华夏中证人工智能ETF基金投资价值分析：DeepSeek引爆AI新周期

Quantitative Models and Construction Methods 1. Model Name: CSI Artificial Intelligence Theme Index - Model Construction Idea: The index selects stocks from companies involved in providing foundational resources, technologies, and application support for artificial intelligence, aiming to reflect the overall performance of AI-themed listed companies[16][20] - Model Construction Process: - Sample Space: The sample space is consistent with the CSI All Share Index, including A-shares and depositary receipts issued by red-chip companies. Exclusions include ST/*ST securities, and securities listed on the STAR Market or Beijing Stock Exchange for less than one or two years, respectively. Other securities must have been listed for over one quarter unless their average daily market capitalization ranks in the top 30 since listing[17] - Investability Screening: Securities with average daily trading volume in the top 80% over the past year are selected[18] - Selection Method: 1. From the investable sample space, companies involved in fields such as big data, cloud computing, machine learning, intelligent chips, etc., are shortlisted 2. The top 50 securities by average daily market capitalization over the past year are selected as index components[19] - Index Adjustment: The index is adjusted semi-annually, with changes implemented on the next trading day after the second Friday of June and December[20] 2. Model Name: AI Industry Profitability Forecast - Model Construction Idea: The model forecasts the revenue and net profit growth of the CSI Artificial Intelligence Theme Index to evaluate its profitability and growth potential[27] - Model Construction Process: - Revenue Forecast: - 2025E: 9,723 billion RMB (+22.62% YoY) - 2026E: 11,198 billion RMB (+15.17% YoY) - 2027E: 12,954 billion RMB (+15.68% YoY) - Net Profit Forecast: - 2025E: 816 billion RMB (+89.59% YoY) - 2026E: 1,076 billion RMB (+31.86% YoY) - 2027E: 1,334 billion RMB (+23.95% YoY)[27] Model Backtesting Results 1. CSI Artificial Intelligence Theme Index - Revenue Growth: - 2025E: 22.62% - 2026E: 15.17% - 2027E: 15.68%[27] - Net Profit Growth: - 2025E: 89.59% - 2026E: 31.86% - 2027E: 23.95%[27] Quantitative Factors and Construction Methods 1. Factor Name: Market Capitalization Distribution - Factor Construction Idea: The factor evaluates the distribution of market capitalization among index constituents to assess concentration and risk[24] - Factor Construction Process: - Stocks with market capitalization >1,000 billion RMB account for 58.43% of the index weight - Stocks with market capitalization between 500–1,000 billion RMB account for 18.77% - Stocks with market capitalization between 200–500 billion RMB account for 22.37% - Stocks with market capitalization <200 billion RMB account for only 0.43%, with no stocks below 100 billion RMB[25] 2. Factor Name: Industry Exposure - Factor Construction Idea: The factor identifies the industry distribution of index constituents to highlight thematic focus[22][23] - Factor Construction Process: - Primary Industry Exposure: - Electronics: 35.88% - Computers: 30.96% - Communications: 24.72% - Tertiary Industry Exposure: - Integrated Circuits: 29.53% - Network Equipment: 24.30% - General Computing Equipment: 9.90%[23] 3. Factor Name: Conceptual Exposure - Factor Construction Idea: The factor evaluates the thematic relevance of index constituents based on their exposure to AI-related concepts[26] - Factor Construction Process: - TMT: 94.63% - Technology Leaders: 92.36% - AI+: 66.91% - AI Applications: 44.71% - AI Computing Power: 51.66%[26] Factor Backtesting Results 1. Market Capitalization Distribution - Weight Distribution: - >1,000 billion RMB: 58.43% - 500–1,000 billion RMB: 18.77% - 200–500 billion RMB: 22.37% - <200 billion RMB: 0.43%[25] 2. Industry Exposure - Primary Industry Weights: - Electronics: 35.88% - Computers: 30.96% - Communications: 24.72%[23] - Tertiary Industry Weights: - Integrated Circuits: 29.53% - Network Equipment: 24.30% - General Computing Equipment: 9.90%[23] 3. Conceptual Exposure - Conceptual Weights: - TMT: 94.63% - Technology Leaders: 92.36% - AI+: 66.91% - AI Applications: 44.71% - AI Computing Power: 51.66%[26]