Alibaba's Chip Spinoff: A Flow Analysis of Capital and Price

The immediate catalyst is clear: news of a potential chip unit IPO sparked a sharp pre-market move. Alibaba's shares rose more than 5% in Thursday's pre-market trade following the Bloomberg report. This sets the stage for a flow analysis of the capital and price action that follows.

The stock's current context is one of significant undervaluation relative to its peak. Trading at $126.06 with a market cap of $302.86B, it sits well below its 52-week high of $192.67. This price action suggests the market is pricing in substantial risk, making any positive catalyst like a spinoff a potential source of re-rating.

A key valuation catalyst is the institutional view. Morgan Stanley reiterated an Overweight rating and $180.00 price target on the stock. The firm sees the spinoff and broader AI developments as positive signals for commercialization, which could support a higher sum-of-the-parts valuation. This analyst target implies a 42% upside from current levels, providing a concrete flow target for the stock.

The Capital Flow: IPO and Chip Revenue

The primary flow from the spinoff is capital, aimed at navigating U.S. sanctions. The IPO plan is a direct response to the ongoing tensions between the U.S. and China, especially in semiconductors. By listing the chip unit, Alibaba seeks to raise funds independently, reducing its exposure to export controls and tariffs that have recently targeted key competitors.

The chip's standalone profit-generating capability is already being demonstrated. T-Head's latest Zhenwu PPU features 96 GB of HBM2e memory and has been deployed in clusters serving over 400 customers, including major state-owned enterprises. This hardware, positioned between Nvidia's A800 and H20, is now being used for AI training and inference, creating a direct revenue stream from a high-demand, domestically developed product.

A new revenue model is emerging from the AI business group. The newly formed Alibaba Token Hub Business Group will focus on digital assistants powered by AI models that consume tokens. Morgan Stanley notes this group will likely have a token-based revenue and cost structure, similar to pure-play foundation model companies. This shift from bundled cloud services to a usage-based model could unlock a new, scalable profit center.

The Valuation and What to Watch

The immediate valuation impact hinges on the sum-of-the-parts model. Morgan Stanley assigns a high-end component value of $19 per share to Qwen. a figure not included in its mid-range sum-of-the-parts estimate. This suggests the market could begin to price in the standalone value of Alibaba's AI models and the new token-based business group, providing a direct flow target for the stock's re-rating.

Three key watchpoints will determine the success of this flow. First is the IPO timeline for T-Head, which remains uncertain. Second is chip sales volume, as the unit must prove it can expand beyond the Alibaba ecosystem to external customers. Third is the AI token monetization rate, which will dictate the profitability of the new Token Hub Business Group.

The new Zhenwu PPU chip is central to this thesis. It features 96 GB of HBM2e memory and is positioned as a domestic alternative to Nvidia's H20. The chip has already been deployed in clusters serving over 400 customers, including major state-owned enterprises, creating a tangible revenue stream from a high-demand product.