Dear Gold Digger,
On January 29, silver printed above $121 an ounce. The next session it fell more than 30%, its worst single session in more than forty years.
Seven months later it trades around $65.50. Silver opened the year at $71.59, so it is down roughly 7% since January 1. Gold is up about 3% over the same stretch. The ratio sits near 67 to 1, which is a polite way of saying the market has decided silver is the one that got carried away.
Most investors took a lesson from that January weekend. I think they took the wrong one.
Look at What Was Actually Holding the Metal Up
Run the calendar back.
In April of 2025, an executive order opened a Section 232 investigation into critical minerals imports. That November, the USGS added silver to the critical minerals list, which pulled the metal inside a probe that was already running. For the better part of a year, traders moved bullion into American vaults ahead of a tariff that might land, and London’s free float got thin enough that one-month lease rates spiked toward 39% in October against a normal 0.3% to 0.5%.
That was a customs bid. It had nothing to do with solar panels, deficits, or the money supply.
On January 15, the administration announced no tariffs on processed critical minerals and told its officials to negotiate supply agreements instead. Silver fell about 3% on the news, to right around $90 an ounce.
And then it ran to $121 anyway.
Read that sequence twice, because it is the whole story of the top. The physical reason to hold the metal in New York was withdrawn on January 15, and the price added another third over the next two weeks on nothing but flow. Momentum with no cargo underneath it. That is precisely the kind of move that unwinds in a single session, and on January 30 it did.
I Told You Silver Wasn’t Confirming. Here’s What I Owe You.
Three days ago I wrote you about the September 16 Fed projections, and buried in the honest part of that letter was a line I want to come back to.
Silver is not confirming. Down on the year, ratio at 67.6, lagging when it should be leading.
That was accurate and it was incomplete. Saying a metal is lagging tells you nothing about whether the lag is a verdict or an opportunity. So let me do the work in public, starting with the case against myself.
The Bull Case Got Smaller While You Weren’t Looking
Here’s the part most silver letters won’t print.
In February, the Silver Institute forecast a 2026 market deficit of 67 million ounces and called it the sixth consecutive annual shortfall. By the time the full World Silver Survey landed in April, that figure had been revised to 46.3 million ounces. Still a deficit. Roughly a third smaller than the number sitting in everybody’s slide deck eight weeks earlier.
Solar is worse. Photovoltaic silver demand is forecast at 151 million ounces this year, down 19% from 186.6 million in 2025. Not because the world is building fewer panels. Because manufacturers got very good at using less silver per cell, and silver-coated copper powder now cuts loadings by 30% to 50% at similar performance. The solar story that powered this thesis from 2021 through 2024 is running in reverse.
Recycling hit 197.6 million ounces last year, the highest in 13 years. Industrial fabrication is forecast down 3%. UBS cut its 2026 deficit estimate to 60 to 70 million ounces from roughly 300 million, an 80% haircut, and told clients silver would trade broadly sideways.
And if you have seen a 215 million ounce deficit quoted anywhere this year, the Silver Institute’s own survey says 46.3. Somebody printed a number that the source document contradicts, and it has been circulating ever since. Delete it from your notes.
That’s the bear case, delivered straight, without the softening adjective the average newsletter would slip in front of it.
One caveat before I move on, and I’m flagging it as an estimate rather than gospel. Silver’s demand mix is quietly rotating while solar shrinks. Electrical and electronics fabrication is the biggest industrial bucket at roughly 423 million ounces this year, and analysts working off that base project data center consumption climbing past 42 million ounces annually by 2027, better than 10% of the category. The Silver Institute doesn’t publish a data center line item, so treat that as an outside calculation rather than a survey number. But hold the shape of it: 42 million ounces is very close to the whole 46.3 million ounce deficit the survey forecasts for this year, and it is being routed into chip packaging, power modules, and thermal interfaces by an industry that does not check the silver price before building.
Now Do the Other Side of the Ledger
The average silver price rose 42% in 2025, to just over $40 an ounce.
Global mine production this year is forecast at 844.1 million ounces. Last year it was 846.6 million. That is a decline of three tenths of one percent.
Think about that for a second. The single most powerful signal a commodity market can send, a 42% price increase in twelve months, went out to every mining company on the planet, and the supply response was less than nothing.
Compare that to any other market you know. Shale operators answer a price move like that in nine months. Semiconductor fabs answer it with a capex announcement inside a quarter. Even cattle ranchers answer it, slowly, but they answer it.
Silver miners didn’t answer it. And the reason is the most important fact in this letter.
Three Out of Four Ounces Are Somebody Else’s Decision
Only about 26% of the world’s silver comes out of a primary silver mine.
The rest arrives as a by-product. Lead and zinc mining throws off 29.4% of global supply. Copper mining throws off 28%. Gold mining throws off another 15.9%.
So when silver rips 42%, the message lands in the inbox of a zinc executive whose board is compensated on zinc, whose mine plan is engineered for zinc, and whose silver revenue is a rounding entry his CFO calls a credit. He is not going to redesign a mine because a metal he doesn’t manage had a good year. He would be fired for trying.
That’s the tell. In silver, the price signal reaches the wrong people. Roughly three quarters of world supply is decided by copper prices, zinc prices, and gold prices, and by permitting offices, and by nothing else.
Which means the ordinary mechanism that heals a commodity shortage, high prices calling forth new ounces, is mostly disconnected in this market. It doesn’t work. It has never worked. And no revision to a deficit forecast changes it by a single ounce.
The 762 Million Ounces That Don’t Come Back
Here is what those years of shortfall did.
Cumulatively, 2021 through 2026, the silver market has run a 762.1 million ounce deficit. That is roughly eleven months of total global mine production, consumed out of above-ground stocks that took decades to build.
The Silver Institute’s own language for where that leaves us: the market “has clearly entered an era of reduced stocks,” and from here “liquidity will generally be thinner, lease rates more volatile and price moves likely to be larger.”
We’ve already watched what that looks like twice. Those October lease rates were the first episode, with spot trading above futures while the borrowing market seized. In February of this year, one-month rates jumped again to 6.3% annualized. Neither event was caused by a forecast revision. Both were caused by somebody needing metal in a hurry and discovering the shelf was thin.
Meanwhile ordinary buyers keep showing up. Coin and bar demand is forecast at 257.6 million ounces this year, up 18% from 217.7 million in 2025. Retail is buying the drawdown while institutions sell the forecast.
And LBMA vaults held 28,213 tonnes as of July, worth about $52.7 billion, which sounds enormous until you remember what portion of it is already spoken for by exchange-traded products and cannot move without somebody redeeming shares.
Now the Honest Part
I have been early on silver before, and early is expensive.
The bear case here is serious, and I’m not going to caricature it. Thrifting is permanent. Once a solar manufacturer engineers a third of the silver out of a cell, that silver never comes back, and the same engineers are now pointed at electronics. If industrial demand keeps sliding 3% a year while recycling grinds higher, the deficit closes on its own and silver spends a long time doing nothing while you pay the storage bill.
The 2026 mine disruptions don’t rescue the story either. Add up the Endeavour blockade in Mexico, Peru’s 9% June decline, and the weather at Los Pelambres, and you get about 1.1 million ounces. That is 2.3% of this year’s forecast deficit. I could have written a very exciting paragraph about supply disruptions there. It would have been noise.
And note that Wall Street’s price call has been wrong in both directions. UBS cut its deficit estimate by 80%, then set a September target of $85. Silver is at $65.50. The desk that turned bearish on the fundamentals was still too bullish on the price.
Silver is half an industrial metal, it trades with more beta than gold in both directions, and it is down on the year for reasons that have nothing to do with my argument. Anyone who tells you the setup is clean is selling you something.
What I’d Do From Here
None of this is a trade alert, and I don’t know what silver does between now and Christmas. It’s a positioning note. Here’s how I’m thinking about it.
Own the metal before you own the leverage. Physical first, in your hands or in allocated storage where you can name the bars. The whole argument above is an argument about ounces existing, and a paper claim on an ounce is not the same instrument as the ounce.
Size it like the volatile thing it is. Silver handed investors a 30% loss in one session this year. That will happen again. Position sizes that survive a January are the only ones that get to be right about a December.
Buy the boredom. The January top was made in public, at maximum enthusiasm, with lease rates screaming. The setup I’m describing is being built now, quietly, while the consensus writes silver off as the metal that already had its moment.
And stop watching the deficit number. It was never the argument. The argument is that 74% of the world’s silver supply doesn’t take orders from the silver price, and no committee, no forecast revision, and no permitting reform fixes that inside a decade.
The shakeout is the setup. That has been true in this metal for fifty years, and January didn’t repeal it.
Silver has spent seven months teaching investors to look away. The ounces are keeping their own count.
That instinct, buying the thing the market has stopped looking at, is the whole job over at The Wealth Advisory. And right now it has me pointed at something with no connection to metals at all.
For forty years this company has been the most boring kind of rich. It owns the warehouses, more than a billion square feet of them, better than twice the footprint of Manhattan, and the tenants pay the rent like clockwork. Then, without any announcement, it assembled a power pipeline approaching the size of six nuclear reactors, on land it already owns, in the exact markets where the data centers are going up. Wall Street still prices it as a pile of concrete.
Same mistake as silver. Different address.
The Most Boring Company in the AI Buildout »
To owning what’s real,
Jason Williams
Senior Investment Strategist, Gold World