fuel the company buys. If crude falls and refined diesel does not, the $12 does not arrive as savings. If crude falls and the company has already filled storage at the old price, the $12 arrives late. Cheaper crude helps only when cheaper fuel reaches the mine. That sentence is the whole crude story as it applies to gold shares.
Hold the scale next to the gold price so it does not get inflated. Gold in late September was trading a little above $4,000, after a hard month, not at the $1,800 of an old model. Twelve dollars an ounce is 0.3 percent of a $4,000 price. It will not, alone, decide a miner’s year. What decides the year is whether gold is still rising enough to cover every sticky cost at once. Fuel. Grades. Royalties. Labor. Power. The Market Ear’s point is not that diesel bankrupts the sector. It is that diesel is a cost the metal does not have, and that the cost is sticky precisely when the metal’s momentum is in question. A small sticky cost matters more when the big price stops helping.
Leverage was the gift. It is also the risk.
Miners did not imagine their lead. The Market Ear says they massively outperformed gold earlier in 2026. That is what operating leverage does in a rising gold price. Costs are partly fixed. Each extra dollar of gold, above those costs, falls harder to profit than it does to the owner of a bar. The shares are a geared claim. Gears work in both directions. When the gold price stalls or falls, the same fixed costs take a larger bite, and the share gives back more than the metal.
The note says that giveback has started, and it ties the turn to a surge in U.S. real yields. Goldman, in The Market Ear’s telling, has flagged gold miners among the sectors hit hardest by the latest rates move. The chart to watch, on that argument, is not gold alone. It is the ratio of the big-miner fund, GDX, to the bullion fund, GLD. The LSEG chart printed with the piece shows the 10-year yield leaping in the latest window while that miner-to-bullion line turns down. I will not pretend to read a level off a small axis. The shape is the fact the chart was published to show. Yields up. Miners losing their lead over the metal.
A yield does not buy diesel either. It changes the discount. Gold pays no interest. When real yields jump, holding the bar costs more in foregone income. Miners are equities, so they get that hit and a second one. A higher discount rate cuts the present value of a mine’s future ounces. A stall in the gold price cuts the ounces’ value itself. A sticky fuel bill cuts the margin between the two. Three blades. The bar only feels the first. That is why a rates shock can hurt miners more than gold even before anyone mentions Hormuz. The fuel story does not replace the rates story. It arrives on top of it, which is why the timing is awkward. The relative trade is rolling over for a monetary reason at the same moment the cost reason is failing to ease.
There is a fair reply, and it belongs in the piece. If gold starts ripping again, operating leverage can pay the fuel bill and have cash left over. The Market Ear says this plainly. A rising gold price is still the strongest friend a miner has. The cleaner-trade claim is conditional. It is the case where bullion goes sideways, or loses momentum, and diesel stays expensive. In that case the miner keeps a cost the bar does not have, and loses the rising price that used to hide the cost. Investors who skip the if will turn a conditional into a religion. Religions are how people sell the low in miners and then watch leverage work upward without them.
What cleaner does and does not mean
Cleaner is not safer, and it is not a ticker. A bullion fund such as GLD is a claim on metal, with fees, a sponsor, and a structure. It does not have a fuel bill. It also does not have a mine’s option on a new deposit, a cost cut, or a gold price that runs. A miner fund such as GDX is a basket of businesses. Some hedge fuel. Some buy it once a year. Some burn little diesel because they are on a grid. Some have copper or silver credits that rise when gold does not. Treating GDX as one diesel bill is as crude as treating GLD as a costless promise. The useful distinction is the one in the theme. One asset’s main risk is the gold price and the real yield. The other’s main risks include those, plus the fuel, the grade, the royalty, the country, and the balance sheet.
A single share can be cleaner than the basket or dirtier. Agnico’s Nunavut litres are a known, dated cost. That can be easier to underwrite than a vague fear of diesel. You can put 130 million litres and a 30 percent overrun into a model. You cannot do that with a slogan. Barrick’s $12 an ounce per $10 of oil is also a model input, not a scare. The work is to ask, for the company you actually hold, three questions. What share of costs is diesel. What share of that diesel is already priced. What gold price does the plan assume, and what happens if gold is flat for a year while that diesel price stays. If you cannot answer, you do not yet know whether you own a bar with leverage or a fuel contract with a logo.
Remote Canadian mines make the question sharper, not because Canada is the only place diesel matters, but because distance and ice make the contract visible. A grid-connected mill in a city does not sealift a year of fuel. A Nunavut pit does. Investors in Canadian gold stocks who assume every ounce has the same cost curve are mixing Agnico’s tank with a heap leach on a power line. Read the fuel note in the quarterly report. It is often a paragraph. This month that paragraph is the paragraph.
What would change the theme
The theme breaks in plain ways. Refined-product exports climb from 58 percent toward the crude recovery, and diesel prices at the mine gate fall. Companies that have not yet bought their year of fuel buy it cheaper. Companies that have already bought it work through the expensive litres and do not replace them at the old price. Or gold rises enough that $12 an ounce, or a 30 percent fuel overrun on 10 percent of costs, is noise next to the margin. Any of those can happen. None of them is what the September 29 flow data showed. Crude was back. Products were not. The Nunavut purchase was already done.
The theme also breaks if the yield move reverses and the miner-to-bullion ratio turns up again. Then the monetary lead can return even with a sticky fuel bill, because leverage on a rising gold price is large and the fuel bill is not the whole cost. Watch the ratio, and watch the 10-year, and do not let a fuel essay talk you out of a tape that has changed. The Market Ear’s instruction was to watch GDX against GLD. That is a relative chart. It does not tell you whether both will rise or both will fall. A miner can beat gold and still lose money. Gold can beat the miners and still be a bad price. Relative and absolute are different bets. Know which one you meant.
A third break is operational. Grades improve. Royalties do not rise. A mine connects to cheaper power. Agnico has backed northern projects that would cut diesel use, including a small wind-and-battery plan at Hope Bay and a longer hydro idea for the Kivalliq. Those are years, not this sealift. They matter to a five-year holder. They do not reprice the litres already bought through 2027. An investor with a five-year clock and an investor with a winter clock are not having the same argument. The fuel bill is a winter fact. The power line is a decade hope. Do not use the hope to deny the fact, and do not use the fact to deny the hope.
The mistake this chart invites
The chart of flows invites a simple trade. Oil is fixed, so sell the fear, so miners’ costs will fall, so buy the miners. The product number blocks that trade. The sealift blocks it again. The yield chart blocks it a third time. Costs fall when fuel at the mine falls, not when a crude tanker count heals. Shares beat gold when the gold price is doing the work leverage needs. Right now the crude count has healed more than the fuel market, and the yield chart is the one The Market Ear says is taking the miners’ lead away.
The other simple trade is just as bad. Diesel is sticky, so sell every miner and hold only metal. That trade ignores how small diesel is next to a gold price that is still historically high, ignores hedges, ignores by-product credits, and ignores the chance that gold rises again. It also ignores that miners already had their run. Giving back a lead is not the same as giving back the whole bull market. The Market Ear says the relative trade has started rolling over. Started is not finished. A rollover can be a pause. It can also be the start of leverage working downward. You do not know which from one chart. You know the condition that would make the downward path likely. Gold without momentum. Diesel without relief.
Sit with that condition instead of a ticker. If you wanted gold and bought miners because they were moving faster, you bought leverage and a fuel bill you may not have priced. If you wanted a business and bought miners because the ore is real, the fuel bill is a line in the model, not a reason to pretend you hold a bar. If you wanted neither and only wanted a story, the 98 and the 58 are not a story. They are a warning that the barrel in the headline is not the litre in the tank.
The close
Middle East crude exports are near their pre-war pace. Refined fuel exports are not. JPMorgan’s September 29 count was 98 percent for crude and 58 percent for products, in a war that has not ended. Gold miners do not burn the crude export. They burn the product, on a delay, and sometimes on a contract that is already signed. Agnico’s Nunavut purchase covers most of the diesel through the 2027 sealift at about 30 percent over the 2026 budget. Barrick’s own table says $10 on the oil price is about $12 an ounce of cost when it reaches the mine. It does not reach the mine just because a tanker count recovered.
The miners had the leverage on the way up. Real yields have been taking that lead back, which is why the miner-to-bullion ratio is the chart to watch. Fuel is the extra fact. The bar does not pay it. The mine does. If gold keeps rising, the mine can still win. If gold goes sideways and diesel stays sticky, the metal is the cleaner hold, because it has no fuel bill. That is the theme. It is a condition, not a command. The investor’s job is to know whether the gold they meant to own is a bar or a tank.
Important information
This article is for information and education only. It is not investment advice and not a recommendation to buy, sell, or hold gold, oil, GLD, GDX, Barrick, Agnico Eagle, or any other security. Miners and metal can both fall. Past outperformance does not predict the next move. Fund shares have fees and risks that are not the same as holding a bar or a mine.
Flow figures are from JPMorgan’s September 29, 2026 note, as reported by Bloomberg and others, including comments by Natasha Kaneva. Other volume estimates have been lower. Barrick’s oil sensitivity is from its second-quarter 2026 outlook table: $70 WTI, and about $12 an ounce on cash costs per $10 a barrel. Agnico’s diesel figures are from the company’s own disclosure on the 2026 Nunavut sealift: about 70 percent and 130 million litres, about 30 percent above the 2026 budget after hedges, diesel about 10 percent of operating costs. The Market Ear’s September 30 note supplied the frame and the GDX/GLD observation. Chart levels that are not printed in the text are not invented here. Company plans can change. This article does not consider any reader’s finances.