The world's 50 biggest mining stocks were worth $2.26 trillion when September closed. They had just lost $264 billion. That is the second-largest monthly drop in a ranking that MINING.COM has kept since the end of 2019. Only March was worse, and March was the month gold fell away from its record.
September did not invent a new story. It took back about three quarters of August's record $357 billion gain. It did so for the same reason the gain arrived. Gold. New York bullion futures slid from $4,441 an ounce at the end of August to $4,158 at the end of September. That is a 6.4% retreat. It leaves the ounce below where it started the year.
This piece has one idea for anyone who owns these stocks. Do not prepare for the $264 billion. Prepare for the metal that pays your bill, and then for the news that is yours alone. The ranking fell as one number. The stocks did not fall as one trade. A holder who treats them as one trade is sizing a word.
Nothing here is advice to buy or sell any stock, fund, or metal. Miners can go to zero. A wide margin last quarter is not a wide margin next quarter. A red month is not a sale. A green year is not a purchase.
Four tapes, one headline
Frik Els published the tally on October 4, 2026. The figures are market values at the September 30 close, taken from each firm's main listing and turned into dollars. A percent in this ranking is a change in dollar value. It is not the local share-price move. That gap matters if you live in Johannesburg, Sydney, or Shanghai and your screen shows a different red.
Sort the month before you sort your book. Gold miners gave back $79 billion, a 12.7% fall, and not one of the fifteen gold names finished higher. Copper producers lost $44 billion, a 7.6% fall, while the metal itself ended three cents below where it began. Iron ore stayed stuck below $100 a tonne, and Rio Tinto took a hit that gold cannot explain. Lithium lost more than a fifth and, by the end of the month, had only a small gain left for 2026.
Silver fell 9%. Platinum-group metals sank deeper into the red for the year. Those are real tapes. They are not the $264 billion. The bulk of the bulk was gold giving back a blow-off, plus a handful of company wounds that the ounce did not cause.
The macro picture, as the ranking puts it, all worked against a metal that pays you only if the price goes up. On September 16 the Federal Reserve raised rates for the first time since 2023. A global bond selloff pushed yields to their highest since 2008. The dollar firmed. Gold has no dividend to hide in. When the rate story turns, the ounce moves first, and the miners move more.
What a normal gold drop looks like
Learn the gearing before you call a stock broken. The ounce fell 6.4%. The gold group fell 12.7%. That is about two dollars of stock for one dollar of metal. It is the old shape of this business. Costs are mostly fixed. A change in the ounce falls through to the margin. A group move near twice the metal is the machine working. It is not, by itself, a reason to panic or a reason to buy.
The gold miners that supplied $138 billion of August's gain gave up $79 billion of it in September. Fifteen gold companies sit in the fifty now. None rose. Newmont and Agnico Eagle each lost a double-digit billion. That was enough to push Agnico back under the $100 billion mark it had cleared only a month earlier. AngloGold Ashanti, the leader on the way up in August, fell 16.4% on the way down and shed $9.4 billion.
Put those names in the gold bucket, not the scandal bucket. A senior that falls with the group is telling you about the ounce. Your prep is simple. Write down the gold price you need for the cash flow you underwrote. Then ask whether $4,158 still clears it. If it does, September was a mark, not a thesis. If it does not, the thesis was the August price, and August is gone.
The $100 billion club shrank from seven names to six. Freeport-McMoRan fell 7.6% to $100.5 billion and stayed in by a hair. A round number is not a moat. Agnico did not become a different mine because it slipped under the line. Freeport did not become a safe mine because it stayed over it. Do not let a club do your work.
When the drop is no longer the metal
The sharp falls had causes stacked on top of the ounce. Those are the ones that need a file, not a chart of bullion.
Kinross lost 21.3% after it cut 2026 and 2027 output guidance by about 8% from the midpoint. The firm blamed a run of winter storms at La Coipa in Chile and weaker grades at Round Mountain in Nevada. The shares dropped more than a tenth in a single session. An 8% cut in guided ounces is not a 6.4% move in gold. It is fewer ounces at whatever gold you get. Prepare by reading the new guide against the old one. Do not net it into "gold was down."
Shandong Gold fell 27.8%, the worst move in the whole fifty. Its board cut the 2026 mined-gold target to between 1.16 million and 1.22 million ounces, from a plan of at least 1.58 million. Zhongjin Gold, its Beijing peer, lost 17.7%. A target cut of that size is a volume story. Price did not do it. If you own the stock, your question is why the plan broke, and whether the new range is a floor or a first cut.
A useful rule sits in the gap between 12.7% and 28%. Inside the group move, assume the ounce until a filing says otherwise. Outside it, assume the company until the filing says the ounce. Kinross, Shandong, and Gold Fields, below, are outside it.
The bid that taxed the buyer
The month's most expensive decision belonged to Gold Fields. The Johannesburg miner made an unsolicited approach for Northern Star worth A$38.7 billion, or $27 billion. A deal would have made the world's second-biggest gold producer. On September 28 the Australian board rejected it as "highly opportunistic."
Gold Fields' shares fell 12% in Johannesburg that day and 21% over the month. Market value dropped $8.6 billion. The firm is said to be weighing a larger cash piece to tempt the target back. That "said to be" is a report, not a term sheet. Do not trade the rumor of a sweeter bid as if the board had signed it.
Northern Star was the best gold name in the ranking, and it was still down. It fell 3.1%. The bid premium cushioned it against the metal. That split is the lesson. If you own the target, a live approach can be a floor. If you own the bidder, the premium is your loss until cash, stock, and the other board say otherwise. September punished the firm that reached. It spared, relatively, the firm that was reached for.
Prep for the next approach is not "gold M&A is back" or "gold M&A is dead." One bid died in public. The buyer's stock paid for the attempt. Before you underwrite another one, write down who issues the shares, who receives them, and what happens to each price if the answer is no. Gold Fields just ran that test for you.
Copper stood still. The stocks did not.
Copper's producers lost $44 billion between them. A 7.6% fall looks almost orderly next to gold's 12.7%. The metal itself finished at $6.56 a pound, or $14,500 a tonne, three cents below where it began. In between it set a fresh record. A stalled U.S. tariff plan drained London and Shanghai warehouses. Then the White House hesitated, and the gains came back out.
Flat is not calm. A metal that prints a record and ends unchanged has already shown you two regimes in one month. Prepare for the warehouse and the tariff headline, not for the monthly close alone. A book that marks copper only on the last day will miss the move that already happened, and the one that can happen again if the tariff plan revives or dies for good.
Southern Copper fell 3.0% to $171 billion and ended the quarter in second place for the first time. The Mexican-Peruvian producer first pushed Rio Tinto off the number-two spot in January. It lost the spot in the March selloff. It passed Rio briefly in August and closed that month $600 million behind. By the end of September the gap was $9.9 billion the other way. Rio was down 8.9% and $15.8 billion lighter, with iron ore stuck below $100 a tonne.
Rank is not quality. Southern Copper did not become the better business because Rio's iron ore mark slipped. It became the larger stock. If your reason for owning either name was "number two in the world," update the reason. The chair moved. The mines did not swap.
Teck Resources held its ground better than most, down 3.4%, while it waits on final regulatory decisions in China and South Korea. Those decisions stand between Teck and the Anglo Teck merger. A stock that is waiting is a clock, not a metal. Prepare by knowing the next vote, the next filing, and what the price already assumes about a yes.
A death, a strike vote, and the biggest dollar loss
BHP suffered the largest dollar loss in the ranking. It fell $26.4 billion, or 11%. The trigger was not a quiet tape in copper or iron. A worker was killed at Escondida on September 23. The world's largest copper mine suspended operations. In the same week, supervisors voted for strike action.
That is a different prep from a gold chart. Safety stops are not macro. A closed pit does not care that copper ended the month flat. An 11% mark on the biggest diversified miner was large enough to swing a country. Canada moved back ahead of Australia as the most valuable mining address, $488 billion to $484 billion, because BHP's fall outweighed what happened to the Canadian names. The United States was third at $307 billion.
Do not allocate by flag off a one-month photo. August was a photo finish the other way. September reversed it with one company's loss. A portfolio that adds Australia because the pie chart says so is buying last month's accident. Know the mine, the metal, and the event. The pie will follow.
Only one company in the fifty finished higher. Amman Mineral, the Indonesian copper and gold producer, gained 1.3%. It plans to double copper cathode output this year from a rebuilt smelter. A green print in a red month is a fact. It is not a system. And the same ranking says Amman is still more than 40% below its peak. The only winner is not the repaired stock. Do not hunt the lone plus sign and call it a method.
A restart headline is not a restart
First Quantum lost 19.2% in September, almost all of it on the last day. A three-minister commission in Panama recommended that the state negotiate a restart of Cobre Panama. The terms would have to pay for the mine's eventual close without the state footing the bill. That was one of 17 recommendations after an audit by Swiss consultants. Any deal would require an end to $27 billion in pending arbitration claims. The report also bars any extension of the operating period or any expansion of the site. The decision now rests with President José Raúl Mulino.
The Toronto shares fell as much as 36% on September 30 before trading was halted. They closed 15% lower on the day. Scotiabank's Orest Wowkodaw called the selling a buying opportunity. BMO's Matthew Murphy noted that the 2023 court ruling struck down the contract, not mining itself, which leaves room for a negotiated restart that could run for decades. Those are analyst views. They are not a decree.
The stakes are large enough to misread. Before the 2023 shutdown, Cobre Panama supplied about 40% of First Quantum's revenue and about 1.5% of the world's copper. The firm is processing stockpiled ore under a government-approved plan and expects 30,000 to 40,000 tonnes of copper from it this year. An expansion toward 100 million tonnes of ore a year, big enough to rank among the top three pits, was already under way before the shutdown. The panel's bar on expansion sits on top of that history. It does not erase it. It also does not reopen it.
Prepare for this name as a political file, not as a copper beta. A halted tape is not a term sheet. A buy-the-drop note is one desk's opinion about a president's choice. If you cannot say who must sign, what claims must die, and what the mine is forbidden to become, you are not prepared. You are hoping. Hope is not a position size.
Lithium left the list
The battery metal was the month's worst tape by a long way. Carbonate futures in Guangzhou fell 22.5% to 122,800 yuan a tonne, about $18,300. That is the lowest close since the first week of January. It is 39% under the two-year high above 200,000 yuan, about $29,500, set in mid-May. A metal that was up almost 70% for the year at the peak now shows a gain of less than 4%.
Supply news did not save it. China's biggest lithium mine, CATL's Jianxiawo, went back on care and maintenance at the start of the month after regulators revoked its environmental approval. A lost mine should have tightened the market. It did not set the price. What broke the tape was a change in method at a Chinese price agency. Reported stockpiles more than doubled, to 175,000 tonnes. The price fell 14% in three days. The selling did not stop there.
That is the prep lithium holders did not want. A disruption is not a floor if the inventory number jumps for a bookkeeping reason. Know which figure you are underwriting: a closed pit, a stockpile, or a price series that can be rebuilt. September showed the stockpile, and the way it is counted, beating the shutdown.
The equities fared worse than the metal. Albemarle and Ganfeng Lithium each lost more than a fifth of their value in September. Both now trade at less than half their 52-week highs. Albemarle is down 25% in 2026. Ganfeng is down 32%. Three lithium producers sat in the ranking at the end of August. SQM, itself 17.9% lower and 36% off its high, is the only one left.
That is a long fall from the craze. Six lithium names sat in the fifty at the late-2022 peak, when carbonate in China neared 600,000 yuan a tonne, about $84,000. Their chairs were taken by two gold miners. Endeavour Mining entered at 49th. Alamos Gold returned at 50th, a month after it had slipped below the line. The price of admission fell to $13.6 billion, from $15.6 billion in August. That is the lowest cutoff since June, and far under the $19.5 billion it took to make the fifty at February's peak.
Here is the quiet change. If you use the top 50 as a map of "mining," the map rewrote itself. Lithium weight left. Gold weight at the bottom came in. You did not place the trade. The cutoff did. For a sense of the distance, the fiftieth company was worth $3.4 billion at the Covid low in March 2020, and $6.3 billion as recently as the end of 2024. The club is still rich. It is not a fixed club. Prep includes knowing that your benchmark can change metals while you are reading the headline.
China's loss, and a gold-copper blend
Every mining address lost ground in September. China lost the most. The six Chinese companies in the ranking dropped $43 billion between them, 15% of their value. Zijin Mining alone accounted for $19 billion of that fall.
Zijin fell 14.2% to $115 billion. It fell in step with Chinese peers as gold slid. Shandong was down 27.8%. Zhongjin was down 17.7%. CMOC was down 13.4%. Zijin is now as much a gold stock as a copper stock. First-half gold output rose 15%, to just over 1.5 million ounces, while copper production fell 6%. Its listed gold unit added to that tilt by taking control of rival Chifeng Gold in a $2.6 billion deal. The stock is down 7.5% for the year and about a third under its 52-week high.
Do not file Zijin under "copper" because that is what it was in an old note. The ounces moved. The deal moved. A blend changes the gearing. More gold means more of September's rate story, and less of copper's flat close. Before you add to a blended miner, split last year's metal mix from this year's. The label on the sector pie is not the mix in the mill.
The Chinese names also have the longest road back to their highs. Shandong Gold is 58% below its peak. Zhongjin Gold, Amman Mineral, China Northern Rare Earth, and Jiangxi Copper are all more than 40% adrift. South32, Anglo American, and Teck are within 8% of theirs. "Cheap versus the high" is not one idea across the fifty. For some names the high was a spike you should not anchor to. For others you are a few percent from a price the market just paid. Know which.
A bad month inside a year that is still up
September's $264 billion is the second-worst month in the series, behind only the $434 billion that vanished in March as gold left its record. Even so, the third quarter closed $107 billion higher than it opened. The ranking is still worth $118 billion more than at the end of last year. A holder who prepares only for September will sell a year that has not broken. A holder who prepares only for the year will ignore a month that just showed the gearing.
The long view is flatter still, and it cuts both ways. The fifty were worth $698 billion at the Covid low in March 2020. They took 42 months to regain the $1.68 trillion peak of the 2022 spike. They then doubled in the thirteen months to January 2026. At $2.26 trillion the list sits $486 billion, or 18%, under the February record.
The gap to the highs is wider at the company level. Had every name closed September at its own 52-week high, the ranking would be worth $2.97 trillion. That is $709 billion more, or 31%. Most of those peaks are not old history. Of the 47 companies with a full year of trading, 33 set their 52-week high in the first quarter, when gold ran to $5,420 an ounce. Ten more set theirs in late August or the first week of September. Among them were BHP, Southern Copper, Newmont, Freeport-McMoRan, Glencore, and Anglo American.
More than thirty of the fifty have set all-time highs at some point in 2026. You are not looking at a forgotten sector. You are looking at a sector that was at records this year, gave a large piece back in one month, and is still well above the crash and the last cycle's peak. Preparation is the refusal to use only one of those facts.
Gold at $5,420 and gold at $4,158 are both 2026 prices. A model built on the first number is a model of a peak, not of a base. A model that treats $4,158 as a disaster ignores that the metal is still the reason the list is above $2 trillion. Write the price you used. If it was the first-quarter high, September did you a favor by marking it down. If it was $4,158 or lower, the month did not break the case. It tested the gearing.
How to prepare, if you refuse the single number
Write the metal at the top of the page. Gold, copper, iron ore, lithium, or a blend with weights. If you cannot name it, you do not have a mining position. You have a ticker that used to be in a list.
For gold, compare the stock's month to the group's 12.7%, not to zero. Near that line, you are holding the ounce, and the ounce was down 6.4% for reasons you can name: a rate hike, yields at a 2008 high, a firmer dollar. Far below that line, open the filing. Kinross cut guidance. Shandong cut a target. Gold Fields paid for a bid that was refused. Those are three different files. They do not share a fix.
For copper, do not take comfort from a flat month. The close was $6.56 and a record happened in between, on a tariff story that then stalled. Southern Copper's rise to second place was Rio's iron ore mark as much as it was copper. Teck is a merger clock. BHP is a death at Escondida and a strike vote, and that single loss moved Australia behind Canada. Flat metal did not stop any of that.
For First Quantum, separate the analyst notes from the signature. Wowkodaw and Murphy think the drop overshot a legal path that still exists. Mulino has not signed a restart. The panel tied any deal to the end of $27 billion in claims and to a ban on expansion. Until those points are decided, size it as politics. Copper exposure is available in names that do not need a president.
For lithium, assume the inventory number can matter more than the mine. Jianxiawo's shutdown was the kind of headline that used to mean a shortage. A methodology change that doubled reported stocks to 175,000 tonnes did the damage. Albemarle, Ganfeng, and SQM have already shown that the equities can fall harder than the salt. Two of the three August lithium names are no longer even in the fifty.
For the list itself, check the members, not just the total. Endeavour and Alamos came in as lithium went out. The door price fell from $15.6 billion to $13.6 billion. A product or a mental index called "top miners" got more gold and less lithium without your order. If you are matching that list, you changed metals. If you are not, stop using the list's red month as your red month.
For deals, mark both sides. Northern Star fell 3.1% with a bid under it. Gold Fields fell 21% for having made the bid. The next approach will look like strategy in the press release. September's prices already told you the first-day split between hunter and hunted.
For the year, hold two numbers at once. The quarter is still up $107 billion. The ranking is still up $118 billion from last year. It is also 18% under the February record, and the sum of each stock's own high is 31% above the close. A prep that uses only the comfort will be late. A prep that uses only the scar will sell the wrong thing.
Do not prepare by buying the one stock that rose. Amman Mineral's 1.3% gain is real, and so is the gap of more than 40% to its peak. One green cell is not a screen.
The close
September took $264 billion off the fifty largest mining stocks and left them at $2.26 trillion. Three quarters of August's record gain went back out, because gold went from $4,441 to $4,158 and the miners, as usual, fell about twice as hard. Copper and iron ore ended the month about where they began, and still produced billion-dollar stock losses where a mine, a merger, or a death intervened. Lithium fell more than a fifth, left two of three names outside the club, and reminded holders that a stockpile count can beat a shutdown.
The idea for an investor is the sort the headline skips. Price the metal move in one column. Price the company file in the next. Gold's 12.7% is the first column. Kinross, Shandong, Gold Fields, BHP, First Quantum, and the lithium trio are the second. The $264 billion is only the sum. It is not a signal, not a bargain, and not a bill you have to pay in every name you hold.
If the document you need is a new guide, a restarted mine, a closed tariff, or a stockpile series you trust, wait for that document. Until it arrives, the month is a map of four tapes. Trade the tape you own. Do not trade the total.
A note on sources and limits
This account follows Frik Els, "Top 50 mining companies take $264 billion hit as gold trade unwinds, lithium stocks exit," MINING.COM, October 4, 2026. Values are dollar market caps at the September 30, 2026 close on each firm's main listing. Percent changes are changes in those dollar values, not local share prices. Metal marks in the piece include New York gold futures, copper at $6.56 a pound, Guangzhou lithium carbonate futures, and iron ore described only as stuck below $100 a tonne. Company events, including the Northern Star rejection, the Kinross and Shandong guidance cuts, the Escondida death, the Panama commission, and the CATL licence loss, are reported as in that article and the MINING.COM reports it cites. Analyst comments from Orest Wowkodaw and Matthew Murphy are their views, not facts about what Panama will do. The line that Gold Fields may add cash to its bid is attributed as a report, not as a company announcement in this text. This is not investment advice. It is not a forecast of any metal, merger, or election. It is not a statement that any name is cheap or dear.