Author | Zhou Zhiyu
In March 2021, Xiaomi CEO Lei Jun announced the company's entry into car manufacturing, calling it &#34;the last major entrepreneurial endeavor of his life&#34; and expressing his willingness to stake his entire reputation on it. At the time, Xiaomi had just emerged from a downturn in its smartphone business, and the market was skeptical about a smartphone company venturing into car manufacturing.
Four years later, on the evening of March 24th, Xiaomi Group released its full-year performance for 2025. The automotive business achieved annual revenue exceeding 100 billion yuan, and its segment operating profit turned positive for the first time.
Starting from the delivery of the first batch of SU7s in April 2024, Xiaomi took less than two years to complete the journey from &#34;zero to 100 billion yuan in revenue, from loss to profit&#34;— a feat that took Li Auto over four years, and NIO eleven years to touch quarterly profitability in Q4 2025.
However, this annual report carries mixed emotions. The full-year data marks the best year in history, but the fourth quarter faced pressure, with gross profit margins for both smartphones and cars declining. Xiaomi President Lu Weibing stated at the earnings call that evening, &#34;The start is not easy, and performance will face some pressure in the short term.&#34;
Xiaomi's situation is not an isolated case. 2025 has been a rare year of harvest for China's technology and new energy industries— full exemption from purchase tax, ample government subsidies, concentrated breakthroughs in AI large models, and rapid growth in new energy penetration. Almost all leading players have delivered their best data.
Now, almost all players need to answer a question: When these favorable conditions no longer hold simultaneously in 2026, who will be able to withstand the headwinds?
<h2>The Two Sides of Profitability</h2>
The most crucial highlight of Xiaomi's financial report lies in the transformation of its revenue structure.
The automotive segment achieved a full-year gross profit of 25.8 billion yuan. The smartphone segment's gross profit was 20.3 billion yuan. A business less than two years old has surpassed the gross profit of a 15-year-old phone business. Full-year revenue reached 457.3 billion yuan, and adjusted net profit was 39.2 billion yuan, both historical highs. Automotive deliveries reached 411,082 units, a year-on-year increase of 200%.
This is one side.
The other side is in Q4. The adjusted net profit for the first three quarters was 10.7 billion, 10.8 billion, and 11.3 billion yuan, showing a steady upward trend. In Q4, it dropped to 6.3 billion yuan, a year-on-year decrease of 23.7%. The gross profit margin for smartphones in Q4 was only 8.3%, while the gross profit margin for cars dropped from 25.5% in Q3 to 22.7%. IoT's gross profit margin fell from 23.9% to 20.1%. All four lines moved downwards in Q4.
The 8.3% for smartphones needs further explanation. Lu Weibing previously revealed to Wallstreet News that the price of memory has increased, causing the 12+256GB memory price to rise from a low of about 30-plus US dollars to about 130 US dollars, an increase of nearly fourfold. He has now revised his forecast upwards, stating frankly that the increase is &#34;even more aggressive than my original forecast,&#34; and the cycle &#34;might be longer than currently anticipated.&#34;
Products with a higher proportion of storage costs in their Bill of Materials (BOM) are hit the hardest. Xiaomi's average selling price (ASP) for smartphones throughout the year was only 1128 yuan, with Redmi and POCO volume models bearing the brunt. The push towards high-end products is providing some offset, with market share in the 6000 to 10000 yuan range nearly doubling to 4.5%. However, the Q4 data indicates that the speed of premiumization has not kept pace with cost increases.
Competitors have already started raising prices. Lu Weibing said, &#34;I can understand them very well, as everyone is having a difficult time. We will try to hold out first, but if we absolutely cannot, we will have to raise prices too.&#34;
The reason he is not rushing to raise prices is practical: smartphones don't just earn money from hardware. Internet services achieved a full-year gross profit of 28.6 billion yuan with a gross profit margin of 76.5%. This revenue relies on shipping volume. Reducing volume to protect profit margins would shrink the user base.
The automotive gross profit margin for the full year was 24.3%, surpassing BYD, Li Auto, and Great Wall Motor. However, it dropped to 22.7% in Q4, mainly due to a decrease in the delivery proportion of the SU7 Ultra and year-end inventory clearance. The increase in memory prices also had a significant impact on cars, although Lu Weibing noted that while cars use considerable memory, its proportion in the overall vehicle BOM is lower than in smartphones.
Another calculation worth noting: the automotive segment's full-year gross profit was 25.8 billion yuan, with operating expenses at 24.8 billion yuan, leaving only 900 million yuan in operating profit. A revenue of 106.1 billion yuan yielded an operating profit margin of less than 1%. Xiaomi CFO Lin Shiwei emphasized that the full name of this business segment is &#34;Smart Electric Vehicle and other Innovative Businesses (including AI)&#34;— indicating increasing investment in AI and ongoing investment in new ventures.
<h2>Betting on the Next Cycle</h2>
Where has the profit gone?
Last year, Xiaomi repurchased shares worth a total of 6.3 billion Hong Kong dollars. R&amp;D expenditure reached 33.1 billion yuan, a year-on-year increase of 37.8%. Capital expenditure was 18.2 billion yuan, with the automotive segment accounting for 12 billion yuan, a threefold increase from the previous year. Inventories rose from 62.5 billion to 80.9 billion yuan.
Investments will continue to increase in 2026. Lu Weibing confirmed that R&amp;D spending will exceed 40 billion yuan this year, with 16 billion yuan allocated to AI, and a cumulative investment of over 60 billion yuan in AI over the next three years. The annual report projects an even larger figure: cumulative R&amp;D investment exceeding 200 billion yuan over the next five years, starting from 2026.
BYD's full-year R&amp;D expenditure exceeded 50 billion yuan, while Huawei consistently invests over 100 billion yuan. Xiaomi is using its profits to buy a ticket to the next cycle.
Lu Weibing revealed that within this year, a product will be launched that simultaneously features the self-developed chip &#34;Xuanjie O1,&#34; the &#34;HyperOS,&#34; and a self-developed AI large model— an event he calls &#34;the grand assembly.&#34; Xuanjie O1 has already seen an investment of 13.5 billion yuan with a team of 2,500 people. &#34;Chips are essentially a platform capability; mastering this platform allows for the development of many products.&#34;
He reiterated the same assessment at the earnings call. The MiMo-V2-Pro large model boasts over a trillion parameters, ranking fifth globally by brand.
The Agent miclaw on mobile devices is undergoing closed-door testing, which Lu Weibing characterized as &#34;the nascent form of an operating system for the Agent era.&#34;
But what about AI commercialization? He also frankly admitted that it is still a bit too early at present.
Another direction for profit reinvestment is overseas expansion. The annual report shows that revenue from mainland China increased from 58.1% to 67%, while overseas revenue slightly decreased. The automotive business has boosted domestic revenue share but also magnified reliance on a single market. Xiaomi needs incremental growth from overseas markets.
Lu Weibing had previously provided Wallstreet News with a more comprehensive overseas expansion roadmap than the annual report. Excluding North America, the IoT market overseas is three times the size of the domestic market, yet Xiaomi's overseas IoT scale is currently only one-sixth of its domestic scale. &#34;If it reaches the domestic level, it could mean a sixfold increase.&#34;
Lu Weibing believes that 2025 is the inaugural year for overseas expansion of major home appliances, covering 4 major regions and 14 countries. The integrated delivery and installation service model has been successfully implemented in multiple countries. He visited a Xiaomi Store in London and noted that &#34;the products sold are mostly high-end, even the rice cookers are selling very well.&#34; The number of overseas new retail stores, which was 450 at the end of last year, is expected to exceed 1,000 by the end of this year.
Smartphone premiumization is also expanding globally from China. The Xiaomi 17 series, launched in Barcelona in February this year, has an overseas starting price of 999 euros. The Leitz Phone (the overseas version of the Xiaomi 17 Ultra Leica edition) is priced at 1999 euros— exceeding the price of the iPhone 15 Pro Max in Europe. Lu Weibing stated, &#34;Today, Xiaomi has the courage and confidence to challenge the iPhone at a higher price point.&#34;
Automotive exports, though latest to start, hold the highest expectations. The company plans to enter Europe in 2027, starting with challenging markets. Lu Weibing himself is leading the preparation team for overseas expansion.
Mixed sources within Xiaomi concluded after visiting Europe that now is an excellent time for Chinese car manufacturers to expand overseas. Although Xiaomi has not officially entered the European market, some European users are already importing cars from China, handling customs clearance and registration themselves, establishing brand recognition.
How will the market price these endeavors? Xiaomi's business, though complex, is not difficult to calculate. From the Q3 peak to the present, Xiaomi's stock price has nearly halved. A simple calculation: the adjusted net profit for the smartphone x AIoT segment in 2025 is approximately 38 billion yuan. Applying a valuation close to Xiaomi's historical valuation center, it would be around 20 to 22 times P/E. The automotive and AI segment, with 106.1 billion yuan in revenue, would be valued at 1.4 to 1.8 times P/S.
The automotive segment includes not just cars— but also large models, Agents, robots, whole-house smart systems, and 1 billion IoT devices with 750 million monthly active users. If valued at 1x P/S for the automotive business alone, the value of AI and overseas expansion would be largely invisible.
At Xiaomi's current stock price, it is clear that investors have not given much premium.
Behind this are two layers of concerns. One is that profits in 2026 may be further revised downwards, as memory price hikes have not peaked, the impact of reduced government subsidies is just beginning, and it's unclear if the Q4 figure of 6.3 billion yuan represents the new normal. The other is market uncertainty: when will the returns from the money Xiaomi is reinvesting be realized? The &#34;grand assembly&#34; product of chip + OS + large model is expected to launch within the year, but how much incremental growth will it bring? The sixfold growth potential for overseas IoT sounds appealing, but the 1,000 new stores are just starting to be rolled out. Automotive exports will not begin until 2027. Every move points to the future, but none can be revealed immediately.
The R&amp;D plan of 200 billion yuan over five years is essentially a bet on one thing: by the time the memory cycle ends and the industry consolidation is complete, Xiaomi will be at the table, and with more cards in hand than its competitors.
<h2>Favorable Winds Won't Last Forever</h2>
Let's place Xiaomi's Q4 within the industry context.
Over the past decade, China's new energy industry has undergone a long validation process. Around 2015, new startups emerged in concentration, surviving on financing. From 2018 to 2020, the elimination phase began, with Byton, Saleen, and others falling by the wayside. In 2022-2023, HiPhi halted operations, NETA teetered on the brink, and NIO's annual losses exceeded 20 billion yuan, with Li Bin's keyword being &#34;survival.&#34;
By 2025, the situation has reversed. NIO and Xpeng achieved their first quarterly profits in Q4, and Leapmotor achieved full-year profitability. Xiaomi reached annual profitability in less than two years since its first delivery. Almost all surviving leading players have written their best chapter in this year.
However, 2025 is also the year when all favorable conditions peaked simultaneously. Full exemption from purchase tax, the most abundant government subsidies, and the fastest penetration rate. These conditions will not all be met in 2026. Subsidies will be halved, and production capacity will continue to expand— Xiaomi at 550,000 units, Leapmotor at 1 million units, and BYD and Huawei-related brands applying pressure across all price segments.
The increase in memory prices is an underlying current affecting both the new energy and consumer electronics sectors. The surge in demand for HBM from AI has squeezed capacity for consumer products. Some low-capacity categories are already out of stock or even discontinued.
In addition to memory, rising prices of various materials are putting pressure on industry players.
Lu Weibing stated, &#34;After the price increase cycle ends, I believe that many product categories will see a reshuffling of the industrial landscape. Some companies may face extreme operational difficulties during this long price increase cycle, potentially leading to significant losses or even failing to survive.&#34;
He immediately followed up with, &#34;People may start to pay attention to what will happen after the price increase cycle.&#34; His assessment is that &#34;the challenging external environment will inevitably force a lot of innovation.&#34;
The implication of consolidation is that manufacturers with lower levels of premiumization and higher memory costs as a proportion of BOM will be hit the hardest. After this memory cycle, the mid-tier players in consumer electronics will face further culling.
Concurrently, the narrative of technology companies entering car manufacturing has entered its second chapter. The first chapter took ten years and answered questions like &#34;Can they build it, sell it, and make a profit?&#34; Xiaomi and Huawei have both provided affirmative answers with their 2025 data. The second chapter poses different questions: With subsidies receding, costs soaring, and competition intensifying, can this model withstand a full economic downturn cycle?
Lu Weibing was candid at the earnings call: &#34;The challenges this year are very, very significant.&#34;
He also added the latter part: &#34;But I believe that after a year of effort in 2026, our management team will still be able to deliver a report card that we consider acceptable.&#34;
The phrase &#34;acceptable&#34; is a term that only a manager experienced through cycles would use. This is likely the most realistic backdrop for 2026.

01810

XIACY

- In March 2021, Xiaomi's CEO Lei Jun announced his last major entrepreneurial project in car manufacturing, receiving initial skepticism.  
- By March 2025, Xiaomi's automotive sector generated over 100 billion CNY in revenue, marking a rapid transition from loss to profit within two years, compared to competitors that took much longer.  
- However, their Q4 performance showed declining profit margins in both mobile and automotive sectors, raising concerns about future profitability and the sustainability of this growth amid rising costs and market pressures.

Xiaomi Earns 39.2 Billion, Lu Weibing Says "Hold the Line First"