Unpacking China’s cheap battery costs

The number that prompted the episode is a cell price of roughly $50 per kilowatt-hour for lithium iron phosphate cells on China’s spot market, a level Frith’s own forecasts at BloombergNEF had placed in the 2030 to 2035 window and which he says arrived five years early. His first move is to discount it. The spot market is thin. Most volume is electric vehicle demand under long-term contracts, and it is stationary storage, buying on shorter terms, that is most exposed to spot pricing. Cell quality at the bottom of the market is uncertain, with more reports of tier-two, tier-three and B-grade product. And a developer in Europe or the US will not pay that price regardless. The figure is a low watermark that tells you something about the market rather than a price most buyers see, and distinguishing it from cost is the hinge of everything that follows.

The decomposition is the part to carry. For a lithium iron phosphate cell, roughly 70% of the cost is materials, about 20% is operating cost, and about 10% is plant and equipment capital. That split bounds the usual explanations before any comparison is made. Labor sits inside the 20%: about 17% of total cell cost in the US against about 11% in China, a genuine difference that has been shrinking for a decade as lines automate. Power is not very different depending on which state and which province you compare. Environmental rules are laxer in China, but a large exporter like CATL ends up meeting its customers’ environmental requirements even where local law does not demand it. Run those differences through BloombergNEF’s model for a 35 gigawatt-hour plant and the Chinese cell comes out near $68 per kilowatt-hour against about $75 in the US, roughly 8% higher. Capital is the one input with a large headline gap, with CATL’s 35 gigawatt-hour Guangdong plant at about $1.7 billion against Ford’s similarly sized licensed plant at $3.5 billion, but spread across a fifteen-year plant life and billions of cells that gap comes to something on the order of a percentage point per kilowatt-hour. The obvious causes, in other words, cannot carry the weight the popular story puts on them.

What is left is operations, and specifically yield and uptime. Scrapping a cell wastes the materials, the energy and the labor at once, and since materials are about 70% of cost, yield is the lever that moves the biggest term. China has been manufacturing cells since the early 2000s and has a workforce fifteen to twenty years deep in the process, which is how yields ramp quickly there. Equipment makers are also next door, so a breakdown is a phone call rather than a flight, and uptime holds. Frith’s counter-example is Northvolt, roughly two years into cell production and still struggling with yields by media accounts, with BMW reportedly citing cost and yield when it cancelled a contract. Kann draws out the implication, and it is the reversal worth keeping: if the Chinese advantage were cheap labor and weak regulation, the West could never match it, but an advantage made of accumulated skill and supplier proximity is at least the kind of problem a Western manufacturer can work on.

Subsidies get the same treatment, real but re-sized. They took the form of provinces handing over land free or cheap, electric vehicle purchase subsidies conditioned on Chinese-made cells with a share flowing back to the cell maker, and tax breaks, and Frith is explicit that they had a big impact. His qualification is about timing rather than magnitude: this was mostly a 2010 to 2020 story that built the champions now dominating the industry, and the bulk of China’s manufacturing capacity, about 90% of it, was commissioned in the four years from 2020 through 2024. The plants making cells at today’s prices were largely not built under the subsidy regime people invoke to explain those prices. Meanwhile the crash itself has causes that are not structural at all: a chronic slight overbuild of capacity, softening electric vehicle demand outside China, lithium below $20,000 a tonne after peaking above $60,000, destocking of expensive inventory, and manufacturers accepting zero or negative margins to hold market share, with some reportedly selling at the cost of raw materials. Which is why Frith expects the very lowest quoted prices to rise over the next year or two even as average prices keep falling on fundamentals, and why his prescription for the West is not to race China on standard cells but to pair Western intellectual property in new chemistries with Chinese capital and manufacturing know-how, an outlook he offers while acknowledging it may be naive.