In the August auto market report card, the most telling indicator may not be any monthly sales blockbuster hitting 100,000 units, but rather the disappointment of a flagship model.
The Li Auto L9 delivered 4,158 units in August. That figure alone is not bad, but it becomes glaring when viewed over a longer period: it delivered 6,106 units in June, dropping by a third in just two months; cumulative deliveries in the first eight months of this year came to less than 20,000 units, only about half of the same period last year. As the earliest flagship in China to make large-battery range-extended technology a mainstream route, the L9's 72.7 kWh ternary battery represents one of the most respectable application scenarios for ternary cathode materials in the passenger car market. Its temporary slowdown is not just a product cycle issue for Li Auto alone.
In the same month, two other layers of the industry chain were also undergoing developments with deeper implications for ternary. On one side, battery leader CATL was reported to have seen a decline in ternary power battery orders for the coming months, and the industry-level reduction trend is already clear; on the other side, automakers were lining up to "de-CATL-ize," with Xiaomi's Pengcheng series switching entirely to battery cells from Sunwoda and CALB, and Li Auto increasing its stake in Sunwoda to become the second-largest shareholder. Taken together, end-use demand, battery configuration, and supply chain maneuvering all repriced ternary in that month.
August auto market: robust overall volume, growth that bypassed ternary
The CPCA recently released its official August monthly report, and the headline numbers were not bad. Wholesale volume of passenger NEVs reached 1.51 million units, up 16.4% YoY; retail penetration rate hit 65.2%, another record high; exports reached 518,000 units, up 154.7% YoY. The wholesale growth rate was 26.5 percentage points higher than retail growth, meaning the portion that could not be sold domestically was fully absorbed by exports.
But for those working with ternary, the "ternary content" of that growth in August declined.
By powertrain route, wholesale battery electric vehicles reached 1.032 million units, up 25.9% YoY, accounting for 68.3% of NEVs; narrow plug-in hybrids reached 386,000 units, up 3.1%; range-extended vehicles were only 93,000 units, down 10.9% YoY, with their share shrinking to 6.1%. Range-extended was the only one of the three major routes to decline, a fact that has been repeatedly confirmed over the past few months, and August merely reinforced it once again. Range-extended is precisely where large ternary batteries are most concentrated: mainstay models such as the Li Auto L series and the AITO M8 and M9 are basically equipped with ternary packs starting at around 50 kWh, and the new-generation flagships have already been configured to 70 kWh or more. As range-extended volume shrinks, ternary directly loses blood.
Exports have doubled every month this year. In August, battery electric vehicles accounted for 63.4% of NEV exports, and 56.5% of those battery electric exports were A00 and A0 class small cars, compared with only 45.2% in the same period last year. In other words, the surge in exports is largely driven by small LFP vehicles. On the domestic sales side, the best-performing BEV segment is the A0 class, up 12.2% YoY.
These structural shifts, when transmitted to the battery cell level, show up in the data: in August, SMM ternary power battery cell production rose only 3.6% MoM, while LFP power battery cell production rose 14% MoM. LFP's share of power batteries has climbed from 66% in the same period last year to 76%. The growth structure is systematically bypassing ternary.
CATL's ternary reduction: what the market sees is auxiliary materials, but the real cause lies at the end-user level
After news of CATL's ternary order cuts spread, the market quickly settled on auxiliary material shortages as the explanation: since Q3, copper foil supply has been tight, battery makers' output has been constrained, limited materials have been prioritized for energy storage and commercial vehicles, and ternary production schedules have had to be pushed back. However, the auxiliary material shortage treats both technology routes equally, yet the squeezed share is specifically ternary. This direction is not determined by auxiliary materials, but by demand.
The weakness in ternary battery cells reflects the fact that sales of high-end models such as extended-range vehicles have fallen short of expectations. Li Auto L9's cumulative sales in the first eight months were halved YoY; AITO M8 dropped from monthly sales of over 20,000 units last year to around 2,000-3,000 units this year, and the facelift switched entirely to LFP batteries across the lineup; AITO M7 declined MoM, and Seres' August wholesale sales were halved YoY; Avatr and IM, the two brands with the highest ternary share, fell 23.5% and 6.2% YoY respectively in August. These models use exactly the large ternary batteries, and with collective volume declines at the model level, the ternary power battery order base itself is shrinking.
Even if auxiliary material supply were ample, ternary power battery production schedules would be difficult to sustain. Ternary had ample stockpiling in the earlier period, and battery cell manufacturers have already been proactively slowing the pace to digest inventories; CATL's overall Q3 business is not bad either, only the portion allocated to ternary is weak. The auxiliary material shortage brought this round of cuts to the surface ahead of time, conveniently giving the market a ready-made explanation. If the cuts were merely an auxiliary material issue, orders should be replenished after the supply chain recovers; but if it is a demand issue, it will not reverse automatically. This round of cuts is the result of transmission from end-user models all the way to battery cell production schedules, and it will not end just because auxiliary material supply eases, nor will it reverse because of a recovery in the auto market in any given month. The real turning point will have to wait for high-end models to regain volume.
De-CATL-ization: automakers' profit defense
Automakers are collectively trimming their battery supplier lists.
The two most symbolic moves both came in the past week: the Xiaomi Pengcheng series launched on September 7 uses Sunwoda and CALB battery cells across the lineup, with CATL completely absent; three days ago, Li Auto announced a 2.65 billion yuan capital increase in SEVB, taking an 11.17% stake to become the second-largest shareholder, while the all-new L8 launched in June has already switched entirely to Sunwoda battery cells. Going further back, NIO ended its collaboration with CATL in the solid-state battery field and shifted to in-house development, the main XPeng MONA versions use CALB, and some versions of the AITO M6 have brought in Gotion High-tech.
Automakers actually have just two reasons.
The first is profit. In H1 this year, CATL's net profit attributable to the parent was 43.284 billion yuan, while 15 mainstream automakers combined only made just over 21 billion yuan in the same period. One battery maker's profit was twice the total of the entire vehicle industry. For a 200,000-yuan car, the battery cost accounts for 30% to 40%. Automakers are busy with R&D, plant construction, and the entire marketing chain, yet they only earn 3,000 yuan per car, with an industry average net margin of 1.5%. Under this structure, second-tier battery cells are 10% to 15% cheaper than CATL's, and switching to another supplier for volume models can save 2,000 to 3,000 yuan per vehicle.
The second is bargaining power. Lithium carbonate has more than doubled from its low point this year, and cost pressure is already significant. Starting September 1, a 2% consumption tax is levied on lithium batteries, rising to 4% in 2027, while batteries produced and used in-house can avoid or offset the tax. The policy essentially points automakers toward in-house development. A more direct stimulus is delivery security. Orders for the new-generation Li Auto MEGA exceeded expectations, but CATL's 5C ternary lithium battery reserves are about to run out, and customers who lock in orders later can only switch to in-house batteries. Whether the flagship model has enough batteries and which month delivery will be scheduled is not up to the automaker itself.
However, this trend should not be overestimated. High-end flagship models above 300,000 yuan still cannot do without the technical endorsement of the Qilin battery, and automakers are still paying the technology tax they have to pay. And in the year when the calls were loudest, CATL's installation share in China's passenger vehicle market in Q1 actually rebounded to 50.1%, returning to half the market for the first time in five years.
Ternary is an option for specific scenarios
The dilemma of ternary has never been about disappearing, but about becoming a choice for specific scenarios. The substitution by LFP is compressing the applicable boundary of ternary toward the high end, and supply chain migration is amplifying uncertainty on the client side. Support still exists. High-end models such as the AITO M9, NIO ES8, and Zeekr are holding up the base, the increase in vehicle battery capacity is partially offsetting the decline in volume, and mid-to-large vehicles in exports are also providing marginal growth.
Several indicators are worth watching going forward: first, when the extended-range market will stabilize, such as the pace of Xiaomi Pengcheng's volume ramp-up and the success of the Li Auto L9 and AITO M8 model replacements; second, whether the export structure can upgrade from small cars to mid-to-large vehicles; and third, how quickly Sunwoda, CALB, and others can ramp up their own supply chains after taking over ternary orders.
SMM Ternary Analyst Wang Yizhou
