On February 13, 2025, Cai Chongxin confirmed at the World Government Summit (WGS 2025) in Dubai that Alibaba and Apple had reached a cooperation agreement. This news not only highlights Apple's deepening localisation strategy in the Chinese market but also underscores the irreplaceability of China's AI industry chain within the global tech ecosystem.

Combined with the recent global developer frenzy sparked by DeepSeek's open-source large language model, the value logic of China's AI-related listed companies is undergoing a fundamental restructuring.

Core drivers behind the AI asset boom in China

1. Technological breakthroughs and cost revolution: DeepSeek's "Catfish Effect"

China's large language model DeepSeek has rapidly emerged as a key player in the global AI ecosystem thanks to its open-source nature, high performance, and cost-effectiveness.Its inference cost is much lower than that of the international mainstream model, prompting Chinese internet companies to rapidly integrate AI capabilities and forcing chipmakers to enhance competitiveness. This technological leap not only reshapes market expectations for AI commercialisation potential but also attracts European enterprises to shift from OpenAI to the Chinese ecosystem.

2. Capital flows and valuation restructuring

According to Morgan Stanley data, since 2025, southbound funds have flowed into Hong Kong for 47 consecutive weeks, Net inflows from February 6th to 12th reached $3.2 billion. Foreign capital's allocation to Chinese tech stocks has significantly increased, with the MSCI China Index trading at a forward P/E ratio of just 11x—far below India’s 21x and the Nasdaq 100’s multiple. Global hedge funds have recently boosted the total market value of China's onshore and offshore markets by more than $1.30 trillion, contrasting sharply with India’s $720 billion outflow during the same period.

3. Policy support and industrial synergy

Government policies prioritizing AI infrastructure (e.g., computing power networks, data centres) are accelerating domestic substitution. Cases like Beijing’s state-owned enterprises adopting DeepSeek demonstrate AI technology’s penetration from internet sectors into traditional industries, forming a full-chain synergy of "upstream hardware—midstream models—downstream applications."

*Reference：https://news.china.com/socialgd/10000169/20250218/47982442_all.html

AI industrial chain & investment themes

1. Upstream computing power: Valuation restructuring of domestic chips and IDCs

• Chip manufacturing: DeepSeek’s low-cost model boosts demand for domestic GPUs, accelerating alternatives to NVIDIA’s ecosystem (e.g., Huawei Ascend, Cambricon).

• Data Centres (IDCs): surging AI computing needs drive valuation recovery for third-party IDC leaders like Future Data. Huatai Securities predicts domestic IDC vacancy rates will decline, benefiting companies with abundant core-region cabinet resources.

2. Midstream models & toolchains: Long-term benefits of open-source ecosystem

• DeepSeek’s fully open-source strategy lowers entry barriers, fostering a thriving ecosystem of small/mid-sized developers.

• Toolchain enterprises (algorithm optimisation, data labelling) gain incremental markets; Morgan Stanley forecasts a 30% service market expansion by 2025.

3. Downstream applications: Commercialisation explosions from consumer internet to autonomous driving

• Consumer scenes: Meitu Company achieves gross profit margin improvement through AI imaging and design tools. According to Meitu Company's forecast, Net Profit will increase by 52% -60% in 2024, with AI subscription revenue becoming the core driving force.

• Autonomous driving: Expanded L4 pilot policies and deepening partnerships between algorithm firms and automakers (e.g., BYD, NIO) may trigger order surges for lidar and high-precision map providers.

• Enterprise services: Alibaba Cloud and Tencent Cloud optimise public cloud services with AI models; Morgan Stanley expects their capital spending to shift toward AI infrastructure.

*Reference：

1. https://news.qq.com/rain/a/20250218A027VH00

2. https://www.sohu.com/a/849136944_122004016

Risks & Strategies: Balancing short-term volatility and long-term opportunities

1. Short-term risks: The Hang Seng Tech Index has risen 17% recently, with surging call option volumes, signalling potential technical corrections.

2. Long-term opportunities: Goldman Sachs maintains an "overweight" rating on the MSCI China Index, forecasting a 14% rise in 2025; BlackRock sees structural opportunities in AI-driven tech stocks over the next 36 months.

*Reference：https://www.xincai.com/article/neisaue7585168

The "Golden Decade" of China’s AI assets

DeepSeek’s emergence marks China’s shift from "technology follower" to "ecosystem leader." Driven by policy dividends, capital inflows, and technological breakthroughs, AI industry valuation systems are being redefined. On the industrial chain, sectors such as computer/media (AI applications), robotics/electronics/intelligent driving (end-side AI), and domestic computing power chains are expected to benefit.

1. Core infrastructure layer: Computing power + data barriers

2. Vertical domain leaders: Commercialisation acceleration

3. Technology breakthrough beneficiaries: Architectural innovation + ecosystem restructuring

4. Cross-linking with US AI+ sectors

• Ecosystem collaboration: Supply chain partnerships or data sharing (e.g., Medallia integrating DeepSeek).

• Policy catalysis: Accelerated valuation recovery for China’s computing power infrastructure and AI licenses.

• Market sentiment spillover: US AI giants’ earnings beats fuel AI theme chasing, with Chinese stocks acting as "mirror" beneficiaries.

AI+Chips

Generative AI

Autonomous driving

AI+ healthcare

AI+education