On August 25, sources told Chinese media that Jack Ma had spent more than 600 million Hong Kong dollars over the previous several days to buy Alibaba shares in Hong Kong. Stock exchange filings also show that Alibaba Chairman Joe Tsai and CEO Eddie Wu have been purchasing shares as well, adding roughly 200 million Hong Kong dollars’ worth over two consecutive days. Together, Ma and Alibaba’s top executives have now bought more than 800 million Hong Kong dollars in stock.

The buying spree wasn’t random. It followed one of the largest share placements in Hong Kong’s history.

On August 24, Alibaba finalized pricing on a massive new stock offering — 80 billion Hong Kong dollars, priced at 112.7 Hong Kong dollars per share, for a total of 710 million new shares. It is the company’s first equity refinancing since its Hong Kong listing in 2019, and one of the biggest placements the Hong Kong market has ever seen.

Market reaction was sharply divided. Sovereign wealth funds from the Middle East and Europe oversubscribed the offering nearly three times over — in under an hour. Retail investors, however, did the opposite: Alibaba’s stock fell 8.54 percent that same day.

Alibaba says every dollar of that 80 billion is going straight into AI — building out infrastructure and full-stack AI capabilities. This isn’t just a balance-sheet top-up. It is the latest move in a much larger campaign: a three-year, 380-billion-yuan AI investment plan the company announced earlier this year. How this bet plays out could define Alibaba’s position for the next decade — and right now, investors are still weighing the odds.

A Cash Flow Stress Test

On paper, Alibaba isn’t short on cash. As of June 30, 2026, the company held 474.5 billion yuan in cash and liquid investments, with a debt ratio of just 43.23 percent — a balance sheet Fitch rates A with a stable outlook, putting Alibaba on par with Tencent as one of the best-capitalized tech companies in China.

So why opt for equity dilution over debt? The answer lies in how quickly AI is burning through cash.

In the March 2025 quarter, Alibaba’s free cash flow fell 76 percent year-over-year to 3.74 billion yuan — with the company pointing directly to rising cloud infrastructure spending. By June 2025, free cash flow had turned negative, with a net outflow of 18.8 billion yuan. By the April-to-June quarter of 2026, that outflow had widened to 44.7 billion yuan.

Meanwhile, capital expenditure jumped 75 percent year-over-year to 67.7 billion yuan this quarter, while operating cash flow grew just 11 percent to 22.9 billion yuan. In other words: AI cloud growth is still in the investment phase, and operating cash flow simply isn’t keeping up with the pace of spending.

CEO Eddie Wu was blunt about it on the earnings call: AI computing investment follows a capex-first business model. To grow fast, you have to build data centers before demand materializes. Alibaba’s three-year, 380-billion-yuan AI investment plan — announced in February 2025 — had already deployed about 190 billion yuan of that as of June this year.

What’s more telling is that Alibaba chose to raise money by selling stock rather than issuing debt. With 474.5 billion yuan in cash and an A credit rating, debt financing should have been cheap and readily available. Instead, management opted for dilution — likely because they didn’t want interest payments eating further into a bottom line that was already down 75 percent year-over-year.

For Alibaba’s leadership, the AI arms race has a narrowing window, and that window matters more than short-term stock price or dilution. The 80 billion Hong Kong dollars is, in effect, equity traded for time — a bet that near-term earnings-per-share pain is worth a guaranteed seat at the AI computing table for years to come.

Nomura’s take: this placement removes the long-standing uncertainty hanging over Alibaba’s finances. Nomura had actually expected the company to rely more on debt for this round of funding, but a wave of AI-driven capital spending across the industry has widened credit spreads and made new bond issuances less attractive — pushing more big companies toward equity instead. Nomura says the dilution is understandable given the current credit environment, and that the placement fully funds Alibaba’s AI ambitions in one shot.

The Math Behind a 49.5-Billion-Yuan AI Business

The real question investors want answered: what does Alibaba actually get for 80 billion Hong Kong dollars?

Management gave a confident answer. On the August 20 earnings call, Eddie Wu laid out an AI investment payback model for the first time: he said the return on AI computing capex is highly predictable, with a three-year payback period — one that could shrink to as little as two to two-and-a-half years as margins on AI products continue to improve.

The number behind that confidence: Alibaba’s AI-related annualized revenue has now surpassed 49.5 billion yuan, and Wu expects it to approach 10 billion dollars next quarter. AI cloud and computing revenue hit 48.4 billion yuan this quarter, up 45 percent year-over-year — the fastest growth in 22 quarters. AI product revenue specifically came in at 12.4 billion yuan, marking its twelfth consecutive quarter of triple-digit year-over-year growth.

CFO Toby Xu added that as synergies across Alibaba’s core businesses deepen and more AI products are commercialized, the company will have greater financial and strategic flexibility going forward. He noted that AI servers typically pay for themselves within three years; assuming a five-year useful life, that leaves at least two more years of positive free cash flow after breakeven.

Still, not everyone is convinced. Prominent hedge fund manager Michael Burry said publicly after the placement that Alibaba is shifting from a capital-light, high-cash-conversion platform into a capital-heavy model where returns on invested capital are under real pressure. Bloomberg Intelligence analyst Robert Lea went further, arguing that given already-thin AI investment returns, markets aren’t just reacting to dilution — they are starting to see Alibaba less as a tech platform and more as a low-margin infrastructure business.

After the placement was announced, Alibaba’s Hong Kong shares fell as much as 10 percent intraday on August 24, before closing down 8.54 percent at 112.5 Hong Kong dollars, leaving the company with a market cap of 2.16 trillion Hong Kong dollars.

Not everyone agrees the market got it right. Some analysts argue the short-term pain is overblown, and that the focus on dilution misses the bigger picture — Alibaba just strengthened its strategic position. Nomura maintains a “buy” rating, calling Alibaba one of the best-positioned companies in China’s AI ecosystem, with full-stack capabilities spanning foundation models, model-as-a-service, cloud infrastructure, and its own AI chips through subsidiary T-Head. Bank of America also kept its “buy” rating, saying the placement strengthens Alibaba’s balance sheet, diversifies its funding sources, and gives the company the capital cushion it needs as its cloud business and AI investments start to pay off.

A Global Arms Race Nobody Wants to Lose

Alibaba isn’t alone in this. Around the same time as its earnings report, practically every major Chinese tech company laid its AI cards on the table.

The week before Alibaba’s earnings, Tencent posted a similarly eye-catching number: capital expenditure of 52.78 billion yuan in the second quarter of 2026, up 176 percent year-over-year and 65 percent from the previous quarter. Tencent’s free cash flow also turned negative for the first time in years, with a net outflow of 13.8 billion yuan. Tencent President Martin Lau said on the earnings call that the spending spike is entirely focused on new AI-native businesses — but he was careful to add that this looks more like a concentrated investment for this year and next, rather than something investors should expect every year going forward.

ByteDance, according to Bloomberg, is reportedly discussing a budget of up to 70 billion dollars for AI data centers and chips in 2026. It is still an internal upper-bound figure, not a finalized budget — but it is enough to turn heads across the industry.

Zoom out further, and Wall Street’s numbers are just as staggering. Based on earnings reports and calls, Microsoft, Alphabet, Amazon and Meta are on track to spend a combined total of more than 700 billion dollars in capital expenditure in 2026. Alphabet posted its first-ever negative quarterly free cash flow as a public company.

According to Dealogic, in just the first five months of 2026, Alphabet, Amazon, Meta, Microsoft and Oracle together issued 159 billion dollars in bonds globally. In June 2026, Alphabet raised its planned equity financing from 80 billion to 84.75 billion dollars — the largest equity raise in corporate history.

Morgan Stanley’s chief global economist, Seth Carpenter, estimates that data center and related infrastructure capex will exceed 3 trillion dollars between 2025 and 2028 — and so far, only about a quarter of that has actually been put to use. Goldman Sachs projects 4 to 8 trillion dollars will flow into AI infrastructure over the next five years.

Different companies, different paths — but they are all trading today’s profits for a seat at tomorrow’s table. Because in this particular race, falling behind costs more than overspending ever could. As Goldman Sachs Global Research put it: the engine driving this entire revolution isn’t earnings or stock prices. It is FOMO — and fear of missing out has proven to be a more powerful motivator than either.