Alibaba and the Cave of Treasures: Will the Company Become Asia's Amazon?

Alibaba raised $10.2 billion through a new share offering in Hong Kong, with nearly three times the number of shares subscribed: sovereign wealth funds from the Middle East, Europe, and Asia lined up to invest, and demand exceeded supply by a factor of 2.8. But the broader market sees the company differently: immediately after the offering was announced, the stock dropped 10% and is still trading 6% below the pre-sale price. Why are institutional investors buying while retail investors are selling?

On Wednesday, August 26, Alibaba announced the completion of the largest secondary offering in the history of the Hong Kong Stock Exchange. The company raised HK$80 billion ($10.2 billion) through the issuance of new shares at HK$112.70 per share—an 8% discount to the market price at the time of the announcement.

The terms of the offering prohibited direct participation by U.S. investors. This narrowed the pool of buyers and may have made it more difficult to resell the shares quickly. But even with this restriction, demand was overwhelming: more than $28 billion in bids, 40% of which came from sovereign wealth funds and long-term institutional investors.

The broader market, however, did not react favorably to Alibaba. The stock fell 10% on the day the secondary offering was announced and has since recovered only partially. Yes, the secondary offering diluted existing shareholders’ stake by about 3%, but certainly not by 10%.

The most vocal skeptic is Michael Burry—the real- life inspiration for the main character in the movie *The Big Short*. In a Substack post, Burry announced that he had sold his position in Alibaba, calling the stock overvalued.

In his view, the stock price would have to “fall by half” for him to become interested again. “I can’t endorse the stock split,” added Burry, implying that he is not enthusiastic about the dilution of existing shareholders’ stakes. Instead of Alibaba, he has built a “large position” in its competitor, JD.com.

Why are these investments necessary? The fact is that Alibaba is in the midst of the largest investment cycle in its history. In early 2025, the company announced plans to invest 380 billion yuan ($56.4 billion) in AI infrastructure between 2026 and 2029. By mid-2026, half of that amount had already been spent. The $10.2 billion raised from the additional share offering is also earmarked for these purposes—AI development, the creation of its own AI chips, and the construction of data centers.

In its recently released report for the second quarter of 2026, the company stated that capital expenditures rose 75% year-over-year to 67.68 billion yuan ($10 billion). As a result, net income fell by 75%, even though revenue rose by 9% to 268.95 billion yuan ($40 billion).

CEO Eddie Wu makes no secret of his ambitions. At the Yunqi Conference in September 2025, he stated that, in the end, there will be only 5–6 supercompanies left in the world—providers of AI infrastructure—and Alibaba intends to be one of them. The company’s goal is $100 billion in revenue from cloud and AI services by 2030 (possibly sooner).

During the quarterly earnings conference call, Wu explained the reasoning: “To capitalize on this future growth, we first need to make capital investments to build the necessary computing capacity.”

Yes, Alibaba is still primarily an e-commerce company. According to its quarterly report, revenue from its marketplaces and related services grew 4% year-over-year in the second quarter to $30.3 billion, while the AI division generated $7.1 billion, but its growth rate was quite different—45%.

If we look at earnings before interest, taxes, depreciation, and amortization (EBITDA), the difference in trends is even more striking: e-commerce saw a 1% decline, while AI and computing power saw a 133% increase.

It's no surprise that management sees AI as a promising area.

The problem with Alibaba is that, unlike its American competitors such as Anthropic and OpenAI, its flagship Qwen series AI models are mostly open-source—that is, free to download and use.

Qwen’s dominance in the open-source world is undeniable: over 3 billion downloads and more than 300,000 derivative models created by developers worldwide based on Qwen. This makes Qwen the largest open-source AI ecosystem, surpassing Meta and Google combined. But how can you make money from it?

First, Model-as-a-Service (MaaS). Although the models are free to download, Alibaba offers them through its Alibaba Cloud platform, providing not only computing power but also convenient tools for deploying, training, and operating AI. In this case, customers pay for using the cloud.

Alibaba Cloud is already China’s largest cloud provider and ranks 4th or 5th globally alongside Oracle, making it a natural partner for companies looking to leverage AI, particularly the Qwen family. According to the Chinese technology portal 36Kr, the company’s revenue from MaaS grew 15-fold in the first five months of 2026, thanks to the explosive popularity of AI agents. Morgan Stanley estimates annual recurring revenue from MaaS at 16 billion yuan ($2.4 billion).

Second, a revenue-sharing arrangement with large users may be announced soon. According to exclusive information from Reuters on August 7, Alibaba plans to follow the example of the startup Moonshot and require large commercial users (with annual revenue exceeding $20 million from using the model) to share their profits. Moonshot has set a rate of up to 30%, while Alibaba is negotiating its own terms. This is a proven freemium model from Silicon Valley: give away the AI for free, but charge for its active commercial use.

Third, its own AI chips. In May, Alibaba announced that it had already shipped 560,000 of its Zhenwu AI chips to more than 400 customers across 20 industries. While they currently lag behind the best American models in terms of performance, they are not subject to U.S. export controls and will help improve the overall margins of the cloud business. Development of the next generation of chips will begin in the second half of 2026, with significantly higher computing performance.

Investment banks are confirming this optimism. In a report dated August 21 (available on Oninvest), Morgan Stanley named Alibaba a “Top Pick” with an “Overweight” rating and a price target of $180 (a 38% increase from the current price).

In a report dated August 21, JPMorgan estimated the internal rate of return (IRR) on Alibaba’s AI projects at 22% after taxes, assuming a cost of capital of 10%, with a payback period of 2.9 years—thus confirming the company’s management’s calculations. JPMorgan’s price target is $210 by December 2026 (a 61% increase). According to their assessment, the 12% cloud base margin actually underestimates the profitability of a mature cloud technology business, which is closer to 20%.

Barclays raised its price target to $200 (from $195) while maintaining its “Overweight” rating. Analysts noted that the 45% growth in Alibaba’s cloud AI service marked the ninth consecutive quarter of accelerating growth, and the AI division’s EBITDA rose to 11.6% from approximately 9% in the previous three quarters.

It’s interesting to compare Alibaba with Amazon. While such comparisons are always somewhat arbitrary, they do share some similarities—more than twenty years ago, Amazon was also a marketplace that decided to venture into the cloud business. Jeff Bezos invested hundreds of millions of dollars in data centers and infrastructure, and Wall Street criticized him for it. For example, in 2015, the investment portal Motley Fool called Amazon a “bottomless money pit” and presented calculations showing that the company was spending billions of investors’ money on its AWS cloud service—at a loss to both the company and its investors—instead of focusing on its core business.

AWS currently generates $45.6 billion in operating profit—that’s nearly 60% of the company’s total operating profit of $80 billion for 2025. Amazon’s stock is trading at all-time highs, and the company is worth more than $2 trillion. What do you think of that, Motley Fool?

Alibaba finds itself in a similar situation today: massive investments in infrastructure, falling short-term profits, and market skepticism. But the company's founders believe in it.

According to Reuters, following the announcement of a secondary offering, Alibaba Chairman Joe Tsai and CEO Eddie Wu jointly purchased HK$202 million ($26 million) worth of the company’s stock. Jack Ma, the company’s chief founder and spiritual leader (though he currently holds no official position at the company), invested even more heavily—HK$600 million ($77 million). In English, this is called “skin in the game.” When management buys shares with their own money, it signals to the market that they believe in the chosen strategy. And Jack Ma has thus made it clear that he fully supports them.

Of course, purchases by founders do not guarantee success. Bloomberg notes that Xiaomi is trading 25% below the price at which founder Lei Jun bought shares in November. Tencent Holdings Ltd. also struggled to win back investors in 2022 and 2023, despite a large-scale share buyback program.

This article was AI-translated and verified by a human editor