Cells turned, materials did not
Battery cell makers entered a recovery cycle in 2025-26. Samsung SDI followed +75.7% in 2025 with +7.4% so far in 2026, Panasonic Energy accelerated from +36.2% to +107.0%, and CATL has logged three straight positive years since its +74.5% rebound in 2024, adding +39.4% in 2025 and +6.8% in 2026. As shown in EVs are not driving the battery cycle, ESS is, the driver is ESS and AI datacenter storage demand.
Yet cathodes, the cost tier directly above cells, have not followed. In 2026 EcoPro BM is at -49.3%, POSCO Future M -32.2%, L&F -24.7% and Ronbay -8.8%. Korean and Chinese alike, the whole cathode tier is negative. One step up from cells, the cycle breaks.
Why propagation breaks
First, customer concentration destroys bargaining power. L&F's top two customers are 100% of revenue; CATL alone was 54.5% of Ronbay's in 2024, with the top five at 79.1%; Samsung SDI and SK On together were 91.4% of EcoPro BM's through the third quarter of 2025. The first thing a recovering cell maker does is demand price cuts and adjust inventory, so cell profit recovery does not become materials profit recovery. Profit allocation between tiers is set by bargaining power, and a supplier with two customers has none.
Second, order headlines differ from revenue. POSCO Future M's cumulative cathode orders reach KRW 92tn, roughly KRW 52tn tied to LG Energy Solution and KRW 40tn to Samsung SDI, but only KRW 2.8tn of its KRW 13.8tn GM contract, about 20%, was actually executed. The EV chasm exposed the gap between nominal long-term contract values and actual shipments. EcoPro BM's KRW 43.9tn Samsung SDI contract and KRW 10.1tn SK On contract trade at the same discount: however large the order book, the customer's plant utilization sets the pace of execution.
Third, the ESS cycle runs on different chemistry. ESS and Chinese EVs run on LFP, while Korea's three cathode majors specialize in high-nickel NCM/NCA, which accounts for 95.7% of EcoPro BM's revenue. The demand driving the cell recovery simply does not use their flagship products. That is why L&F and POSCO Future M are racing into LFP conversion, and why the gap persists until that conversion completes.
The exception proves the rule
The only cathode name in the green is Umicore at +3.9%, for a paradoxical reason. Its +111.6% surge in 2025 had nothing to do with a battery recovery: the market rewarded the CORE strategy of shrinking battery investment and retreating to catalysis and recycling. Being better defended the further it sits from batteries suggests, in reverse, that the recovery has not yet reached the cathode tier.
China lived this cycle a step earlier. Ronbay collapsed -63.6% in 2023, more than a year before Korean cathode makers, and recovered earlier as a lagged follower of CATL's rebound. And in 2026, while CATL holds at +6.8%, Ronbay sits at -8.8%, unable to follow. A customer holding more than half of your revenue squeezes the margin out of any recovery, and the same cell-to-materials compression shows up in the Chinese chain.
Re-testing the thesis in mid-2026
The materials chain on this wiki has grown considerably since this piece was first written, so run the numbers again. In 2025 cathodes did bounce alongside cells: EcoPro BM +77.1%, POSCO Future M +56.5%, L&F +43.4%, Ronbay +47.7%. In 2026 the two tiers split again, and four of the five cathode nodes, all but Umicore, are negative.
The honest crack sits inside the cell tier itself. LG Energy Solution is at -19.3%, SK Innovation (SK On) at -6.6% and BYD at -12.5%, so the 2026 cell recovery is selective, favoring the names with ESS and AI storage exposure. Yet viewed as customer-supplier pairs, the thesis gets sharper. Below Samsung SDI (+7.4%) sits EcoPro BM at -49.3%; below LG Energy Solution (-19.3%) sits L&F at -24.7%; below CATL (+6.8%) sits Ronbay at -8.8%. In every pair the cathode trades below its own customer. However much or little the cell recovers, the material above it captures less than that.
Verification conclusion
The idea that gains propagate along the value chain holds in the battery chain only down to cells and breaks at materials. Three variables cut the propagation: customer concentration, contract execution rates, and chemistry mismatch. The next checkpoints are the tiers above, precursors and lithium (EcoPro Materials, POSCO Holdings, Albemarle, SQM). Whether the broken cycle never reached the raw-material tier at all, or moves on lithium's own separate price cycle, is a question whose test, now that the raw-material tier is on the map, runs in At the materials tier, the customer sets the clock.