The most important number here is the cloud segment's growth rate. In the quarter ended June 2026, AI Cloud and Compute Services grew 45%, its fastest pace in more than five years, and now makes up roughly 18% of Alibaba's total revenue, up from about 13.5% a year earlier. Profit from that segment (measured on an adjusted basis) more than doubled compared to the prior year, showing that the business is not just growing but getting more efficient as it scales. AI-related product revenue alone reached about $1.8 billion in that quarter, equal to roughly a third of Alibaba Cloud's revenue from outside customers. Management has said it wants AI to eventually make up more than half of all cloud revenue within about a year.

To pay for the expansion, Alibaba spent about $10 billion on capital expenditures in that one quarter alone, a 75% jump from the year before. That is a big reason free cash flow has swung to a large negative number recently. Reported profit and earnings-per-share both fell well short of what analysts expected. Looking further out, over the next three to five years, most estimates expect Alibaba's overall revenue to keep growing in the high single digits to low double digits per year, with the cloud and AI business growing much faster than the rest of the company and becoming a bigger share of total profit over time. Management has set a longer-term goal of surpassing $100 billion combined from cloud and AI revenue within five years.

Our View: We see this mix of numbers as bad for near-term reported profit but good for the multi-year growth story. The fact that cloud revenue growth hit a multi-year high while cloud profit more than doubled is the strongest evidence supporting our Buy rating, because it shows AI spending is already turning into real, profitable revenue rather than just being an expensive bet. We are less sure exactly how long the current period of negative cash flow will last, since that depends on how quickly this spending slows down.

Growth Drivers and Strategic Bets

Cloud Computing Alibaba Cloud has about a 36% share of the domestic AI cloud computing market, more than double that of its nearest Chinese competitor. The size alone should allow Alibaba to provide pricing/performance that undercuts smaller players (something we view as an authentic advantage).

Artificial Intelligence Alibaba's family of AI models, called Qwen, ranks among the most widely used Chinese AI models based on several usage and developer-adoption metrics. And while competitors are experimenting with selling AI chips/models/API access to developers, Alibaba makes many of its largest Qwen models "open-weight." Open-weight allows developers to download and freely run these models. Alibaba builds developer loyalty and creates a natural funnel of future cloud customers who will eventually pay to run those models on Alibaba's servers. Fast commerce, which offers lightning-fast local delivery, has been seeing compound annual growth rates of between 45% and 56% in recent quarters. Meanwhile improvements in advertising revenues are cushioning the blow from weaker overall shopping activity. To be sure, one assumption we are making is that we think Alibaba is overspending on AI today before there is sufficient demand from paying customers. We expect meaningful conversion of spending into outside-cloud customer revenues to begin around fiscal year 20:28, not this year. Alibaba's indigenous chip design unit, T-Head, is helping the company become less reliant on Western chipmakers such as Nvidia, which are unable to sell their latest and greatest chips in China due to U.S. export restrictions. It's smart to hedge its exposure, but we'll believe that T-Head can fully replace Nvidia at scale someday. Outside of China, Alibaba is tiny relative to Amazon Web Services, Microsoft Azure, and Google Cloud, and lags far behind them in trust among large non-Chinese enterprise customers. Alibaba also faces heavy domestic competition from PDD, JD. com, ByteDance's Douyin, and Meituan, all of which are pouring money into AI.

Our View: We think Alibaba Cloud. Our view that Alibaba can maintain its lead position in China's cloud market is fairly confident, while our view about Alibaba Cloud's international prospects is more speculative. We believe Qwen's open-weight approach creates future revenue and competitive-position advantages for Alibaba. We are counting on historical industry trends to play out again regarding Alibaba Cloud internationally; AWS, Azure, and Google Cloud have too much of a lead in trust and installed infrastructure outside of China to commit much on that front.

Capital Allocation and Balance Sheet

Alibaba went from having strongly positive free cash flow two years ago to negative free cash flow today primarily due to an enormous increase in capital expenditures investing in AI data centers and chips. Alibaba financed this spending partially by selling $10.2 billion worth of new stock in Hong Kong in late August 2026, the largest stock sale ever in Hong Kong. Alibaba plans to use all of the proceeds from the sale to continue investing in AI infrastructure. Because of this stock sale, existing shareholders were diluted by about 3.6%, and each share now owns a slightly smaller percentage of the company.

Chairman Joe Tsai and CEO Eddie Wu also bought shares around the time of the stock sale worth over $25 million combined. Tsai's stake is now around 1.44% of the company. Insider ownership across executives and board members is around 12.5% of the company. Alibaba does have a small annual dividend, but it yields less than 1% at current prices. Share buybacks will likely be on hold for the time being considering how much capital is going towards AI expansion.

Our Take: Funding this buildout with mostly stock rather than debt is prudent and conservative given the circumstances, even though it dilutes shareholders in the near term because it does not put additional financial risk on the company while cash flow is negative. Tsai and Wu buying shares alongside the company is a very mild vote of confidence in the plan, but the amounts they spent are trivial compared to the $10.2 billion being raised, so we wouldn't place too much emphasis on this factor alone.

Competitive Benchmarking