Investors have a habit of forgiving conviction calls that come too early. Michael Burry has spent the past two years testing how far that patience extends with Chinese tech stocks, buying and dumping the same names in cycles most traders would find dizzying.
This time, the timing worked in his favor almost immediately.
Scion Asset Management built Alibaba (BABA) into its largest holding in mid-2024, then liquidated the entire position in the first quarter of 2025 and bought bearish put options against, along with JD.com, Baidu, and PDD, according to Benzinga’s review of SEC filings.
By April, he had reversed again, disclosing on Substack that he had opened a new Alibaba stake and added meaningfully to JD.com.
That reversal did not last long. Burry now says he moved his entire Alibaba position into JD.com months ago, initially planning to shift capital back within a month or two.
That plan is off. In a post on X (the former Twitter), Burry said he will not flip any of it back to Alibaba, calling share issuance “a new paradigm again for BABA” and warning that the company’s return on invested capital will keep falling as a result.
Alibaba priced its offering hours after Burry spoke
Alibaba said on Sunday, Aug. 23, that it planned to sell HK$80 billion, or roughly $10.2 billion, in a Hong Kong share placement to fund its artificial intelligence buildout, according to Reuters.
The deal marks the largest primary follow-on offering ever by a Hong Kong-listed company, ranking third-largest globally this year, behind only Alphabet and Intel.
Related: Michael Burry says Nvidia rival is quietly getting serious
The company issued 710 million new shares priced at HK$112.70 each, an 8.4% discount to the prior close, Reuters noted. That discount matters because steep pricing signals a company prioritizing speed of capital over shareholder-friendly terms, exactly the dynamic Burry flagged.
Demand was not the problem. The offering was nearly three times oversubscribed, with total demand reaching roughly $28 billion, including close to $6 billion from sovereign wealth funds and long-only investors.
Strong demand and a falling stock price can coexist when existing holders are the ones absorbing the dilution.
Alibaba shares opened down roughly 8% in Hong Kong trading Monday, Aug. 24, falling as much as 10% intraday.
For a stock that had been one of the better-performing large caps in Chinese tech this year, an 8% single-day drop on a widely anticipated fundraise says something about how the market is now pricing dilution risk against AI ambition.
Alibaba’s own numbers complicate Burry’s argument
The case against Burry’s framing comes from Alibaba itself.
On its earnings call the prior week, management said it had already committed close to half of its three-year AI capital spending plan. It added that the expected payback period on that spending had narrowed to roughly two-and-a-half years, down from three, according to a BigGo Finance report.
That detail rarely makes it into coverage built around Burry’s quotes, but it is the crux of the disagreement. Burry is arguing that repeated share issuance destroys per-share value faster than the underlying AI investments can generate returns.
More Manager Buy/Sells:
- Michael Burry increases his bet against popular chip giant
- Warren Buffett reveals he broke his own investing pattern
- Mark Cuban bets on MLB with Athletics minority stake
Alibaba is arguing the opposite, that the payback window is shrinking and the capital is buying market position that would cost more to build later.
Both claims can be true, depending on the time horizon an investor chooses. Burry’s ROIC concern is a near-term accounting reality. Alibaba’s payback claim is a forward bet on China’s AI infrastructure race still being winnable from here.
A financing pattern is emerging across Chinese AI spending
What makes this moment bigger than one investor’s trade is the financing method itself.
Alphabet and Intel are the only two companies globally that have raised more through a single primary follow-on offering this year, according to Reuters, and both did so to fund the same kind of capital-intensive AI infrastructure buildout now driving Alibaba’s raise.
Chinese tech companies have historically leaned on retained earnings, debt, or Hong Kong listing proceeds rather than dilutive follow-on placements at this scale.
If Alibaba’s raise is replicated by Baidu, JD.com, or Tencent as their own AI infrastructure bills come due, the market may need to start treating share dilution as a standing cost of staying competitive in AI, not a one-time event tied to a single earnings cycle.
That is the number worth watching long after this week’s headlines fade.
Related: Bill Gates has made $2.3 billion on Michael Burry’s former favorite