Milan Adams says the bull market ended in the spring. The quote agreed. The ounces did not. Gold near $4,112 and silver near $56 are a crash in the price, not proof that the vault got fuller, the lease got easy, or the mine got built.
This is not a recommendation to buy or sell gold, silver, a miner, a royalty, or a fund that holds either metal. It is a reading of a piece published on ZeroHedge the evening of October 16, 2026, under the Tyler Durden byline and authored by Milan Adams via Preppgroup. The piece stacks a price collapse on a physical deficit and asks which one you believe. A quote can fall while a vault stays thin. A vault can be thin and the quote can still fall further. You can lose money on either reading. Central-bank figures are estimates. One of them was revised, in 2026, by nearly four-fifths. Treat the direction as sturdier than any single tonne count.
Here is the only idea. A lower price is not a larger pile. Gold ran to a record in the last days of January 2026. One spot reading printed $5,608. The LBMA benchmark was about $5,405. COMEX futures settled at $5,318.40. Three figures for one metal, nearly $300 apart, which is the piece’s reminder that the price depends on which price you mean. By the middle of October the metal was near $4,112, about 27 percent under that January spot peak. Silver touched $121.62 on January 29 and was back around $56, down 53 percent, more than half. The people who had waited three years for the complex to deflate could tell themselves the trade had unwound. The piece says the ounces did not get the memo. In the second quarter, as the price fell, central banks bought 288.9 tonnes, the highest second-quarter total on record and 62 percent above the year before. They bought through a quarter in which gold lost 14 percent. Silver’s projected deficit for 2026, in the Silver Institute’s World Silver Survey, is 46.3 million ounces, against 40.3 million the year before. A sixth straight annual deficit. Mine production flat. The shortfall widened while the price was cut in half. The price rationed the enthusiasm. It did not refill the pile.
What the screen did
A screen is a fast witness and a bad historian. From a January spot print of $5,608 to an October quote near $4,112 is a bear market in the only unit most people check. Twenty-seven percent is not a dip you explain with a bad week. Silver’s 53 percent is a different animal. The gold-to-silver ratio fell below 45 in the squeeze, a level the piece says is rarely visited outside outright stress, and by autumn it had drifted back toward 70, close to the long-run average. Silver is money on Fridays and an input on Mondays. When both moods break, the input falls harder, because factories can wait or thrift, and the tourists leave faster than the central banks. If the ounces got more plentiful, the quote did its job. If they did not, the quote did a different job. It shook out the people who needed a higher number to stay.
The winter in London is the piece’s better evidence. Pressure built in the last weeks of 2025 and broke in January. Silver rose 25 percent in a fortnight at one point. The high of $121.62 arrived as a squeeze. Lease rates for physical metal spiked. Forward curves inverted. The exchange-for-physical spread between London and New York widened until the two venues had stopped pricing the same commodity. A squeeze clears because a high price does three things. It destroys marginal demand. It pulls hoarded metal out of private hands. It attracts new shorts. In 2026, the piece says, the first two worked weakly. Factories kept buying. Private holders in Asia largely held. What unwound the squeeze was the third function. Futures and options pressed the paper price down until the physical premium vanished. Speculative positioning is not a vault. When the speculator is gone, the chart looks healed. The bar may still be spoken for.
The bid that does not have to trade
Between April and June of 2026, central banks bought 288.9 tonnes. The World Gold Council’s own reporting, as the piece uses it, says institutions kept adding through a 14 percent quarterly decline. That is the opposite of buyers chasing a rally. A quarterly total does not prove the buying clustered in the weakest weeks. It proves that, across the quarter, the bid that does not publish a ticker was still removing metal. Official gold, once vaulted, sits. It does not come back to make a price on the way down. The quote is set by the metal that is free to move. The official bid is busy making less of it free.
The piece then does the honest thing with the number everyone wanted to believe. The Council’s first estimate of first-quarter demand was 244 tonnes, reported widely as proof of conviction. In the second-quarter release that figure was revised to 57 tonnes. A swing of nearly 190 tonnes, about four-fifths of the original. The revision landed without the noise that greeted the first print. Anyone who states last quarter’s official buying with confidence is using a number that may not survive the next release. The direction is the part the piece is willing to keep. A four-year pattern, surveys that still say they intend to buy, and a second quarter that was a record even after the revision habit is priced in. The 244 is not a holding. It was a draft.
The surveys are the piece’s other leg. The Council’s ninth Central Bank Gold Reserves Survey, published in 2026, found that 89 percent of central bankers expected global gold reserves to rise over the next twelve months, and a record 45 percent expected their own holdings to grow. An HSBC survey the same year put the share planning additions near 40 percent, and a majority of those said current prices would not stop them. Reserve managers are supposed to buy weakness and sell strength. In 2026, the piece says, they inverted the habit and knew they were doing it. Morgan Stanley research circulated with a different kind of finding. Gold had come to represent a larger share of global central-bank reserves than United States Treasuries, the first such reading since 1996. Not a price target. A demotion. The instrument that anchored the post-war reserve was not abandoned inside official portfolios. It was passed.
Do not promote 288.9 tonnes into a floor. The piece puts London’s free float, the metal that can actually settle a trade, somewhere between 3,000 and 4,000 tonnes, inferred rather than published. Against a float of that size, roughly 289 tonnes in ninety days is on the order of 8 percent of the tradable stock, absorbed by buyers who will not sell it back. The piece says to treat that as an order of magnitude. Not every official tonne comes out of the float. Some was already in a vault. The point survives the caveat. What matters is not the gross stock of gold in the world. What matters is the stock that can be borrowed, shipped, and delivered. Official vaults grow by shrinking that stock.
The piece restores one old sentence and one newer one. John Pierpont Morgan, before the Pujo Committee of the House in 1912, was asked to define money. Money is gold, and nothing else. Alan Greenspan, in the 1966 essay Gold and Economic Freedom, wrote that in the absence of the gold standard there is no way to protect savings from confiscation through inflation. The piece reads that line, in 2026, as a description rather than a theory. You do not have to accept either sentence to use the tonne count. A tradition you do not add to is a story. A bid that takes 288.9 tonnes in a quarter when the price is falling is a bid. It does not have to hit a retail screen. It has to take bars.
Silver’s deficit did not get the memo
Silver is where the piece stops being a reserve sermon. The World Silver Survey 2026, produced by Metals Focus for the Silver Institute, projects a global shortfall of 46.3 million ounces this year, against 40.3 million in 2025. Sixth consecutive annual deficit. Mine production flat. A competing figure of 215 million ounces has circulated in secondary coverage. The piece says the two numbers do not reconcile, and that the Institute’s own headline is the smaller one. The direction is not in dispute in either version. Cumulative deficits since 2021 have drawn above-ground stocks down. The deficits are a matter of record. The claim that they must produce a lasting shortage, and a higher price, is an inference. It assumes demand keeps growing and supply keeps failing. Name it as an inference.
The mechanics are why silver and gold are not one trade. Gold, once mined, is nearly all still with us. Its “deficits” are reallocations. The metal can be found at a price. Somewhere between half and 60 percent of annual silver consumption, the piece says, enters uses where the metal is dispersed in amounts too small to recover. A panel holds perhaps fifteen to twenty grams. A connector holds a fraction of one. When the product is scrapped, the trace usually goes with it. Roughly half of each year’s mine supply leaves the monetary system. Deficits do not merely tighten a market. They convert a monetary asset into an industrial one, until the residual stock is small enough to squeeze. That, the piece says, is what January was. The price cleared $121.62. The ratio fell through 45. Then the paper sellers did their work, and the consumed ounces did not come back.
Keith Neumeyer, chief executive of First Majestic Silver, has made a version of this case for a decade, including in interviews with Kitco News. He has said the metal should trade between $100 and $200. In 2026, having touched $121.62 and retreated, the piece says he looked less eccentric than he had in 2016. His line, repeated often enough that the piece prints it, is that the market is going to wake up one day and there will not be any silver. That is a miner talking his book. Treat it as a book. It is also a description of a deficit that does not clear by press release. A lower price rations the next order. It does not unmake the last panel.
Solar is the end market, and the thrift is real. Manufacturing absorbed 186.6 million ounces in 2025. Photovoltaic demand fell 6 percent that year and was forecast to fall a further 19 percent in 2026, to roughly 151 million ounces, which the piece calls the largest annual decline on record for silver’s biggest industrial customer. Total industrial offtake slipped 3 percent in 2025, to 657.4 million ounces, the first drop since the pandemic, and was forecast to fall another 3 percent in 2026, to roughly 639.6 million. The destruction of marginal use is what a high price is for. The piece’s claim is that the destruction has not produced a pile. The industrial number is falling. The annual deficit, 46.3 million against 40.3 million, is not. Those can both be true if mine supply is flat and the monetary bid, or the other uses, still outrun what the mines ship. Do not quote 457 million. The figure is 657.4. A dropped digit turns a contraction into a fiction.
The clock on a cure is long, and the piece does not pretend it is a slogan with a single year on it. Average grades have fallen. More rock must be moved for the same metal, at a moment when energy and capital are expensive. Permitting in several important places now routinely exceeds a decade from discovery to first production. Financing costs sit above the levels of the 2010s. Shareholders who lived through two decades of value destruction do not rush to build. A data centre is financed in about eighteen months. A solar farm in about a year. A plant is retooled in about two. The demand cycle, the piece says, runs at roughly one-fifth the length of the supply cycle, and it has been running that way since 2021. Anyone who treats the drop from $121.62 to $56 as proof that supply won has skipped the only lag that matters. The price can fall 53 percent and the mine that would answer it can still be a drawing.
The float, the lease, and the tariff that did not land
London’s float is the metal that can be borrowed. The lease rate is what you pay to borrow it. The exchange-for-physical spread is what you pay when the bar in one city is not the contract in another. In January those readings said the two cities had stopped pricing the same thing. A market drowning in silver does not pay up for immediacy. A market scared of its own paperwork does. The collapse that followed does not, by itself, mean the float was refilled. It can mean the leveraged buyer was removed and the bar was not.
The tariff story in the piece is the opposite of a duty that arrived. In November 2025, silver and the platinum-group metals were added to the United States Geological Survey’s list of critical minerals. That designation opened an investigation under Section 232 of the Trade Expansion Act of 1962, the statute that lets a president restrict imports on national-security grounds. The administration had ninety days. The market spent the interval pricing a tariff on refined silver entering the United States. A tariff would pull metal toward America and away from London, the same sort of drain the piece says hit the London float in 2021.
On January 14, 2026, the president issued a proclamation declining, for the time being, to impose tariffs on processed critical minerals and their derivatives. The agencies were pointed toward trade talks. Two further proclamations appeared in the Federal Register on January 20. A duty was imposed on certain categories. Precious metals were, for now, left outside. Silver had already been excluded from the reciprocal tariff schedules the previous September, alongside other critical minerals then under investigation. The piece’s point is not that a tariff took the bars. The point is that the market spent a winter pricing the possibility, the squeeze happened inside that window, and the statute is still on the shelf. A deferred duty is not a bar. It is a veto that was not used. The price is still a signal. The signal now includes a review that can be reopened. Call that a political option. Do not call it a tariff that landed on January 14. It did not.
The unit those old lines assumed
The piece’s fiscal arithmetic is doing a specific job. It is not a poem. The United States, on the estimate of the Committee for a Responsible Federal Budget, ran a federal deficit of roughly $2 trillion in fiscal 2026. Interest expense was about $1.1 trillion, a record 3.4 percent of gross domestic product, and the second-largest line in the budget, ahead of defense and ahead of Medicare. The Congressional Budget Office, on its own conventions for net interest, projects a figure nearer $1 trillion for the year, rising from 3.3 percent of GDP to 4.6 percent by 2036 if nothing dramatic changes. A bond pays you in a unit the same borrower is stretching to meet the interest. Gold, in the Greenspan line the piece chooses, is the refusal of that stretch. A deficit of that size does not make gold go up next Tuesday. It makes the old insult, that gold does nothing, incomplete. The bond does something. It pays you in the unit being diluted.
There is a trap on the other side. If every deficit is a reason to own metal, then metal has no price that is too high and no decline that is information. The January spot print of $5,608 was, for a while, that error in public. People who bought the unit-of-account story at the top did not buy a hedge. They bought a story at the price of a story. The 27 percent decline is the tuition. The deficit did not vanish because the tuition was charged. Both facts fit in one portfolio. The deficit is a reason the hedge exists. The price is the reason you do not own it at any number a chart can print.
The piece distinguishes this from 1980 on purpose. That rally coincided with double-digit inflation, an oil shock, a dollar in retreat, and a leveraged attempt by Nelson Bunker Hunt and William Herbert Hunt to corner silver. Gold peaked near $850 on January 21, 1980. Silver touched about $50 and collapsed in late March, when the Hunts could not meet a margin call. Silver Thursday. Paul Volcker’s Federal Reserve had already driven the funds rate toward 20 percent. What the piece says is different now is the bid. Balance sheets that do not intend to sell, and physical consumption that cannot be switched off like a futures position. That is a harder thing to deflate than two brothers on margin. It is also not a guarantee. The Hunts were a position. A central bank can be a position too, if the surveys are talk and the next revision is another four-fifths.
The models that cannot share a page
Goldman Sachs research, as the piece has it, put gold at $4,900 by the end of 2026. J.P. Morgan’s commodity team went further, an average of $6,000 in the final quarter and $6,300 by the end of 2027, before a round of downward revisions pulled the consensus toward $4,500 by year-end. On silver, J.P. Morgan forecast $63 in the fourth quarter, a level the metal had already traded through before it retreated. BullionVault’s aggregate of several forecasting models put silver above $80 by December. Those paths cannot all be true. They are arguments about whether January was a blow-off that broke the bull, or a spike inside a bull the crash did not kill.
The piece’s lean is the second. Its evidence is the mismatch, not the highest target. A 27 percent drawdown in gold and a 53 percent drawdown in silver, against a second-quarter official bid of 288.9 tonnes, a silver deficit that widened from 40.3 million ounces to 46.3 million, a lease market that showed a real squeeze in January, and a tariff review that scared the float and then declined to strike. Forecasts that only model the screen are describing the screen. Forecasts that only model the shortage are describing a vault the screen has not agreed to price. Know which model you are in before you bet the spread.
A target is not a timeline. “By December” and “by the end of 2026” are crowded phrases in a piece published in the middle of October. A metal can be scarce and late. It can be scarce and cheaper for another year if the marginal seller is a fund, not a mine. The decade-long permit clock cuts both ways. Scarcity does not have to rally on a journalist’s schedule. Abundance does not have to arrive on a miner’s hope. J.P. Morgan’s $63 silver call had already been exceeded and then abandoned by the price. That is the whole warning about a calendar.
Three endings, and the one the headline prefers
The piece sets out three endings and refuses false precision on the odds.
Managed re-rating. Gold and silver grind higher over several years, with drawdowns like the one in 2026. Mines respond. New supply arrives in the early 2030s. The deficit narrows. The metals settle on a higher plateau without a crisis. The piece calls this the least dramatic outcome and, on the evidence, the most probable. It requires official buying to continue and industrial demand to continue. Both, it says, are already in place. The solar thrift is the objection sitting inside this ending. Photovoltaic demand is forecast down 19 percent. If thrift outruns the monetary bid, the plateau is lower than the grind assumes. The piece still ranks this path first.
Disorderly squeeze. A tariff decision, a policy error, or an inventory draw repeats January, into a smaller float and thinner dealer books. The price overshoots. Factories buy at any level to keep lines running. The dislocation moves into margins, then components, then consumer prices. This ending needs no new invention. It needs a repetition of something that already happened once this year. The January 14 proclamation, by declining to tariff the metals, removed one trigger and left the statute. A later decision can be the trigger the first one was not.
Rupture. A failed auction, a currency crisis, a disorderly repricing of sovereign debt. Gold rises severalfold rather than by a modest percentage. Official buying becomes a scramble. The reserve architecture is abandoned formally, not only in practice. The piece calls this the least likely and the most consequential, the scenario in which the phrase “precious metals shock” earns its name. The fiscal numbers, $2 trillion and $1.1 trillion of interest, are the dry tinder for this ending. They are not the match.
A fourth ending should be named so it does not sneak in as destiny. The paper market keeps working. Lease rates stay calm. The tariff stays declined. The next central-bank print revises down the way the first quarter did, from 244 tonnes to 57. The deficit of 46.3 million ounces narrows because solar’s 19 percent cut is larger than the piece’s balance allows, and a later year flips to surplus. Then there is no reckoning. There is a bear market that was a bear market. The piece is written against that ending. It is not allowed to ban it. The argument, the piece says honestly, rests on three legs: sovereign accumulation, physical deficit, and fiscal deterioration. One leg can fail and the case can survive. All three failing in the same decade would be close to a refutation.
What to do with a crash that did not refill the vault
You do not have to share the climax to use the distinction.
First, separate the quote from the float. For gold, write down tonnes that survive revision, not the draft. 288.9 in the second quarter is the print the piece is willing to stand on. 244 in the first quarter is the print it tells you not to trust. For silver, write down 46.3 million ounces against 40.3 million, and the industrial path from 657.4 million to a forecast 639.6 million. If the deficit flips and the official bid becomes a seller, the lower price is a larger pile. If they do not, the lower price is a different fact. It is the price of a claim. Claims are allowed to be cheap. Bars are allowed to be scarce.
Second, do not let a house target pick your size. The stack in this piece runs from a consensus pulled toward $4,500, through Goldman’s $4,900, up to a J.P. Morgan path of $6,000 and $6,300 that was already being revised. The width is the information. A position that only works if $4,900 arrives on time is a position on a calendar. Scarcity can be right and the calendar wrong for longer than a margin account lasts. January showed that being early and being levered is one trade, and the second one sets the exit. Silver at $63 was a forecast the price had already passed and then left behind. Silver above $80 by December, in the BullionVault aggregate, is another calendar. Hold it as a scenario. Do not hold it as a debt the market owes you.
Third, split the metals. Gold’s case here is official demand, a reserve share that Morgan Stanley finds larger than Treasuries for the first time since 1996, and a fiscal bill that makes the Greenspan line feel current. Silver’s case is a deficit that consumption created, widened in the survey from 40.3 million ounces to 46.3 million, and that a tariff review can still disturb, offset by an industry cutting photovoltaic use by a forecast 19 percent. The crash punished silver more because silver had the squeeze, the tourists, and the substitutes. The next move does not owe you a rhyme. A portfolio that holds them as one idea will be right about the reserve and wrong about the factory, or the reverse.
Fourth, watch the plumbing, not the adjective. Lease rates. Whether the January inversion is a memory or a live spread. Whether the Section 232 review stays a declination or becomes a duty. Whether the next World Gold Council quarter looks like 288.9 tonnes or like a number that later becomes 57. Those series can falsify the piece. A headline about a reckoning cannot.
A claim is not a bar
Read the reckoning as a settlement problem and it gets smaller and more useful. Most of what the public calls owning gold or silver is a claim. A future. An unallocated account. A fund. A miner’s share, which is a claim on a cost curve, a permit, and a price, not on a coin. Those claims are liquid because someone else stands between you and the bar. They are cheap because they do not all ask for delivery on the same morning.
The crash from $5,608 to $4,112, and from $121.62 to about $56, happened inside that claim market. Stops, margin, funds that had to sell what they could sell. The piece’s complaint is that the audience then treated the claim market as a census of the metal. It is an auction of the metal allowed to move, plus a larger auction of promises. When official buyers take a few hundred tonnes in a quarter and do not bring them back, they shrink the pile the promises refer to. When a factory disperses silver it cannot recover, it does the same from the other side. When a tariff review pulls metal toward one city and then declines to strike, it moves the promise of a drain without yet creating one.
A reckoning, if the word deserves to be kept, is the morning a marginal holder needs a bar, in a form a refinery will accept, in the city the contract named, on a date that is not later. None of this requires gold at $5,608 again. It requires a delivery the float cannot make at the price on the screen. Investors who have only owned the claim experience this as a spread or a headline. Investors who needed the bar experience it as an absence. Selling a claim adds no ounces. It changes the name on the promise. The price can fall by half and the vault can be the same vault. Silver’s 53 percent is allowed to be that kind of fall. Gold’s 27 percent is allowed to be that kind of fall.
There is a mining version of the same mistake. A producer’s ounces in the ground are not the float. They are not bars until they are mined, refined, and not already spoken for by a stream, a hedge, or a government. A lower price cuts the producer’s cash flow now and cuts the supply response on a clock that exceeds a decade. The crash is worse for the mines than for the shortage. The shortage is a stock. The mine is a flow that just got less willing. Selling both because the quote fell is risk control when the deficit is closing. It is a donation to the next buyer when the deficit is widening and the equities have been marked to a claim market that does not deliver.
What would make this reading wrong
The reading is wrong if the October quotes are not the quotes. The piece says gold near $4,112 and silver around $56. A $5,608 spot, a $5,405 LBMA print, and a $5,318.40 futures close are three contracts wearing one month. Twenty-seven percent and 53 percent depend on which print you call the peak. The direction does not. A reader who needs the exact percentage more than the direction is arguing with a chart.
The reading is wrong if the second-quarter tonnes are revised the way the first quarter was. 288.9 is a record second quarter in the piece. 244 became 57. If the next release does that to the second quarter, the official bid was a draft, and the lower price and a larger float may be the same event. The piece warns you about this and then leans on the number anyway. You are allowed to lean less hard than the piece does.
The reading is wrong if silver’s thrift is larger than the balance. Photovoltaic demand down 6 percent in 2025 and a forecast 19 percent in 2026, industrial offtake from 657.4 million ounces toward 639.6 million, are the piece’s figures. Engineers paid to remove silver will remove more if the price says to. A competing deficit of 215 million ounces does not reconcile with 46.3 million. If the smaller headline is the one that breaks, the shortage was the larger fiction, or the smaller one. The piece picks the Institute’s headline and says the direction agrees. A reader should keep both numbers in view until they do.
The reading is wrong if the deferred tariff is treated as a duty, or if the declination is treated as the end of the review. The proclamation of January 14 declined to impose the tariffs for the time being. Precious metals were left outside. A later proclamation can change that. Until it does, the bars are not being held by a new duty. They are being held, if they are being held, by the official bid, the industrial use, and the thinness of the float.
The reading is wrong if the fiscal arithmetic is asked to do the price’s job. A $2 trillion deficit and a $1.1 trillion interest bill explain why a monetary metal has a constituency. They do not explain $5,608, and they do not forbid $4,112. People who used the deficit as a reason to buy the top have already been answered. The answer was the drawdown.
The idea, once
Milan Adams says it ended in the spring. Gold’s January 2026 cluster, $5,608 spot, about $5,405 on the LBMA benchmark, $5,318.40 on COMEX, gave way to a middle-of-October market near $4,112, about 27 percent lower. Silver gave way from $121.62 on January 29 to about $56, down 53 percent. That is a crash. It is not, on the rest of the piece, a refill. Central banks bought 288.9 tonnes from April to June, through a 14 percent quarterly decline. The 244-tonne first quarter was revised to 57. Do not cite the draft. Silver’s deficit is projected at 46.3 million ounces against 40.3 million, a sixth annual deficit, with mine supply flat. Solar used 186.6 million ounces in 2025, fell 6 percent, and is forecast down another 19 percent, to about 151 million. Industrial offtake was 657.4 million ounces, not 457, and is forecast at 639.6 million. The January 14 proclamation declined to tariff the metals. It did not impose the duty. Goldman’s $4,900, J.P. Morgan’s $6,000 and $6,300 before the cut toward $4,500, and BullionVault’s silver aggregate above $80 cannot share a page. Greenspan’s 1966 line and a $1.1 trillion interest bill are the fiscal leg. Neumeyer’s line is a producer’s line. Morgan Stanley’s finding is about reserve share, not a price.
One object is the quote. It fell. The other is the pile. The piece says it did not come back. A lower price is not a larger pile. Own that distinction, or do not own the trade. The reckoning, if it comes, will be the day a claim asks for a bar and the bar is in a vault that does not answer to the screen.