Editorial Article
Four Times the Market, Half the Material Share
Four point oh three times. That is what the EV battery market is forecast to do between 2025 and 2033 — USD 91.7 billion to USD 369.5 billion, a 19.03 per cent baseline CAGR, 21.53 per cent optimistic, 16.53 per cent conservative. It is the largest number in this entire chain of reports and the one most often quoted second-hand. It is also the number that tells you least, because it is a value of a product whose unit price falls every year.
That is not a quibble. It is the whole analytical problem with the parent market, and it has a consequence most material suppliers have not priced.
A four-fold value claim is a much larger volume claim
If the price per kilowatt-hour of a finished pack were flat, USD 91.7 billion becoming USD 369.5 billion would mean 4.03 times the gigawatt-hours. It is not flat. Cells fell 20 per cent in 2024 alone to about USD 115/kWh, and the direction of travel in this industry is not in dispute.
Take a deliberately modest deflation rate — 4 per cent a year on realised pack price, one fifth of the 2024 cell decline. Over eight years that multiplies price by 0.7214. Divide it out and the market’s 4.03 times of value implies about 5.59 times the volume.
So the headline is simultaneously a 4× revenue statement and a 5.6× tonnage statement, and the gap between them — around 39 per cent of the implied unit growth — is the part that gets absorbed by deflation. Every materials supplier reading the 19.03 per cent as their own growth line is over-reading it by a factor of one and a half, before any question of market share.
And the biggest material inside it grows at less than half the container’s rate
This is where the arithmetic gets uncomfortable. The LFP cathode active material market is USD 8.30 billion in 2025, rising to USD 15.99 billion by 2033 — an 8.54 per cent CAGR. Set that against the 19.03 per cent the battery market claims and a pattern appears:
| LFP cathode as a share of EV battery market value | |
| 2025: USD 8.30bn of USD 91.70bn | 9.05% |
| 2033: USD 15.99bn of USD 369.50bn | 4.33% |
| Change | −4.72 percentage points |
The cathode market roughly doubles. The market that contains it roughly quadruples. The share halves.
That is not because cathode is losing a technology race — LFP is the chemistry gaining share in exactly the period covered. It is because value is migrating up the chain, into cells, modules, packs, thermal systems, electronics and the compliance apparatus around them, while the tonne of powder at the bottom behaves like what it is: a manufactured commodity with a documented deflation rate. The same asymmetry shows up in every material report in this chain. Purified wet-process phosphoric acid deflates 2.48 per cent a year in revenue per tonne; sodium-ion electrolyte 9.60 per cent a year; sodium-ion active materials 7.42 per cent a year. Four independent models, four separate deflation lines, one direction.
The peak year is a factory list
Now the part of the forecast that is easy to miss because it is written as a supporting sentence rather than a warning. The acceleration in this model runs from 18.27 per cent in 2027 to a peak of 23.67 per cent in 2029, then decelerates. The stated basis for that inflection is capacity: announced expansions from Samsung SDI, Volkswagen, LG-Honda, Umicore and Albemarle, with most facilities starting production between 2024 and 2027, “aligning with the modeled acceleration in 2027–2029.”
Read that again as a matter of evidence rather than as a citation. The shape of the forecast — its steepest section — is not derived from demand. It is derived from a list of plants and the dates they are scheduled to start. Those two things are not the same, and the difference is the single most common error in battery forecasting: announced capacity is the cheapest document in the industry to produce, and it is not a shipment.
The same mislabelling shows up closer to the ground. In the sodium-ion materials work, two published figures for 2025 output differ by 5.8 times — 8.1 GWh against 47.2 GWh — and the resolution is not a choice between institutions but an arithmetic test: at 2,450 tonnes of cathode per gigawatt-hour, 47.2 GWh requires 115,640 tonnes of cathode, which is very nearly the 120,000 tonnes the same compilation reports for 2026. The figure and its own source belong to different years. Capacity, shipment and installation are three different numbers wearing the same unit.
Four chokepoints the model does not price
The conservative scenario at 16.53 per cent names its own reasons: upstream supply concentration, Chinese equipment dependency, and regulatory compliance hurdles. Those are stated qualitatively in the forecast and they deserve to be quantified, because three of the four are dated.
Cobalt. The Democratic Republic of Congo produces about 70 per cent of global supply. Its regulator ARECOMS imposed an annual export cap of 96,600 tonnes for 2026 and 2027. In July 2026 an administrative failure on the customs platform threatened up to 20,000 tonnes worth USD 1.1 billion, with 60 to 75 per cent of companies judged unlikely to meet the quota deadline. Cobalt has risen about 160 per cent since February 2025 to roughly USD 26/lb — USD 57,320 a tonne. A 2026 multilayer shock-propagation study found cobalt demand is inelastic in the short term while processing is highly concentrated, which is the formal way of saying a local administrative failure propagates globally and abruptly.
Graphite. China controls more than 90 per cent of spherical and synthetic anode material. A December 2023 export-licensing regime sharply cut spherical graphite shipments in early 2024, and by 2025 less than 1 per cent of uncoated spherical graphite was produced outside China. That is not a concentration statistic. It is a single point of failure with a licensing switch on it.
Lithium refining. Mining is spread across Australia, South America and Africa; conversion is not. China holds roughly 60 to 70 per cent of lithium chemical refining capacity. The mine is not the chokepoint. The plant that turns the mine’s output into a specification is.
Nickel. Indonesia went from 31.5 per cent of global supply in 2020 to about 60 per cent in 2024 after a raw-ore export ban pulled in Chinese-backed refining investment — and the state is now tightening control of the sector that investment created. A jurisdiction that can create a supply chain in four years can reprice it in one.
February 2027 sits inside the peak
The EU Battery Regulation’s battery passport obligation lands in February 2027, and recycled-content targets begin in 2031. The forecast’s own note is that these add fixed costs not factored into the growth rates.
Where they land matters more than that they exist. February 2027 is the year the acceleration begins; 2031 is the year after the window closes. So the compliance cost arrives during the steepest part of the curve, and the recycled-content requirement arrives after the model has already stopped looking. A fixed cost per tonne of documentation, traceability and audit is not a large number for a cathode producer shipping a hundred thousand tonnes. It is an existential number for the marginal one — which is precisely the mechanism by which the cathode share in the table above gets squeezed from the bottom rather than competed away from the top.
What would change this
One: if the pack price proves stickier than 4 per cent a year of decline — say deflation slows to 1 or 2 per cent as cells become a smaller share of a larger, more integrated pack — then the volume implied by 4.03 times of value falls, the material suppliers’ tonnage problem shrinks, and the whole deflation argument weakens. That is a real possibility and it is checkable against transacted pack prices, not list prices.
Two: if cathode’s share of battery value stops falling in the published accounts of the majors over the next two reporting cycles, then the migration of value upward is a forecast artefact rather than a structural fact, and the material layer is a better place to be than this analysis says.
Until one of those breaks, the parent report should be read for what it is: a very large number describing a container, growing at a rate driven partly by a factory schedule, holding materials that are individually cheaper every year. It is the denominator of this entire chain. It is not a growth rate you can borrow — and the four reports underneath it are where the actual economics sit.
This analysis draws on the EV battery market report for the container, and on the LFP cathode active material, purified wet-process phosphoric acid, sodium-ion electrolyte and sodium-ion materials work for the contents.
Frequently Asked Questions
How large is the global EV battery market?
USD 91.7 billion in 2025 to USD 369.5 billion by 2033 — a 19.03% baseline CAGR, with an optimistic case of 21.53% and a conservative case of 16.53%. Growth is front-loaded, peaking at about 23.67% in 2029 before decelerating.
Why does LFP cathode grow slower than the EV battery market?
LFP cathode active material is USD 8.30 billion in 2025 and USD 15.99 billion in 2033, an 8.54% CAGR. That is less than half the 19.03% the battery market claims, so cathode's share of EV battery value falls from 9.05% to 4.33% — the cathode market roughly doubles while the market containing it roughly quadruples.
What drives the forecast's 2027–2029 acceleration?
Announced capacity expansions from Samsung SDI, Volkswagen, LG-Honda, Umicore and Albemarle, with most facilities starting production between 2024 and 2027, are the stated basis for the modelled acceleration in 2027–2029. Capacity announcements are not shipments — the same conflation produces the 47.2 GWh versus 8.1 GWh disagreement in 2025 sodium-ion output.
Which upstream chokepoints matter most?
Cobalt: the DRC supplies about 70%, with an ARECOMS cap of 96,600 tonnes for 2026–27 and 20,000 tonnes worth USD 1.1 billion at risk in July 2026. Graphite: China holds more than 90% of spherical and synthetic anode output, with under 1% of uncoated spherical graphite made outside China by 2025. Lithium refining: China holds 60–70%. Nickel: Indonesia went from 31.5% in 2020 to about 60% in 2024.