Quantitative Models and Factor Construction Quantitative Models and Construction Methods - **Model Name**: Maximized Factor Exposure Portfolio (MFE) **Model Construction Idea**: The MFE portfolio is designed to maximize the exposure of a single factor while controlling for various constraints such as industry exposure, style exposure, stock weight deviation, and turnover limits. This approach ensures that the factor's predictive power is tested under realistic portfolio constraints, making it more applicable in practice [39][40]. **Model Construction Process**: The MFE portfolio is constructed using the following optimization model: $ \begin{array}{ll} max & f^{T} w \\ s.t. & s_{l} \leq X(w-w_{b}) \leq s_{h} \\ & h_{l} \leq H(w-w_{b}) \leq h_{h} \\ & w_{l} \leq w-w_{b} \leq w_{h} \\ & b_{l} \leq B_{b}w \leq b_{h} \\ & \mathbf{0} \leq w \leq l \\ & \mathbf{1}^{T} w = 1 \end{array} $ - **Objective Function**: Maximize single-factor exposure, where \( f \) represents factor values, and \( w \) is the stock weight vector. - **Constraints**: 1. **Style Exposure**: \( X \) is the factor exposure matrix, \( w_b \) is the benchmark weight vector, and \( s_l, s_h \) are the lower and upper bounds for style exposure. 2. **Industry Exposure**: \( H \) is the industry exposure matrix, and \( h_l, h_h \) are the lower and upper bounds for industry deviation. 3. **Stock Weight Deviation**: \( w_l, w_h \) are the lower and upper bounds for stock weight deviation. 4. **Constituent Weight Control**: \( B_b \) is a binary vector indicating benchmark constituents, and \( b_l, b_h \) are the lower and upper bounds for constituent weights. 5. **No Short Selling**: Ensures non-negative weights and limits individual stock weights. 6. **Full Investment**: Ensures the portfolio is fully invested with \( \mathbf{1}^{T} w = 1 \) [39][40][41]. **Model Evaluation**: The MFE portfolio is effective in testing factor performance under realistic constraints, making it a practical tool for portfolio construction [39][40]. Quantitative Factors and Construction Methods - **Factor Name**: DELTAROE **Factor Construction Idea**: Measures the change in return on equity (ROE) over a specific period to capture improvements in profitability [16]. **Factor Construction Process**: $ \text{DELTAROE} = \text{ROE}_{\text{current quarter}} - \text{ROE}_{\text{same quarter last year}} $ Where ROE is calculated as: $ \text{ROE} = \frac{\text{Net Income} \times 2}{\text{Beginning Equity} + \text{Ending Equity}} $ [16]. **Factor Evaluation**: DELTAROE is a profitability factor that has shown strong performance in multiple sample spaces, including CSI 300, CSI 500, and CSI A500 indices [17][19][24]. - **Factor Name**: Pre-expected PEG (Pre-expected Price-to-Earnings Growth) **Factor Construction Idea**: Incorporates analysts' earnings growth expectations to evaluate valuation relative to growth potential [16]. **Factor Construction Process**: $ \text{Pre-expected PEG} = \frac{\text{Forward P/E}}{\text{Expected Earnings Growth Rate}} $ Where forward P/E is based on analysts' consensus earnings estimates [16]. **Factor Evaluation**: This factor has demonstrated strong predictive power in growth-oriented sample spaces such as CSI 300 and CSI A500 indices [17][24]. - **Factor Name**: DELTAROA **Factor Construction Idea**: Measures the change in return on assets (ROA) over a specific period to capture improvements in asset efficiency [16]. **Factor Construction Process**: $ \text{DELTAROA} = \text{ROA}_{\text{current quarter}} - \text{ROA}_{\text{same quarter last year}} $ Where ROA is calculated as: $ \text{ROA} = \frac{\text{Net Income} \times 2}{\text{Beginning Total Assets} + \text{Ending Total Assets}} $ [16]. **Factor Evaluation**: DELTAROA has shown consistent performance across multiple indices, including CSI 1000 and public fund-heavy indices [22][26]. Factor Backtesting Results - **DELTAROE**: - CSI 300: Weekly excess return 0.75%, monthly 2.28%, YTD 8.04% [17]. - CSI 500: Weekly excess return 0.07%, monthly 0.59%, YTD 6.67% [19]. - CSI A500: Weekly excess return 0.68%, monthly 3.61%, YTD 9.20% [24]. - **Pre-expected PEG**: - CSI 300: Weekly excess return 0.72%, monthly 2.10%, YTD 7.22% [17]. - CSI 500: Weekly excess return 0.15%, monthly 1.34%, YTD 9.62% [19]. - CSI A500: Weekly excess return 0.85%, monthly 2.07%, YTD 10.35% [24]. - **DELTAROA**: - CSI 300: Weekly excess return 0.44%, monthly 2.27%, YTD 7.10% [17]. - CSI 1000: Weekly excess return 0.66%, monthly 1.57%, YTD 8.57% [22]. - Public Fund Index: Weekly excess return 0.66%, monthly 1.57%, YTD 8.57% [26].

Quantitative Factors and Construction Factor Name: Stability of Turnover Rate (STR) - **Factor Construction Idea**: The STR factor is designed to evaluate the stability of daily turnover rates. It aims to identify stocks with stable turnover rates, as opposed to focusing solely on low or high turnover rates. This approach addresses the limitations of traditional turnover rate factors, which may misjudge stocks with high turnover but significant future returns [1][8]. - **Factor Construction Process**: - The STR factor is constructed using daily turnover rate data. - The stability of turnover rates is calculated, inspired by the Uniformity of Turnover Rate Distribution (UTD) factor, which measures turnover rate volatility at the minute level. - The STR factor is then adjusted to remove the influence of common market styles and industry effects, ensuring a "pure" factor signal [8]. - **Factor Evaluation**: The STR factor demonstrates strong stock selection capabilities, even after controlling for market and industry influences. It is considered an effective and straightforward factor [6][8]. Traditional Turnover Rate Factor (Turn20) - **Factor Construction Idea**: The Turn20 factor calculates the average daily turnover rate over the past 20 trading days. It assumes that stocks with lower turnover rates are more likely to outperform in the future, while those with higher turnover rates are more likely to underperform [6][7]. - **Factor Construction Process**: - At the end of each month, the daily turnover rates of all stocks over the past 20 trading days are averaged. - The resulting values are neutralized for market capitalization to eliminate size effects [6]. - **Factor Evaluation**: While the Turn20 factor has historically performed well, its logic has limitations. Specifically, stocks with high turnover rates exhibit significant variability in future returns, leading to potential misjudgments of high-performing stocks within this group [7]. --- Backtesting Results of Factors STR Factor - **Annualized Return**: 40.75% [9][10] - **Annualized Volatility**: 14.44% [9][10] - **Information Ratio (IR)**: 2.82 [9][10] - **Monthly Win Rate**: 77.02% [9][10] - **Maximum Drawdown**: 9.96% [9][10] - **July 2025 Performance**: - Long Portfolio Return: 1.29% [10] - Short Portfolio Return: -0.02% [10] - Long-Short Portfolio Return: 1.32% [10] Turn20 Factor - **Monthly IC Mean**: -0.072 [6] - **Annualized ICIR**: -2.10 [6] - **Annualized Return**: 33.41% [6] - **Information Ratio (IR)**: 1.90 [6] - **Monthly Win Rate**: 71.58% [6]

Quantitative Models and Construction Methods Barra Style Factors - **Model Name**: Barra Style Factors - **Construction Idea**: The Barra style factors are designed to capture the performance of different market styles, such as size, value, growth, and profitability, through specific factor definitions[4][14] - **Construction Process**: - **Size Factor**: Measures the market capitalization of stocks - **Value Factor**: Captures the book-to-market ratio of stocks - **Growth Factor**: Reflects the growth potential of stocks - **Profitability Factor**: Based on earnings expectations[4][14] - **Evaluation**: These factors are widely used in the industry to analyze market trends and style rotations[4][14] --- Open-source Trading Behavior Factors - **Factor Name**: Ideal Reversal Factor - **Construction Idea**: Identifies the strongest reversal days by analyzing the average transaction size of large trades[5][15] - **Construction Process**: 1. Retrieve the past 20 trading days' data for a stock 2. Calculate the average transaction size per day (transaction amount/number of transactions) 3. Identify the 10 days with the highest transaction sizes and sum their returns (M_high) 4. Identify the 10 days with the lowest transaction sizes and sum their returns (M_low) 5. Compute the factor as $M = M_{high} - M_{low}$[43] - **Evaluation**: Captures the microstructure of reversal forces in the A-share market[5][15] - **Factor Name**: Smart Money Factor - **Construction Idea**: Tracks institutional trading activity by analyzing minute-level price and volume data[5][15] - **Construction Process**: 1. Retrieve the past 10 days' minute-level data for a stock 2. Construct the indicator $S_t = |R_t| / V_t^{0.25}$, where $R_t$ is the return at minute $t$, and $V_t$ is the trading volume at minute $t$ 3. Sort minute-level data by $S_t$ in descending order and select the top 20% of minutes by cumulative trading volume 4. Calculate the volume-weighted average price (VWAP) for smart money trades ($VWAP_{smart}$) and all trades ($VWAP_{all}$) 5. Compute the factor as $Q = VWAP_{smart} / VWAP_{all}$[42][44] - **Evaluation**: Effectively identifies institutional trading patterns[5][15] - **Factor Name**: APM Factor - **Construction Idea**: Measures the difference in trading behavior between morning (or overnight) and afternoon sessions[5][15] - **Construction Process**: 1. Retrieve the past 20 days' data for a stock 2. Calculate daily overnight and afternoon returns for both the stock and the index 3. Perform a regression of stock returns on index returns to obtain residuals 4. Compute the difference between overnight and afternoon residuals for each day 5. Calculate the statistic $\mathrm{stat} = \frac{\mu(\delta_t)}{\sigma(\delta_t) / \sqrt{N}}$, where $\mu$ is the mean, $\sigma$ is the standard deviation, and $N$ is the sample size 6. Regress the statistic on momentum factors and use the residual as the APM factor[43][45][46] - **Evaluation**: Captures intraday trading behavior differences[5][15] - **Factor Name**: Ideal Amplitude Factor - **Construction Idea**: Measures the structural differences in amplitude information between high and low price states[5][15] - **Construction Process**: 1. Retrieve the past 20 trading days' data for a stock 2. Calculate the daily amplitude as $(\text{High Price}/\text{Low Price}) - 1$ 3. Compute the average amplitude for the top 25% of days with the highest closing prices ($V_{high}$) 4. Compute the average amplitude for the bottom 25% of days with the lowest closing prices ($V_{low}$) 5. Compute the factor as $V = V_{high} - V_{low}$[48] - **Evaluation**: Highlights amplitude differences across price states[5][15] - **Factor Name**: Composite Trading Behavior Factor - **Construction Idea**: Combines the above trading behavior factors using ICIR-based weights to enhance predictive power[31] - **Construction Process**: 1. Standardize and winsorize the individual factors within industries 2. Use the past 12 periods' ICIR values as weights to compute the composite factor[31] - **Evaluation**: Demonstrates superior performance in small-cap stock pools[32] --- Backtesting Results of Models and Factors Barra Style Factors - **Size Factor**: Return of 0.64% in July 2025[4][14] - **Value Factor**: Return of 0.59% in July 2025[4][14] - **Growth Factor**: Return of 0.16% in July 2025[4][14] - **Profitability Factor**: Return of -0.32% in July 2025[4][14] Open-source Trading Behavior Factors - **Ideal Reversal Factor**: - IC: -0.050 - RankIC: -0.061 - IR: 2.52 - Long-short monthly win rate: 78.3% (historical), 66.7% (last 12 months) - July 2025 long-short return: 0.47%[6][16] - **Smart Money Factor**: - IC: -0.037 - RankIC: -0.061 - IR: 2.76 - Long-short monthly win rate: 82.2% (historical), 91.7% (last 12 months) - July 2025 long-short return: 1.78%[6][19] - **APM Factor**: - IC: 0.029 - RankIC: 0.034 - IR: 2.30 - Long-short monthly win rate: 77.4% (historical), 58.3% (last 12 months) - July 2025 long-short return: 1.42%[6][23] - **Ideal Amplitude Factor**: - IC: -0.054 - RankIC: -0.073 - IR: 3.03 - Long-short monthly win rate: 83.6% (historical), 75.0% (last 12 months) - July 2025 long-short return: 3.86%[6][28] - **Composite Trading Behavior Factor**: - IC: 0.067 - RankIC: 0.092 - IR: 3.30 - Long-short monthly win rate: 82.6% (historical), 83.3% (last 12 months) - July 2025 long-short return: 2.13%[6][31]

- The report aims to track large purchases and net active purchases using transaction detail data[1] - The indicators used are the proportion of large order transaction amounts and the proportion of net active purchase amounts[7] - The proportion of large order transaction amounts reflects the buying behavior of large funds[7] - The proportion of net active purchase amounts reflects the active buying behavior of investors[7] - The top 5 stocks with the highest average proportion of large order transaction amounts over the past 5 days are: Sobute, China Railway Industry, Tibet Tianlu, Poly United, and China Power Construction[4][9] - The top 5 stocks with the highest average proportion of net active purchase amounts over the past 5 days are: Weixing Co., HNA Holdings, Kaili Medical, Liaogang Co., and Hengyi Petrochemical[4][10] - The top 5 industries with the highest average proportion of large order transaction amounts over the past 5 days are: Banking, Real Estate, Petroleum and Petrochemical, Transportation, and Coal[4] - The top 5 industries with the highest average proportion of net active purchase amounts over the past 5 days are: Media, Textile and Apparel, Computers, Electronics, and Light Manufacturing[4] - The top 5 ETFs with the highest average proportion of large order transaction amounts over the past 5 days are: China Agricultural Theme ETF, E Fund CSI 300 Medical and Health ETF, Huabao CSI Medical ETF, Bosera SSE STAR 100 ETF, and Guotai CSI Livestock Breeding ETF[4][15] - The top 5 ETFs with the highest average proportion of net active purchase amounts over the past 5 days are: Penghua CSI Subdivision Chemical Industry Theme ETF, GF SSE STAR 50 ETF, Harvest CSI Rare Metals Theme ETF, E Fund Guozheng Robotics Industry ETF, and Harvest CSI Software Services ETF[4][16]

Quantitative Models and Factor Construction Quantitative Models and Construction Methods - **Model Name**: Maximized Factor Exposure (MFE) Portfolio **Model Construction Idea**: The MFE portfolio is designed to test the effectiveness of individual factors under realistic constraints, such as industry exposure, style exposure, stock weight limits, and turnover constraints. This approach ensures that the factors deemed "effective" can genuinely contribute to return prediction in the final portfolio[41][42]. **Model Construction Process**: The MFE portfolio is constructed using the following optimization model: $ \begin{array}{ll} max & f^{T} w \\ s.t. & s_{l} \leq X(w-w_{b}) \leq s_{h} \\ & h_{l} \leq H(w-w_{b}) \leq h_{h} \\ & w_{l} \leq w-w_{b} \leq w_{h} \\ & b_{l} \leq B_{b}w \leq b_{h} \\ & \mathbf{0} \leq w \leq l \\ & \mathbf{1}^{T} w = 1 \end{array} $ - **Objective Function**: Maximize single-factor exposure, where \( f^{T} w \) represents the weighted exposure of the portfolio to the factor \( f \), and \( w \) is the stock weight vector. - **Constraints**: 1. **Style Exposure**: \( X \) represents the factor exposure matrix for stocks, \( w_b \) is the benchmark weight vector, and \( s_l, s_h \) are the lower and upper bounds for style factor exposure[42]. 2. **Industry Exposure**: \( H \) is the industry exposure matrix, and \( h_l, h_h \) are the lower and upper bounds for industry deviations[42]. 3. **Stock Weight Deviation**: \( w_l, w_h \) are the lower and upper bounds for stock weight deviations relative to the benchmark[42]. 4. **Constituent Weight**: \( B_b \) is a binary vector indicating whether a stock is part of the benchmark, and \( b_l, b_h \) are the lower and upper bounds for constituent weights[42]. 5. **No Short Selling**: Ensures non-negative weights and limits individual stock weights to \( l \)[42]. 6. **Full Investment**: Ensures the portfolio is fully invested with \( \mathbf{1}^{T} w = 1 \)[43]. - **Implementation**: 1. Define constraints for style, industry, and stock weights. For example, for CSI 500 and CSI 300 indices, industry exposure is neutralized, and stock weight deviations are capped at 1%[45]. 2. Construct the MFE portfolio at the end of each month based on the constraints[45]. 3. Backtest the portfolio, accounting for transaction costs (0.3% per side), and calculate performance metrics relative to the benchmark[45]. **Model Evaluation**: The MFE portfolio effectively tests factor performance under realistic constraints, making it a robust tool for evaluating factor predictability in practical scenarios[41][42]. Quantitative Factors and Construction Methods - **Factor Name**: DELTAROA **Factor Construction Idea**: Measures the change in return on assets (ROA) compared to the same quarter in the previous year, capturing improvements in asset utilization efficiency[16]. **Factor Construction Process**: $ DELTAROA = ROA_{current\ quarter} - ROA_{same\ quarter\ last\ year} $ Where \( ROA = \frac{Net\ Income}{Total\ Assets} \)[16]. **Factor Evaluation**: DELTAROA is a growth-oriented factor that has shown strong performance in multiple sample spaces, particularly in the CSI A500 index[19][25]. - **Factor Name**: Standardized Unexpected Earnings (SUE) **Factor Construction Idea**: Measures the deviation of actual earnings from expected earnings, standardized by the standard deviation of expected earnings, to capture earnings surprises[16]. **Factor Construction Process**: $ SUE = \frac{Actual\ Earnings - Expected\ Earnings}{Standard\ Deviation\ of\ Expected\ Earnings} $[16]. **Factor Evaluation**: SUE is a profitability factor that performs well in growth-oriented indices like CSI 1000 and CSI A500[19][23][25]. - **Factor Name**: One-Year Momentum **Factor Construction Idea**: Captures the trend-following behavior of stocks by measuring price momentum over the past year, excluding the most recent month[16]. **Factor Construction Process**: $ Momentum = \frac{Price_{t-12} - Price_{t-1}}{Price_{t-1}} $ Where \( t-12 \) and \( t-1 \) represent the stock price 12 months and 1 month ago, respectively[16]. **Factor Evaluation**: Momentum is a widely used factor that has shown consistent performance in large-cap indices like CSI 300 and CSI 500[19][21]. Factor Backtesting Results - **CSI 300 Sample Space**: - **Best-Performing Factors (1 Week)**: Single-quarter revenue growth, DELTAROA, single-quarter ROE[19]. - **Worst-Performing Factors (1 Week)**: Three-month volatility, one-month volatility, three-month turnover[19]. - **CSI 500 Sample Space**: - **Best-Performing Factors (1 Week)**: One-year momentum, standardized unexpected revenue, standardized unexpected earnings[21]. - **Worst-Performing Factors (1 Week)**: SPTTM, single-quarter SP, dividend yield[21]. - **CSI 1000 Sample Space**: - **Best-Performing Factors (1 Week)**: Three-month reversal, standardized unexpected revenue, single-quarter surprise magnitude[23]. - **Worst-Performing Factors (1 Week)**: Dividend yield, one-month volatility, BP[23]. - **CSI A500 Sample Space**: - **Best-Performing Factors (1 Week)**: DELTAROA, standardized unexpected earnings, single-quarter ROA[25]. - **Worst-Performing Factors (1 Week)**: Three-month volatility, one-month turnover, one-month volatility[25]. - **Public Fund Heavyweight Index Sample Space**: - **Best-Performing Factors (1 Week)**: One-year momentum, standardized unexpected revenue, expected net profit QoQ[27]. - **Worst-Performing Factors (1 Week)**: Dividend yield, one-month volatility, three-month volatility[27].

Quantitative Models and Construction Methods - **Model Name**: Maximized Factor Exposure Portfolio (MFE) **Model Construction Idea**: The MFE portfolio is designed to maximize the exposure of a single factor while controlling for various constraints such as industry exposure, style exposure, stock weight deviation, and turnover limits. This approach ensures that the factor's predictive power is tested under realistic portfolio constraints, making it more applicable in actual investment scenarios [39][40]. **Model Construction Process**: The MFE portfolio is constructed using the following optimization model: $ \begin{array}{ll} max & f^{T}\ w \\ s.t. & s_{l}\leq X(w-w_{b})\leq s_{h} \\ & h_{l}\leq H(w-w_{b})\leq h_{h} \\ & w_{l}\leq w-w_{b}\leq w_{h} \\ & b_{l}\leq B_{b}w\leq b_{h} \\ & \mathbf{0}\leq w\leq l \\ & \mathbf{1}^{T}\ w=1 \end{array} $ - **Objective Function**: Maximize single-factor exposure, where \( f \) represents factor values, and \( w \) is the stock weight vector. - **Constraints**: - **Style Exposure**: \( X \) is the factor exposure matrix, \( w_b \) is the benchmark weight vector, and \( s_l, s_h \) are the lower and upper bounds for style exposure. - **Industry Exposure**: \( H \) is the industry exposure matrix, and \( h_l, h_h \) are the lower and upper bounds for industry deviation. - **Stock Weight Deviation**: \( w_l, w_h \) are the lower and upper bounds for stock weight deviation. - **Component Weight Control**: \( B_b \) is a binary vector indicating benchmark components, and \( b_l, b_h \) are the lower and upper bounds for component weights. - **No Short Selling**: Ensures non-negative weights and limits individual stock weights. - **Full Investment**: Ensures the portfolio is fully invested (\( \mathbf{1}^{T}\ w=1 \)) [40][41]. **Model Evaluation**: The MFE portfolio effectively tests factor efficacy under realistic constraints, making it a robust tool for factor validation in enhanced index strategies [39][40]. --- Quantitative Factors and Construction Methods - **Factor Name**: Three-Month Reversal **Factor Construction Idea**: Measures the reversal effect by calculating the return over the past 60 trading days, assuming stocks with recent underperformance may outperform in the future [17]. **Factor Construction Process**: $ \text{Three-Month Reversal} = \text{Cumulative Return over the Past 60 Trading Days} $ **Factor Evaluation**: Demonstrates strong performance in certain index spaces, such as CSI 1000 and CSI A500, but underperforms in others like CSI 500 [17][22][25]. - **Factor Name**: One-Year Momentum **Factor Construction Idea**: Captures the momentum effect by excluding the most recent month and calculating the cumulative return over the prior 11 months [17]. **Factor Construction Process**: $ \text{One-Year Momentum} = \text{Cumulative Return over the Past 11 Months (Excluding the Most Recent Month)} $ **Factor Evaluation**: Performs well in CSI 500 and CSI 1000 spaces but shows mixed results in other index spaces [17][21][23]. - **Factor Name**: Standardized Unexpected Earnings (SUE) **Factor Construction Idea**: Measures the deviation of actual earnings from expected earnings, standardized by the standard deviation of expected earnings [17]. **Factor Construction Process**: $ \text{SUE} = \frac{\text{Actual Earnings} - \text{Expected Earnings}}{\text{Standard Deviation of Expected Earnings}} $ **Factor Evaluation**: Consistently performs well across multiple index spaces, indicating its robustness as a predictive factor [17][22][25]. - **Factor Name**: Delta ROE (DELTAROE) **Factor Construction Idea**: Measures the change in return on equity (ROE) compared to the same quarter in the previous year [17]. **Factor Construction Process**: $ \text{DELTAROE} = \text{Current Quarter ROE} - \text{ROE from the Same Quarter Last Year} $ **Factor Evaluation**: Demonstrates strong predictive power in CSI 500 and CSI A500 spaces, with moderate performance in other index spaces [17][21][25]. --- Factor Backtesting Results - **Three-Month Reversal**: - CSI 300: Weekly excess return 0.66%, monthly excess return 0.65%, YTD excess return 3.01% [19]. - CSI 500: Weekly excess return 0.79%, monthly excess return 1.17%, YTD excess return 4.07% [21]. - CSI 1000: Weekly excess return 1.09%, monthly excess return 1.40%, YTD excess return 0.38% [23]. - CSI A500: Weekly excess return 1.08%, monthly excess return 0.36%, YTD excess return 3.64% [25]. - **One-Year Momentum**: - CSI 300: Weekly excess return 0.46%, monthly excess return 0.36%, YTD excess return -1.85% [19]. - CSI 500: Weekly excess return 1.26%, monthly excess return 1.18%, YTD excess return 2.77% [21]. - CSI 1000: Weekly excess return 1.45%, monthly excess return 1.73%, YTD excess return 0.26% [23]. - CSI A500: Weekly excess return 0.74%, monthly excess return 0.87%, YTD excess return -2.03% [25]. - **SUE**: - CSI 300: Weekly excess return 0.51%, monthly excess return 2.15%, YTD excess return 3.03% [19]. - CSI 500: Weekly excess return -0.41%, monthly excess return 0.13%, YTD excess return 2.86% [21]. - CSI 1000: Weekly excess return -0.08%, monthly excess return 2.77%, YTD excess return 4.41% [23]. - CSI A500: Weekly excess return 0.47%, monthly excess return 1.63%, YTD excess return 2.04% [25]. - **Delta ROE (DELTAROE)**: - CSI 300: Weekly excess return 0.26%, monthly excess return 2.27%, YTD excess return 5.32% [19]. - CSI 500: Weekly excess return 0.58%, monthly excess return 2.49%, YTD excess return 4.03% [21]. - CSI 1000: Weekly excess return -1.15%, monthly excess return 0.74%, YTD excess return 3.01% [23]. - CSI A500: Weekly excess return 0.52%, monthly excess return 2.82%, YTD excess return 5.13% [25].

身处变局，结构求新 2025 年 06 月 26 日 金融工程 2025 中期策略展望 ➢ 黄金量化：对抗风险，继续持有。从驱动因素来看，当下财政因素继续占主 导。特朗普施政主张对通货膨胀与财政赤字有进一步扩张可能，但实施进度有待 观察。当下全球地缘政治风险与经济政策不确定性仍高，对促发黄金上涨有一定 作用。技术面测算当下已经历盘整积累支持目标价到 3885 美元/盎司。 ➢ 行业量化：轮动加速，机会分散。胜率赔率一致性继续减弱，短期轮动速度 或提高。从生命周期来看，长期配置成长期行业有更高的预期收益。生命周期模 型显示一级行业整体成长性不足，成长性整体分散在三级行业，其中基础化工细 分领域扩张特征明显。 ➢ 利率量化：寻觅资产，关注融资。国债利率跌破 2%，稳定在 1.6%-1.7%低 位，景气度的回升和信贷脉冲的回升抑制了利率的进一步下行。当下通胀资产持 有意愿再度疲软，叠加房地产 ROE 转负、高 ROE 行业杠杆率偏低，导致利率缺 乏上行动能。利率或延续底部震荡，突破需依赖高 ROE 高杠杆行业对资金需求 的拉动。 ➢ 风格量化：成长萌芽，关注预期。当下Δg 和Δgf 都有扩张，成长性的稀缺 度抬升， ...

证券研究报告 | 2025年06月21日 多因子选股周报 估值因子表现出色，中证 1000 增强组合年内超额 12.61% 核心观点 金融工程周报 国信金工指数增强组合表现跟踪 因子表现监控 以沪深 300 指数为选股空间。最近一周，预期 EPTTM、单季 EP、EPTTM 等因子表现较好，而一年动量、高管薪酬、非流动性冲击等因子表现较差。 以中证 500 指数为选股空间。最近一周，BP、预期 BP、预期 EPTTM 等因 子表现较好，而一年动量、三个月机构覆盖、非流动性冲击等因子表现较差。 以中证 1000 指数为选股空间。最近一周，BP、一个月换手、三个月波动等 因子表现较好，而一年动量、三个月机构覆盖、单季 ROE 等因子表现较差。 以中证 A500 指数为选股空间。最近一周，单季 EP、预期 EPTTM、预期 PEG 等因子表现较好，而三个月反转、一年动量、一个月反转等因子表现 较差。 以公募重仓指数为选股空间。最近一周，预期 EPTTM、单季 EP、预期 PEG 等因子表现较好，而一年动量、三个月机构覆盖、预期净利润环比等因子表 现较差。 公募基金指数增强产品表现跟踪 目前，公募基金沪深 300 指 ...

- Model Name: RPV (Renewed Correlation of Price and Volume); Model Construction Idea: The RPV factor integrates intraday and overnight information by dividing price and volume into four quadrants, effectively identifying the reversal and momentum effects of price-volume correlation factors through the monthly IC mean; Model Construction Process: The RPV factor is constructed by combining the best representatives of intraday and overnight price-volume correlations, incorporating "trading volume" information in the form of correlation, and completing information integration; Model Evaluation: The RPV factor is novel and effective[1][6][7] - Model Name: SRV (Smart Relative Volume); Model Construction Idea: The SRV factor splits intraday price changes into morning and afternoon changes, calculates the "smart" indicator by minute, and uses the correlation coefficient between the afternoon "smart" turnover rate and afternoon price changes; Model Construction Process: The SRV factor combines the more effective intraday price-volume correlation factor and the overnight price-volume correlation factor, where the turnover rate is replaced by the turnover rate of the last half-hour of the previous day, which has a higher proportion of informed trading; Model Evaluation: The SRV factor performs better than the RPV factor[1][6][7] Model Backtest Results - RPV Model, Annualized Return: 14.69%, Annualized Volatility: 7.75%, IR: 1.90, Monthly Win Rate: 72.79%, Maximum Drawdown: 10.63%[1][7][10] - SRV Model, Annualized Return: 17.48%, Annualized Volatility: 6.50%, IR: 2.69, Monthly Win Rate: 75.74%, Maximum Drawdown: 3.74%[1][7][10] Factor Construction and Evaluation - Factor Name: RPV; Factor Construction Idea: The RPV factor integrates intraday and overnight information by dividing price and volume into four quadrants, effectively identifying the reversal and momentum effects of price-volume correlation factors through the monthly IC mean; Factor Construction Process: The RPV factor is constructed by combining the best representatives of intraday and overnight price-volume correlations, incorporating "trading volume" information in the form of correlation, and completing information integration; Factor Evaluation: The RPV factor is novel and effective[1][6][7] - Factor Name: SRV; Factor Construction Idea: The SRV factor splits intraday price changes into morning and afternoon changes, calculates the "smart" indicator by minute, and uses the correlation coefficient between the afternoon "smart" turnover rate and afternoon price changes; Factor Construction Process: The SRV factor combines the more effective intraday price-volume correlation factor and the overnight price-volume correlation factor, where the turnover rate is replaced by the turnover rate of the last half-hour of the previous day, which has a higher proportion of informed trading; Factor Evaluation: The SRV factor performs better than the RPV factor[1][6][7] Factor Backtest Results - RPV Factor, Annualized Return: 14.69%, Annualized Volatility: 7.75%, IR: 1.90, Monthly Win Rate: 72.79%, Maximum Drawdown: 10.63%[1][7][10] - SRV Factor, Annualized Return: 17.48%, Annualized Volatility: 6.50%, IR: 2.69, Monthly Win Rate: 75.74%, Maximum Drawdown: 3.74%[1][7][10]

一、券商金股股票池上月回顾 2025年5月，曼卡龙、潮宏基、科兴制药等券商金股股票的月度上涨幅度靠前。 2025年5月，国元证券、华西证券、华创证券收益排名前三，月度收益分别为 10.78%、7.30%、5.70%，同期偏股混合型基金指数收益1.06%，沪深300指数收益 1.85%。 2025年以来，东北证券、华鑫证券、东兴证券收益排名前三，年度收益分别为 45.28%、40.52%、39.15%，同期偏股混合型基金指数收益3.37%，沪深300指数收 益-2.41%。 二、券商金股股票池中选股因子表现 最近一个月，EPTTM、盈余公告后跳空超额、剥离涨停动量表现较好，单季度营 收增速、单季度净利润增速、SUE表现较差； 今年以来，总市值、剥离涨停动量、SUE表现较好，预期股息率、波动率、BP表 现较差。 三、券商金股股票池本月特征 截至2025年6月3日，共有43家券商发布本月金股。在对券商金股股票池进行去重 后，总共有286只A股。 从绝对占比来看，本期券商金股在机械(9.56%)、医药(9.04%)、电子(8.01%)、基 础化工(7.75%)、食品饮料(6.46%)行业配置较高。 从相对变化来看，本 ...