Group 1 - The report emphasizes the importance of predicting net profit growth for constructing an ideal stock portfolio, suggesting that a focus on growth factors can enhance stock selection effectiveness [5][6][10] - A forward-looking test was conducted on the ideal portfolio, which showed that selecting stocks with known future net profit growth can significantly outperform the overall market [7][12][14] - The ideal portfolio based on known net profit growth achieved an annualized return of 13.91% with a Sharpe ratio of 0.54, compared to the market's 6.69% return and 0.26 Sharpe ratio [10][11] Group 2 - The report outlines four perspectives for screening stocks based on expected net profit growth, which include using the latest financial report data, historical performance, and acceleration in net profit growth [18][30][40] - The first perspective, which adjusts predictions based on the latest financial report, achieved a prediction success rate of 91.94% [30] - The second perspective focuses on the stability of Return on Equity (ROE) to enhance prediction accuracy, achieving a success rate of 82.55% [34][36] Group 3 - The report indicates that the stock screening pool consists of approximately 600-800 stocks, with an average prediction success rate of 85.10% for net profit growth [60] - The screening pool's total market capitalization and circulating market capitalization are comparable to the CSI 1000 index, ensuring a robust performance across different market conditions [64][65] - The selected stocks consistently outperformed the CSI 1000 index, particularly during bearish market years [68] Group 4 - The report discusses the performance of growth factor-based stock selection, showing that the strategy outperformed the CSI 1000 index in most years since 2016 [80] - The growth factor selection process involved further filtering based on historical growth factors, leading to improved portfolio performance [75] - The report highlights the importance of aligning industry weightings with the CSI 1000 index to mitigate sector biases in stock selection [76] Group 5 - The report introduces a methodology for converting binary classification of net profit growth into continuous probability predictions using logistic regression, which enhances the accuracy of stock selection [96][100] - The analysis indicates that a threshold of 70% provides an optimal balance between prediction accuracy and the number of stocks selected [100][101] - Long-term performance of the strategy shows significant outperformance compared to the CSI 1000 equal-weighted index [103]