Why Cobalt's Price Barely Reacts to Cobalt Demand
Most cobalt supply exists because someone was mining copper or nickel. That is why cobalt output responds to copper and nickel prices more than its own.
Cobalt is the clearest case on this site of a metal whose supply economics are not really its own, because most cobalt exists as a by-product, covered on this material’s geology page: its production volume is substantially decided by decisions made in the copper and nickel markets, not the cobalt market.
Putting a number on the concentration
“Highly concentrated” is usually said about cobalt supply as a qualitative label. It does not have to stay qualitative. Economists measure market concentration with the Herfindahl-Hirschman Index, the same tool competition regulators use to decide whether a merger creates too much concentration in an industry:
Running this on cobalt’s real 2024 mine production shares (DR Congo 76%, Indonesia 9.7%, Russia 3.0%, Canada 1.6%, Philippines 1.3%, Australia 1.2%, Cuba 1.2%, all other producers combined 6.1%) gives:
For reference, US antitrust merger guidelines treat a market above 2,500 as “highly concentrated.” Cobalt’s mine-supply HHI is more than double that threshold, using real, already-published production shares, not a hypothetical. The “all other producers” bucket is treated here as a single ~6% block for simplicity, which if anything understates true concentration slightly, since it is really several smaller producers, not one.
Why a copper price crash can shrink cobalt supply
When copper or nickel prices fall enough to threaten a mine’s economics, operators can scale back or idle production regardless of what cobalt is doing, since cobalt was never the primary reason the mine was economic in the first place. This means cobalt supply can genuinely tighten even while cobalt demand and price are rising, an unusual dynamic that a metal mined for itself would not show, and one that makes cobalt supply forecasting dependent on modelling copper and nickel market conditions, not cobalt’s own market alone.
A genuinely volatile price history
Cobalt has experienced some of the sharpest price swings of any critical raw material in recent history: a dramatic price spike around 2017 to 2018 driven by battery demand expectations, followed by an equally sharp decline as new supply, much of it from expanded DRC copper-cobalt operations, came online faster than demand grew. One simple way to quantify how much a price series actually moves around its own average is the coefficient of variation:
Applied to this site’s own published 2020-2024 annual average cobalt price series ($31,300, $51,100, $63,600, $34,100, $26,500 per tonne), the coefficient of variation works out to roughly 34%, a large number for a five-point annual series and a quantitative signal of exactly the boom-bust pattern described above. This is a real, recurring feature of by-product commodity economics: because cobalt supply cannot respond quickly and precisely to cobalt-specific price signals, the market tends to overshoot in both directions rather than settle smoothly.
Demand risk isn’t uniform either: the LFP substitution
Supply concentration is only half the resilience picture. Demand for cobalt is being reshaped by a chemistry choice happening entirely on the demand side, and it is not happening evenly around the world. LFP (lithium iron phosphate) cathodes use no cobalt and no nickel at all, and by 2025 LFP had overtaken nickel-based chemistries (NMC and NCA, both covered on the application-wise page above) to account for roughly half of global EV battery deployment.
The effect on cobalt demand is not symmetric across regions. A market moving toward LFP is a market quietly opting out of the DRC-concentrated supply risk quantified above, since a cobalt-free cathode carries none of that exposure. A market still overwhelmingly on NMC/NCA carries the full weight of it. This is exactly the kind of regionally uneven exposure a country-level or global-average resilience model would miss if it treated “cobalt demand” as one number rather than a demand pool splitting apart by region in real time.
Sources: by-product supply dynamics and historical price volatility are widely documented in commodity market analysis (Cobalt Institute, Benchmark Mineral Intelligence); HHI and CV are computed directly from this project’s own published production and price data; the 2,500 HHI threshold is the US Department of Justice/FTC Horizontal Merger Guidelines convention; LFP adoption shares are widely reported in battery chemistry market tracking (SNE Research, Benchmark Mineral Intelligence, IEA Global EV Outlook).