Quantitative Models and Construction Methods - Model Name: Expected High Dividend Portfolio Model Construction Idea: The model aims to construct a portfolio with high expected dividend yield by leveraging historical dividend data, fundamental indicators, and short-term factors like reversal and profitability[5][10][16] Model Construction Process: 1. Dividend Yield Calculation: - Phase 1: Calculate dividend yield based on annual report profit distribution announcements - Phase 2: Predict and calculate dividend yield using historical dividend data and fundamental indicators[5][10] 2. Screening Process: - Exclude suspended and limit-up stocks from the CSI 300 constituents[15] - Remove the top 20% of stocks with the highest short-term momentum (21-day cumulative return)[15] - Exclude stocks with declining profitability (quarterly net profit YoY growth < 0)[15] 3. Final Selection: - Rank the remaining stocks by expected dividend yield - Select the top 30 stocks with the highest expected dividend yield and construct an equally weighted portfolio[11] Model Evaluation: The model demonstrates strong historical performance with significant excess returns and controlled drawdowns, making it a robust strategy for high-dividend stock selection[13] Model Backtesting Results - Expected High Dividend Portfolio: - Cumulative Return: 358.90% - Cumulative Excess Return (vs CSI 300 Total Return Index): 107.44% - Annualized Excess Return: 8.87% - Maximum Rolling 1-Year Drawdown of Excess Return: 12.26% - Monthly Excess Win Rate: 60.19%[13] Quantitative Factors and Construction Methods - Factor Name: Expected Dividend Yield Factor Factor Construction Idea: Predict future dividend yield by combining historical dividend data, fundamental indicators, and short-term influencing factors[5][16] Factor Construction Process: 1. Calculate historical dividend yield based on profit distribution announcements[5][10] 2. Predict future dividend yield using fundamental indicators and historical dividend patterns[5][10] 3. Incorporate two short-term factors: - Reversal Factor: Accounts for short-term price reversals - Profitability Factor: Reflects the company's earnings performance[5][16] Factor Evaluation: The factor effectively identifies high-dividend stocks and serves as a reliable input for portfolio construction[16] - Factor Name: Red Dividend Timing Framework (Composite Signal) Factor Construction Idea: Combines multiple single-factor signals to assess the market's outlook on dividend stocks[25][28] Factor Construction Process: 1. Evaluate five single-factor signals: - Inflation: PPI YoY (High/Low) - Liquidity: M2 YoY (High/Low) - M1-M2 Gap: Scissors Difference (High/Low) - Interest Rate: US 10-Year Treasury Yield (High/Low) - Market Sentiment: Dividend Stock Turnover Ratio (Up/Down)[28] 2. Assign binary signals (1 for bullish, 0 for bearish) to each factor 3. Aggregate the signals into a composite indicator[28] Factor Evaluation: The framework provides a systematic approach to timing dividend stock investments, though the September 2025 signal suggests a cautious stance[25][28] Factor Backtesting Results - Expected Dividend Yield Factor: - August 2025 Portfolio Average Return: 5.69% - Excess Return (vs CSI 300 Index): -4.80% - Excess Return (vs CSI Dividend Index): +4.70%[5][16] - Red Dividend Timing Framework (Composite Signal): - Latest Signal (September 2025): 0 (Neutral)[25][28]
金工定期报告20250902:预期高股息组合跟踪