Alibaba's CXMT Stake Hits ~20x Return After Memory Chip IPO

Alibaba's investment memo just got a very public mark-to-market. According to the South China Morning Post, the company's roughly 5 percent stake in newly listed Chinese memory maker CXMT is now worth more than 140 billion yuan, against about 7.6 billion yuan Alibaba has put in since 2021. That works out to nearly a 20-times paper return, calculated on CXMT's Tuesday closing market capitalisation of around 3.14 trillion yuan, or about US$464 billion.

The number is eye-catching, but the more interesting read is what it says about the shape of Alibaba as an investor. The SCMP frames CXMT as one node in a portfolio that also includes about 36 percent of Kimi-maker Moonshot, taken for roughly US$800 million per Alibaba's fiscal 2024 filings, plus minority positions in Zhipu (which trades overseas as Z.ai) and MiniMax, both of which listed on the Hong Kong Stock Exchange in early 2026. Baichuan and 01.AI are also in the book, though the reporting notes those two have shifted away from foundational AI models.

Taken together, the pattern looks less like a conglomerate and more like a house fund. Alibaba is buying influence across the AI stack, memory underneath and model companies on top, without trying to own any of them outright. That's a very different posture from the control-acquisition era, and a comfortable one to hold while the frontier is still moving and Beijing is watching who owns what.

The honest caveat is that 140 billion yuan is a first-day-of-trading mark, not a cheque that has cleared. Memory prices, U.S. export controls on Chinese DRAM, and any Beijing-side view on whether Alibaba is even allowed to trim a strategically important holding all sit between the paper gain and a realised one. What the reporting doesn't give you is the lockup terms on the CXMT stake, the structure of any supply agreements between CXMT and Alibaba Cloud, or how the model-company positions are being carried on Alibaba's own books.

If CXMT's valuation holds, though, the read for the rest of the sector is that quiet minority bets on Chinese hard-tech infrastructure are the trade that has paid this cycle, not splashy platform acquisitions. Whether Alibaba can repeat the trick with Moonshot, Zhipu and MiniMax is the more interesting question for the next twelve months.