China Auto Sales in October: BYD Leads the Pack, Exports Become Growth Engine
nashnova research
October NEV retail hit roughly 1.07 million units at a 66.9% penetration rate, with BYD topping the chart at 490,000 units; as domestic growth slows, exports and autonomous driving are becoming the variables that define the next leg.
Has the market peaked? What's behind the slowdown?
CPCA estimates October passenger-vehicle retail at about 1.6 million units, down 28.9% year-on-year; NEV retail at roughly 1.07 million, penetration 66.9%.
This means → NEV penetration is closing in on a 70% ceiling. Each incremental point gets harder — a high base and a high penetration rate squeeze growth from both sides.
BYD sold 490,200 units in October, up 11% year-on-year, with a Jan–Oct total of 3.62 million. The leader still grows, but the industry's overall pace has clearly downshifted — divergence is widening.
Among the EV startups, who's pulling ahead?
Leapmotor delivered 106,100 units globally in October, up 51% year-on-year, marking four straight months above 100,000. Jan–Oct cumulative: 773,000 units, up 65.9%; its two-millionth vehicle is expected off the line in November.
XPeng delivered 50,200 units, a single-month record, with 13,326 overseas — up 166%. Xiaomi crossed 50,000 for the first time; its "Pengcheng" range-extender series, launched in September, contributed over 20,000 in its first full delivery month.
In plain terms = the startups have split into a "100k club" and a "50k tier." Leapmotor leads on both scale and growth rate, opening a clear gap.
How far along are the legacy automakers in going electric?
SAIC hit 258,600 NEV sales in October, up 25.12%, with a 60.14% NEV penetration rate. Geely sold 140,800 BEVs, up 43%; including PHEVs, penetration reached 69.3%.
Great Wall's NEV sales were 49,500 units, penetration 41.1% — all three legacy makers are now above 40%, well past the "half-hearted" phase.
This means → the line between legacy automakers and startups is blurring. The competition is no longer "ICE vs. EV" — it's whose electric product wins.
Why are exports being called the "growth engine"?
Jan–Sep auto exports reached 8.205 million units, up 65.8% year-on-year. NEV exports hit 4.006 million, up 1.3×, accounting for nearly 50%. In September alone, NEVs were 54.2% of exports — above half for four consecutive months.
October single-month exports: Chery led at 194,700, BYD at 170,200, Geely at 107,200 (up 158%). For the EU — the highest-certainty growth market — Jan–Sep shipments reached 1.83 million, already exceeding last year's full-year total; September alone was up 128%, and PHEV exports surged 372%.
In plain terms = the volume that can't be absorbed at home is being redirected overseas. But exports aren't free money — tariffs, FX, and localization costs erode margins on three fronts. Every unit lost domestically needs nearly two extra units overseas just to break even.
How are automakers navigating the tariff wall?
CKD — completely knocked down, shipping disassembled kits for local assembly — now accounts for 34.6% at Leapmotor, 41.9% at Great Wall, and 42.1% at SGMW. Local assembly has become the practical route around tariffs.
This reflects a shift from "exporting finished cars" to "exporting capacity." But CKD is a bridge, not a destination — the real watershed is whether localization rates and per-vehicle gross margins can turn positive.
This means → exports can rescue volume, but whether they rescue profit depends on who builds a complete overseas supply chain first and pushes localization costs down.
Autonomous driving and robots — real growth or just a new narrative?
2026 is being called the L3 mass-production year: the first L3 vehicles with dedicated plates are running commercial pilots in Beijing and Chongqing; two mandatory national standards for intelligent connected vehicles take effect on January 1, 2027. City NOA — urban navigation-assisted driving — and auto-parking have already trickled down to the ¥150k–200k mainstream segment.
Robotics is accelerating in parallel: BYD's "Yao-Shun-Yu" robot project has reached its seventh-generation prototype; XPeng's robotics arm closed a $900 million+ Series A in August at a post-money valuation above $6.3 billion; Tesla's Optimus has begun early-stage production at its Fremont plant.
In plain terms = smart driving has moved from "does it exist?" to "does it work well?", and robots have moved from slide decks to prototypes. Whether these two lines convert into profit is the key test for any re-rating of industry leaders.
市场有风险，内容仅供研究参考，不构成投资建议。