开源量化评论（119）：优选金股30组合的年度回顾及最新持仓

• Model Name: 优选金股 30 组合 - Model Construction Idea: The 优选金股 30 组合 aims to track the latest market trends and dynamically capture high-elasticity stocks by relying on analysts' new recommendations[3][4] - Model Construction Process: 1. Divide all gold stocks into new and repeated gold stocks based on whether they are newly recommended or repeated from the previous month[4] 2. Select the top 30 stocks with the highest performance expectations among the new gold stocks each month[20] 3. Weight the selected stocks based on the number of recommendations by analysts[20] - Model Evaluation: The model has consistently outperformed the 中证 500 index for nine consecutive years, with an annualized return of 23.1% and an annualized excess return of 21% since 2017[2][12] - Model Backtest Results: - 2017: 优选金股 30 return 33.5%, 中证 500 return 0.6%, excess return 32.9%[15] - 2018: 优选金股 30 return -13.0%, 中证 500 return -33.3%, excess return 20.3%[15] - 2019: 优选金股 30 return 53.3%, 中证 500 return 27.1%, excess return 26.2%[15] - 2020: 优选金股 30 return 87.6%, 中证 500 return 20.5%, excess return 67.1%[15] - 2021: 优选金股 30 return 28.7%, 中证 500 return 15.5%, excess return 13.2%[18] - 2022: 优选金股 30 return -19.8%, 中证 500 return -20.7%, excess return 0.8%[18] - 2023: 优选金股 30 return 11.2%, 中证 500 return -7.3%, excess return 18.5%[18] - 2024: 优选金股 30 return 15.4%, 中证 500 return 5.3%, excess return 10.1%[18] - 2025: 优选金股 30 return 52.3%, 中证 500 return 30.4%, excess return 21.9%[18] - Factor Name: 业绩超预期指标 (SUE) - Factor Construction Idea: The SUE factor measures the degree to which a company's actual financial performance exceeds analysts' expectations, which is generally associated with better stock returns[26][27] - Factor Construction Process: 1. Convert cumulative analyst forecast data into single-quarter values[27] 2. Compare single-quarter forecast values with actual values using the formula: $S U E={\frac{R_{t}-E_{t}}{\sigma(R_{t}-E_{t})}}$ where $R_{t}$ is the actual value, $E_{t}$ is the forecast value, and $\sigma(R_{t}-E_{t})$ is the standard deviation of the difference between actual and forecast values[26] 3. Integrate earnings forecasts, earnings bulletins, and periodic reports, and forward-fill factor values during financial report vacuum periods[27] - Factor Evaluation: The SUE factor shows strong selection effects in new gold stocks, indicating that stocks with higher performance expectations tend to perform better[27] - Factor Backtest Results: - New Gold Stocks: Annualized return 13.6%, annualized volatility 24.7%, return-volatility ratio 0.55, maximum drawdown 38.5%[26] - Repeated Gold Stocks: Annualized return 8.4%, annualized volatility 23.9%, return-volatility ratio 0.35, maximum drawdown 45.0%[26]