For most of 2026, silver has been gold with a shorter fuse. It rose more on the way up. It fell more on the way down. This month it found a third gear. It fell while its own bull case rose. That is the useful fact in a Goldman note that a ZeroHedge column, on the morning of Friday, October 9, turned into a comfort story for the bears and then a dare to buy the discomfort.
Hold the two writers apart. Robert Quinn, on Goldman’s metals desk, titled the note “Silver Futures: Stuck.” He describes flows. He does not, on the column’s own reading, issue a buy. Tyler Durden reads the same four charts and goes further. He says a market this short, this hedged, and this bored has usually been a better buy than a sell. He says silver does not need good news. It needs the bad news to stop. Those are not the same document. One is a positioning map. The other is a bet that the map means a turn.
The idea is one sentence. Silver cannot use its own story while the dollar is still writing the price. A washed-out book is a condition. It is not a signal. This piece is not a recommendation to buy or sell silver, a miner, a fund, or a future.
The morning the stories split
The column is dated Friday, October 9, 2026, and it is written off Thursday’s tape. Silver was down another 1.8 percent, to $59.22. Gold edged up, to $4,122. The two metals did not even share a sign. That is the caffeine problem in reverse. Silver did not just fall harder than gold. It fell while gold rose.
Quinn’s line, as the column quotes it, is plain. Commitment of Traders data through the end of September showed a strong bearish turn in speculative positioning. Through early October, some structural demand barometers improved. Macro headwinds were still formidable. The silver price stayed stuck, with negative flows arguably still in force. Read that as a man refusing a slogan. The industrial case got better. The price did not. He does not then declare the industrial case dead. He says the macro still won the week.
The industrial case, in this note, is not a vague hope about “electrification.” Quinn tracks two Goldman equity baskets as live proxies. One is a data-center basket, a stand-in for the AI buildout. The other is a solar basket, because photovoltaic cells are one of the large industrial uses of silver. Over the prior six months, both baskets had kept a positive link with changes in managed-money gross longs in silver. When the stocks that stand for the use of the metal rose, the futures longs tended to rise with them.
From September 29 to October 7, the link broke. The data-center basket rose 4.2 percent. The solar basket rose 2.4 percent. Silver futures fell 1.4 percent. Over the same stretch, Quinn says, U.S. real rates edged higher and the dollar rose 0.8 percent. He notes the history that fits the break. Stronger dollar and higher real-rate stretches have coincided with managed money cutting gross longs. The baskets did their old job. The dollar did a louder one.
That is the whole week, stripped of personality. The stocks that are supposed to speak for silver’s demand went up. The metal went down. A buyer who says “solar is working, so silver must work” is describing a rule that failed in the only week that matters to a person who is long now. Rules that fail in the current week are not dead forever. They are not a reason to add.
Who sold, and what kind of selling it was
Start with the futures book, because that is the part of the market that can be counted. In the week to September 29, Quinn writes that Managed Money, Other, and Non-Reportable combined to sell a net $1.6 billion of silver futures. The column calls it the largest weekly amount since February. The split matters more than the total. About $800 million was liquidation of old longs. About $800 million was new shorts. This was not only tired bulls leaving. Half of it was fresh bets that the price would fall.
Goldman’s reading of the same CFTC data put managed-money net length 29,600 lots shorter than a year earlier. Speculative net length, in dollar terms, had been cut roughly in half from the January peak. The column puts that peak near $24 billion and the later figure near $12 billion. The price took the same road. A blow-off high above $115 in late January. Under $60 by this week. A holder who owned the metal and the speculative length lost both. That is not diversification. That is one trade wearing two names.
Put the round trip in scale before you admire the short. From above $115 to under $60 is a decline of nearly half. A market that has already done that is not “early” in a bear story. The bears have been paid. The question Quinn is circling is whether they have been paid in full, or whether the next dollar of short is still the right dollar. The column wants the first answer. The note, by the column’s own admission, will not give it. Quinn is careful not to make the call.
There is a leftover from the other side of the year. On August 21, Goldman’s desk noted clients buying three-month silver digitals struck at $80 and $90. The column does the arithmetic from $59.22. Those tickets need a rally of about 52 percent, in roughly six weeks, to pay. A digital is not a coin in a drawer. It expires. A client who bought the dream of $90 in August now needs a move that the September sellers were paid to bet against. Both books can be held by serious people. They cannot both be the base case for the same date.
The last time the book was this empty
The column says the street has seen this film, and recently. On August 6, Quinn noted that managed-money net length sat at a 3 percent two-year rank, and that short-term momentum had just flipped. That flip, in the desk’s framework, forced the start of a maximum unwind by the trend followers, the CTAs. From July 28 to September 8, by Quinn’s tally as the column retells it, the December contract jumped 15.2 percent. Managed money bought $1.8 billion, and most of that was new longs. The washout was the fuel. The rally was the fire. Then the fire went out, and the CTAs went short again.
The new short is large. Trend-follower net length in silver went from about $1.2 billion long at the start of September to about $1.4 billion short. The column calls that the shortest in at least a year. The swing is about $2.6 billion in five weeks. A strategy that buys what is rising and sells what is falling does not have a view on solar panels. It has a view on the slope. The slope has been down. The machines are short because the rule says to be short.
That is why a CTA extreme is both a warning and a trap. The warning is mechanical. If the slope turns, the same rule that built the short will buy it back, and it will not wait for a new essay about the dollar. The trap is the reason people remember July and forget the months that did not rhyme. A short trend book can stay short while the trend stays down. “They are max short” is a description of a position. It is not a date. The July-to-September rip happened. It also ended. Anyone using it as a promise is using one outcome as a law.
Hedged, bored, and paying up to be scared
The options market tells a second story, and it does not match a market that is about to explode in either direction. Quinn says discretionary sentiment soured. Normalized 25-delta put-call skew richened into the top decile of the past two years. In plain words, investors are paying more, relative to calls, for puts than they have at almost any point in two years. They want protection against a fall more than they want a ticket to a rise.
At the same time, three-month implied volatility fell to around 33. At the January peak it was above 100. The column says the new figure is the lowest in about a year. Nobody expects a large move. The people who do expect a move expect it lower. The column’s translation is fair as a translation. It is not fair as a forecast. A market that is bored and skewed for a fall has, at times, been the setup before a rise. It has also been the setup before a market that stayed boring, and before a market that fell and made the puts look wise. The skew says where the fear is. It does not say the fear is wrong.
Hold those two facts together, because the column wants them to point one way and they do not have to. Low volatility means the price of a big move is cheap. Rich put skew means the cheapness is not evenly shared. Downside is the expensive side. A person who buys silver here because “vol is cheap” is buying a quiet market. A person who buys puts because “skew is rich” is paying up for a fear that many people already share. Neither trade is the CTA short. Neither trade is the industrial basket. A book can be crowded in protection, crowded in trend shorts, and still be right about the next month if the dollar keeps rising.
The mirror, and who is actually in charge
Mid-September was the opposite picture, and Quinn had a note for that too. The column cites “Silver Futures: Hawkish Enough?” Silver dropped 4.7 percent while the data-center basket plunged 6.9 percent. Industry leaders were questioning the pace of the AI build. In that week, the demand proxy and the metal agreed. The AI story was the problem, and silver behaved as if it had heard.
This month they do not agree. The AI story, on the basket, is fine. Solar is fine. Silver still cannot rally. The column’s conclusion is blunt. The Fed and the dollar are in charge. That is a stronger sentence than Quinn’s, but the chart he is leaning on supports the direction of it. Managed-money gross longs and an inverted dollar index have moved almost together since August. Gross longs slid from roughly $6.9 billion in early September to about $5 billion as the dollar index pushed to 102.5, the highest level of the past six months. When the dollar rose, the silver longs shrank. Tick for tick is the column’s phrase. Even if you sand that down to “close enough to matter,” the week of October 7 is not a mystery. Real rates up. Dollar up. Silver down. Baskets ignored.
A silver investor who wants a different driver has to explain that chart, not the solar story. The solar story had its chance in the window the note measures. It lost. Losing one window is not a permanent verdict on photovoltaic demand. It is a verdict on which price was in the driver’s seat from September 29 to October 7. The driver’s seat is the thing a position has to survive.
The pillar Quinn will name, and the headline that hit the same morning
Quinn does point at a crack, and he points with a caveat the column prints in his words. He acknowledges uncertainty over the Middle East. Then he notes that Goldman’s foreign-exchange research thinks one pillar of the resistance to silver, the rising dollar, could stall. The FX team is cautious on near-term dollar prospects. The reasons, from a Tuesday note the column attributes to Stuart Jenkins, Michael Cahill, and colleagues, are these. The September dollar rally was tied in large part to U.S. stocks outperforming. The trade-weighted dollar set new highs for the year. After a sharp move, dollar positioning is stretched in several major pairs. Recent Fed talk has stressed patience on further tightening. Stretched plus patient is the case for a pause in the dollar, not the case for a new silver bull market.
The caveat arrived the same morning, and it is not a footnote. Brent jumped more than 5 percent, to above $105, on a report that the White House had asked the Pentagon to draw up strike options against Iran. Ten-year yields moved toward 5.3 percent. The dollar rose again. Wednesday’s FOMC minutes did not help the metal. Most officials, the column says, thought another hike would likely be appropriate by year end. Goldman’s economists still expect that hike in December. The 30-year yield, in the column’s telling, sat at 5.67 percent.
Read the two Goldman desks side by side, because they are not a single voice. The FX desk is cautious on the next leg of dollar strength. The economists still expect a December hike. Quinn says the dollar pillar could stall, and he says the Middle East is uncertain. Thursday’s oil spike is what uncertainty looks like when it is no longer a word. A war premium in Brent, a higher yield, and a firmer dollar are the macro headwind he already named. They do not have to last. They also do not have to vanish before the next COT report. A silver position that needs the dollar to stall is short the oil headline whether or not the holder thinks of it that way.
The column says the bad news for bonds may become good news for anything that cannot be printed, and that the asymmetry is to the upside. That is Durden, not Quinn. It is a possible outcome. It is not the note. A Fed that is still debating hikes, a long bond near 5.67 percent, and a Brent market above $105 are not a backdrop in which “the absence of bad news” can be assumed. The absence is the whole trade. If the next Iran headline is a de-escalation that sticks, the FX desk’s caution gets a cleaner test. If the headline reverses by the Friday close, as the column itself warns these headlines can, the test is noise. Silver does not get to skip the test because the CTA book is short.
Paper is stuck. The bars are not in the same place.
The futures chart is a New York story. The bars are a geography story, and Goldman’s commodity strategists have been telling it for a month. Lina Thomas and Daan Struyven warned that tariff fears pulled so much metal into the United States that they expect much of it to stay trapped there. Available inventories outside the United States are tight. If investor demand returns, they see room for a repeat of the volatility of the second half of 2025 and the first half of 2026. The column adds a market crack you can see without a model. On October 6, a ZeroHedge post flagged a J.P. Morgan comment from Willig. Traders are less willing to trade New York silver against London silver. An arbitrage that people will not run is a market that can show two prices for one metal. That is not a rally. It is a pipe getting narrow.
A narrow pipe cuts both ways. Tight metal outside the United States can make a small buying impulse travel a long way. It can also mean the visible New York stock is a bad guide to the rest of the world. A squeeze needs a buyer who cannot wait, and a holder who will not lend. It does not need a solar basket. It also does not arrive on a schedule because a strategist wrote the word “tight.” Physical tightness is a condition, like the CTA short. It becomes a price only when someone is forced.
The column’s materials-desk color is smaller and more concrete. Overnight, China bought the dip. The desk calls the China physical bid the key. India, it says, is absent. Positioning is put at 2 or 3 out of 10. The question the desk asks, typos and all, is why a person who thinks rates have stopped falling would not buy precious metals. That question assumes the thing Quinn says is still in doubt. If rates have not stopped, the question is the wrong question. China’s bid is a real bid if it continues. One overnight is not a program.
The column also says exchange-traded funds were buying gold and selling silver, citing Bloomberg. If that reading is right, the public vehicle for “precious metals” was splitting the complex in the same direction as Thursday’s tape. Gold in. Silver out. A holder of a mixed metal fund was not making a silver decision. The fund was making it for them. And the column notes, as its own napkin, a gold-silver ratio near 70 with gold at $4,122 and silver at $59, up from about 66 in early September. The cheaper metal, on that ratio, kept getting cheaper relative to gold. A rising ratio is what it looks like when silver is the one that cannot clear the macro.
One claim in the column should be labeled as the column’s, not as a fact this piece will carry. It says the People’s Bank of China added gold again in September and, “as we noted,” buys about twice what it reports. Official reserve data are what officials report. A claim that the true buying is double is an allegation with a history in this market. It is not a line item you can audit from the silver note. China’s central bank, the column itself says, does not buy silver. A gold bid in Beijing is not a silver bid. Do not borrow it.
What “unstuck” would actually require
Quinn’s map has four corners, and the column lists them cleanly. Speculators dumped $1.6 billion in a week, half of it new shorts, and dollar-length is about half of January’s. CTAs are about $1.4 billion net short, the shortest in at least a year. Options skew is in the top decile of two years, and implied volatility is near a one-year low. The demand proxies rose, and inventories outside the United States are tight on Goldman’s earlier work. A market can be all of those things and still fall if the dollar and real yields rise again next week.
So the unblock is not a better solar headline. Solar already rallied. The unblock Quinn is willing to mention is a stall in the dollar. Goldman’s FX group thinks that stall is plausible because the move has been sharp and the positioning is stretched, and because the Fed’s recent language has been patient. Plausible is not scheduled. The same week, oil spiked on a strike-options report, yields pushed higher, and the minutes kept a year-end hike in the conversation. A stall that requires the oil headline to fade, the yields to stop rising, and the Fed to sound less willing is three stalls. Silver needs all three only if you accept that the dollar is the price. If you do not accept that, you are back to the broken rule. The baskets rose. The metal fell. Hoping the rule returns is not a position. It is a wish that last week did not happen.
There is a second unblock, and it is uglier. A squeeze does not need the dollar to roll over if the shorts are forced. A CTA book that is shorter than it has been in a year will buy if the model turns. A put-skew that is already rich does not, by itself, force anyone. The force comes from price. A fast rise makes the trend rule flip and makes the shorts pay. The July setup did that, and then the rally was given back in spirit by the September dump. A forced buy is a trade. It is not a new fair value. Investors who need a fundamental story to hold the position after the squeeze will be looking for one in a market that just showed them the fundamental proxies losing to the dollar.
There is a third path, and the bears are allowed to be right. Brent stays elevated. Yields stay high. The dollar does not stall. The $1.6 billion of selling was the start, not the capitulation. Friday’s COT data, which the column says it will check, can show more shorts rather than a turn. In that path the comfort of the bears is not a problem. It is the correct read. “Bears have never been this comfortable” is a headline. Comfort can be complacency. It can also be the feeling of people who have the trend, the yields, and the oil market on their side. You do not fade comfort just because it is comfort.
What this is not
It is not a signal that the industrial use of silver has failed. Data centers and solar stocks rose in the window. The use case the baskets stand for did not roll over. What failed was the idea that the use case sets the futures price on a week when real rates and the dollar are rising. Industrial demand can be real and still be a small voice in a macro week. A miner who sells silver into that week gets the macro price, not the solar stock price.
It is not proof that the short is a buy. The column says the asymmetry is up. Quinn does not. The difference is the whole risk. A crowded short can be fuel. It can also be early. Half of the September week’s selling was new shorts, not just old longs giving up. New shorts are people who chose the position with the price already far below $115. They are not only trapped bulls. Some of them will be wrong. Calling all of them trapped is how a positioning chart becomes a story.
It is not a gold signal wearing a silver costume. Gold rose on the morning silver fell. Funds, the column says, were buying one and selling the other. The ratio widened. A person who wanted “precious metals” got a split decision. Silver’s extra fall is the industrial metal failing to collect its industrial reward, plus the monetary metal failing to collect gold’s. That is a worse combination than a simple beta. It means the reason you own silver, if the reason is “it wins when the solar story wins,” did not work. And the reason you own it as gold’s cheaper cousin did not work either, because the cousin went up.
It is not a reason to treat a three-month $90 digital as a forecast. Those options were a client expression in August. From $59 they need a move of about half to finish in the money, and they do not have much calendar left. Expired hope is not support. It is paper that dies on a date.
What a mining investor should do with a stuck metal
A silver miner does not sell a data-center basket. It sells ounces, or it hopes to. The ounce on Thursday morning was $59.22 in the column’s print, not $115. A mine plan written near the blow-off is a different business from a mine plan that has to live at half that price. Costs did not halve because the Comex length halved. Diesel, power, labour, and steel do not mark themselves to managed-money longs. A company that looked rich at $100 silver can look ordinary at $60 if the cost guide was set in the party. The futures dump does not show up in the mill. The price does.
The equity will still exaggerate the metal. It always has. A CTA short of $1.4 billion can squeeze and drag the shares with it for a few sessions. A fresh leg down in the dollar, or a fresh spike in yields, can do the opposite. Neither move is a change in the ore. Investors who buy the miner because the futures book looks washed out are buying a derivative of a positioning trade. They should know the exit. The exit is not “when solar stocks confirm.” Solar stocks already confirmed, and the metal fell. The exit is whatever happens to the dollar and to the trend rule that is short.
Developers and explorers are further from the ounce and closer to the financing window. A stuck silver price, with volatility the lowest in a year, is a dull tape for raising money. A violent squeeze is a better tape and a worse reason. Windows open because prices move, and they punish the people who sell shares into the move as if the move were a study. Physical tightness outside the United States, if Thomas and Struyven are right, can make the price jump without making a project bigger. The tonnes in the ground do not grow because London is tight and New York is full. The share price might. Those are different events, and the second one can reverse before a drill starts.
The practical filter is short. Know whether you own the bar, the future, the option, or the share. The bar does not expire and does not get a margin call. The future does. The August digitals do, on a clock of weeks. The miner has costs the bar does not have, and a multiple that swings with the CTA book. A single “silver is stuck, so buy something” is four trades pretending to be one. The column is about futures. Most of the damage and most of the snapback will show up there first. The shares will be louder. Louder is not smarter.
What would prove this reading wrong
The reading is wrong, on the driver, if silver rallies hard while the dollar keeps rising and real yields keep rising. Then the September rule really did break, and the baskets were early rather than ignored. A metal that can rise into a stronger dollar is a metal with a buyer the macro does not explain. Official demand will not be that buyer in silver. The column is clear that China’s central bank is a gold buyer, not a silver buyer. A private physical bid, the China dip-buying the desk noted, could be. One overnight does not prove it. A month of it, with the dollar still firm, would.
The reading is wrong, on the trap, if Friday’s COT and the ones after it show the new shorts adding, and the price falling with them, in an orderly way. Then the $1.6 billion was a step, not a flush. Comfort was not the problem. It was the trend. Low volatility would mean the fall is being priced as a grind, not a crash. Grinds ruin people who bought a squeeze and then waited.
The reading is wrong, on the boredom, if implied volatility spikes because of a real supply break rather than because of a dollar reversal. The Thomas and Struyven warning is exactly that. Metal stuck inside the United States, thin stocks outside it, and an arb that traders will not do. If that pipe clogs, silver can jump for a reason that is not the Fed. The jump would not make Quinn’s dollar chart false. It would add a second driver on top of it. Two drivers are allowed. They are also how a clean story becomes a messy profit or a messy loss. The investor who has room for only one explanation should not be in a market that already has two.
A week is not a regime, and a regime is not a trade
January’s high above $115 and this week’s print under $60 are the regime. The metal gave back a fortune. Speculative length gave back about half its dollar size. That is the backdrop every short and every dip buyer is standing on. Inside it, September’s week of $1.6 billion is one week. Early October’s split, baskets up and silver down, is ten days. A regime can contain a dozen weeks that look like turns and are not. The August 6 washout was a turn for six weeks and 15 percent. Then the length came back out. People who bought the rhyme in early September and held it as a belief are the longs who became the $800 million of liquidation. The rhyme was real. The hold was the error.
The column says it will check Friday’s COT to see if the late-September dump was capitulation or the start. That is the right question, and it cannot be answered on Thursday morning. Capitulation is a word that gets used while the selling is still happening, because it flatters the buyer. It is visible later, if the selling stops and the price stops making new lows. Until then it is a hope with a chart attached. Quinn’s word is better. Stuck. Stuck means the flows are still negative, the macro is still a headwind, and the demand proxies are not enough. Stuck is not “about to rally.” Stuck is “unable to leave.”
Unable to leave is a different claim from “the bears are too comfortable.” Comfort is a feeling about other people. The dollar index at a six-month high is not a feeling. Real rates edging up are not a feeling. Brent above $105 on a strike report is not a feeling. A short CTA book is a feeling only if you treat it as destiny. Treated as a position, it is a risk to the bears and a risk to anyone who stands in front of it while the trend is intact. Both risks are live. A Pulitzer is not required to say so. Arithmetic is.
The idea, once
On Thursday, October 9, 2026, in the account this piece is built from, silver was $59.22, down 1.8 percent, while gold was $4,122 and up. From September 29 to October 7, a Goldman data-center basket rose 4.2 percent and a solar basket rose 2.4 percent. Silver futures fell 1.4 percent. The dollar rose 0.8 percent, and real rates edged up. In the week to September 29, speculative accounts net sold $1.6 billion of silver futures, split evenly between long liquidation and new shorts. The CTA book swung from about $1.2 billion long to about $1.4 billion short. Options skew is in the top decile of two years, and three-month implied volatility is near 33, the lowest in about a year. Quinn’s note is called “Silver Futures: Stuck.” He says a dollar stall is possible, and he says the Middle East is uncertain. The same morning, Brent jumped above $105 on an Iran strike-options report, yields pushed up, and the Fed minutes left a year-end hike on the table.
The industrial bull case showed up. The price still fell. A short, hedged, bored market can snap higher if the dollar stalls, as it did for a while after the August washout. It can also keep falling if the dollar does not. The crowd’s comfort is not the analysis. The driver is. Silver does not get to rally on its own story while the dollar is still writing the price.
A note on sources and limits
The prices, the flows, the basket moves, the volatility and skew comments, the CTA estimates, the August digital-option reference, the July-to-September rally of 15.2 percent, the FX-desk caution, the Brent move, the FOMC characterization, and the physical-market comments are taken from Tyler Durden’s October 9, 2026 column on the Quinn note “Silver Futures: Stuck,” including the passages it attributes to Robert Quinn, to Goldman FX research, to Lina Thomas and Daan Struyven, and to a materials-desk remark. Where the column goes past the note, and says the asymmetry is to the upside, this piece treats that as the column’s judgment, not as Quinn’s. The gold-silver ratio near 70 is the column’s napkin math from $4,122 and $59.22. The claim that official Chinese gold buying is twice the reported figure is the column’s, and it is not adopted here. Interval quotes and basket levels will not match every screen, because they are a morning’s marks, not a fix.
Nothing here is investment advice or a solicitation to buy or sell any metal, future, option, fund, or mining share. Positioning extremes fail in both directions. Futures and options can wipe out a stake that a bar would have survived. Mining shares can fall while a squeeze is being called, and they can rise for a day without a change in the ore. Readers should read the primary note and the primary data, and should speak with a licensed adviser before any decision.