Bank of America: Want AI Exposure? Own Resource Stocks
Bank of America's strategy team says the AI trade isn't just about software and chips. The heavy buildout of data centers is pulling hard on energy and raw materials-especially copper. For investors, resource equities may offer cheaper, more diversified exposure than crowded mega-cap tech.
AI's growing appetite for copper
"Artificial intelligence devours commodities," wrote BofA's team led by Michael Hartnett. Data centers require vast amounts of power and wiring; copper is essential across power cables, transformers, servers, EVs, and renewables-everything feeding the AI stack.
BloombergNEF projects a 6 million-ton global copper shortfall by 2035 as demand outruns supply. AI-linked copper use is set to average 400,000 metric tons per year over the next decade, peaking at 572,000 tons in 2028, with cumulative data center demand topping 4.3 million tons. According to Bloomberg, copper can account for nearly 6% of a data center project's capital expenditure.
Supply strain and price outlook
BNEF sees copper supply reaching 29 million tons by 2035-still behind forecast demand. With tighter balances ahead, copper prices could reach $13,500 per ton by 2028. That backdrop favors miners and related suppliers as margins expand on stronger pricing.
How to position
- Prioritize copper-focused miners and diversified miners with meaningful copper exposure.
- Consider royalty/streaming companies for lower operating risk and commodity-linked upside.
- Use broad resources or copper-focused ETFs to avoid single-asset risk.
- Complement core tech holdings with a measured allocation to resources for balance and valuation support.
What to watch
- Data center build rates, grid interconnection backlogs, and utility capex plans.
- Copper inventories across LME/COMEX/SHFE and treatment/refining charges (TCRCs) as signals of tightness.
- Project pipelines, permitting timelines, and disruptions at major mines.
- Policy incentives for domestic supply, recycling capacity, and substitution risk (e.g., aluminum in certain applications).
Risks to stress-test
- Commodity price volatility and USD strength.
- Delays, cost overruns, or geopolitical issues at key assets.
- Weaker global growth or China property softness affecting base metals demand.
- Faster-than-expected recycling or efficiency gains that soften demand growth.
Bottom line
AI infrastructure is material-intensive, and copper sits at the center. Pairing core tech with resource equities can create a more balanced AI exposure at more reasonable valuations. Size positions prudently and add on weakness to manage cyclicality.
For broader context on metals and the energy transition, see BloombergNEF's coverage. Finance teams exploring practical AI adoption can review curated AI tools for finance to enhance research and workflow.
