
Investors who held XPeng over the past year are down 61.7%, including dividends. If you had bought on 30 October 2025, that loss would feel especially sharp given how many forecasts then leaned on rapid revenue expansion, margin improvement into positive territory, and a rich 40.8x future P/E within three years. Which of those specific assumptions about profitability, international expansion, and premium pricing did the original case fail to fully allow for?
This theme extends beyond XPeng. See which of 94 robotics and automation stocks may still merit a closer look.
The shares cost US$22.95 at the start of the period, and anyone looking at XPeng then had to pick between two very different storylines.
The bullish narrative pointed to a Fair Value of US$26.29, a price implied by assumptions that revenue could rise around 31.6% a year and profit margins shift from a loss into a 4.6% gain within about three years. Supporters highlighted AI hardware, smart-driving software, and international expansion.
The more cautious case used a Fair Value of US$18.06 and focused on risks. That view emphasized global EV rivalry, price wars, and geopolitical barriers that could limit overseas growth and pressure XPeng’s margins.
XPeng emphasized a bullish storyline with mass-produced robotaxi hardware, the GX and G9L AI SUVs, global production in Austria, and humanoid robot lines moving toward volume output. That supported the international and physical AI expansion angle. The numbers, however, showed a tougher profitability picture. Revenue rose from CNY 18,274.405m to CNY 19,743.593m, while net margin moved from a loss of 2.6% to a loss of 6.8%, so the margin improvement assumption remained unproven and the evidence was mixed.
The takeaway is straightforward: when an argument depends on future profitability, it helps to focus on the reported net margin and how it changes relative to any growth in sales. Product launches and partnerships may contribute to revenue, but the thesis only really shifts once those new lines begin to move margins toward the level initially projected.
XPeng now trades at US$9.04, and the selected Narrative’s Fair Value sits above that level based on its own assumptions about the business. The argument leans on Physical AI, where smart EVs, robotaxis, and humanoid robots all reuse one in house tech stack.
For that higher figure to make sense, a buyer today would be assuming XPeng eventually earns materially stronger margins from these AI platforms than it has so far.
"XPeng's shift from a pure EV maker to a broader Physical AI group that reuses its in-house chips, models and software across vehicles, Robotaxi and humanoid robots is expected to create multiple revenue streams from a single technology stack and improve long term returns on R&D investment."
Not everyone reads the same price the same way. → See the higher figure this Narrative lands on, and how it gets there
This company's disappointment is already part of the story. Your next idea could come from looking where the price and the possibilities still seem far apart. Here are three companies priced below our estimates.
Those are three of them. See the full list of 33 companies trading below our fair value estimates →
Have feedback on this article? Concerned about the content? Get in touch with us directly. Alternatively, email editorial-team@simplywallst.com