The 2026 battery-metals divergence

Four materials all sold to battery makers. Four wildly different 2026 price stories. The word "battery metals" is a category label that hides more than it reveals — the divergence between them has strategic implications.

The four paths, one category

These four share essentially one demand driver (EV battery production + grid storage). Yet their price behaviour, supply structures, and framework positions are all different.

Why the divergence

Same demand, different chemistry response

The four materials respond differently to the LFP vs NCM chemistry mix shift that's the biggest structural force in batteries:

• Lithium: BOTH LFP and NCM use it. LFP's rise (20% → 55%

of global EV battery chemistry 2020-2024) is carbonate- heavy; NCM is hydroxide-heavy. Net demand: strongly positive either way. Lithium is chemistry-agnostic.

• Cobalt: LFP uses ZERO cobalt. NCM uses progressively

less (8-1-1, 9-0.5-0.5 chemistries). Net demand per EV: falling. Cobalt total demand flat-to-negative through 2030.

• Nickel: NCM + NCA heavy; LFP zero. **Net demand per

EV: falling** as LFP share rises. But nickel's non- battery demand (~70% stainless steel) dominates; battery downside is modest vs total.

• Graphite: ANODE (not cathode). Every EV battery uses

graphite anode regardless of cathode chemistry. LFP anodes even use slightly more graphite than NCM. Net demand: strongly positive regardless of chemistry.

So LFP share rising is:

• Lithium: neutral-to-positive
• Cobalt: strongly negative
• Nickel: modestly negative (small share of nickel's total demand)
• Graphite: strongly positive

Same category, different supply discipline

Layer the framework's discipline regime question on top of demand:

• Lithium: concentrated supply (Chile+Australia+China ~85%

mining, China ~65% refining). Discipline = cost-curve (marginal producer economics) + Chilean state (SQM lease). Disciplined enough to support rally.

• Cobalt: extremely concentrated supply (DRC >50%).

Discipline = byproduct nature (nickel/copper economics determine cobalt output). Some supply discipline but structural demand is the bigger story.

• Nickel: concentrated supply (Indonesia 50%+). Discipline

= Indonesian state quota policy 2024+. Strong + explicit.

• Graphite: concentrated supply (China >75%). Discipline

= Chinese export-control policy Dec 2024. Strong + explicit.

Result: same demand driver, 4 framework cells

From docs/minerals/FRAMEWORK.md:

• Lithium: Structural bull (cyclical recovery phase)
• Cobalt: Asymmetric tail-risk (declining demand +

concentrated supply = skewed return distribution)

• Nickel: Structural bull (via state discipline)
• Graphite: Policy-premium bull (via state discipline

+ structural demand)

Four framework cells, one demand sector. This is the strongest case in the atlas against treating "battery metals" as a monolithic theme.

What this means for investors

If you believe in the EV growth story:

• Lithium + graphite are the cleanest exposures (both

benefit from LFP + NCM)

• Nickel is a secondary beneficiary (primary driver is

stainless steel, not EVs)

• Cobalt is NOT an EV beta — its demand is being

disintermediated by LFP chemistry

If you believe in the chemistry migration to LFP:

• Long lithium carbonate specifically (not hydroxide)
• Long graphite (EV volumes still rise; anode share per

EV stable)

• Short cobalt on pure demand math (offset by DRC supply

events)

• Underweight nickel from the battery side (non-battery

demand carries the material)

If you believe in Chinese export-control escalation:

• Long graphite producers (Syrah, Anovion, Novonix —

non-Chinese SPG capacity)

• Neutral cobalt (DRC Chinese-linked mines partially

affected but not controlled)

• Long lithium non-Chinese refiners (Albemarle, Arcadium)

If you believe in commodity super-cycle:

All four benefit, but lithium and graphite are the purer plays because they have no substitution-out risk from chemistry migration.

What could break which narrative

• Lithium: a major new Australian or Chilean mine

ramping faster than expected; a breakthrough in sodium-ion batteries (commercial Na-ion displaces LFP at low end)

• Cobalt: full LFP penetration in EVs (currently ~55%;

~75%+ would crush per-unit demand further); major Chinese battery chemistry innovation

• Nickel: Indonesian quota discipline breaks; LFP share

in premium EVs grows faster than expected

• Graphite: silicon-anode commercialisation (Sila, Group14,

OneD Battery Sciences; currently tiny but growing); fundamental substitution risk on a 5-10 year horizon

Silicon-anode as the long-horizon graphite threat

One material not yet in the atlas deserves a mention here: silicon-anode technology would displace some graphite from EV batteries in the 2028-2035 horizon. Silicon has 10x the energy density of graphite as an anode material, but swells during charging (~300% vs graphite's ~10%), which has kept it out of commercial batteries at scale.

Companies addressing this (Sila, Group14, Amprius, OneD) are at pilot-commercial scale. If one commercialises at cost parity with graphite, graphite anode demand growth flatlines after 2030. Worth watching as a kill-criterion for the graphite structural-bull thesis.

Cross-references

• docs/minerals/materials/lithium.md
• docs/minerals/materials/cobalt.md
• docs/minerals/materials/nickel.md
• docs/minerals/materials/graphite.md
• docs/minerals/FRAMEWORK.md — four cells illustrated
• docs/minerals/ANALYSES/chinese-export-control-playbook.md

— graphite's policy context

• docs/minerals/ANALYSES/ai-capex-materials-cluster.md —

orthogonal demand cluster (AI capex ≠ battery demand)