The trust holds 81.7 million pounds and a net asset value near $7.4 billion. The units closed below that metal. China cut its Treasuries to $591 billion and kept buying gold. The customer who stocks a vault is not the customer who used to set the price.
This is not a recommendation to buy or sell a trust, a miner, a bar, or any fund. It is not a price target. It is not a view that a stockpile makes you safe. A pound that is locked up can still fall in price. A share can trade below the pound it claims. You can lose money on a real deficit. A buyer who will not sell is a fact about supply. It is not a promise about your entry.
Here is the only idea. Just in time set the price. Just in case takes the metal off the market. John Ciampaglia, chief executive of Sprott Asset Management, put it in those words on Alex Deluce’s Gold Telegraph. He is not the chief executive of Sprott Inc. A just-in-time chain is cheap because the fuel, the part, and the bar are supposed to arrive when the line needs them. A just-in-case chain leaves metal idle on purpose. Idle metal is a cost. It is also a bid that does not come back as supply next month. If you mark a commodity off the old screen, you are marking a market that still assumes the metal returns. A growing share of the buying is being done by people who have said they will not return it.
Two clocks, one rock
He drew the split on energy, because energy makes it obvious. Oil, liquefied gas, and pipeline gas move all the time. A ship, a pipe, a cargo. Stop the ship and the price jumps the same week. That is a just-in-time market. Nuclear fuel is the other clock. Once the fuel is in the core, he said, the station can run about two years without a new delivery. Another load can sit on site, working through the steps of the fuel cycle. You cannot disrupt that station the way you disrupt a gas import. The metal was bought years before the light stays on.
He put a number on why a grid notices the difference. A nuclear station, in his telling, runs about 92 percent of the time. Wind and solar do not. He said you cannot run a heavy industrial economy on a source that works about a quarter of the time and that you do not control. He used a blackout in Spain as the picture. Treat that as his picture, not as a settled cause. Grids fail for more than one reason. The point he wants is narrower. Variable power is cheap on a spreadsheet until the hour it is not there. Base load is a machine that is there. The machine eats uranium on a calendar, not on a spot screen.
That calendar is why he thinks the last few years changed the bid. He lines up two shocks, 2022 and 2026, with the oil shock of the 1970s. Then, he said, a squeeze in oil pushed governments to build most of the nuclear stations that still run. Now, he said, the shocks have pushed the same governments back toward energy they can store and run through a bad month. He does not claim this was his original uranium thesis. When Sprott entered, demand looked flat. China was building reactors. The West was closing about as many. The deficit was a low price and missing supply. The policy reversal, he said, arrived later, and it made the case longer.
A book that does not come back
The cleanest object in the interview is not a forecast. It is a book of uranium that, by his account, does not rejoin the market. Sprott did not merely change the manager of an old fund. On April 28, 2021, Sprott Asset Management signed an agreement with Uranium Participation Corporation. Shareholders would become unitholders of a new trust, and the old company would sit underneath it. He remembered the announcement as May, and he put the courtship at about two and a half years. Shareholders approved the arrangement on July 7, 2021. Sprott dates the trust’s inception to July 19, 2021. Rick Rule, he said, pushed the idea in 2018, when the spot price had fallen to about $18 a pound and then only into the $20s. The best mines were on care and maintenance. A new fund was hard to sell. So they bought an old one.
He said the book then was about 18 million pounds, and that the vehicle had grown from about $600 million. The filing is tighter. At the end of March 2021, Uranium Participation reported 16,269,658 pounds of U3O8 and 300,000 kilograms of uranium as UF6, with a market value of about C$665 million. Those are not the same unit, and the dollar figure is Canadian. Sprott’s own page, as of October 16, 2026, shows 81,697,348 pounds of U3O8. The net asset value of the trust was $7,436,223,962. The market value of the uranium inside it was $7,352,761,282. He told Deluce the vehicle had grown to over $7 billion in five years. The page supports that top end. He also said they had raised about $4 billion since inception. A raise and a net asset value are not the same number. Price did part of the climb from the $20s to about $90. Both figures can sit in the same sentence if you do not add them.
The rule of the book is the theme. Unlike Sprott’s gold, silver, and platinum-and-palladium vehicles, this one does not hand metal back to a shareholder. He said they have never sold a pound. He said they have never lent a pound. The uranium sits in long-term storage. Sprott’s fact sheet as of September 30, 2026 names the keepers: Cameco in Canada, ConverDyn in the United States, and Orano in France. A holder of the gold trust can redeem units for a London bar, and Sprott stores that metal at the Royal Canadian Mint. Uranium title is not a bar you can have mailed to the house. The trust sequesters. Sequester means the pound leaves the commercial float and does not automatically return when the spot price looks rich.
That is a bid with no offer attached. It is also not a magic premium. On October 16 the units closed at $18.81 against a net asset value of $21.61. Sprott’s own figure for that gap is a discount of 12.95 percent. The metal and the ticket were not the same price on the same day. An investor who hears “physical uranium” and pays the market price is not paying the spot price of the pound, and is not owed a bar. The trust can be a real withdrawal of supply and a bad purchase on the same afternoon. Do not let the slogan close that gap.
He was surprised by who showed up. A prospectus for $300 million, filed in August 2021, was used up in three weeks, he said. The old vehicle had taken about 15 years to reach $600 million. They reloaded for a billion dollars and, in his memory, that was gone in about six months. He called the buyers a quiet group of uranium bulls the firm had not known was there. Today he puts the holder base at about 70 percent institutions and family offices. Education and reported trades, he said, are why an opaque market let them in. Opacity is still the mineral’s native state. A trust that publishes its pounds does not make the whole fuel cycle transparent. It makes one inventory legible.
The price rose, and the hole did not close
He does not argue that $90 fixed the mine. A broker index for October 16, published by Uranium Markets, put the price at $90.13 a pound, with a bid of $89.75 and an offer of $90.50. That is one desk’s print, not the official month-end. Cameco’s month-end average of the UxC and TradeTech indicators for September 30 was $89.63 spot and $96.50 long-term. Write-ups of that Cameco series call $96.50 the highest long-term reading in it, a level also printed in August. His “$90” is the spot neighborhood. The contract price is a different print. Utilities do not live only on the spot screen. An investor who lives only on it will misread both.
What the repricing did, in his account, is wake the easy pounds. He said brownfield restarts added about 50 million pounds a year. Every idled plant that could come back, came back. The structural deficit, he said, is still there. The pounds that were spoken for stayed spoken for. The pounds that required a new hole in the ground did not appear on the same clock.
The constraint, he said, is not geology. Uranium is not rare in the way a headline likes. The constraint is a large, viable deposit, and the years between a permit and a pound. He called the Athabasca basin in Saskatchewan the Saudi Arabia of the metal. The permits did move this year, and the record is specific. On February 19, 2026, the Canadian Nuclear Safety Commission issued Denison Mines a licence to prepare the site and construct the Phoenix in-situ mine at Wheeler River. Denison called it the first large-scale Canadian uranium mine approved for construction in more than 20 years. In early March the same commission issued NexGen Energy a licence to prepare the site and construct Rook I, on the Patterson Lake peninsula. That licence does not authorize operation. The Saskatchewan government said these were the first Canadian uranium mines approved for construction since 2004. He said most mines now under construction were found in the last cycle and then sat 10 or 12 years. NexGen’s own line is that the team has been preparing since the Arrow discovery 12 years ago. Construction slips. Permits slip. A construction licence is not a pound. Demand, in his view, will not wait with the same patience, because the demand is now a set of national plans, not a single utility’s reload.
The plans are the part that was not in the 2018 file. India as well as China. A West that had talked about closing stations and then watched the politics of a dark grid. Life extensions. Stations in the United States that had been switched off and are being talked back on, which he said had not happened before. Half-built plants that someone now wants to finish. A tech firm, in his example, offering a utility a 20-year power contract at about twice the wholesale price if the utility upgrades the turbines and sells the extra power. Take the multiple as his description of a deal, not as a term you can audit from this interview. The direction is the point. The buyer of electricity is now helping pay to keep a reactor, and a reactor is a 60-year customer for fuel. A spot price does not describe a 60-year customer.
He was blunt about the people who would have to build it. Specialized welders. Mining engineers. University programs he said had emptied. A workforce that left when the work left for cheaper countries, and a generation that did not replace it. A permit can be stamped in a year. A welder is not stamped. If the bull case needs new fuel-cycle plants in the West inside a decade, he thinks it can be done, and he thinks it will be late, costly, and short of hands. China’s edge, he said, is cost, standard designs, and speed on big infrastructure. The West’s edge is not speed. It is the admission, finally, that its own rules were part of the delay.
The government is trying to become that buyer
He then lifted the uranium rule and set it on the rest of the table. For a long time the cheap place did the processing. Rare earths, enrichment, copper smelting. Globalization called that efficiency. It was efficiency. It was also a single pipe. When the pipe became a weapon, governments that had preached the market started copying the stockpile.
He does not pretend the copy is elegant. Free capital, he said, still finds the better project. Government money has a mixed record. What is happening now, in his words, is industrial policy in a trade fight. Loans. Offtake. A price floor if the price is crashed on purpose. He used rare earths and a cobalt project in Idaho. China, he said, can flood a small market, take a loss, and leave the other country’s mine unbuilt. A small market can still sit inside a large industry. In the line the program used to bill the interview, he said a cutoff would cripple $6 trillion of the global economy. That is his figure. It is not a table you can audit from the conversation. The usable point is smaller. The metal market can be tiny. The factory that stops without the metal is not.
Project Vault is his American example, and he was careful with it. A public-private pile. Firms put money in. The vault buys the metal they name. On a bad day they take their metal out. He called it infancy. He said the first slice of money was about $12 billion, and that real resilience across many metals would cost much more. That matches the shape of the program, not a warehouse already full. The Export-Import Bank approved a loan of up to $10 billion in February 2026. The design beside it was about $2 billion of private capital. Glencore and Mercuria later committed $500 million each. A copper cargo of 12,000 tonnes has been described. A ceiling, a commitment, and a cargo are three objects. He is not claiming the $12 billion is already rock. He is claiming a new type of buyer has a budget, and that the budget is still small relative to the list.
The list is the tell. He said the U.S. Geological Survey’s critical-minerals set now covers a huge part of the periodic table. A few years ago, he said, it was hard for officials to call copper or uranium critical, because the assumption was that some other country would keep shipping. That assumption is the just-in-time assumption. Naming the metal critical does not mine it. It does mark it as something a government may buy, store, and refuse to treat as a normal offer. He thinks that shift raises the cost curve. Idle inventory is less efficient than a pipe. Less efficient, in a commodity, means a higher price floor than the old model allowed, if the inventory is actually bought. If it is only named, you have a speech.
China is running the same trade from the other side
He does not think the West invented this. He thinks China has been doing it with quieter tools, and is now doing it in public. The hinge he picks is 2022, when Western governments froze a large block of Russian reserves inside the dollar system. A keyboard, he said, taught every reserve manager that a Treasury is a permission. After that, the question was what you hold that a keyboard cannot freeze.
Gold is the answer they can show. He said China used to go long stretches without reporting buys. Now it reports often, and he reads the reporting as a message. The published record supports the direction, and it is more exact than the adjective. The World Gold Council, using the central bank’s own report, said the People’s Bank of China added 23 tonnes in September 2026, the largest monthly addition since October 2023. That was the 23rd straight reported month, a streak that runs from November 2024, after a pause earlier that year. Holdings reached 2,410 tonnes, 8.5 percent of official reserves. At the end of 2024 the same source put the stock at 2,280 tonnes, 5.5 percent of reserves. The tonnes rose by 130. The share rose by three points. Most of that share jump is the gold price, not the buying. The United States still holds about 8,133 tonnes. China’s published pile is not a rival to that stock. Deluce said many of his guests think the real buying is larger than the report. A belief is not a tonne. Use the tonnes you can cite. Treat the rest as a reason to watch, not as a number in a model.
The Treasury side is also sharper than his round figure. He said China’s stock of U.S. government debt had fallen from about $1.3 trillion to almost half. The peak near $1.3 trillion is the old high, not last year’s number. The Treasury’s own tally for August 2026 was $590.7 billion, down $27.3 billion from July, the lowest since 2008. That is less than half the old peak, not a gentle trim. He said they are not merely letting bonds mature. He said they are selling, and recycling the money into gold, mines, and pipes. He named uranium in Namibia and nickel in Indonesia as the kind of rock he means. Those are examples of a strategy, not a full list of deeds you should treat as closed transactions from this conversation.
Put the two ledgers together and the theme is the same trade in two directions. A Western government is trying to buy metal it used to assume would arrive. A large Eastern creditor is trying to swap paper claims for metal and mines it can run. Neither buyer is a fabricator who will sell the surplus next Tuesday. Both are building a reserve. The spot price still prints every day as if the only buyers were fabricators.
Gold is the same idea without a reactor
He was asked if official buying through a pullback felt like an old cycle. He said no. The structure, for him, is the freeze, the fracture, and a reserve asset that is not a promise. Individuals are late to a move he thinks governments started. He said it is astonishing how many institutions still own little or no gold. He said it is not a crowded trade. He also said Western, Indian, and Chinese investors have already raised their holdings inside gold funds. Those two lines can both be true. A pension can still be empty while an exchange-traded fund takes in ounces. Empty is not the same as early. The ounces in the funds are the crowded part. The mandates that still hold none are the part he thinks has not arrived.
He was specific about the private buyer in China, and you should hear it as a constraint, not as a romance. A saving culture. Property, in his view, is dead as a store of wealth. Crypto is not allowed. The local stock market has not been a refuge. Gold is allowed, familiar, and outside the claim that just got frozen somewhere else. Indian buyers, he said, have long kept gold and silver outside a banking system they do not fully trust. The new piece is that some of that demand now sits in funds, not only in jewelry and small bars. A fund is easier to buy. It is also easier to sell on a bad Tuesday. Do not confuse a cultural preference for metal with a promise that the fund holder will behave like a central bank.
The central bank is the buyer who fits the theme. It publishes a purchase in a down month. It does not need a quarterly performance fee. It is exchanging a paper reserve for a bar. That can coincide with a falling price, which is what Deluce said was underway. A falling price and a continuing official bid are not a contradiction if you accept the split. The price is set by whoever is willing to trade today. The official bid is set by whoever is willing to hold through today. Ciampaglia’s uranium book is the same split, with a published pound count instead of a vault photograph.
What you can mix up
The equity is not the commodity. He said copper and uranium themselves have been resilient while many of the related shares have sold off. Cost of capital, inflation, and a conflict that feeds energy prices are, in his account, knocking the stocks around more than the rocks. A miner is a business with a permit, a workforce, and a discount rate. A pound in a trust is a pound. In a violent tape they do not have to move together. If you buy the share because you believe the rock, you still own the share.
The spot is not the term. Uranium near $90 and a term indicator near $96 are both “high” against $18 and both unfinished against the cost of a new mine, if you accept his worry that the next generation of projects still needs a higher price to justify the wait. He did not give you the incentive price. Do not invent one. The interview’s point is that a price which restarted brownfields has not, in his view, solved greenfields. Brownfield is the mine you already dug. Greenfield is the decade.
The vault is not the import. A government program that may lend to manufacturers, and a trader who delivers a cargo, are not the same as a country that already imports the metal for ordinary use. He said Vault is early and underfunded relative to the ambition. Believe the adjective. An earmark is a budget line. A budget line does not mill a magnet.
The speech is not the welder. He thinks Western governments have admitted they slowed their own mines. Admission is cheap. A four-year permit that becomes a one-year permit still needs a crew, a mill, and a power line. Canada’s mineral endowment, which he tied to the way Sean Boyd, chair of Agnico Eagle, talks about the country’s moment, is rock in the ground. Rock in the ground is not a supply chain. His own firm’s uranium had to be bought from the existing float because a new mine could not be stood up on a fund’s timetable. That is the lesson inside the origin story. When they could not build, they bought what already existed. Governments that talk about building are about to learn the same sequence, or they are about to skip it and simply bid for what already exists. Bidding for what exists is a price. It is not a new pound.
What would make this reading wrong
The reading is wrong if the sequestered uranium comes back. A trust that has never sold can change its policy. A loan of pounds, if it ever happened, would put supply back into the float. He said neither has occurred. The day a sale is disclosed, the “bid with no offer” line is weaker. Watch the pounds, not the slogan. Sprott posts them.
The reading is wrong if brownfield pounds plus a few new Athabasca mines close the deficit before the reactors that exist only on paper are actually ordered. Ambition is not a fuel contract. He said the industry moves slowly. A slow industry can also mean the demand he is counting does not arrive. Life extensions can slip. A tech power contract can be a press release. A station that was off can stay off. If the demand side is mostly speeches, the $90 price is a speculative bid sitting on a smaller hole, and the just-in-case story is early by a decade rather than early by a year.
The reading is wrong if governments stockpile in name and not in metal. Project Vault can remain a loan ceiling. A critical-minerals list can grow while inventories do not. A price floor that is never paid is a sentence. Just in case raises the cost curve only when someone actually pays to hold idle metal. Count deliveries. Do not count strategies.
The reading is wrong if China’s Treasury sales and gold buys are a normal reserve tweak, not a structural exit from paper. The August figure is real. The motive is an inference. A country can sell bonds because the price fell, because the yuan needed support, or because it wants mines and bars. He picks the third. The tonnes and the bond line will tell you more over the next year than another interview will. If the gold buying stops when the price rises, it was a trade. If it continues when the price falls, it is closer to his reserve story. September was a down-cycle test only if you accept that prices were weak. He and Deluce said they were. One month is not a regime.
The reading is wrong if you needed one ticker to own the whole shift. There is no such ticker. A physical trust, a developer in Saskatchewan, a smelter, a gold bar, and a magnet maker are five different ways to be wrong. Institutions that own no gold can keep owning no gold. The trade he calls uncrowded can stay uncrowded for a reason you do not like, including a long bond market that, as he closed, nobody wants to finance. He said buyers will not take the long debt. That is a cost-of-capital problem for every mine that still has to be built. A higher floor for the metal and a higher hurdle for the mine can arrive together. The rock can be right and the share can still be early, or expensive, or both.
The idea, once
Ciampaglia’s point is not that every commodity is in a supercycle and you should own the basket. He is a manager with a product to explain. Hear the product as a mechanism. A uranium trust holds 81.7 million pounds, with a net asset value near $7.4 billion, and he says it has never sold a pound and never lent one. The units can trade at a discount to that metal, and on October 16 they did. The mechanism is the withdrawal. The price of the unit is a separate decision.
He thinks governments are trying to copy the mechanism across a list of metals that used to be someone else’s problem to ship on time. The copy is young. Twelve billion dollars of facility, in his framing, will not stock a periodic table. China, on the numbers you can check, has cut its Treasury holdings to $590.7 billion and reported official gold holdings of 2,410 tonnes after 23 straight months of buying. The tonnes rose by 130 from the end of 2024. The reserve share rose much more, because the price did. That is not the end of the dollar. It is a creditor swapping a slice of paper for a slice of rock, in public.
Just in time set the price, because the metal was supposed to come back. Just in case takes it off the market and leaves it there. If you cannot say whether the buyer in front of you is a factory or a vault, you do not know which clock you are paying for. The factory’s clock is a shipment. The vault’s clock is a year in which that shipment does not return.