Summary: Morgan Stanley argues that control of the entire artificial intelligence technology stack - from semiconductors through cloud services to foundation models and applications - will be a key differentiator among tech firms over the long term. On that basis, the bank elevated Alibaba to its top pick among China's major internet platforms, pointing to proprietary chips, extensive cloud infrastructure and a growing family of models and applications as strategic advantages despite short-term earnings headwinds.
Morgan Stanley's analysts laid out their thesis in a note that emphasized full-stack ownership. "Ownership of the full AI stack - chips, cloud infrastructure, models, and applications - will differentiate long-term AI winners from laggards," the analysts wrote.
The bank singled out Alibaba as standing out among China's large internet companies. It upgraded the company to its top pick, replacing Tencent, and noted that this change comes even as Alibaba faces near-term pressure on earnings.
At the center of the bank's case is Alibaba's semiconductor arm, T-Head. Morgan Stanley described T-Head as a central pillar of Alibaba's strategy, arguing that developing proprietary AI chips can cut dependence on third-party suppliers, allow hardware to be tuned for specific workloads and improve cost efficiency in an inference-heavy AI market like China.
The analysts added that in-house chips can help companies expand computing capacity during demand spikes, limit exposure to export restrictions and reduce regulatory risk by aligning with national priorities on technological self-sufficiency.
Alibaba ties its chip work into its cloud business through AliCloud, creating an integrated platform that pairs semiconductor development with large-scale cloud infrastructure. The bank highlighted the company's foundation models, including the Qwen family, and the application-layer services that operate across Alibaba's ecosystem.
"Alibaba stands out for its in-house chips (T-Head), cloud infra (AliCloud), SOTA open-weight models (Qwen), and consumption centric applications (Qwen apps)," Morgan Stanley's analysts wrote. They said this vertical integration "enables Alibaba to optimize the entire AI value chain and monetize AI across both infrastructure and applications."
Morgan Stanley also pointed to the expanding scale of China's AI semiconductor market. The bank projects the country's AI chip total addressable market could reach $67 billion by 2030, and it estimated that domestic supply could potentially cover about 76% of demand as local capacity grows.
For investors considering Alibaba, the note acknowledged the trade-off between the company's strategic positioning in AI and its near-term earnings pressures. The bank's upgrade to top pick reflects a longer-term view centered on the economics of owning multiple layers of the AI stack.
Separately, the article included a description of an AI-driven stock evaluation tool that assesses Alibaba among thousands of companies using more than 100 financial metrics. The description stated that the AI identifies stocks with attractive risk-reward profiles based on fundamentals, momentum and valuation, and it cited past winners highlighted by the tool, including Super Micro Computer (+185%) and AppLovin (+157%).