Quantitative Models and Construction Methods 1. Model Name: Sentiment Model - **Model Construction Idea**: The sentiment model is designed to measure the strength of market sentiment using factors related to limit-up and limit-down stocks[14] - **Model Construction Process**: The model uses factors such as the proportion of net limit-up stocks, next-day returns of limit-down stocks, proportion of limit-up stocks, proportion of limit-down stocks, and high-frequency board-hitting returns. These factors are aggregated to calculate a sentiment score, with a maximum score of 5. The sentiment score for the current period is 0[14][18] - **Model Evaluation**: The sentiment model indicates that the market sentiment remains low, reflecting weak investor confidence[14][18] 2. Model Name: Moving Average Strength Index - **Model Construction Idea**: This model evaluates the strength of market trends by calculating the moving average strength index based on secondary industry indices[14] - **Model Construction Process**: The moving average strength index is calculated using the performance of secondary industry indices. The current market score is 181, which corresponds to the 62.50th percentile since 2023[14] - **Model Evaluation**: The index suggests that there is still significant room for downward movement in the market[14] 3. Model Name: High-Frequency Capital Flow Model - **Model Construction Idea**: This model uses high-frequency capital flow data to generate buy and sell signals for major broad-based indices[14] - **Model Construction Process**: The model tracks the capital flow trends of indices such as CSI 300, CSI 500, CSI 1000, and CSI 2000. The signals for all indices are currently negative, indicating a bearish outlook[14][18] - **Model Evaluation**: The model shows that all major broad-based indices have turned negative, reflecting weak market conditions[14][18] --- Model Backtesting Results 1. Sentiment Model - Sentiment score: 0 (out of 5)[14][18] 2. Moving Average Strength Index - Current score: 181 (62.50th percentile since 2023)[14] 3. High-Frequency Capital Flow Model - CSI 300: Negative signal - CSI 500: Negative signal - CSI 1000: Negative signal - CSI 2000: Negative signal[14][18] --- Quantitative Factors and Construction Methods 1. Factor Name: Factor Crowding Index - **Factor Construction Idea**: The factor crowding index measures the degree of crowding in specific factors, which can serve as a warning for factor inefficiency[19] - **Factor Construction Process**: The index is calculated using four metrics: valuation spread, pairwise correlation, long-term return reversal, and factor volatility. The composite score is derived from these metrics. For example, the crowding scores for small-cap, low-valuation, high-profitability, and high-growth factors are 0.06, -0.31, -0.01, and 0.28, respectively[19][20] - **Factor Evaluation**: The crowding index provides insights into the potential inefficiency of factors due to excessive capital allocation[19] --- Factor Backtesting Results 1. Factor Crowding Index - Small-cap factor crowding score: 0.06 - Low-valuation factor crowding score: -0.31 - High-profitability factor crowding score: -0.01 - High-growth factor crowding score: 0.28[19][20]

Quantitative Models and Construction Methods 1. Model Name: Sentiment Model - **Model Construction Idea**: The sentiment model is designed to measure the strength of market sentiment using factors related to limit-up and limit-down stocks[14] - **Model Construction Process**: The model incorporates factors such as the proportion of net limit-up stocks, next-day returns of limit-down stocks, proportion of limit-up stocks, proportion of limit-down stocks, and high-frequency board-hitting returns. These factors are aggregated to generate a sentiment score, with a maximum score of 5. The sentiment score for the current period is 0[14][18] - **Model Evaluation**: The sentiment model indicates weak market sentiment, as reflected by the score of 0[14][18] 2. Model Name: Moving Average Strength Index - **Model Construction Idea**: This model evaluates the strength of market trends by calculating the moving average strength index based on secondary industry indices[14] - **Model Construction Process**: The moving average strength index is calculated using the performance of secondary industry indices. The current market score is 181, which corresponds to the 62.50th percentile since 2023[14] - **Model Evaluation**: The model suggests that the market still has significant downside potential[14] 3. Model Name: High-Frequency Capital Flow Model - **Model Construction Idea**: This model uses high-frequency capital flow trends to generate buy and sell signals for major broad-based indices[14] - **Model Construction Process**: The model tracks high-frequency capital flows and generates signals for indices such as CSI 300, CSI 500, CSI 1000, and CSI 2000. The signals for all indices are currently negative, indicating a bearish outlook[14][18] - **Model Evaluation**: The model shows a bearish signal across all major indices, reflecting weak market conditions[14][18] --- Model Backtesting Results 1. Sentiment Model - Sentiment score: 0 (out of 5)[14][18] 2. Moving Average Strength Index - Current score: 181 (62.50th percentile since 2023)[14] 3. High-Frequency Capital Flow Model - CSI 300: Negative signal - CSI 500: Negative signal - CSI 1000: Negative signal - CSI 2000: Negative signal[14][18] --- Quantitative Factors and Construction Methods 1. Factor Name: Factor Crowding Indicator - **Factor Construction Idea**: The factor crowding indicator measures the degree of crowding in specific factors, which can serve as a warning for factor underperformance[19] - **Factor Construction Process**: The indicator is calculated using four metrics: valuation spread, pairwise correlation, long-term return reversal, and factor volatility. These metrics are aggregated to produce a composite crowding score for each factor. For example: - Small-cap factor crowding score: 0.06 - Low-valuation factor crowding score: -0.31 - High-profitability factor crowding score: -0.01 - High-growth factor crowding score: 0.28[19][20] - **Factor Evaluation**: The crowding scores indicate varying levels of crowding across factors, with low-valuation and high-profitability factors showing negative scores, suggesting potential underperformance[19][20] --- Factor Backtesting Results 1. Factor Crowding Indicator - Small-cap factor crowding score: 0.06 - Low-valuation factor crowding score: -0.31 - High-profitability factor crowding score: -0.01 - High-growth factor crowding score: 0.28[19][20]

Quantitative Models and Construction Methods - **Model Name**: Sparse Auto Encoder (SAE) **Model Construction Idea**: The model aims to compress high-dimensional features into low-dimensional sparse coding while ensuring the reconstructed features retain most of the original information. It also incorporates autoregressive loss and sparsity penalties to enhance robustness and reduce overfitting [7][8][9] **Model Construction Process**: 1. **Encoding**: Compress input features into sparse coding $ \text{code}_{i}=\text{Encoder}(x_{i}) $ Here, $ x_{i} $ represents input features, and $ \text{code}_{i} $ is the compressed sparse coding [8] 2. **Decoding**: Reconstruct features from sparse coding $ \hat{x}_{i}=\text{Decoder}(code_{i}) $ $ \hat{x}_{i} $ represents reconstructed features, which should closely resemble $ x_{i} $ [8] 3. **Prediction**: Predict future index returns using hidden layer features $ \hat{y}_{i}=\text{Predictor}(res_{i}) $ $ \hat{y}_{i} $ represents the predicted future returns [8] 4. **Loss Function**: Combines prediction error, reconstruction error, and sparsity penalty $$ Loss=\frac{1}{N}\sum\nolimits_{i=1}^{N}\left(\mathcal{J}(y_{i},{\hat{y}}_{i})+\lambda_{1}\mathcal{L}\left(x_{i},{\hat{x}}_{i}\right)+\lambda_{2}SparseLoss(code_{i})\right) $$ $ \mathcal{J} $ measures prediction error, $ \mathcal{L} $ measures reconstruction error, and $ SparseLoss $ applies sparsity penalties using KL divergence or vector norms [9][12] **Evaluation**: The model effectively selects features, enhances robustness, and learns the "true" patterns of index movements [11] - **Wavelet Transform for Noise Reduction** **Construction Idea**: Decompose time-series data into multiple components to isolate noise and retain meaningful information [19][20] **Construction Process**: 1. Select parent wavelet $ \varphi $ and mother wavelet $ \psi $ $ \varphi_{jk}=2^{-j/2}\varphi(2^{-j}-k) $ $ \psi_{jk}=2^{-j/2}\psi(2^{-j}-k) $ Parent wavelet captures low-frequency trends, while mother wavelet captures high-frequency fluctuations [19][20] 2. Reconstruct time-series data using wavelet coefficients $$ x(t)=\sum\nolimits_{k}s_{j,k}\varphi_{j,k}+\sum\nolimits_{k}d_{j,k}\psi_{j,k}+\ldots+\sum\nolimits_{k}d_{1,k}\psi_{1,k} $$ Coefficients $ S_{J,k} $ and $ d_{j,k} $ are calculated as: $ S_{J,k}=\int\varphi_{J,k}x(s)ds $ $ d_{j,k}=\int\psi_{J,k}x(s)ds $ [20] **Evaluation**: Reduces overfitting risks by filtering out noise and retaining meaningful components [21] Model Backtesting Results - **SAE Model** **Performance on CSI 500 Index**: - Multi-strategy annualized return: 43.86% - Long-only annualized return: 23.30% - Short-only annualized return: 16.68% - Sharpe ratio: 2.07 (multi-strategy), 1.39 (long-only), 1.28 (short-only) - Maximum drawdown: -14.00% (multi-strategy), -16.04% (long-only), -14.30% (short-only) [29][33][34] **Performance on CSI 1000 Index**: - Multi-strategy annualized return: 51.21% - Long-only annualized return: 26.00% - Short-only annualized return: 20.01% - Sharpe ratio: 1.41 (long-only), 1.27 (short-only) - Maximum drawdown: -22.08% (long-only), -19.85% (short-only) [43][46][47] **Performance on CSI 2000 Index**: - Multi-strategy annualized return: 32.40% - Long-only annualized return: 32.56% - Sharpe ratio: 1.62 (long-only) - Maximum drawdown: -25.59% (long-only) [55][56] **Performance on CSI All Share Index**: - Multi-strategy annualized return: 18.74% - Long-only annualized return: 18.83% - Sharpe ratio: 1.26 (long-only) - Maximum drawdown: -16.95% (long-only) [55][56] Quantitative Factors and Construction Methods - **Input Features** **Construction Idea**: Use common technical indicators and derived metrics from daily K-line data as model inputs [16][18] **Construction Process**: 1. **Technical Indicators**: - RSI: $ RSI=(N\text{-day absolute closing price increase})/(N\text{-day absolute closing price decrease}) $ - OBV: $ OBV=\text{sum of closing price change signs} \times \text{turnover rate} $ - MACD: $ DIF=12\text{-day EMA}-26\text{-day EMA} $ $ DEA=DIF\text{'s 9-day EMA} $ $ MACD=DIF-DEA $ [16][17] 2. **Derived Metrics**: Rolling averages, relative positions of moving averages, volatility metrics, and other derived indicators [16][18] **Evaluation**: The feature set is comprehensive but not optimized, as no additional filtering was applied to avoid overfitting [18] Factor Backtesting Results - **RSI, OBV, MACD** **Performance**: Incorporated into the SAE model, contributing to the overall strategy performance across indices [16][18] Key Observations - The SAE model performs better on smaller-cap indices like CSI 2000 and CSI 1000 compared to CSI 500, indicating its effectiveness in smaller market segments [62] - Multi-strategy returns are balanced between long and short positions, with no significant bias toward either direction [42][54] - The model's robustness and sparsity design mitigate overfitting risks and enhance generalization across different market conditions [11][21] - Setting appropriate thresholds for signal generation improves strategy stability and reduces transaction costs [66]

1. Report Industry Investment Rating - Not provided in the content 2. Core Viewpoints - The report presents several ETF strategy indices constructed by Huabao Research and tracks their performance and positions on a weekly basis, aiming to help investors convert quantitative models or subjective views into practical investment strategies [11] 3. Summary by Relevant Catalog 3.1 ETF Strategy Index Tracking - **Overall Performance**: The table shows the performance of various ETF strategy indices last week. The Huabao Research Quantitative Windmill ETF Strategy Index had the highest weekly excess return of 2.56%, while the Huabao Research SmartBeta Enhanced ETF Strategy Index had the lowest weekly excess return of -2.76% [12] 3.1.1 Huabao Research Size Rotation ETF Strategy Index - **Strategy**: It uses multi - dimensional technical indicator factors and a machine - learning model to predict the return difference between the Shenwan Large - Cap Index and the Shenwan Small - Cap Index. It outputs weekly signals to predict the strength of the indices in the next week and determines positions accordingly to obtain excess returns [13] - **Performance**: As of 2026/1/30, the excess return since 2024 was 29.34%, the excess return in the past month was 5.89%, and the excess return in the past week was - 1.86%. The index's positions include 50% in the CSI 500ETF and 50% in the CSI 1000ETF [13][17] 3.1.2 Huabao Research SmartBeta Enhanced ETF Strategy Index - **Strategy**: It uses price - volume indicators to time self - built Barra factors and maps timing signals to ETFs based on their exposure to 9 major Barra factors to achieve market - outperforming returns. The selected ETFs cover mainstream broad - based index ETFs and some style and strategy ETFs [17] - **Performance**: As of 2026/1/30, the excess return since 2024 was 20.15%, the excess return in the past month was - 2.11%, and the excess return in the past week was - 2.76%. The index's positions are mainly in several science - innovation and growth - style ETFs [17] 3.1.3 Huabao Research Quantitative Windmill ETF Strategy Index - **Strategy**: It starts from a multi - factor perspective, including the grasp of medium - to - long - term fundamentals, tracking of short - term market trends, and analysis of the behavior of various market participants. It uses valuation and crowding signals to indicate industry risks and multi - dimensionally digs out potential sectors to obtain excess returns [20] - **Performance**: As of 2026/1/30, the excess return since 2024 was 51.39%, the excess return in the past month was 6.51%, and the excess return in the past week was 2.56%. The index's positions are mainly in commodity - related and financial - related ETFs [20][25] 3.1.4 Huabao Research Quantitative Balance ETF Strategy Index - **Strategy**: It adopts a multi - factor system, including economic fundamentals, liquidity, technical analysis, and investor behavior factors, to construct a quantitative timing system for trend analysis of the equity market. It also builds a prediction model for market large - and small - cap styles to adjust the equity market position distribution and obtain excess returns through comprehensive timing and rotation [24] - **Performance**: As of 2026/1/30, the excess return since 2024 was - 10.24%, the excess return in the past month was 0.48%, and the excess return in the past week was - 0.36%. The index's positions include bonds and equity - based ETFs [24][27] 3.1.5 Huabao Research Hot - Spot Tracking ETF Strategy Index - **Strategy**: It uses strategies such as market sentiment analysis, tracking of major industry events, investor sentiment and professional opinions, policy and regulatory changes, and historical analysis to track and dig out hot - spot index target products in a timely manner, constructing an ETF portfolio that can capture market hot spots and providing short - term market trend references for investors [27] - **Performance**: As of 2026/1/30, the excess return in the past month was 6.21%, and the excess return in the past week was 3.21%. The index's positions are mainly in commodity, Hong - Kong - stock, and short - term financing ETFs [27][30] 3.1.6 Huabao Research Bond ETF Duration Strategy Index - **Strategy**: It uses bond market liquidity and price - volume indicators to screen effective timing factors and predicts bond yields through machine - learning methods. When the expected yield is below a certain threshold, it reduces the long - duration positions in the bond investment portfolio to improve long - term returns and drawdown control [30] - **Performance**: As of 2026/1/30, the excess return in the past month was 0.40%, and the excess return in the past week was 0.14%. The index's positions are mainly in bond - related ETFs [30][33]

Group 1 - The market sentiment indicator as of January 30 is at 2.6, a slight increase from 2.35 the previous week, indicating overall stability in sentiment with a bullish model perspective [2][9]. - The price-volume consistency indicator remains high, suggesting a strong correlation between market attention and stock price movements, reflecting an active market sentiment [13][16]. - The trading volume of the entire A-share market increased by 9.44% week-on-week, with an average daily trading volume of 30,632.46 billion yuan, indicating a slight recovery in market activity [19]. Group 2 - The short-term score rankings show that the oil and petrochemical, construction materials, non-ferrous metals, light industry manufacturing, and communication sectors are leading, with both oil and petrochemical and construction materials scoring 98.31, the highest among sectors [43][44]. - The industry crowding indicator shows a positive correlation with weekly price changes, with high crowding sectors like oil and petrochemical leading in gains, while low crowding sectors like commercial retail and environmental protection lag behind [46][50]. - The model indicates a preference for small-cap and growth styles, with the 5-day RSI showing a rapid decline relative to the 20-day RSI, suggesting potential weakening of signals in the near term [43][53].

Core Viewpoint - The market may present certain structural opportunities in the upcoming week, despite a generally pessimistic market sentiment indicated by technical models [1][2]. Market Overview - Last week (January 26-30, 2026), the Shanghai Composite Index rose by 1.13%, while the CSI 300 Index increased by 0.08%. Conversely, the CSI 500 Index fell by 2.56%, and the ChiNext Index decreased by 0.09% [3]. - The current overall market PE (TTM) stands at 23.3 times, which is at the 82.0% percentile since 2005 [3]. Quantitative Indicators - The liquidity shock indicator for the CSI 300 Index was 5.07 on Friday, slightly lower than the previous week (5.09), indicating current market liquidity is 5.07 times the average level over the past year [2]. - The PUT-CALL ratio for the SSE 50 ETF decreased to 0.89 from 0.98, suggesting a decline in investor caution regarding the short-term performance of the SSE 50 ETF [2]. - The five-day average turnover rates for the Shanghai Composite Index and Wind All A were 1.75% and 2.49%, respectively, indicating increased trading activity [2]. Macro Factors - The RMB exchange rate fluctuated last week, with onshore and offshore rates increasing by 0.07% and 0.27%, respectively [2]. - The official manufacturing PMI for China in January was reported at 49.3, lower than the previous value (50.1) and below the consensus expectation (50.18) [2]. Seasonal Trends - Historical data since 2005 shows that major indices have a high probability of rising in the first half of February, with average gains of 2.85% for the Shanghai Composite, 3.61% for the CSI 300, 5.34% for the CSI 500, and 4.65% for the ChiNext [2]. Technical Analysis - The SAR indicator for the Wind All A Index broke downwards on January 20 but rebounded upwards on January 23 [2]. - The market score based on moving average strength is currently at 172, placing it at the 59.50% percentile for 2023 [2]. - The sentiment model score is at 0 (out of 5), indicating a negative trend signal [2]. Industry Crowding - The industry crowding levels are relatively high in telecommunications, non-ferrous metals, comprehensive sectors, electronics, and basic chemicals, with notable increases in electronics and defense industries [4].

Investment Rating - The industry rating is "Overweight," indicating an expected increase of over 10% relative to the benchmark index in the next 6 to 12 months [16]. Core Insights - The market is currently in an upward trend, with the core observation being whether the profit-making effect is positive. The market trend line is near 6800 points, and the profit-making effect is at the zero value edge, suggesting the potential for a halt in the upward trend [5][8]. - The short-term outlook indicates continued outflows from broad-based ETFs, particularly the CSI 300 ETF, which is exerting pressure on the index. A significant drop in non-ferrous metals has also dampened short-term risk appetite [7][8]. - The industry trend configuration model suggests waiting for reversal signals in the liquor and real estate sectors, while the TWO BETA model continues to recommend the technology sector, focusing on rebound opportunities in commercial aerospace [6][8]. Summary by Sections Market Overview - The WIND All A index is in an upward trend, with a significant distance of 6.77% between the short-term (20-day) and long-term (120-day) moving averages, indicating a continued upward trend [5][9]. - The market experienced a decline of 1.59% last week, with small-cap stocks (CSI 1000) down 2.55% and mid-cap stocks (CSI 500) down 2.56%. The CSI 300 saw a slight increase of 0.08%, while the SSE 50 rose by 1.13% [2][7]. Valuation Metrics - The PE ratio of the WIND All A index is at the 90th percentile, indicating a high valuation level, while the PB ratio is at the 50th percentile, suggesting a moderate valuation level [9][11]. Positioning Recommendations - The report recommends a 70% allocation to absolute return products based on the WIND All A index, reflecting a cautious but optimistic stance on market conditions [9][10]. - The performance trend model highlights the importance of focusing on the computing power-related industry chain and suggests waiting for significant volume reductions in high-performing cyclical sectors such as industrial non-ferrous metals and chemicals [6][14].

证券研究报告/金融工程定期报告 2026 年 02 月 01 日 分析师：吴先兴 执业证书编号：S0740525110003 Email：wuxx02@zts.com.cn 量化择时周报：趋势指标进入边缘位置，由重仓位到重结构 分析师：王鹏飞 执业证书编号：S0740525060001 Email：wangpf@zts.com.cn 1、《量化择时周报：牛市格局仍在延 续，主题投资重回主线》2026-01-25 2、《沪深 300 增强策略本周超额收益 3.90%》2026-01-25 3、《净利润断层策略本周绝对收益 1.99%》2026-01-18 报告摘要 趋势指标进入边缘位置，由重仓位到重结构 请务必阅读正文之后的重要声明部分 相关报告 上周周报（20260125）认为：尽管 ETF 份额的持续下降，对市场有短线压力，但在 每日近 3 万亿金额的成交下，预计影响也较为有限，市场上行趋势仍将延续。最终 WIND 全 A 在 ETF 份额持续下降的影响和周五周期股大幅回调的影响下，全周下跌 1.59%。市值维度上，上周代表小市值股票的中证 1000 下跌 2.55%，中盘股中证 500 指数下跌 2.56 ...

Quantitative Models and Construction Methods 1. Model Name: Convolutional Neural Network (CNN) for Price-Volume Data Modeling - **Model Construction Idea**: The model leverages convolutional neural networks to analyze standardized graphical representations of price-volume data, aiming to predict future price trends and map learned features to industry thematic indices[79][81] - **Model Construction Process**: - Standardize price-volume data into graphical formats for each stock within a specific time window[79] - Apply convolutional neural networks to extract features from these graphical representations[79] - Map the extracted features to thematic industry indices, such as the CSI Petrochemical Industry Index, CSI Subdivision Chemical Industry Theme Index, and others[81] - **Model Evaluation**: The model effectively identifies industry themes based on price-volume data and provides actionable insights for sector allocation[79][81] --- Model Backtesting Results 1. CNN Model - **Thematic Indices Configured**: - CSI Petrochemical Industry Index (h11057.CSI)[81] - CSI Subdivision Chemical Industry Theme Index (000813.CSI)[81] - CNI Oil & Gas Index (399439.SZ)[81] - CSI Oil & Gas Resources Index (931248.CSI)[81] - CNI Nonferrous Metals Index (399395.SZ)[81]

Quantitative Models and Construction Methods 1. Model Name: Sentiment Model - **Model Construction Idea**: The sentiment model is designed to capture the market's emotional state by analyzing various market indicators[2][14] - **Model Construction Process**: - The sentiment model score is calculated based on the proportion of stocks hitting their daily limit up or down, and the subsequent returns of stocks that hit their limit down the previous day[14] - The sentiment model score is 0 out of 5, indicating a negative sentiment[14] - **Model Evaluation**: The sentiment model indicates a pessimistic market sentiment[2][14] 2. Model Name: High-Frequency Capital Flow Model - **Model Construction Idea**: This model uses high-frequency capital flow data to generate buy and sell signals for major indices[2][14] - **Model Construction Process**: - The model tracks the capital flow trends for major indices such as CSI 300, CSI 500, CSI 1000, and CSI 2000[14] - Signals are generated based on the direction of capital flows: positive for buy signals and negative for sell signals[14] - For the past week, the signals were positive for CSI 300 and CSI 500, and negative for CSI 1000 and CSI 2000[2][14] - **Model Evaluation**: The model indicates that CSI 300 and CSI 500 are in a buying cycle, while CSI 1000 and CSI 2000 are in a selling cycle[2][14] Model Backtesting Results Sentiment Model - **Sentiment Model Score**: 0 out of 5[14] High-Frequency Capital Flow Model - **CSI 300**: Positive signal[14] - **CSI 500**: Positive signal[14] - **CSI 1000**: Negative signal[14] - **CSI 2000**: Negative signal[14] Quantitative Factors and Construction Methods 1. Factor Name: Small Market Cap Factor - **Factor Construction Idea**: Measures the degree of crowding in small-cap stocks[18] - **Factor Construction Process**: - The factor's crowding degree is calculated using valuation spreads, pairwise correlations, long-term return reversals, and factor volatility[18] - The composite score for the small market cap factor is 0.05[19] - **Factor Evaluation**: The crowding degree of the small market cap factor has decreased[18] 2. Factor Name: Low Valuation Factor - **Factor Construction Idea**: Measures the degree of crowding in low-valuation stocks[18] - **Factor Construction Process**: - The factor's crowding degree is calculated using the same metrics as the small market cap factor[18] - The composite score for the low valuation factor is -0.28[19] - **Factor Evaluation**: The crowding degree of the low valuation factor is relatively low[18] 3. Factor Name: High Profitability Factor - **Factor Construction Idea**: Measures the degree of crowding in high-profitability stocks[18] - **Factor Construction Process**: - The factor's crowding degree is calculated using the same metrics as the small market cap factor[18] - The composite score for the high profitability factor is 0.20[19] - **Factor Evaluation**: The crowding degree of the high profitability factor is moderate[18] 4. Factor Name: High Growth Factor - **Factor Construction Idea**: Measures the degree of crowding in high-growth stocks[18] - **Factor Construction Process**: - The factor's crowding degree is calculated using the same metrics as the small market cap factor[18] - The composite score for the high growth factor is 0.53[19] - **Factor Evaluation**: The crowding degree of the high growth factor is relatively high[18] Factor Backtesting Results Small Market Cap Factor - **Crowding Degree**: 0.05[19] Low Valuation Factor - **Crowding Degree**: -0.28[19] High Profitability Factor - **Crowding Degree**: 0.20[19] High Growth Factor - **Crowding Degree**: 0.53[19]