Silver crashed 55 percent from its January record, landed exactly on the .618, and has turned up 8 percent in a week.
Silver hit an all-time high above $121 on January 29, then lost 38 percent in a single day when a hawkish Fed chair nomination triggered a wave of forced long liquidation. Peak to trough it fell about 55 percent, down to roughly $55.
That $55 low is not random. Drawn from the 2022 base near $14 up to the January peak, the .618 retracement sits at $55.06. Price tagged it and turned. Silver is back at $62.45, up 13.5 percent off the low and up 8.06 percent in the last week alone.
My view is the full move back up runs to at least $80, which is the .382 retracement at $80.44 and about 29 percent above here. The first test on the way is the .50 at $67.75. Worth knowing: the futures forward curve independently prices silver near $78 to $80 by 2030.
THIS IS NOT FINANCIAL ADVICE. The $55 support, the $80 target and every fib level in this piece are my own analysis and my own opinion, nothing more. I am not a financial advisor. Silver fell 38 percent in a single day this year, so treat every number here as one trader's read and do your own research before you risk a dollar. NFA · DYOR.
Every band on this chart is a fib level. Price stopped dead on one of them.
What a .618 actually is, without the mysticism.
After anything runs up hard and then falls, traders want to know how far the fall goes before buyers step back in. Fibonacci retracements are just fixed percentages of the original move, and the .618 means the price gave back 61.8 percent of everything it gained. It is not magic. It works partly because a very large number of traders and algorithms are watching the same line and placing orders there, which makes it real whether or not you believe the maths.
The .618 is the deepest level a trend can retrace and still be considered intact. Below it, the move is usually treated as broken rather than resting. So $55 was the line where silver either held its four-year uptrend or lost it. It held, and it held on the first touch. That is the difference between a pullback and a top.
Where the levels sit and what each one is worth.
| Fib level | Price | From $62.45 | What it is |
|---|---|---|---|
| Peak | $121.50 | +94.6% | Jan 29 all-time high |
| .236 | $96.13 | +53.9% | Full recovery territory |
| .382 | $80.44 | +28.8% | My target for the move |
| .50 | $67.75 | +8.5% | First real test on the way up |
| .618 | $55.06 | −11.8% | The low. Tagged and held. |
| .786 | $37.00 | −40.8% | Where the thesis is dead |
The crash was positioning, not demand. That distinction is the whole trade.
Silver did not fall 38 percent in a day because the world stopped needing it. It fell because a hawkish Fed chair nomination hit a market that was crowded with leveraged longs after a parabolic run, and the exits were too small for everyone. Forced liquidation does not care about supply and demand.
Six straight years short. The shortfall adds up to an entire year of world mine supply.
Between 2017 and 2020 the silver market ran a surplus of about 132 million ounces. Since 2021 it has been in deficit every single year, and the cumulative shortfall now stands at 762 million ounces. To put that in scale, that is roughly equal to one full year of global mine production that the world consumed without replacing.
| Year | Balance (Moz) | What happened |
|---|---|---|
| 2017–2020 | +132.0 total | Four surplus years, the market was oversupplied |
| 2021 | −83.7 | First deficit as industrial demand reset higher |
| 2022 | −254.0 | The worst year on record |
| 2023 | −200.1 | Second largest shortfall |
| 2024 | −137.9 | Deficit narrowing but still severe |
| 2025 | −40.3 | Sharp narrowing |
| 2026F | −46.3 | Sixth consecutive deficit year, forecast |
| 2021–2026F | −762.3 | About one year of global mine output |
Read that table down the right side. The deficit peaked in 2022 at 254 million ounces and has shrunk every year since, to roughly 46 million now. That is about 82 percent smaller than the peak. The shortfall is real and it is six years long, but it is narrowing, not widening, and that is exactly what you would expect after prices spiked and manufacturers started designing silver out. Anyone selling you the deficit story without that sentence is selling you half a chart.
Most silver is not mined on purpose. Roughly 70 percent comes out of the ground as a byproduct of copper, lead and zinc operations. That means the decision to produce silver is made by a copper company responding to the copper price, not by anyone responding to silver. A higher silver price does not reliably produce more silver, which is why these deficits persist rather than self-correct the way they would in a normal commodity.
Every AI server has silver in it. Nobody was modelling that five years ago.
Silver is the most electrically and thermally conductive metal there is, which is why it ends up in the places where an AI rack cannot afford resistance or heat: connectors, relay contacts, circuit traces, switchgear, power distribution and thermal management. As racks get denser and power-hungrier, the silver content per unit goes up rather than down.
180 grams is 5.8 troy ounces, so a 50,000 server build consumes about 290,000 ounces before you add any networking, cooling or power gear. On its own that is small. The number that counts is the aggregate: data centre silver demand of 20 to 40 million ounces a year is between 43 and 86 percent of the entire 46 million ounce deficit forecast for 2026. AI is not a rounding error in this market any more. It is roughly the size of the shortfall itself, and it is growing 15 to 25 percent a year.
| Source | Annual demand | Note |
|---|---|---|
| Total industrial | ~700–750 Moz | The bulk of all silver consumption |
| Electronics and electrical | 465.6 Moz in 2024 | A record; includes circuit boards, connectors, 5G |
| Solar PV | 200+ Moz | Every panel uses silver paste to carry current |
| AI data centres | 20–40 Moz | The newest pillar, growing 15–25% a year |
Unlike gold, which is hoarded and recycled almost indefinitely, a large share of industrial silver is consumed and never comes back. It is spread in microscopic quantities across billions of devices at concentrations too low to economically recover. Every year of deficit permanently removes metal from the system rather than moving it between vaults.
Three ways to own the bounce, and they are not equally levered.
| Ticker | Price | Today | YTD | 1 Year | Fee |
|---|---|---|---|---|---|
| SLV · the metal | $56.07 | +4.14% | −12.96% | +65.01% | 0.50% |
| SIL · miners | $83.77 | +6.59% | +0.44% | +71.88% | 0.65% |
| SILJ · juniors | $27.75 | +6.28% | +0.29% | +87.36% | 0.69% |
Look at the year to date column. The metal is down 13 percent while both miner funds are flat to slightly positive. The miners have already decoupled and started pricing a recovery that the metal has not delivered yet. On today's move they ran at roughly 1.5 times SLV, which is the leverage working in the right direction.
Neither miner fund is a pure silver bet. SIL is 22 percent Wheaton Precious Metals, a royalty company, which damps the very operating leverage most people buy SIL for. And SILJ, despite the junior label, holds Wheaton, Franco-Nevada, Royal Gold, Pan American, plus KGHM and Boliden, which are copper and base metal businesses. A large slice of the junior fund is neither junior nor silver.
What kills this trade.
Supporting the bounce
- The .618 held on first touch and price is 13.5 percent off the low with a +8.06 percent week behind it.
- Miners are already leading, flat year to date against a metal down 13 percent. That divergence usually resolves in the metal's favour.
- The forward curve prices $78 to $80 by 2030, so the futures market independently agrees with the destination even if it disagrees on timing.
- Sixth consecutive supply deficit with byproduct mining unable to respond quickly to price.
Against it
- A 38 percent single-day crash leaves scars. Leveraged longs who were carried out do not come back quickly, and that thins the bid on rallies.
- Demand destruction is real. Manufacturers who engineered silver out at $121 do not re-engineer it back in at $62.
- Still down 12.4 percent year to date and 20.5 percent over three months. One good week does not reverse that.
- Silver is a Fed trade as much as a metal trade. The same catalyst that broke it in January still exists, and rate expectations can move it faster than any deficit.
The level that invalidates this is $55. A clean break and close below the .618 means the four-year trend is broken rather than resting, and the next shelf is the .786 down at $37. That is the risk you are taking to play for $80, and it is worth roughly 12 percent against 29 percent in favour.
Johnny's read.
Want the entries, stops, and sizing?
Join the Discord to find out! →Silver futures chart and performance figures from the author's own TradingView terminal, COMEX SI front month SIU2026, Aug 5, 2026 · SLV, SIL and SILJ prices, returns, expense ratios and holdings via stockanalysis.com, Aug 5, 2026 · silver spot pricing cross-checked against JM Bullion and Trading Economics, Aug 5, 2026 · supply deficit and industrial demand data from The Silver Institute World Silver Survey 2026 · January crash reporting via Investing News Network and Carbon Credits. Compiled Aug 5, 2026.