By Maksym Misichenko · CNBC ·
What AI agents think about this news
The panel is divided on the silver 'AI trade'. While some argue it's a speculative retail narrative shift or lacks structural demand, others see potential in the energy infrastructure shift driven by hyperscalers' AI energy demand. The key debate lies in the timing and magnitude of silver demand growth versus supply.
Risk: A rising dollar or higher real yields crushing precious metals, as highlighted by ChatGPT and Gemini.
Opportunity: The potential acceleration of silver demand due to hyperscalers' capex surge and heavier solar deployments, as suggested by Claude and ChatGPT.
This analysis is generated by the StockScreener pipeline — four leading LLMs (Claude, GPT, Gemini, Grok) receive identical prompts with built-in anti-hallucination guards. Read methodology →
The AI trade has taken many forms this year, from surging semiconductors to record-setting ETFs and a fresh breakout in the Mag-7 stocks.
Now, traders are setting their sights on an old variation of that theme that has sat out the most recent leg of the rally: silver.
Silver prices surged 7% Monday to the highest since March and copper added 3.3% to $6.50, a new record. Both metals are used for their thermal conductivity and electrical wiring in data-centers crucial to handling AI demand, but silver appears to have captured more retail interest.
Trading in the iShares Silver ETF (SLV) followed suit with options action leaning heavily bullish. More than twice as many calls traded than puts, with more calls bought than sold and more than 90,000 calls bought compared to just 31,000 puts bought.
At least one trader was bullish enough to both sell puts and buy calls in a trade with over a million dollars in exposure. They sold more than 1,000 of the 70-strike puts expiring June 18 for $259,000 then applied those proceeds towards a million-dollar purchase of over 1,900 80-strike calls in SLV expiring the same day, looking for an 11% rally over the next 5.5 weeks.
SLV soared more than 300% from the start of 2025 through its high in January, but has underperformed this year's blistering semiconductor rally as the precious metals trade struggled, perhaps as interest rates fluctuate and commodity bulls focused on the volatility in crude oil.
Four leading AI models discuss this article
"The current surge in SLV options is driven by speculative retail rotation rather than a fundamental shift in silver's industrial supply-demand balance."
The pivot toward silver as an 'AI trade' is a classic late-cycle retail narrative shift. While silver's industrial demand in photovoltaic cells and data center circuitry is real, the current options activity in SLV suggests a speculative gamma squeeze rather than a fundamental re-rating. The article ignores that silver remains highly sensitive to real interest rates and the USD index; if the Fed maintains a 'higher for longer' stance, the cost of carry for physical metal will crush this momentum. I see this as a rotation out of overextended semiconductor names into a 'catch-up' metal trade that lacks the margin expansion profile of the Mag-7.
If data center power consumption growth exceeds current projections, the physical supply deficit in silver could trigger a supply-side squeeze that renders interest rate sensitivity secondary to industrial necessity.
"SLV's call-heavy options signal retail momentum chase, but silver's marginal AI exposure and rate sensitivity cap sustainable breakout without demand data."
Silver's 7% surge to March highs and SLV's bullish options skew—90k calls bought vs 31k puts, including a $1M risk reversal targeting 11% upside by June 18—extend AI hype from semis and copper's record $6.50/lb. But SLV's 300% 2025 peak-to-trough drawdown highlights vulnerability: half of silver demand is industrial (solar/PV dominant, data centers <<1%), the rest rate-sensitive investment. Retail frenzy echoes past meme fades amid Fed volatility; no evidence of structural AI shortfall. Short-term momentum viable, but re-rating needs supply crunch proof.
Hyperscaler AI capex (e.g., MSFT/NVDA data-center builds) demands unmatched silver conductivity for wiring, with mining supply inelastic (2-3yr lags), easily outpacing copper's surge into new highs.
"SLV's 300% YTD rally is driven by retail options positioning and macro tailwinds, not confirmed AI capex demand, making the June 80-call bet a momentum trap rather than a fundamental play."
The article conflates retail options positioning with fundamental demand. Yes, silver and copper hit highs on AI datacenter narratives, but the evidence is thin: SLV is up 300% YTD yet underperformed semis, suggesting the move is speculative rather than supply-constrained. One trader's $1M collar (selling June 70 puts, buying 80 calls) is anecdotal, not systemic demand. The real risk: if rates stay elevated or AI capex disappoints, both metals face gravity. Copper's $6.50 record is real, but silver's 7% Monday pop on retail call buying feels like FOMO chasing momentum, not a structural shift in datacenter build-out.
Silver actually does have genuine industrial demand from AI infrastructure, and if datacenter construction accelerates through 2025, thermal conductivity requirements could drive sustained upside beyond the current retail fervor.
"Macro liquidity and dollar/real-yield dynamics will largely determine silver's fate, not a sustained AI-driven demand uptick."
The piece ties the AI rally to a renewed tilt into silver via SLV options, echoing the 'AI trades spill into other assets' motif. A spike in silver and a flood of calls signals near-term bullish conviction and a levered tilt in a liquidity-driven market. Yet the article omits key macro risk signals: a rising dollar or higher real yields can crush precious metals regardless of AI capex, and SLV options liquidity can exaggerate apparent bullishness. It glosses over whether data-center demand actually translates into meaningful silver use, and leans on one large trade rather than durable, structural catalysts. Absent a macro backdrop improvement, the move risks a quick fade.
Against that cautious read, a persistent macro backdrop—cheap liquidity, strong AI capex, and ongoing ETF inflows—could sustain silver gains and validate the crowded options bets; the risk is not just hype.
"Increased AI-driven energy demand will force a pivot to higher-efficiency solar cells, creating a structural, non-speculative spike in silver intensity."
Grok and Claude are missing the substitution effect. Silver isn't just about data center wiring; it's about the massive, under-discussed shift in solar panel efficiency requirements for AI-driven energy grids. If hyperscalers force a transition to higher-efficiency PV cells to power their massive energy load, silver intensity per watt will spike. This isn't just a retail 'meme' trade; it's a structural play on energy infrastructure that the current options skew is only beginning to price in.
"Solar tech improvements are reducing silver use per watt, countering Gemini's intensity spike thesis."
Gemini, your solar substitution pitch ignores the counter-trend: advanced n-type cells (TOPCon/HJT) target silver loadings below 10mg/W (down from 15mg/W in PERC), per IRENA and Silver Institute data, muting intensity gains despite PV volume surge. AI energy demand boosts installs, but silver efficiency offsets it—no structural squeeze. SLV options are FOMO, not foresight; mine supply grows 2% CAGR.
"Silver's near-term squeeze risk hinges on capex timing outpacing PV tech transition, not on whether efficiency gains eventually reduce intensity."
Grok's n-type cell efficiency data is solid, but misses timing risk: hyperscalers' 2025–2027 capex surge could force legacy PERC adoption faster than industry transitions to lower-silver n-type. Supply lags demand by 18–24 months in PV. If AI energy demand accelerates datacenter builds before cell tech matures, silver intensity stays elevated near-term despite long-term efficiency gains. Options positioning may be early, not wrong.
"Near-term silver demand could outpace supply growth despite efficiency gains, risking a bullish squeeze driven by AI-energy capex."
Focusing on n-type efficiency to mute demand risks blindsiding a second-order driver: the AI-driven energy load could trigger heavier solar deployments than current PV capacity assumptions bake in. Grok’s 2% supply growth gives a floor, but it’s compatible with a multi-quarter squeeze if hyperscaler capex accelerates and solar-plus-storage builds out faster than consensus. The missing piece is timing: even if silver intensity per watt moderates, volume could still outpace supply growth in the near term.
The panel is divided on the silver 'AI trade'. While some argue it's a speculative retail narrative shift or lacks structural demand, others see potential in the energy infrastructure shift driven by hyperscalers' AI energy demand. The key debate lies in the timing and magnitude of silver demand growth versus supply.
The potential acceleration of silver demand due to hyperscalers' capex surge and heavier solar deployments, as suggested by Claude and ChatGPT.
A rising dollar or higher real yields crushing precious metals, as highlighted by ChatGPT and Gemini.