Summary of Key Points from the Conference Call Industry Overview - The focus of the conference call is on the **Chinese power utilities sector**, particularly in relation to **AI Data Centers (AIDCs)** and the integration of renewable energy as outlined in the **15th Five-Year Plan (FYP)** [2][9]. Core Insights and Arguments - The **15th FYP** emphasizes the integration of renewable energy with AI computing, promoting **"compute-power integration"** to support AIDCs [2][9]. - AIDCs are expected to significantly increase their power consumption, potentially reaching **9% of China's total power demand by 2030**, driven by the rise of AI technologies like **OpenClaw** [3][20]. - The **energy infrastructure** in China is more advanced compared to the US, providing a competitive edge in the AI sector due to lower electricity prices and a more efficient power grid [4][30]. - The report recommends a **Buy** rating for **Longyuan Power (916 HK/001289 CH)** and **China Power International (2380 HK)**, with target prices raised to reflect new FX assumptions [5][62]. Potential Growth and Consumption Projections - Power consumption at AIDCs is projected to grow **5x** by 2030, with a potential impact on total power consumption reaching **0.5%** in a bull-case scenario [3][26]. - The report outlines a scenario analysis for token requests from overseas users of Chinese AI models, estimating that power consumption could account for **0.002%** of total power demand in 2026 under a base-case scenario [20][26]. Competitive Advantages - China's energy infrastructure is characterized by a **stronger power generation and transmission network**, which is less constrained compared to the aging infrastructure in the US [4][45]. - Chinese AI models are more cost-effective, with lower average token prices compared to US models, making them attractive to overseas users [31][32]. Risks and Challenges - Several risks are associated with the export of Chinese AI tokens, including regulatory risks, GPU constraints, and potential geopolitical tensions that could affect market access [33][29]. - The actual energy consumption may vary based on the architecture of AI models and the computing power of GPUs used by overseas clients [29]. Company-Specific Insights - **Longyuan Power** is positioned to benefit from the anticipated surge in demand for AI-driven green electricity, with a focus on wind and solar power generation [56]. - **China Power International** aims to reduce its coal capacity exposure to below **10%** by 2025, focusing on expanding its renewable energy investments [60][61]. Financial Projections - Longyuan's revenue is projected to grow from **RMB 31.37 billion** in 2024 to **RMB 36.34 billion** by 2027, with a net profit increase from **RMB 6.38 billion** to **RMB 7.64 billion** over the same period [63]. - China Power's target price is set at **HKD 3.90**, reflecting an upside potential of **17.5%** from current levels [62]. Conclusion - The integration of renewable energy with AI computing in China presents significant investment opportunities, particularly for companies like Longyuan and China Power, which are well-positioned to capitalize on the growing demand for sustainable energy solutions in the AI sector [5][61].