Tearing apart the buffer that hides the deficit

Silver.

The financial press spent the last six months telling you a story about solar panels. Silver demand from photovoltaics dropped 19% this year. Manufacturers are thrifting — using less paste per cell, redesigning layouts, printing thinner lines. Bloomberg ran it as a demand headwind. CNBC followed. The consensus take crystallized: silver is losing its biggest growth driver.

They’re reading the label on one valve while the whole system drains.

Here’s what nobody put in the headline: the deficit widened. Solar used 19% less silver, and the shortfall still grew by 15%. That’s not a demand story. That’s a supply story the demand headlines are hiding. Total supply is contracting 2% this year while coin and bar demand is surging 18% to its highest level since 2022. The solar thrifting narrative is accurate in isolation. It’s also completely irrelevant to the structural picture — like celebrating that a sinking ship used less fuel today.

The real mechanism is simpler and uglier. The world has been drawing down above-ground silver stocks to cover these deficits for six straight years. 762 million ounces. Gone. Not recycled back in — consumed, exported, locked in private vaults. The buffer that made the market look balanced is being eaten alive.

And the price everyone stares at on the COMEX screen is the last thing that will reflect it.

Numbers.

Mine production in 2026 is projected at roughly 820 million ounces. Recycling adds another 200 million. Other supply — net government sales, producer hedging — contributes roughly 40 million more. Total supply: about 1.06 billion ounces. Total demand: approximately 1.11 billion ounces. The gap — 46.3 million ounces — gets filled by pulling metal out of vaults, ETFs, and above-ground stockpiles. That’s how it’s worked since 2021. The problem is the stockpile isn’t infinite.

At COMEX, registered silver — the metal actually available for delivery against futures contracts — stood at around 95 million ounces as of mid-July. Down more than 75% from the pandemic-era peak. Open interest on those same contracts represents around 545 million paper ounces. The coverage ratio is roughly 18 percent. For every ounce sitting in a deliverable vault, roughly six ounces of paper claims exist against it.

In London, LBMA vault holdings totaled 28,082 tonnes — roughly 903 million ounces — at the end of June. Sounds like a lot until you subtract the silver backing ETFs and ETCs stored in those same vaults. The free float is significantly smaller than the headline number. And it’s 23.5% below the 2021 peak.

Here’s the supply side that makes it structural: 75% of all silver mined on this planet comes out of the ground as a byproduct. It’s a hitchhiker on copper, zinc, lead, and gold operations. You can’t “ramp up” silver production. You’d have to ramp up zinc production, and zinc has its own problems right now. Mine output peaked at 900 million ounces in 2016 and has never recovered. Price doesn’t fix that. Geology does, eventually. Eventually isn’t this year.

The deficit isn’t dramatic in any single year. That’s why it’s so dangerous. It’s quiet. It’s cumulative. And 762 million ounces later, the cushion is threadbare.

January told you everything you needed to know, if you were watching the right screen.

Silver hit $121.62 on January 29th — an all-time high. Then it crashed more than 30% in roughly thirty hours. The COMEX tape looked like a cardiac event. Margin hikes, forced liquidations, leveraged paper sold faster than physical buyers could absorb. The financial press called it a bubble bursting.

On the Shanghai Gold Exchange, physical silver traded at a 12% premium over COMEX the same week. During the crash low, that premium reportedly blew past 50%. Paper silver was in freefall. Physical silver in the world’s largest consuming market was screaming scarcity.

I watched the January crash in real time. I’ve seen enough of these to know what the tape looks like when leveraged longs get carried out. What I hadn’t seen before was the physical market completely ignoring the paper market’s panic. That’s new. That’s the weak link — not in silver itself, but in the price-discovery mechanism that’s supposed to reflect its scarcity.

COMEX is 95% cash-settled. The price on your Bloomberg screen is mostly a derivatives construct. The actual metal — the stuff solar manufacturers in Jiangsu need, the stuff data center builders in Virginia need for switchgear and busbars and high-cycle connectors — that metal is getting harder to source. The five largest U.S. hyperscalers have earmarked $736 billion for data center buildout in 2025 and 2026. Every one of those facilities needs silver for things copper can’t do. Nobody is modeling this demand in their silver forecasts.

The gold-silver ratio sits at 69:1. Historically, anything above 80 has been a screaming buy signal for silver. At 69 it’s not screaming. But it’s muttering in a market where the cumulative physical deficit has no modern precedent.

The managed money crowd reduced their net long exposure in mid-July. Primarily by adding fresh shorts, not just long liquidation — but either way, lighter positioning into a tightening physical market. I’ve seen this setup before. The paper gets cleaned out. The physical stays tight. Then something forces the reconciliation, and it isn’t gentle.

The sequence is mechanical. I’ve watched it run in palladium, in nickel, in rhodium. Silver is bigger and more liquid than any of those, which means it takes longer to break. It also means the break is louder.

The buffer has been absorbing 40 to 50 million ounces of deficit per year for six years. At some point — and nobody knows the exact ounce — the buffer stops absorbing and starts rationing. That’s when premiums don’t just widen. They detach. Shanghai already showed you the preview in January.

If a delivery failure hits COMEX — or even the credible threat of one — the first thing that happens is a scramble for registered metal. That 95 million ounces of deliverable inventory gets bid on simultaneously by parties who assumed it would always be there. Second wave: ETF redemptions reverse. Instead of selling paper silver, funds start buying physical to cover. Third wave: industrial users with just-in-time procurement realize just-in-time doesn’t work when the vault is empty, and they panic-buy forward.

Where does capital go? Not into the broad precious metals ETFs — SLV and its peers are part of the problem, not the solution, because their metal sits in the same vaults being drawn down. The edge, if there is one, is in primary silver miners — the 25% of production that actually comes from dedicated silver operations where management can respond to price. Companies with unhedged production, operating mines in stable jurisdictions, and reserves that aren’t dependent on base metal economics to justify extraction.

They’re the levered play on a reconciliation between the paper price and the physical reality.

Silver at $59 with a 762-million-ounce hole in the buffer and an 18% delivery coverage ratio is not a price that reflects the physical market. It’s a price that reflects a derivatives market that hasn’t been forced to settle in metal yet. In my experience, “yet” does most of the work in that sentence. The timing is unknowable. The direction, if the deficit persists, is not.