The worst profits, but the best story.

Yesterday, after the close of trading in Hong Kong stocks, Alibaba released its fourth-quarter and full-year financial results for fiscal year 2026.

My initial reaction after browsing was contradictory. On one hand, the group's annual revenue surpassed the RMB 1 trillion mark for the first time, reaching RMB 1,023.67 billion, with external revenue growth in the cloud business hitting a nine-quarter high.

On the other hand, Q4 revenue fell short of expectations, with operating profit turning into a loss of RMB 848 million, compared to a profit of RMB 28.465 billion in the same period last year; adjusted EBITA plummeted by 84% year-on-year; non-GAAP net profit shrank to just RMB 86 million; and net free cash flow outflow reached RMB 17.3 billion. Alibaba's core business profitability is undergoing its most severe contraction in recent years.

Tencent, which also released its financial results on the same day, reported a Q1 Non-IFRS net profit of RMB 67.9 billion and free cash flow of RMB 56.7 billion. Alibaba's revenue exceeded Tencent's by nearly RMB 50 billion, yet its core business profit lagged far behind. When comparing the two report cards, the financial consequences of their vastly different strategic choices become immediately apparent.

This financial report cannot be simply defined as "good" or "bad." It more closely resembles a mid-transformation report, documenting the tangible costs a trillion-dollar giant is paying as it bets on the future.

01

The "Disappearance" of Profits

Is a Strategic Active Choice

The sharp contraction in profits this time essentially reflects Alibaba's proactive shift of core e-commerce earnings onto the balance sheet of AI infrastructure.

The financial results show that Q4 free cash flow turned into a net outflow of RMB 17.3 billion, with a full-year net outflow of RMB 46.609 billion, compared to a net inflow of RMB 73.87 billion in the previous fiscal year. The RMB 120 billion cash flow gap largely points to investments in data centers and computing power infrastructure.

CFO Xu Hong broke down the company's financial position on the earnings call: Currently, the company holds approximately $38 billion in net cash, or about $59 billion if excluding debt maturing beyond five years. Combined with its strong ability to raise capital in the markets, this is "sufficient to support this unprecedented round of capital expenditures."

Wu Yongming used a more vivid metaphor: AI is more like manufacturing. To generate more revenue, two types of factories must be built simultaneously—training factories and inference factories. He emphasized that currently, "almost none of the servers are idle" and predicted that given the highly certain demand growth over the next three to five years, the investment return path is "very clear."

Notably, Wu Yongming also signaled on the call that to achieve the goal of "annual cloud and AI commercialization revenue exceeding $100 billion in the next five years," Alibaba Cloud's future computing power center assets will be more than ten times those held before the AI boom in 2022, with capital expenditures over the next three years "likely far exceeding the previously committed RMB 380 billion over three years."

To clarify: Profit decline itself is not positive news, but in a market where computing power supply lags far behind demand, only companies bold enough to heavily invest in infrastructure during the early cycle stages have a chance to seize the initiative in the next phase of growth.

Before the earnings call, the market saw "profits disappearing"; after the call, it realized this represented large-scale, forward-placed capital expenditures with clear return expectations.

This is the first layer of logic behind the stock price reversal.

02

Alibaba's Corporate Nature Is Undergoing Structural Change

What the market re-evaluated in this financial report is not just the growth rate of Alibaba's AI business but also the very nature of the company itself, which is transforming from an e-commerce platform into an AI computing power infrastructure company.

Several dimensions of evidence support this judgment.

At the revenue structure level, Alibaba Cloud's external commercialization revenue grew by 40%, hitting a nine-quarter high. AI-related product revenue reached RMB 8.971 billion, accounting for over 30% of external revenue for the first time. Cloud's share of the group's consolidated revenue rose from 11.8% in fiscal year 2025 to 15.4% in fiscal year 2026, a 3.6 percentage-point increase in one year.

Wu Yongming also predicted that AI revenue will exceed 50% of cloud revenue within the next year, officially becoming the primary growth engine for cloud income.

At the underlying hardware level, T-Head's self-developed "Zhenwu" GPU chip has cumulatively shipped 470,000 units at scale, with over 60% serving external commercial clients. More than 30 automotive companies and intelligent driving solution providers now use Zhenwu PPUs for R&D on Alibaba Cloud, with actual usage exceeding 100,000 cards.

At the business model level, Alibaba disclosed its AI model and application services ARR (Annual Recurring Revenue) for the first time using SaaS company standards, surpassing RMB 8 billion, with projections to exceed RMB 10 billion in the June quarter and RMB 30 billion by year-end. Globally, ARR is often seen as a key signal that commercialization visibility is clear enough to warrant ARR-based valuation models.

Putting these three layers together: A company with self-developed chips, self-developed large models, and a self-built cloud platform is shifting its primary growth engine from e-commerce GMV to AI computing power services ARR, with both chip shipments and external client counts having reached scaling (large-scale) stages.

This is no longer a story of "an e-commerce company doing a bit of AI"—Alibaba is turning AI infrastructure into its core business.

Of course, this doesn't mean Alibaba will abandon e-commerce. E-commerce remains the fundamental source of cash flow, with Taobao and Tmall Group's customer management revenue growing 8% year-on-year and same-day retail revenue surging 57% year-on-year. However, the acceleration of growth is shifting from e-commerce to cloud and AI, which is the deeper reason the market is willing to pay a premium for this financial report.

03

Understanding Alibaba's Position in Competitors' Strategies

To fully grasp the true implications of Alibaba's financial report, we must view it within the competitive landscape.

ByteDance's AI capital expenditures exceeded RMB 150 billion in 2025 and are planned to rise to RMB 200 billion in 2026. Doubao now has 345 million monthly active users, with daily token calls exceeding 120 trillion. Tencent's AI-related capital expenditures reached RMB 37 billion in Q1, up about 63% quarter-on-quarter, with AI new products dragging profits down by about RMB 8.8 billion.

All three giants are simultaneously using core business profits to fund AI infrastructure, but their paths diverge significantly—Alibaba resembles a "power plant," heavily investing in underlying computing power; Tencent acts more like a "transmission network," grafting AI applications onto its existing ecosystem; ByteDance adopts a hybrid approach, with both application-layer products like Doubao and large-scale computing power deployments.

Pony Ma's self-assessment at the shareholder meeting was candid: "A year ago, we thought we'd boarded the ship, then realized it was leaking. Now we feel we're standing on it but still can't sit down comfortably—we hope it speeds up." A media analysis article that day pointed out that in AI, Alibaba and Tencent "are no longer earning the same kind of money," with Alibaba Cloud's growth engine shifting from traditional cloud computing to models, inference, and AI infrastructure.

A natural consequence of this divergence is that Alibaba's valuation model is undergoing a transformation.

In the past, the market valued Alibaba based on e-commerce GMV and CMR growth; now, it is attempting to reprice the company using new metrics such as computing power scale, AI ARR, and self-developed chip penetration.

After the financial report, JPMorgan maintained an "Overweight" rating with a U.S. stock target price of $200; Citigroup analysts wrote that while AI technology investments are dragging down profits, "management indicates this is an active strategic choice aimed at seizing huge market opportunities."

Meanwhile, an external variable that is easily overlooked but should not be ignored is chip controls.

In 2026, the U.S. shifted its H200 chip exports to China to "conditional release," capping shipments to China at no more than 50% of U.S. domestic sales, requiring third-party performance verification before shipping, mandating customers prove non-military use, and imposing a 25% revenue-sharing fee on sales.

Simultaneously, the U.S. House of Representatives passed the Remote Access Security Act, attempting to extend export controls to cloud computing services. This "controlled open" geopolitical environment underscores the strategic value of Alibaba's self-developed chips—the Zhenwu GPU is not just a technological option but also a supply chain security option.

As one analysis noted, China's self-sufficiency rate in AI chips is expected to exceed 80% by 2027. Against this backdrop, Alibaba's self-developed chip shipment growth and external client penetration will gradually become an independent variable for capital markets to assess its long-term value.

04

How Far Is the Turning Point from Materialization?

Beyond AI, the financial report and earnings call records reveal several other major spending areas.

First is same-day retail, with Taobao Flash Sales representing one of Alibaba's most aggressive business expansions over the past year. However, this high growth comes with high subsidies, and sustained investment in Flash Sales has directly dragged down profit margins for the Taobao and Tmall Group. Second is the defense and offense in e-commerce user acquisition, with continued investments in price competitiveness and user experience at Taobao and Tmall.

Pursuing all three fronts simultaneously has an immediate impact on cash flow. Previously, Wu Yongming stated that Alibaba would continue to "execute a strategy focused on e-commerce and AI" and make "firm investments" in new businesses. The word "firm" itself implies that the pain period for profits is not yet over.

Looking back, the sharp stock price fluctuations before and after this financial report essentially reflect a market recalibration of Alibaba. Pre-call selling was based on inertial thinking (inertial thinking), while post-call buying stemmed from a reevaluation of the company's underlying logic. In a market where computing power supply lags far behind demand, companies that proactively secure infrastructure positions may gain pricing power.

Alibaba is the only giant simultaneously investing across five layers: self-developed chips, large models, cloud computing, AI applications, and e-commerce scenarios. This battle line is longer than any peer's, implying greater imagination space but also greater resource consumption.

Wu Yongming said AI data center investments have "very certain returns," but certainty does not equal realization.

Xu Hong said consumer business cash flow will "trend positively over the next two years," but this depends on how quickly Flash Sales losses narrow and international e-commerce turns profitable.

Management said margin improvements would become visible "within the next 1-2 years" and "in the next quarter or two," meaning the market still needs several quarters of data for cross-verification.

This time, the market voted "yes" with the stock price. But it is growing increasingly discerning toward narratives of "trading today's profits for tomorrow's growth." Between "yes" and "confirmation" lies an entire timeline of performance delivery.

— END —