BofA Securities has reiterated a "Buy" rating on Alibaba-W (09988, BABA.US) with a US$175 price target for its US-listed shares and a HK$172 target for its Hong Kong-listed shares, following the company's flagship Apsara Conference held this week. Alibaba's CEO Eddie Wu reaffirmed the company's long-term commitment to artificial intelligence, emphasizing continued investment in AI models, chips, and cloud infrastructure as AI advances toward machine intelligence.
The bank observed robust and rapidly accelerating customer demand, noting Alibaba's clear strategic commitment to AI agents and the ongoing development of recursive self-improvement (RSI) architecture. The company is demonstrating increasingly autonomous and self-reinforcing full-stack capabilities across infrastructure, models, platforms, and applications. This conference reinforced BofA's positive outlook on Alibaba, as its industry-leading full-stack AI ecosystem enables value capture across multiple layers of the AI stack, with cloud driving near-term monetization and AI services while applications and international expansion support longer-term growth.
The next potential catalyst is the quarterly update for the period ending September, expected before the October blackout period. The bank noted Alibaba's plan to operate over 20 gigawatts of global data center capacity by 2032, highlighting management's confidence that medium-to-long-term AI computing demand will significantly exceed current supply, implying an average annual addition of roughly 2 gigawatts. BofA estimates each gigawatt carries a construction cost of approximately US$30 billion to US$35 billion, but expects a substantial portion to be financed through leasing and other partnership arrangements, limiting Alibaba's direct capital expenditure burden. As a result, the bank maintains its capex forecast of RMB 230 billion to 250 billion for fiscal years 2027-2028.
This 20-gigawatt target broadly aligns with management's guidance of US$100 billion in external AI and cloud revenue by fiscal 2031 (approximately calendar 2030). Alibaba's semiconductor arm, T-Head, unveiled its latest-generation AI accelerator, the Zhenwu V900, which management describes as China's highest-performing AI chip. The chip delivers three times the performance of the Zhenwu M890 and can scale to clusters of up to 500,000 chips, supporting frontier model training and inference. Alibaba's management expects T-Head shipments to rise significantly in coming years, targeting an increase in self-developed chips' share of Alibaba Cloud's deployed computing capacity from under 10% currently to 40%-50% over time. The bank views this as a key driver for improving supply chain control and enhancing cost efficiency across AI infrastructure.
On the Tongyi roadmap, the company plans to further advance the Qwen Max series, with Qwen 4.5/5 targeting 5 trillion to 10 trillion parameters, while also addressing broader deployment needs through Flash, Plus, and local models. Qwen 4 is expected to launch imminently, with upgraded multimodal capabilities spanning image, video, audio, and world models.
Comments