Quantitative Models and Construction Diffusion Index Model - **Model Name**: Diffusion Index Model [5][27] - **Model Construction Idea**: This model is based on the principle of price momentum, aiming to capture upward trends in industry performance. It selects industries with strong upward momentum for allocation [26][27]. - **Model Construction Process**: - The model calculates the diffusion index for each industry, which reflects the proportion of stocks in the industry with positive momentum. - Industries are ranked based on their diffusion index values, and the top-performing industries are selected for portfolio allocation. - Example rankings as of September 12, 2025: Comprehensive (0.99), Banking (0.969), Communication (0.951), Steel (0.95), Non-ferrous Metals (0.947), Retail (0.934) [27][28]. - **Model Evaluation**: The model has shown mixed performance over the years. While it captured trends effectively in 2021, it struggled during market reversals in 2024. It has been tracking performance for four years [26][27]. - **Testing Results**: - 2025 YTD excess return: 4.53% [25][30] - September 2025 average weekly return: 2.79%, excess return over equal-weighted industry index: 0.88% [30] GRU Factor Model - **Model Name**: GRU Factor Model [6][33] - **Model Construction Idea**: This model leverages GRU (Gated Recurrent Unit) deep learning networks to process high-frequency volume and price data, aiming to identify industry rotation opportunities [38]. - **Model Construction Process**: - The GRU network is trained on historical minute-level volume and price data to generate industry factors. - Industries are ranked based on their GRU factor scores, and the top-ranked industries are selected for allocation. - Example rankings as of September 12, 2025: Comprehensive (4.56), Construction (3.8), Real Estate (3.6), Textile & Apparel (0.08), Comprehensive Finance (-0.07), Home Appliances (-0.16) [6][34]. - **Model Evaluation**: The model performs well in short-term scenarios but struggles in long-term or extreme market conditions. It has faced challenges in capturing excess returns in 2025 due to concentrated market themes [33][38]. - **Testing Results**: - 2025 YTD excess return: -7.37% [33][36] - September 2025 average weekly return: 1.70%, excess return over equal-weighted industry index: -0.23% [36] Backtesting Results of Models Diffusion Index Model - **YTD Excess Return**: 4.53% [25][30] - **September 2025 Weekly Average Return**: 2.79% [30] - **September 2025 Weekly Excess Return**: 0.88% [30] GRU Factor Model - **YTD Excess Return**: -7.37% [33][36] - **September 2025 Weekly Average Return**: 1.70% [36] - **September 2025 Weekly Excess Return**: -0.23% [36]

Market Overview - The Shanghai Composite Index rose by 0.41% today, with a trading volume of 728.77 million shares and a total transaction value of 878.71 billion yuan, an increase of 9.03% compared to the previous trading day [1] Industry Performance - The banking sector led the gains with an increase of 1.65%, followed by steel at 1.47% and comprehensive industries at 1.41% [1] - The sectors with the largest declines included beauty and personal care, down 2.23%, non-ferrous metals, down 1.17%, and textiles and apparel, down 0.97% [1] Key Stocks - Notable gainers included Shanghai Pudong Development Bank, which rose by 2.89%, and Liugang Co., which surged by 10.02% [1] - In contrast, significant decliners included companies like Feiya Group, which fell by 7.61%, and Nanshan Holdings, down 6.31% [1] ETF Insights - The Food and Beverage ETF (515170) has seen a slight decline of 0.35% over the past five days, with a P/E ratio of 19.83 and a net outflow of 796,000 yuan [3] - The Gaming ETF (159869) increased by 3.68% in the last five days, with a P/E ratio of 42.76 and a significant net outflow of 110 million yuan [3] - The Cloud Computing 50 ETF (516630) experienced a minor increase of 0.08%, with a high P/E ratio of 99.46 and a net inflow of 405,000 yuan [4]