The December 28 Silver Crash Wasn't an Accident — Inside the Five-Year Supply Squeeze
The December 28 Silver Crash Wasn't an Accident — Inside the Five-Year Supply Squeeze
On December 28, silver hit $83.90 per ounce. An all-time high. The next trading session, nearly 10% of that price vanished. Leveraged traders were force-liquidated, the chart broke like a cliff. On the surface it's "just volatility," but I don't think it was an accident. And the most important fact is that this is the third time the same playbook has run.
What Actually Happened — CME Hiked Margins at the Top
The rally started around $30 per ounce at the beginning of the year. By December 26 it pushed past $75, and on the 28th it hit $83.90. That's roughly +163% year-to-date. Moves of that size are driven by real supply and demand, not random speculation.
The crash trigger was CME Group, the Chicago Mercantile Exchange. They operate COMEX, the world's largest silver futures market. On December 29—right at the peak—they raised margin requirements.
In futures trading, margin is the security deposit you post to control a much larger position. A typical $20,000 margin lets you control roughly $400,000 of silver. When CME suddenly said you now needed $25,000, anyone who couldn't add $5,000 immediately got a margin call: forced liquidation. When thousands of traders are forced to sell simultaneously, the price falls off a cliff.
Timing was decisive. The week between Christmas and New Year is the thinnest liquidity window of the entire year. Peak price + empty market + margin hike = flash crash. You can decide whether that timing was intentional. The mechanics are the same either way.
Why I'm Confident — This Is the Third Time
The same playbook has now run three times.
| Year | Peak | Drawdown | Trigger |
|---|---|---|---|
| 1980 | $50 | -80% | COMEX 'Rule 7' — new positions banned, 50% margin |
| 2011 | $49 | -48% | Five margin hikes in two weeks |
| 2025 | $84 | -10% (ongoing) | Two margin hikes in two weeks |
1980 was the Hunt Brothers cornering attempt unwound—the famous Silver Thursday. 2011 was post-GFC panic buying liquidated. So what's different about 2025?
Structural industrial demand. That's the entire difference. The 1980 and 2011 rallies were speculative. The 2025 price floor is supported by inelastic industrial demand that doesn't care what the paper price is.
You Can Crush the Paper Price, but You Can't Print Physical Silver
The silver market has been in a supply deficit for five consecutive years. The 2025 deficit is estimated at roughly 230 million ounces. That's the entire annual output of Mexico, one of the world's top producers.
Three demand pillars are all accelerating.
① Solar — 29% of industrial demand. It was only 11% a decade ago. Each panel uses about 20g of silver paste, and there's no commercially viable substitute. Some analyses suggest that by 2050, 85–90% of all known reserves could be consumed by solar alone.
② EVs — A battery EV uses 25–50g of silver, about 70% more than an ICE vehicle. Battery management, power electronics, charging systems—all silver-heavy. Automakers won't halt production over silver prices; they'll just pay up.
③ AI infrastructure — This category didn't exist in 2011 or even 2020. AI servers and data centers rely on silver for high-performance electrical connections. It's brand-new demand stacked on top of solar and EV demand.
Two additional supply shocks just landed.
First, China's export restrictions. China controls about 60% of the world's refined silver supply. To protect their own industrial stockpile, they've effectively shut down export licenses. Elon Musk has tweeted publicly about silver shortages—the man who builds EVs, solar panels, and AI chips is worried about silver supply. That itself is the signal.
Second, inventory collapse. Shanghai silver inventories are down 86% from 2020 levels. London and COMEX are drawing down in parallel. At current usage rates, the major hubs hold only about 30 days of usable silver.
The Paper-Physical Decoupling
The Shanghai premium tells you the paper market is lying. A normal premium is $5–10 per ounce. Recently it has stretched to roughly $35. Translation: physical buyers in Asia are ignoring the paper price and paying massive premiums to secure actual metal.
The paper market can be manipulated with margin hikes. The physical market cannot. This decoupling can't last forever. Either paper prices catch up to physical reality, or the paper market becomes irrelevant.
Gold-Silver Ratio and Where Price Could Go
The gold-silver ratio is the number of silver ounces required to buy one ounce of gold. The historical mineral ratio is roughly 15:1, and most monetary systems anchored around that level too.
In 2020 the ratio peaked above 125. It now sits at about 58. Silver has caught up substantially but is only at the long-run average. If the ratio reverts to its historical mean and gold holds at $4,400, the implied silver price is roughly $129.
How I'd Stack Exposure
This is a starting point for your own research, not a recommendation.
- Physical silver — If you believe physical pricing will eventually drive paper pricing, physical is priority one. American Silver Eagles and Canadian Maple Leafs offer the best liquidity. Bars make sense at higher dollar amounts. Store at home or in a reputable vault. Avoid bank safe-deposit boxes.
- Streaming companies — WPM, FNV, RGLD. They finance miners upfront in exchange for the right to buy future production at a fixed low price. Lower volatility, fixed-cost structure, much more resilient in drawdowns. They barely flinched on December 28 while retail-favored miners got destroyed.
- Tier-1 majors — Newmont (NEM), Barrick (GOLD), Agnico Eagle (AEM). All-in sustaining costs around $15/oz so they stay profitable through corrections. Avoid the single-project juniors that Reddit gets excited about.
- Dry powder — Paper-market manipulation will keep volatility high. Keep some cash to buy engineered dips.
FAQ
Q: Some argue the December crash was just profit-taking, not margin manipulation.
A: Both happened. Profit-taking pulled the trigger; the margin hike accelerated forced liquidation. The signal is the timing—two margin hikes at the all-time high during the thinnest liquidity week of the year. The exact same pattern showed up in 1980 and 2011.
Q: Can't I just buy SLV and call it done?
A: If your core thesis is that paper-market exposure is the risk, then SLV is paper. That said, it's easy to trade. I'd anchor core exposure in physical, use ETFs for trading, and layer streamers/miners for leverage.
Q: What if China lifts export restrictions?
A: Short-term shock is possible. But China only loosens if domestic demand drops or strategic reserves get full. Neither is visible right now. Even if they loosen, a five-year structural deficit doesn't vanish overnight.
Q: At a gold-silver ratio of 58, am I already late?
A: The 80/50 rule says buy above 80, sell below 50. At 58 we're near the bottom of the buy zone but still in it. Mean reversion + industrial demand + supply deficit + China restrictions all point the same direction.
December 28 was a shakeout. The fundamentals didn't shake.
More in this Category
Buy-and-Hold Is Dead: What -85% to -99% Crashes Tell Us About 2026
Buy-and-Hold Is Dead: What -85% to -99% Crashes Tell Us About 2026
Over the past 12 months PayPal fell 85%, Rivian 92%, Beyond Meat 99.7%. Here's why "buy a famous company and bury it" no longer works in 2026 — and what changed.
Wall Street Doesn't Hold Anymore: How to Follow the Money's Footprints
Wall Street Doesn't Hold Anymore: How to Follow the Money's Footprints
Institutions moving hundreds of billions of dollars can't hide. Here's how to read the footprints they leave in price action, volume, and sector ETF relative strength.
Single Stock vs Basket Investing: Which Survives 2026?
Single Stock vs Basket Investing: Which Survives 2026?
In the AI era, guessing which company wins is far harder than guessing which industry rises. Here's a side-by-side of basket investing vs single-name bets — and when each makes sense.
Next Posts
February 28, 2022: The Day the Dollar Got Switched Off, and the Sovereign Gold Rush That Followed
February 28, 2022: The Day the Dollar Got Switched Off, and the Sovereign Gold Rush That Followed
Since Russia's $300B reserves were frozen, central bank gold buying has run at roughly 60 tons a month versus 17 tons pre-2022. I trace the four-year shift that almost nobody is pricing in.
The $29 Trillion Gold Stock and Why New Supply Is Effectively Captured by Central Banks
The $29 Trillion Gold Stock and Why New Supply Is Effectively Captured by Central Banks
About $0.5T of new gold is mined each year, but central banks buying ~60 tons a month are absorbing nearly all of it. I unpack the price-discovery mechanics of an asset whose supply cannot scale.
Physical, ETFs, or Miners — How I Decide Which Form of Gold to Hold, and the Three Mistakes to Avoid First
Physical, ETFs, or Miners — How I Decide Which Form of Gold to Hold, and the Three Mistakes to Avoid First
A 10-15% allocation is usually a reasonable starting point, but the bigger question is the form you hold it in. I compare physical, ETFs, and gold miners across six dimensions.
Previous Posts
Trump's Grid Executive Order - $1.4 Trillion Just Started Moving
Trump's Grid Executive Order - $1.4 Trillion Just Started Moving
Trump's 'Strengthening the US Grid Reliability and Security' executive order pairs with $1.4 trillion in grid rebuild spending, funneling cash directly to infrastructure contractors who already sit on $44B backlogs.
Data Centers Are Leaving the Grid - Fuel Cells vs Small Nuclear
Data Centers Are Leaving the Grid - Fuel Cells vs Small Nuclear
Google, Meta, and Oracle aren't waiting 5-10 years for grid connections — they're building their own power plants. Here's how Bloom Energy's fuel cells stack up against Oklo's small modular reactors.
Grid Picks and Shovels - Six Names Across Build, Components, and Metals
Grid Picks and Shovels - Six Names Across Build, Components, and Metals
AI is the picks-and-shovels for data centers. The grid is the picks-and-shovels for AI. And grid contractors, components, and raw materials are the picks-and-shovels for the grid. Here are six names that map directly onto the $1.4T rebuild.