Related articles
TMTPOST -- The American depositary receipts (ADRs) of Xpeng Inc. jumped 13% to $22.25 on Wednesday, outperforming the U.S. stock market as the benchmark S&P 500 index fell 1.6%. Xpeng shares logged their best day since March 10 after Chinese electric vehicle (EV) maker reported stronger-than-anticipated quarterly sales and guidance and signaled it is right on the track of breakeven this year.
Credit：Xpeng
Xpeng』s top line for the quarter ended March 31 smashed Wall Street anticipation. Revenue surged 141.5% year-over-year (YoY) to RMB15.81 billion ($2.18 billion), compared with analysts』 expected $2.2 billion. The company』s sales accelerated even in the traditional seasonal slowdown in the auto market, highlighting the drive of national subsidies from the Chinese government to simulate domestic spending. Total revenue for from October to December rose 59.4%, bringing the growth rate of the whole year to 33.2%.
Xpeng delivered a new record of 94,008 vehicles for the March quarter, representing an increase of 330.8%. Deliveries popped 2.7% from the previous quarterly record set three months ago. Revenue from vehicle sales grew 159.2% YoY to RMB14.37 billion, following a 66.8% YoY increase for the December quarter.
「Despite seasonality for auto sales, our quarterly deliveries hit a new historical high, making us the top-selling automaker among emerging EV companies. Positive market feedback strengthened our confidence in our three-year product cycle,」 Xpeng CEO He Xiaopeng said in a statement on Wednesday.
Xpeng』s bottom line also beat estimates. On non-GAAP basis, diluted net loss per American depositary share (ADS) for the first quarter came in at RMB0.45 ($0.06), down from loss of RMB1.49 for the preceding quarter. Non-GAAP diluted net loss per ordinary share was RMB0.22 ($0.03), versus loss of $0.20 for the fourth quarter, whereas analysts projected a per-share loss of $0.21. Net loss narrowed 51.8% YoY to RMB660 million.
Gross margin increased 2.7 percentage points YoY to record high 15.6%, up from 14.4% for the prior quarter. Vehicle margin stood at 10.5%, compared with 5.5% a year ago and 10.0% for the last quarter of 2024. Xpeng co-president and vice chairman Hongdi Brain Gu credited the performance to 「significant improvements in cost reduction」, noting vehicle gross margin increased for seven consecutive quarters.
Looking forward, Xpeng expected the second quarter would see deliveries to be between 102,000 and 108,000 with a YoY increase range of 237.7% and 257.5%, suggesting further acceleration and a new record. The company felt confident that it would maintain the triple-digit increase in sales. Total revenue is anticipated to be between RMB17.5 billion and RMB18.7 billion, up between 115.7% to 130.5% YoY.
CEO He touted the outlook of vehicle launches, development of in-house chips and artificial intelligence (AI) adoption. 「「We are just beginning to unleash our growth potential. I believe our strong product cycle, global expansion and accelerated adoption of physical AI technologies, will fuel strong and sustainable growth for Xpeng,」 said He.
In an interview in February, He said Xpeng expected to be able to breakeven for a single quarter sometime in 2025. That was the first time for Xpeng management to disclose a timeline for reaching breakeven. He on an earnings call on Wednesday reaffirmed the expectation. 「We are very confident that we will not only meet our target of more than doubling sales this year, but also will achieve profitability in the fourth quarter of 2025 and large-scale free cash flow for the whole year,」 He said.
Entering the second quarter, Xpeng launched X9 in April, and the model soon became the best-selling full-electric MPV in China, with 1,603 units delivered in its debut month. It also held the top spot in weekly all-electric MPV sales for four consecutive weeks. Xpeng is scheduled to unveil the Max version Mona M03 on May 28 and debut G7 SUV in June. He said G7 is expected to become 「a highly competitive product in the RMB 250,000 SUV segment.」
With the upcoming Mona M03 Max, smart assisted driving powered by high computing power and large AI models will be made available for the first time in a RMB150,000 ($20,805)-level vehicle. He commented that will be a milestone in democratizing AI in the auto industry. smart assisted driving powered by high computing power and large-scale AI models will be made available for the first time in a RMB 150,000 (USD 20,805)-level vehicle.
Xpeng will unveil the next-generation P7 in the third quarter, and Kungpeng EV will enter mass production in the fourth quarter.
Xpeng began building out its AI infrastructure last year to support its AI foundation model, and has established the first 10,000-card intelligent computing cluster in the auto industry in China. In parallel, Xpeng's autonomous driving training dataset is expected to reach 200 million clips by the end of this year.
The Guangzhou-based company has long been investing in proprietary chip development. It achieved its first successful tape-out of the Xpeng Turing AI chip, a custom-designed chip tailored for large AI models. It planned to mass produce the Turing chip and start vehicle integration in the second quarter. More vehicles will be equipped with the chip in the third quarter, He said.
Xpeng's humanoid robot IRON made its debut at the Auto Shanghai show in April, featuring its advanced capabilities in voice interaction, walking, and fine motor control, including highly dexterous finger movements.
CEO He noted that the current IRON is merely the previous-generation prototype. The company believed the on-device computing power of IRON will be significantly enhanced as the Xpeng Turing AI chip will also be deployed in the robot. Moreover, the large model powering IRON robot will bypass transitional architectures and adopt directly Xpeng』s VLA architecture, which is consistent with the company』s foundation model, and it will leverage its shared cloud-based AI infrastructure for the Training.
Xpeng aims to launch a mass-production–ready humanoid robot by 2026, achieving industry-leading levels of intelligence and leveraging production-scale data to drive its rapid evolution.
