The Argument

Your friend says the silver deficit is a paper artifact — strip out ETFs and speculative demand, and the physical market is balanced. You say it's a genuine structural supply shortage that puts a hard floor under prices.

The data says you're both partially right, but you're more right on the timeline that matters.

The Numbers: Let's Kill the Ambiguity

The World Silver Survey 2025 (published by Metals Focus for the Silver Institute) provides the definitive supply/demand balance. This is the report your friend likely read. Here's the 2024 data from the balance table (p.9):

Supply (2024): 1,015.1 Moz

Demand (2024): 1,164.1 Moz

2024 Market Deficit: 148.9 Moz

Note: ETF/ETP inflows of ≈62 Moz are tracked separately from the physical balance in the Survey methodology but represent additional demand on the physical market.

Your Friend's Argument: The Ex-Investment Calculation

Here's the math your friend is implicitly doing, and it's not wrong:

Non-investment demand (industrial + jewelry + silverware + photography): 680.5 + 208.7 + 54.2 + 25.5 = 968.9 Moz

Total supply: 1,015.1 Moz

Balance ex-investment: +46.2 Moz surplus

Strip out coins, bars, and ETFs entirely — pretend no one buys silver as an investment — and supply exceeds fabrication demand by 46 million ounces. Your friend's reading of the data is arithmetically correct for 2024.

This is the strongest version of his argument. The deficit IS driven by investment demand. If every ETF liquidated tomorrow and coin/bar demand went to zero, the physical fabrication market would be in modest surplus.

Why Your Friend Is Still Wrong About What Matters

The ex-investment surplus of 46 Moz sounds comfortable until you examine three things:

1. Investment demand doesn't just disappear — it's structurally embedded

Your friend's thought experiment assumes investment demand can go to zero. It can't. Here's why:

Coins and bars (190.9 Moz in 2024) represent physical metal removed from circulation. In 2024, this was actually a five-year low — down 22% year-over-year — and the market was still in 149 Moz deficit. The 2024 coin/bar figure was depressed by:

• US investment demand crashing 46%
• Double-digit declines across Western markets

Even at this cyclical trough in physical investment, the deficit was massive. In 2025, coin/bar demand is forecast to recover to 204.4 Moz. ETF inflows have surged to 187 Moz through mid-2025.

The realistic floor for investment demand isn't zero — it's roughly 150-180 Moz (coins, bars, and baseline ETF holdings). That alone consumes most of the 46 Moz "ex-investment surplus" and then some.

2. The 46 Moz surplus is razor-thin relative to the market

Total demand is 1.16 billion ounces. The ex-investment "surplus" is 46 Moz — that's 4% of total demand. A single year of 5% growth in industrial demand (about what we've been seeing) adds 34 Moz of consumption, nearly eliminating the cushion.

For context:

• Industrial demand grew from 508.2 Moz (2020) to 680.5 Moz (2024) — a 34% increase in four years
• Photovoltaic demand alone was 197.6 Moz in 2024
• Electronics & electrical demand (including PV) hit 460.5 Moz

The ex-investment surplus is one bad year away from vanishing on industrial demand alone.

3. The cumulative deficit has already consumed the inventory buffer

This is the structural argument that your friend's static analysis misses.

The Silver Institute reports the silver market has run consecutive annual deficits since 2021, with a cumulative shortfall of approximately 580 Moz through 2024 — equivalent to roughly 8.5 months of mine supply. Including 2025's estimated deficit (≈118 Moz), the five-year cumulative draw approaches 700 Moz. This metal has been pulled from above-ground inventories. The evidence:

COMEX silver: Total warehouse stocks fell from ≈600 Moz (2020) to ≈382 Moz. Of that, only ≈113 Moz is registered (deliverable against futures contracts). In the first week of January 2026 alone, 33.45 Moz was withdrawn — 26% of registered inventory in seven days.

Silver lease rates: Exploded to ≈8% in January 2026, versus historical norms of <1%. This is the physical market screaming tightness.

Above-ground inventory coverage: Depleted from ≈22 months of demand coverage (2020) to ≈12.3 months (2024). The 2011-2020 silver depression occurred because there were 22 months of buffer inventory absorbing deficits without price pressure. That buffer is now roughly half consumed.

Your friend's argument works in a single-year snapshot. It fails when you account for the cumulative drawdown that has already occurred.

The Supply Side: Genuinely Constrained

The Byproduct Problem

This is the single most important structural fact about silver supply: approximately 72% of mined silver is a byproduct of copper, lead, zinc, and gold mining.

From the World Silver Survey 2025, primary silver mines produced roughly 227 Moz — about 28% of total mine supply. The remaining 593 Moz came from base metal and gold mines where silver is an incidental byproduct.

What this means: Silver mine supply does not respond meaningfully to higher silver prices. A copper miner's decision to expand production is driven by copper prices, not silver prices. When silver goes from $20 to $50, SCCO doesn't suddenly mine more copper to get more silver. The silver just comes along for the ride.

From SCCO's 10-K (2025): Silver produced 20,984 thousand ounces in 2024, up 14% — but this was driven by higher copper throughput, not silver price response. SCCO specifically notes silver represented only "5.1% of our sales." Their capex decisions are driven by copper, not silver.

BHP's Q4 2025 call confirmed the same dynamic: they tout being a "sizable gold, silver, uranium producer" but only in the context of "significant byproduct credits" — silver is a cost offset, not a strategic product.

Mine Production Has Peaked and Plateaued

Global silver mine production:

• 2016 peak: ≈900 Moz
• 2024: 819.7 Moz
• 2025E: 835.0 Moz

Despite silver prices roughly doubling from 2020 ($20) to 2024 ($28 average), mine production is still 7% below 2016 levels. The modest 2% increase expected in 2025 comes mainly from Mexico (Juanicipio ramp-up) and Australian lead/zinc mines.

From Hecla's Q3 2025 earnings call, CEO Robert Krcmarov stated directly: the silver market "faces fifth consecutive year supply shortages rising industrial demand." The CEO of the largest US silver producer is telling you — on a recorded earnings call — that supply is short.

Where Is New Supply Coming From?

The top primary silver miners are growing, but mostly through M&A — consolidating existing ounces, not discovering new ones:

• Hecla (HL): Greens Creek produced 8.5 Moz (2024), down from 9.7 Moz (2023). Lucky Friday at ≈1.2 Moz/quarter. Keno Hill still pre-commercial, targeting 345-385 tpd by 2027. Total company: 16.2 Moz (2024).
• First Majestic (AG): 3.7 Moz silver in Q2 2025 (up 76% YoY on Gatos integration). Guiding 30-32 Moz silver equivalent for 2025. But CEO Neumeyer noted production is "limited man hours" — capacity-constrained.
• Pan American (PAAS): Raised silver guidance to 22-22.5 Moz after MAG Silver acquisition (Juanicipio). La Colorada Skarn project won't produce until late decade.
• Coeur Mining (CDE): Guiding 18+ Moz silver for 2025 — a 62% increase over 2024. But nearly all of this growth comes from the SilverCrest acquisition (Las Chispas), not organic mine development. Las Chispas produces ≈1.5 Moz/quarter of high-grade silver.

These are the world's top primary silver miners. Combined, they produce roughly 70-80 Moz — less than 10% of global supply. And the recent production growth (CDE +62%, AG +76%) is almost entirely M&A-driven: existing ounces changing corporate hands, not new ounces entering the market. The global supply number barely moves.

New mine development timeline: 10-15 years from discovery to production. There is no supply cavalry coming in the next 2-3 years.

The Bear Case: What Could Prove Your Friend Right

In fairness, here's what could rebalance the market without prices staying elevated:

Solar Thrifting Is Real and Accelerating

This is the strongest structural bear argument. Solar panels are the fastest-growing silver demand category, consuming 197.6 Moz (2024). But:

• Copper-cored silver paste has evolved from >50% silver (Q2 2023) to 15% silver (Q2 2025) to 10% silver formulations entering production (Q3 2025)
• Silver's share of solar panel cost jumped from 5% (2023) to 14% (2025), creating intense economic pressure to substitute
• BNEF forecasts solar silver demand dropping to 194 Moz (-7% YoY) in 2025
• First-ever decline in global solar installations forecast for 2026 (649 GW, led by China -14%)

If thrifting continues at this pace AND installations plateau, solar silver demand could drop 30-40% over 3-5 years. That's 60-80 Moz of demand destruction — more than enough to eliminate the ex-investment deficit.

Pandora Substitution: The Jewelry Signal

Pandora, the world's largest jewelry brand, announced PANDORA EVERSHINE — platinum-plated jewelry replacing ≈80% of their silver-based assortment. Timeline: 30% in 2027, 20% in 2028, remaining 20% thereafter. Consumer testing showed platinum perceived as more premium.

Pandora's silver purchases are a meaningful fraction of jewelry demand. If other brands follow, jewelry demand (208.7 Moz) could erode by 10-20% over 5 years. That's 20-40 Moz.

Recycling Is Rising

Recycling hit a 12-year high of 193.9 Moz (2024), up 6% YoY. At sustained high prices, industrial scrap recovery and silverware recycling increase. But recycling still only covers ≈17% of total demand and has a relatively low capture rate (≈25-30% of end-of-life silver).

The Verdict: You're Right on the 2-Year Timeframe

Here's the synthesis:

Your friend is technically correct that: The marginal silver deficit in any given year is driven by investment demand. Strip out coins, bars, and ETFs, and the fabrication market shows a small surplus (≈46 Moz in 2024).

Your friend is wrong about the implication because:

1. Investment demand has a structural floor of ≈150-180 Moz (it doesn't go to zero). Even the 2024 five-year-low of 191 Moz coins/bars was enough to create a 149 Moz deficit.

2. The cumulative deficit (≈580 Moz through 2024, approaching 700 Moz through 2025) has depleted the inventory buffer that historically absorbed shortfalls. COMEX registered down to 113 Moz, lease rates at 8%, coverage from 22 months to 12 months.

3. Industrial demand is growing at 5-7% annually, consuming the thin ex-investment surplus year by year. Industrial demand alone is now 83% of mine supply — leaving only 135 Moz of mine production for ALL other uses.

4. Supply cannot respond because 72% is byproduct. Primary miner "growth" is mostly M&A-driven consolidation, not new ounces entering the market.

On a 2-year horizon (your timeframe), you are right. The inventory drawdown has created a physical tightness that wasn't present in prior cycles. Even if investment demand pulled back to 2024's five-year-low levels, the market remains in deficit. The floor under silver prices is genuine and structural.

On a 5-year horizon, your friend's concerns gain weight. Solar thrifting, Pandora-style substitution, and potential recycling expansion could collectively destroy 100-150 Moz of demand — enough to flip the ex-investment balance into meaningful surplus. This is the regime change scenario where investment demand truly becomes the marginal driver rather than industrial tightness.

The Number That Settles It

If you want one number to end the argument, it's this:

Industrial silver demand in 2024: 680.5 Moz. Total mine production: 819.7 Moz.

Industrial demand alone consumes 83% of all silver mined on Earth. That leaves 139 Moz of mine supply for jewelry (209 Moz), silverware (54 Moz), photography (26 Moz), and investment (191+ Moz). The physical market needs ≈480 Moz from mine supply that doesn't exist.

Recycling (194 Moz) fills part of the gap but is nowhere near enough. The deficit is REAL, it's PHYSICAL, and it's been consuming inventories for five consecutive years.

Your friend is right that if investment demand evaporated, the bleed would slow to a trickle. But the bleed has already happened — nearly 600 Moz of inventory is gone through 2024, COMEX registered is depleted to 113 Moz, lease rates are screaming, and there's no supply response coming.

The house has a small leak in the roof. Your friend is arguing about whether the next rainstorm will make it worse. You're pointing out the basement is already flooded.

You're both right. You're more right.