US-China AI BATTLE hinges on rare earths, critical minerals

Clear Street analyst Tim Moore discusses the U.S.-China AI race and competition for rare earths and critical minerals, highlighting strategic pressure points in technology and supply chains. The piece is commentary rather than a market-moving development, with no specific financial figures, policy changes, or corporate actions disclosed.

Analysis

The market is underpricing how quickly the AI race can migrate from software bragging rights to hard-input bottlenecks. If the strategic narrative shifts from model performance to deployment resilience, the marginal dollar rotates toward firms with secure access to power, advanced packaging, semicap tools, and non-China mineral supply chains rather than pure application-layer winners. That is a second-order negative for asset-light AI beneficiaries and a relative positive for the industrial and materials ecosystem that can de-risk supply continuity.

The more important implication is policy latency: governments can subsidize fabs and data centers faster than they can replace concentrated upstream mineral processing. That creates a multi-quarter window where “domestic AI capacity” sounds bullish but actually amplifies dependence on constrained inputs, tighter capex budgets, and longer lead times. In that regime, suppliers with diversified sourcing or Western processing capacity should see pricing power, while Chinese-linked suppliers face higher geopolitical discount rates even before any formal export restriction.

Consensus is likely too focused on headline AI competition and not enough on the embedded commodities squeeze. A trade war over critical minerals would not just raise costs; it would also compress the ROI of AI infrastructure by making buildouts more expensive and less predictable, which can cap multiple expansion for the broader AI complex. The biggest contrarian risk is that the market treats this as a slow-burn theme, when in reality procurement decisions for batteries, magnets, wafers, and data center equipment reprice within weeks once supply risk enters the conversation.

Key Decisions for Investors

• Go long a basket of non-China critical-mineral beneficiaries (for example MP, TECK, FCX) versus short China-exposed industrial metals names on a 3-6 month horizon; the thesis is pricing power and policy support as supply-chain risk gets re-rated.
• Pair long AMAT or LRCX against short a higher-multiple AI application basket over 6-12 months; if AI capex stays intact but build costs rise, picks-and-shovels should outperform software-only beneficiaries with weaker pricing power.
• Buy out-of-the-money calls on utility/data-center power names with constrained capacity over 3-9 months; the hidden winner in an AI arms race is often electricity and grid infrastructure, not the model layer.
• Use any rally in pure-play AI momentum names to reduce exposure or hedge with put spreads for 1-3 months; if critical-mineral headlines intensify, these names can de-rate quickly on margin pressure and capex deferral risk.
• Consider a long FCX/short a China-linked industrial proxy pair for tactical exposure to supply-chain bifurcation; best entry is on a dip after the next geopolitical headline rather than chasing strength.