S07.3 · Materials, Chemicals & Mining
A $410B critical-minerals market where AI-optimized battery chemistry like LFP is destroying demand for the highest-margin cobalt and nickel inputs.
Battery and critical minerals — lithium, cobalt, nickel, graphite, rare earths and related inputs — is a $409.7B market in 2025, forecast to reach $669.8B by 2032 (6.1% CAGR, DataM Intelligence); the IEA's Global Critical Minerals Outlook 2025 has demand more than doubling since 2017. Two things are true at once: structural growth of 6-10%+ a year, and price cycles violent enough to erase the growth story for stretches — lithium fell more than 80% from its 2022 peak through 2023-25 on oversupply. The fact that organizes the whole segment is China's midstream chokehold, 60-90% control of refining capacity across lithium, cobalt, graphite and rare earths. The bottleneck, and the margin, is processing, not extraction.
Upstream is concentrated by geology: the Democratic Republic of Congo supplies roughly 70% of cobalt, Indonesia more than 50% of nickel, and Australia, Chile and China dominate lithium. But extraction itself is fragmented across many operators. It is midstream refining — converting ore and brine into battery-grade material — where China's dominance is structural rather than incidental, and where the highest margins in the value chain sit. The chain runs from upstream mining and brine extraction through midstream refining into downstream cathode and battery manufacturing.
The segment prices in two regimes, and conflating them misreads margin durability. Ore and concentrate trade as a commodity — exchange-linked and cyclical, base-metals behavior. Battery-grade refined material behaves like a specialty product, governed by multi-year customer qualification cycles and offtake premiums rather than spot pricing.
The largest disruption is not happening in mining or refining at all. It is happening in cell chemistry. AI-optimized battery chemistry — above all the shift toward lithium-iron-phosphate (LFP) and sodium-ion formulations that reduce or eliminate cobalt and nickel content per kilowatt-hour — is directly destroying demand for precisely the highest-margin minerals in the chain. LFP already accounts for more than 40% of global EV battery deployment; this is underway, not a scenario. The exposed parties are cobalt- and nickel-rich incumbent suppliers, whose reserves and refining investments were sized to a chemistry mix that AI-accelerated materials discovery is now moving away from.
The commercial consequence is that battery growth is decoupling from the fixed mineral-intensity ratios the industry planned around a decade ago. A "designer battery chemistry" category is forming as AI materials-discovery platforms let battery makers optimize chemistry for cost, supply security and performance simultaneously, rather than accepting whatever chemistry a given mineral supply chain could support. For investors and operators in upstream cobalt and nickel, demand forecasts anchored to historical EV-battery mineral intensity are increasingly unreliable; for lithium and battery-grade refiners able to serve multiple chemistries, the same shift reduces single-chemistry concentration risk. A separate dynamic runs alongside: automakers and battery manufacturers have taken direct positions in upstream supply to secure volume against China's midstream concentration — a supply-security answer to processing risk, not a chemistry-driven shift.