A few nights ago, I walked past an electric car charging in the underground lot of my apartment building. Between the dark pillars, a single green light was blinking quietly. The car was asleep while the vessel inside it stayed awake, receiving drop by drop the electricity a power plant had sent. In the last journeys I saw the places where that electricity is made. This time it is the chain of the vessel that holds it. The battery. Running an electric car, storing the electricity an AI data center will use at night — in the end, both are this vessel's job.
At making battery cells, Korea stands near the top of the world. LG Energy Solution, Samsung SDI, SK On. Though even that standing is not what it was: pressed by China's giants, the three companies' share shrinks a little each year. And when you trace this chain upward, you watch the place where the value pools drift off in an unexpected direction. Making a finished battery well and making serious money from it, it turns out, are different stories.
Finished Cells Have Few Customers
Start with the finished battery. The "paradox of the small wall" we saw in semiconductor bonding repeats itself here. Korea's big three are world-class, but their customers are, in the end, a handful of automakers. With so few customers, however good the cell, you cannot name the price you like. On top of that, with the mass adoption of electric cars running later than expected — the so-called chasm — the whole chain has been catching its breath for several years now.
As an aside: the word chasm, I'm told, comes originally from geology. A deep split in the earth — terrain where the far side is in plain view and you still cannot walk across. For a name attached to this chain it is excessively accurate, and when I first looked up the meaning I couldn't help laughing a little.
One Layer Down — Cathode Materials
So we climb one more layer up the chain. The key material that accounts for a large share of a battery's cost: cathode materials. Here stand Korean companies like EcoPro BM, POSCO Future M, and L&F. But stop at this station, and one of the chain's sad truths comes into view. The value does not pool here either. It leaks further upstream.
To make cathode materials you need minerals like lithium and nickel, and the upstream that mines and refines them is overwhelmingly in China's grip. In particular, the cheap lithium iron phosphate family, where demand is now exploding, is one where China effectively sets the world's standard. Korea's companies, in the middle of the chain, make exquisite materials well — but they can name neither the price of their inputs nor the price of the finished product. From above, Chinese minerals press down on the price; from below, a handful of automakers. It is the fate of the one caught in the middle of the chain.
The Courtyard Is Moving
Even so, a new courtyard is opening on this chain. While electric cars stall, demand is growing fast for the large storage installations — ESS — that hold the electricity AI data centers will use. And as America moves to shut China out of battery materials, a space has opened for Korean-made lithium iron phosphate cathode materials to wedge into. That is why a company like L&F is pouring hundreds of billions of won into building a plant for this new cathode material. On a chain where value leaks away, the companies are hurrying to move toward where the courtyard is going.
What this chain teaches is a footnote to the thesis of Part 3. Value pools at the narrowest link, I said — but that link is not necessarily inside your own country. On this battery chain, the narrowest and most fearsome link is the Chinese factories that refine lithium. Trace the chain up far enough and you arrive, in the end, at the very bottom of everything: minerals.
The People Who Don't Leave
| Company | Employees | Average pay | Average tenure | Revenue per employee | Operating profit per employee |
|---|
| LG Energy Solution (cells) | 12,922 | about 112 million won | 8.2 years | about 1.83 billion won | about 100 million won |
| EcoPro BM (cathode materials) | 1,192+ | about 61 million won | 4.7 years | about 2.1 billion won | about 120 million won |
| L&F (cathode materials) | 1,924 | about 60.4 million won | 4.3 years | about 1.12 billion won | about -81 million won (loss) |
* Per-employee figures are rough values: FY2025 consolidated results divided by disclosed headcount. LG Energy Solution's pay reflects bonuses from the profit-surge years. EcoPro BM's headcount is Korea-only, excluding overseas subsidiaries, so its per-employee values may be somewhat inflated. For L&F, the operating loss (-156.8 billion won) is divided by headcount as is.
What catches the eye in this table is neither the pay nor the revenue but the tenure. 4.3 years, 4.7 years. After the fifteens and seventeens of the earlier chains, you see at a glance how young this industry is. And L&F's 1,924 people are each carrying a loss of some 80 million won on their backs — a loss larger than their own pay. Yet they do not leave; they stay, and draw up the blueprints for the new cathode materials plant. They are holding out through the time of the chasm on the time of people who believe there is a far side. If the timetable of an industry whose wall is finished reads fifteen years of tenure, this is what the timetable of an industry still building its wall looks like. On the day this young tenure turns double-digit, will the chain's wall stand finished along with it? I cannot know. But if, even then, the value is still leaking upstream, this table will remain a somewhat forlorn record.
There are places, though, where it is not the price but the number of customers that decides a chain's fate. The next journey goes to a chain that for a long time had exactly one customer. That one customer was the government, and so it was the government that set the price — defense.
This piece is part of the "Korea Value Chain" series and is an original work. It is an analysis based on public data and industry sources; some figures and assessments are estimates. It is not intended as a direct basis for investment decisions.
By Alexandro Lee · July 2026
댓글
댓글 쓰기