Quantitative Models and Construction Methods 1. Model Name: Fund Selection Strategy Based on Trading Motivation Factor and Stock Spread Income Factor - **Model Construction Idea**: This strategy combines the trading motivation factor and the stock spread income factor to select funds with high stock spread income, active trading motivation, and low likelihood of performance manipulation[2][24] - **Model Construction Process**: - The **trading motivation factor** is derived from fund report data, including fund flows, stock buy/sell amounts, and the proportion of top 20 stocks traded[47] - The **stock spread income factor** is calculated from the stock spread income in the fund's profit statement[47] - The strategy adopts a semi-annual rebalancing approach, adjusting positions at the end of March and August each year, and selects funds from active equity funds after deducting transaction costs[24] - **Model Evaluation**: The strategy has shown long-term outperformance against the Wind Active Equity Hybrid Fund Index, with a fee-adjusted annualized excess return of 3.64% since March 2011[24][28] 2. Model Name: Fund Selection Strategy Based on Fund Manager Trading Uniqueness - **Model Construction Idea**: This strategy evaluates the uniqueness of fund managers' trading behaviors by constructing a network based on their holdings and transactions, aiming to identify funds with distinctive trading styles[3][32] - **Model Construction Process**: - A network is built using detailed fund manager holdings and transaction data - A metric is calculated to measure the uniqueness of each fund manager's trading behavior compared to their peers[48] - The strategy adopts a semi-annual rebalancing approach, adjusting positions in early April and September each year, and selects funds from active equity funds, general stock funds, and flexible allocation funds after deducting transaction costs[32] - **Model Evaluation**: The strategy has demonstrated significant outperformance, achieving a fee-adjusted annualized excess return of 5.66% since its inception[32][36] 3. Model Name: Industry-Themed ETF Selection Strategy Based on Filing Information - **Model Construction Idea**: This strategy leverages the forward-looking information from the public disclosure stage of ETF filing materials to construct an industry-themed filing similarity factor (T+1), aiming to capture market investment hotspots[4][39] - **Model Construction Process**: - The T+1 factor is constructed by calculating the similarity between the indices tracked by newly filed ETFs and existing market indices[48] - The strategy adopts a monthly rebalancing approach, selecting ETFs from industry-themed ETFs with a transaction fee rate of 0.1% per side, using the CSI 800 Index as the benchmark[39] - **Model Evaluation**: The strategy has consistently outperformed the CSI 800 Index since December 2018, with a fee-adjusted annualized excess return of 11.33%[39][44] --- Model Backtesting Results 1. Fund Selection Strategy Based on Trading Motivation Factor and Stock Spread Income Factor - **December 2025 Return**: 1.56% (vs. 3.06% for the benchmark)[28] - **Annualized Return**: 10.85% (vs. 7.33% for the benchmark)[28] - **Annualized Volatility**: 21.62% (vs. 19.97% for the benchmark)[28] - **Sharpe Ratio**: 0.50 (vs. 0.37 for the benchmark)[28] - **Maximum Drawdown**: 48.39% (vs. 45.42% for the benchmark)[28] - **Annualized Excess Return**: 3.64%[28] - **IR**: 0.61[28] - **Excess Maximum Drawdown**: 19.22%[28] - **December 2025 Excess Return**: -1.54%[28] 2. Fund Selection Strategy Based on Fund Manager Trading Uniqueness - **December 2025 Return**: 5.36% (vs. 3.06% for the benchmark)[36] - **Annualized Return**: 13.40% (vs. 7.87% for the benchmark)[36] - **Annualized Volatility**: 19.52% (vs. 18.30% for the benchmark)[36] - **Sharpe Ratio**: 0.69 (vs. 0.43 for the benchmark)[36] - **Maximum Drawdown**: 37.26% (vs. 45.42% for the benchmark)[36] - **Annualized Excess Return**: 5.66%[36] - **IR**: 1.09[36] - **Excess Maximum Drawdown**: 10.84%[36] - **December 2025 Excess Return**: 2.27%[36] 3. Industry-Themed ETF Selection Strategy Based on Filing Information - **December 2025 Return**: 5.84% (vs. 3.31% for the benchmark)[43] - **Annualized Return**: 19.22% (vs. 6.90% for the benchmark)[43] - **Annualized Volatility**: 21.05% (vs. 18.85% for the benchmark)[43] - **Sharpe Ratio**: 0.91 (vs. 0.37 for the benchmark)[43] - **Maximum Drawdown**: 34.89% (vs. 42.96% for the benchmark)[44] - **Annualized Excess Return**: 11.33%[44] - **IR**: 0.64[44] - **Excess Maximum Drawdown**: 19.07%[44] - **December 2025 Excess Return**: 2.53%[44]