Explosive Growth Anticipated As Morgan Stanley Forecasts 72% Five‑Year CAGR For China AI Cloud Market, Alibaba Positioned As Primary Beneficiary

Morgan Stanley projects a rapid expansion of China’s AI cloud market, defined as GenAI‑related IaaS plus MaaS. In a report dated March 16, analysts estimate that, driven by a surge in inference compute demand and accelerated adoption of large‑model applications, the market will grow at a compound annual growth rate of 72% from 2024 to 2029, expanding from RMB 15 billion in 2024 to RMB 218 billion by the end of 2029. GenAI’s share of the IaaS plus PaaS market is expected to rise from 6% in 2024 to 39% in 2029, elevating AI cloud from a peripheral segment to the principal growth engine of the cloud industry.

In assessing competitive positioning, Morgan Stanley designates Alibaba as its top pick and assigns an overweight rating with a U.S. target price of USD 180. The bank highlights Alibaba’s full‑stack approach—spanning in‑house T‑Head chips, GPU infrastructure, the Qwen large model, the Bailian model platform and a broad 2C/2B application ecosystem—as the most comprehensive AI infrastructure proposition in China, analogous to Google’s role in the U.S. market. ByteDance is identified as the most disruptive new entrant through its Volcengine business, and Morgan Stanley projects ByteDance’s 2026 capital expenditure at RMB 25 billion, exceeding Alibaba’s RMB 16 billion and Tencent’s RMB 13 billion.

Morgan Stanley further argues that China’s cloud sector is entering its first price‑increase cycle in two decades. The bank’s sensitivity analysis suggests that a 1% rise in Alibaba Cloud’s contract prices could lift EBITA margins by roughly one percentage point and raise EBITA forecasts by about 11%; a 10% increase in contract pricing could expand EBITA margins by four percentage points. This margin elasticity is presented as a material factor not yet fully reflected in market valuations.

Citing IDC, the report notes that China’s overall public cloud market (IaaS, PaaS and SaaS) was approximately USD 45 billion in 2024 and is forecast to reach USD 105 billion by 2029, a CAGR of 18%. IaaS constitutes the largest portion, representing 56% of the public cloud market, with a 2024 IaaS market size near USD 25 billion and an expected five‑year CAGR of 17%. Within this context, GenAI’s expansion is projected to outpace the broader market, with inference workloads emerging as the dominant growth driver. IDC’s forecast indicates that the share of training workloads in GenAI IaaS will decline from 76% in 2024 to 23% in 2029, while inference demand is expected to grow at a 103% CAGR versus 26% for training.

Empirical usage data underscore the inference surge. ByteDance’s Doubao reported daily token usage rising from 4 trillion in December 2024 to 50 trillion by the end of 2025, an increase of more than twelvefold. MiniMax’s annual recurring revenue rose from USD 100 million in December 2025 to USD 150 million by February 2026, a 50% increase, while token consumption expanded sixfold, partly reflecting rapid adoption of OpenClaw. Morgan Stanley also highlights that China’s public cloud market remains relatively underpenetrated—about 10% of the U.S. market in 2024—and that corporate intent to adopt public cloud services has increased from 58% to 71% in the bank’s CIO survey, indicating substantial upside for penetration.

On competitive dynamics, Morgan Stanley observes a structural reconfiguration. Large private cloud providers are regaining share from state‑owned enterprises and telecom operators after a period of aggressive pricing and guaranteed state‑enterprise contracts. Alibaba Cloud’s IaaS share recovered from 25.5% in H1 2024 to 26.8% by Q2 2025, a reversal attributed to superior AI models, stronger supply‑chain access and faster execution by private firms. ByteDance’s Volcengine has rapidly scaled, achieving a 14.2% share in GenAI IaaS and a 37.5% share in MaaS by mid‑2025, while its overall public IaaS share rose from near zero in 2024 to about 4% in Q2 2025. Baidu, despite offering a full‑stack solution with Kunlunxin chips, Wenxin models and the Qianfan platform, is assessed as trailing the market leaders in scale and model capability, earning a neutral rating.

Price adjustments by global cloud providers are already evident. AWS raised prices for certain machine‑learning EC2 capacity blocks by roughly 15% in January 2026, and Google Cloud announced substantial increases for network, storage and AI infrastructure fees effective May 2026. In China, Wangsu Technology implemented CDN price increases of 35%–40% from February 1, 2026; UCloud raised prices for contract renewals and new customers from March 1, 2026; and Tencent’s AI platform increased prices for some proprietary models by up to 400%, with supply‑chain inflation cited as a primary driver.

Morgan Stanley emphasizes three structural levers that could materially improve cloud profitability. First, the shift from training to inference workloads supports higher‑value, token‑based pricing and better server utilization through parallel processing, enhancing margin profiles. Second, proprietary ASICs materially lower infrastructure unit costs; firms with in‑house chips such as Alibaba (T‑Head) and Baidu (Kunlun) can achieve procurement cost advantages exceeding 50% relative to competitors. Third, extending server depreciation schedules—U.S. hyperscalers have moved to six‑year depreciation—would provide additional margin uplift if adopted more broadly in China, where depreciation currently ranges from three to five years.

Overall, Morgan Stanley concludes that China’s AI cloud market is poised for explosive growth, with Alibaba positioned as the primary beneficiary due to its comprehensive stack, ByteDance as a formidable challenger, and an emerging pricing cycle that could materially re‑rate industry profitability.