Ask AI · Why is the stock price under pressure despite impressive financial results—what is the market worried about?

Recently, Xiaomi released its full-year 2025 performance: total revenue for the full year reached RMB 457.29B, up 25.0% year over year; adjusted net profit was RMB 39.17B, surging 43.8% year over year—both of the two core financial metrics hit historical highs.

Chart source: Hong Kong Exchanges and Clearing Limited (HKEX) - Xiaomi Annual Results Report

Xiaomi’s smart electric vehicle business has moved from losses in its early stage to profitability, achieving a first annual operating profit turnaround to RMB 900 million; full-year deliveries surpassed 410k units, far exceeding the 350k-unit target set at the beginning of the year.

However, another side of this earnings report is also worth examining: in Q4 2025, company revenue was RMB 410k, up 7.3% year over year on a high base compared with the same quarter last year; operating profit fell 23.7% year over year to RMB 350k.

In addition, amid efforts to boost consumer demand, the phone business saw 2025 revenue of RMB 186.44 billion, down 2.8% year over year, while the gross margin fell from 12.6% in 2024 to 10.9%—in Q4, the phone gross margin was only 8.3%.

Over the entire past year, Xiaomi has experienced a combination of multiple interim opportunities: the auto business began to generate scale-driven revenue, while consumer electronics and home appliances industries rebounded under policy tailwinds—especially large home appliances with higher ticket prices, which lifted the company’s overall performance scale.

But when the timeline moves into 2026, the pressure from high sales bases for autos and home appliances, the worsening external storage cost environment, and policies entering a mature phase have all led the market to start worrying whether the company can keep growing above the high base, and to begin provisioning for risks early—its stock price had entered the historical peak range of HKD 60—70 in June 2025. Since the start of 2026, it has instead fluctuated long-term between HKD 30 and 40.

The divergence between earnings and the stock price often reflects the market’s doubt about the sustainability of a company’s growth.

Lower limit: phones; upper limit: making cars

Since Xiaomi officially announced its plan to build cars in 2021, the industry has gradually formed a consensus about Xiaomi: the ceiling is about making cars—whether the car-making business can open up new growth space; the floor is about phones—the valuation support level depends on whether the phone “core business” can hold steady.

In 2025, these two curves showed sharply different trends.

Car-making business has made tremendous progress

Auto business: completed the leap from 1 to 10

In 2025, Xiaomi’s smart electric vehicles and AI, and other innovation business segment’s full-year revenue first broke the one-trillion-yuan threshold, reaching RMB 106.07 billion, up 223.8% year over year; its share of total revenue rose from 9% in 2024 to 23.2%. The company delivered over 410k new vehicles over the full year, with deliveries up 200.4% year over year.

First worth paying attention to is profitability. Xiaomi’s 2025 auto gross margin reached 24.3%, up 5.8 percentage points from 18.5% in 2024—what level is this? Compared with automakers with a delivery scale close to Xiaomi’s— in 2025, Li Auto’s gross margin was 18.7%, and XPeng’s was 18.9%.

It needs to be clarified that in Xiaomi’s auto business, service revenue accounts for less than 3%, so its gross margin reflects the gross margin of the auto business itself, with higher “substance.”

Next, look at average selling price per vehicle. In 2025, Xiaomi’s auto ASP (average selling price) reached RMB 251.2k, higher than RMB 234.5k in the same period last year.

Comparison with peers: Li Auto, affected by increased sales mix from low-price models such as the i6, had a full-year ASP of about RMB 263k; NIO, with a higher brand sales mix driven by lower-priced offerings such as the Firefly and LeDO, had a full-year ASP of about RMB 236k; XPeng’s full-year ASP was RMB 159k.

RMB 250k—RMB 300k is the mainstream price band for mid-to-high-end models. Xiaomi being able to stand firm in this range in itself shows support from product strength and brand strength.

Profit side is also impressive. In 2025, Xiaomi’s car-making and AI, and other innovation business segment first turned from loss to profit on an annual basis. At first glance, a profit of RMB 900 million may not be high as a proportion of total profit—especially when combined with deliveries of over 400k units.

But look at peers in reverse:

In 2025, Leapmotor delivered nearly 600k vehicles with profit of RMB 538 million; in 2025, XPeng delivered nearly 430k vehicles and had profit support from high-margin technology service fees collected for a mass-market brand, but it still posted a loss for the year; for a luxury player like NIO—according to media reports, NIO controlled costs for profitability down to “a packet of tissues” and “a team-building event,” and still lost RMB 14.94 billion for the full year; it finally achieved quarterly profitability in Q4 2025 for the first time. Meanwhile, Li Auto, which used to earn over RMB 116.92B per year, was hit by the price war, and in 2025 it relied on interest-related income to reach profit of RMB 1.1 billion, down 85.8% year over year.

With the backdrop of peers, the “quality” of RMB 900 million in profit is still rising. In 2025, Xiaomi’s auto business profitability was almost second only to Li Auto among new forces.

For Xiaomi’s auto business, the SU7 launched in 2024 gave the company a peak start, allowing it to complete the training from 0 to 1. In 2025, the “single-model record-breaking myth” created by YU7 (over 200k units for big orders within 3 minutes) rapidly solidified Xiaomi’s foundation from 1 to 10.

If the entire company’s strength is gathered, it seems not too difficult for Xiaomi to recreate two more blockbuster models. But as the penetration of its fan base thins out, marketing momentum shows diminishing marginal returns, and Xiaomi’s vehicle product matrix and customer segments keep expanding (family customers, and new energy), Xiaomi is now at the moment of stepping out of its comfort zone.

Phones bleed, but don’t collapse

In contrast to the auto business, the phone business has performed differently.

First is that shipment volumes lagged the overall market. In 2025, the phone × AIoT segment generated revenue of RMB 6.35B, up 5.4% year over year. Of this, the smartphone business generated RMB 186.44 billion for the full year, down 2.8% from RMB 410k in 2024; full-year smartphone shipments were 165 million units, down 2% year over year. Considering that, according to Omdia research into 2025 global smartphone shipments, shipments still grew by 2% year over year—especially in a year when consumption policy tailwinds were present.

However, if you extend the timeframe, these shipment volumes are still far higher than 150 million units in 2022 and 146 million units in 2023, and remain in the historical mid-range—only that it has pulled back from the historical high of 169 million units in 2024.

Second, the trend of gross margin is also not optimistic. In 2025, the company’s full-year smartphone business gross margin was 10.9%, down 1.7 percentage points from 12.6% in 2024. Entering Q4, against the backdrop of a sharp rise in storage hardware prices, the smartphone business gross margin was only 8.3%, down 3.7 percentage points from 12% in the same quarter of 2024.

It needs to be clarified that Xiaomi phone gross margin’s historical peak occurred in 2021 at 11.9%; the current level of 10.9%, while having retreated from a high point, is still higher than 7.2% in 2019 and 8.9% in 2020. What the gross margin reflects is “under pressure” rather than “collapse”—and the company’s strategy to move upmarket has already shown phased results.

Regarding gross margin, company management was candid in the earnings call: both the speed and magnitude of the increase in memory costs exceeded expectations. Xiaomi Group partner and president of the phone department Lu Weibing explained, “This is indeed a huge challenge for Xiaomi. Demand for AI-driven servers has surged, reshaping the supply-demand relationship in the storage industry. I’ve said before that this is a cyclical phase—we need to see it by 2027. But my personal feeling is that this price-increase cycle may be longer than I expected, and the magnitude of the increase may be higher than I expected.”

Smartphones, tablets, and laptops—products that account for the highest share of storage cost—have to bear massive cost pressure. Although Xiaomi has temporarily avoided the risk of supply interruption by leveraging scale advantages with suppliers’ long-term contracts, as well as its previously relatively pessimistic inventory strategy, the decline in gross margin has already become a fact when reflected in the financial statements.

When the phone business “base” loosens, it directly affects the market’s confidence in valuing Xiaomi. As a phone maker that has already proven itself, Xiaomi’s valuation in this area has been fully priced in by the market. Investors’ expectations for its phone business are no longer “whether it can do well,” but rather “whether it can continue to break through on top of a high base.” This is also why, even though the auto business has performed impressively, the stock price is still under pressure—there is a crack in the company’s floor.

There are also positive developments—Xiaomi’s penetration in the mid-to-high-end smartphone market has increased. In 2025, Xiaomi’s share of smartphone sales in mainland China at RMB 3,000 and above reached 27.1%, up 3.8 percentage points from 23.3% in 2024; in the RMB 6,000—RMB 10,000 price tier, market share rose 2.3 percentage points year over year to 4.5%.

In the earnings call, Lu Weibing clearly outlined Xiaomi’s pricing strategy: “The trend of price increases is inevitable. We will do our best to carry it for consumers for a while. When we can’t carry it anymore, it will definitely go up. I hope that users will understand us more then.”

This confidence of “carrying it for a while longer” comes from Xiaomi’s ability to diversify risk across its businesses—home appliance operations are less affected by memory cost fluctuations, and cross-category synergy can help share risk. In addition, a higher proportion of high-end models means Xiaomi has stronger bargaining power in cost pass-through than in the past. When low-end models bear the risk of losing price-sensitive users, the price resilience of high-end models becomes the key buffer to offset cost pressure. This turns “price increases” from a “passive response” into an expression of structural capability.

After catching the last train, Xiaomi’s auto business must take charge early

On March 19 this year, at Xiaomi’s spring launch event. Lei Jun announced two new brand ambassadors for the auto business, Shuqi and Sun Bing—putting the most passion into the car presentation. Then came the large model, and he even “rode along” on the heat of a “lobster” (OpenClaw, a personal AI agent, generally known as a “small lobster” because it uses a lobster as the product’s image; it later broadened to refer to personal AI Agent products). Meanwhile, since there were no new phone products, Xiaomi’s “major” laptop and smart watch refresh—after four years—only got the final ten-odd minutes of airtime with Lei Jun.

After car-making success, the market’s positioning of Xiaomi essentially shifted from being a phone company to being a car company. But in fact, as product iteration cycles and software update cycles are shortened dramatically, in mainstream perception, new energy vehicles have effectively become large, expensive consumer electronics. It’s no different from phones.

In 2026, it’s not Xiaomi’s “big product year”: the SU7 is only a mid-cycle refresh, and the range-extended models are still in the stage of deep disguise. The 550k delivery target represents about a 34% year-over-year increase versus 411k units in 2025, but growth will slow compared with 2025. However, it’s also important to note that 550k means an average of 46k units per month. For a car company that has only mass-produced for two years, this is still a very demanding requirement—especially given the background that China’s new energy vehicle penetration rate is nearing the 50% line at the margin.

In 2025, Xiaomi delivered a response it can be proud of. It proved that amid fierce market competition, the company still has the ability to build blockbuster products and achieve rapid scaling. Especially in 2025, when the price war was intense, Xiaomi’s auto still delivered an overall gross margin of 24.3%—with no one among new forces matching it. And achieving scaled deliveries and annual profitability in the second year of mass production is unprecedented across the entire new forces lineup.

But the rule of the business world is that past success cannot offset future risks. Because the first two “answer sheets” Xiaomi produced were too perfect—so the market has always held it to the standards of a top-tier student. Then, as the phone business shows signs of pressure and the early-year refreshed models were unable to continue the previous “big order myth,” the divergence between performance and valuation becomes completely reasonable.

For Xiaomi, the main topic for 2026 is no longer “whether it can build the next blockbuster,” but “whether it can allocate resources well and control the pace across multiple business lines”—after all, the narrative in the car-making circle is becoming more diversified. Influenced by iterations of advanced driver-assistance technology and the development of AI large models, in the capital market’s evaluation framework for all consumer electronics companies, software and hardware infrastructure such as chips, algorithms, and base large models are being assigned increasingly higher weights. Even an aggressive player like Li Auto has redefined itself as a robotics company.

In this regard, the advanced driver-assistance chips for the “Volvo-like trio” (Li Auto, XPeng, NIO) have already been installed or are preparing to be installed in cars. And reportedly in 2026, Huawei’s advanced driver-assistance systems will be installed in over 100 models from major brands. Domestic intelligent driving solution providers like Momenta have demonstrated advanced driver-assistance system capabilities that don’t lag behind first-tier automakers, and more and more they are empowering joint-venture automakers that had previously fallen behind. By the end of this year, BYD has to build 20,000 fast-charging “flash-charging” chargers… The elimination-round competition is “hell-level” difficulty. In 2026, when cost pressure is high in the phone business and cash-generating ability is tightening, to respond to competition that has fully heated up, Xiaomi has no choice but to put resources into a wider range of areas—proving that it has not only successful product definition capabilities, but can also fully invest in both software and hardware, resulting in breakthroughs across the board.

At the China Development High-Level Forum in March 2026, Lei Jun said that over the next five years, they will invest RMB 200 billion in R&D.

But where will the money come from? In 2026, it’s probably not something you can rely on the phone business’s cash-generating ability to grow another notch. For Xiaomi, the pressure is all transferred to car-making.

How to turn car-making from a company growth story into a more tangible source of cash flow—from scaled deliveries and single-digit revenue creation, to scaling and profits rising together. By using car-making to add another “cash cow” for the company, supporting increasingly diverse investment needs—against the backdrop that it’s hard for peers even to avoid losses—this task looks even more difficult.

When announcing car-making, Lei Jun admitted that this is the “last train” and the final window period for car-making. That’s why Xiaomi’s auto business hasn’t enjoyed the long development runway and long-term capital and market tolerance that other new EV makers have had over 10 years. In just about two years, it has to take charge early.

This is probably the fate that latecomers inevitably have to face.