Copper Is the AI Trade No One Priced In—3 Miners With the Most to Gain

COMEX copper futures reached a record above $6.70/lb in August, supported by projected structural demand from data centers, grid investment, EVs and defense against constrained mine supply and ore grades that have declined roughly 40% since 1991. Copper backwardation has steepened as metal is diverted into U.S. warehouses ahead of possible tariffs, signaling near-term physical tightness outside the U.S. Freeport-McMoRan's first-half net income rose 65% year over year, while Hudbay posted record trailing-12-month adjusted EBITDA; however, tariff-policy uncertainty and execution risk at smaller producers could reverse the trade if spot-to-December spreads normalize.

Analysis

The investable distinction is between a copper price driven by genuine ex-U.S. deliverable scarcity and one driven by tariff-related inventory relocation. FCX has the best near-term earnings torque because its producing asset base converts sustained higher realized prices into cash flow without requiring project execution; HBM offers higher beta but adds Peru/Canada jurisdiction and single-asset concentration risk. SCCO is a less attractive pure expression despite scale: its valuation and Peru/Mexico political exposure leave less room for multiple expansion if copper merely holds current levels rather than accelerating.

Over the next 1-3 months, the key signal is not outright COMEX price but the spot-to-forward spread and regional physical premiums. A normalization of backwardation after a tariff decision would likely compress miners faster than metal, particularly FCX and HBM, because recent equity moves embed a scarcity narrative and elevated operating leverage. Conversely, persistent tightness after policy clarity would validate a more durable deficit thesis and force upward revisions to 2027-28 copper decks, where current sell-side assumptions are likely still materially below spot.

The underappreciated second-order loser is the electrical-equipment supply chain: transformer, cable and grid contractors face working-capital and fixed-price-contract risk before they can pass through higher copper costs. AI infrastructure spending is therefore not a clean one-for-one copper demand catalyst for NVDA; higher physical-layer costs can slow marginal data-center projects even while GPU demand remains strong. At 6-18 months, new domestic processing incentives could benefit U.S.-linked producers, but they also risk redirecting capital toward lower-return projects, making asset quality and permitting status more important than broad copper beta.

Consensus appears too willing to annualize the tightness signal. Warehouse dislocation can produce backwardation without equivalent end-demand acceleration, while a weaker global industrial cycle would hit copper consumption before long-cycle supply responds. The bullish thesis is falsified by a sustained collapse in backwardation alongside declining Chinese import premiums, or by FCX/HBM guiding to cost inflation or volume shortfalls that absorb the metal-price benefit.

Key Decisions for Investors

• Initiate a 3-6 month long FCX / short SCCO pair, sized dollar-neutral: FCX offers superior near-term cash-flow sensitivity and lower political-risk discount, while SCCO is vulnerable to valuation compression if the physical tightness premium fades. Reassess if the pair underperforms by 10% or if SCCO delivers materially stronger production guidance.
• Add HBM only on copper-price or broad-risk pullbacks rather than chasing strength; use a 6-12 month horizon and smaller sizing than FCX given higher operating and jurisdictional volatility. Target a 2:1 upside/downside profile, with a stop tied to a sustained break in copper backwardation rather than a single futures-price print.
• Use COPX for diversified exposure if physical-market confirmation persists through the next tariff-policy decision; it reduces idiosyncratic mine, permitting and ramp risk versus concentrated FCX/HBM positions. Do not increase exposure solely on higher COMEX prices absent continued spot-premium strength.
• Avoid treating NVDA as a copper-adjacent long. Monitor electrical-equipment and grid-contractor margins for evidence that copper inflation is delaying data-center power interconnects; this would be a second-order capex timing risk to the AI infrastructure complex over 6-18 months.
• Set an alert for spot-versus-December copper spread normalization and Chinese physical-premium weakness. If both occur, reduce copper-miner exposure promptly: the likely first move would be multiple compression in FCX/HBM before reported earnings estimates reset.