研究所 分析师:肖承志 SAC 登记编号:S1340524090001 Email:xiaochengzhi@cnpsec.com 近期研究报告 证券研究报告：金融工程报告 《本周微盘股大幅跑输的三个原因— 《微盘股将再次迎来高胜率区间—— 微盘股指数周报 20250803》 - 2025.08.04 《微盘股持续创新高背后的历史意义 《短期上涨动能枯竭，控制仓位做好 防御——微盘股指数周报 20250615》 - 2025.06.16 《为何微盘股基金仓位下降指数却不 断新高？——微盘股指数周报 20250608》 - 2025.06.09 《小盘股成交占比高意味着拥挤度高 吗？——微盘股指数周报 20250601》 - 2025.06.02 金工周报 微盘股成交占比持续回落 ——微盘股指数周报 20250824 ⚫ 投资要点 本周股指期货贴水幅度先收窄后走阔，这意味着基差端仍在亏 损，只不过亏损的幅度开始收窄，斜率开始放缓。中证 1000 指数本 周表现和微盘相当，意味着 beta 端的风险暂时消除。最后是大盘股 虹吸效应仍然存在，同样的泛微盘成交占比也回落至 21%附近，也是 阶段性底部区间。 ⚫ 万得 ...

证券研究报告：金融工程报告 研究所 分析师:肖承志 SAC 登记编号:S1340524090001 Email:xiaochengzhi@cnpsec.com 近期研究报告 《大盘资金流出，中小盘资金回流— —微盘股指数周报 20250810》 - 2025.08.11 《微盘股将再次迎来高胜率区间—— 微盘股指数周报 20250803》 - 2025.08.04 《微盘股持续创新高背后的历史意义 有何不同？——微盘股指数周报 20250727》 - 2025.07.28 《微盘股的流动性风险在哪？——微 盘股指数周报 20250720》 - 2025.07.21 《上证站上 3500 点，后续关注小盘+ 科技成长——微盘股指数周报 20250713》 - 2025.07.14 《"量化新规"或将平稳落地，双均线 法再现买点——微盘股指数周报 20250706》 - 2025.07.07 《现阶段主要矛盾是交易范式之争— —微盘股指数周报 20250629》 - 2025.06.30 《调整仍不充分——微盘股指数周报 20250622》 - 2025.06.23 《短期上涨动能枯竭，控制仓位做好 防御——微盘 ...

Quantitative Models and Construction Methods 1. Model Name: Diffusion Index Model - **Model Construction Idea**: The model is used to monitor the critical points of trend changes in the micro-cap stock index by analyzing the distribution of stock price movements over a specific time window[6][38]. - **Model Construction Process**: The model calculates the diffusion index based on the relative price changes of all constituent stocks in the micro-cap stock index over a given time window. For example, if the horizontal axis is 0.95 and the vertical axis is 15 days, the value of 0.51 indicates that after 5 days (review period from T=20 to T=15), if all stocks in the micro-cap index drop by 5%, the diffusion index value is 0.51[38]. - Current diffusion index value: 0.87 (horizontal axis = 20, vertical axis = 1.00)[38]. - **Model Evaluation**: The model shows that the current distribution is relatively uniform, indicating that the time window has little impact, and the main influence comes from the spatial distribution[38]. 2. Model Name: First Threshold Method (Left-Side Trading) - **Model Construction Idea**: This method triggers a sell signal when the diffusion index reaches a predefined threshold[6][42]. - **Model Construction Process**: - On May 8, 2025, the model triggered a sell signal when the diffusion index reached 0.9850[42]. 3. Model Name: Delayed Threshold Method (Right-Side Trading) - **Model Construction Idea**: Similar to the first threshold method but with a delayed response to confirm the trend[6][44]. - **Model Construction Process**: - On May 15, 2025, the model triggered a sell signal when the diffusion index reached 0.8975[46]. 4. Model Name: Dual Moving Average Method (Adaptive Trading) - **Model Construction Idea**: This method uses two moving averages to adaptively identify trading signals based on the diffusion index[6][47]. - **Model Construction Process**: - On August 4, 2025, the model issued a sell signal based on the dual moving average strategy[47]. --- Model Backtesting Results 1. Diffusion Index Model - Current diffusion index value: 0.87[38]. 2. First Threshold Method - Triggered sell signal at diffusion index value: 0.9850[42]. 3. Delayed Threshold Method - Triggered sell signal at diffusion index value: 0.8975[46]. 4. Dual Moving Average Method - Triggered sell signal on August 4, 2025[47]. --- Quantitative Factors and Construction Methods 1. Factor Name: Momentum Factor - **Factor Construction Idea**: Measures the tendency of stocks to continue their past performance[5][33]. - **Factor Construction Process**: - Weekly rank IC: 0.224 - Historical average rank IC: -0.005[5][33]. 2. Factor Name: Beta Factor - **Factor Construction Idea**: Captures the sensitivity of stock returns to market movements[5][33]. - **Factor Construction Process**: - Weekly rank IC: 0.146 - Historical average rank IC: 0.006[5][33]. 3. Factor Name: Illiquidity Factor - **Factor Construction Idea**: Measures the impact of trading volume on stock price changes[5][33]. - **Factor Construction Process**: - Weekly rank IC: 0.14 - Historical average rank IC: 0.041[5][33]. 4. Factor Name: Unadjusted Stock Price Factor - **Factor Construction Idea**: Reflects the raw stock price without adjustments for splits or dividends[5][33]. - **Factor Construction Process**: - Weekly rank IC: 0.131 - Historical average rank IC: -0.015[5][33]. 5. Factor Name: PE_TTM Reciprocal Factor - **Factor Construction Idea**: Represents the inverse of the price-to-earnings ratio based on trailing twelve months[5][33]. - **Factor Construction Process**: - Weekly rank IC: 0.125 - Historical average rank IC: 0.017[5][33]. --- Factor Backtesting Results 1. Momentum Factor - Weekly rank IC: 0.224 - Historical average rank IC: -0.005[5][33]. 2. Beta Factor - Weekly rank IC: 0.146 - Historical average rank IC: 0.006[5][33]. 3. Illiquidity Factor - Weekly rank IC: 0.14 - Historical average rank IC: 0.041[5][33]. 4. Unadjusted Stock Price Factor - Weekly rank IC: 0.131 - Historical average rank IC: -0.015[5][33]. 5. PE_TTM Reciprocal Factor - Weekly rank IC: 0.125 - Historical average rank IC: 0.017[5][33].

Core Insights - The report from Founder Securities indicates that after a continuous rise in stock prices, the valuation levels of small and micro-cap stocks have significantly increased, yet both absolute and relative valuation levels remain considerably below historical extremes [1][3] Valuation Analysis - Small-cap companies are expected to yield significant excess returns by 2025, with the CSI 2000 index showing a year-to-date increase of 25.3%, outperforming other major indices [2] - The Wande Micro-cap Index has surged over 50%, reaching historical highs, while the overall P/E ratio of the CSI 2000 index is currently at 146 times, and the P/B ratio is 2.75, both being the highest since the index was launched in 2023 [2] - The overall valuation of small-cap stocks is heavily influenced by a few loss-making companies, with the CSI 2000 index reporting a total net profit of 77.5 billion, contrasted by losses totaling 121.1 billion from loss-making firms [2] Relative Valuation Comparison - A comparison of the smallest 20% of A-share companies against the largest 20% shows a median P/E ratio of 1.17 times and a median P/B ratio of 1.09 times, indicating that relative valuation levels are consistent with those of the CSI 2000 constituents and even lower [3] - Despite the increase in stock prices, the valuation levels of small and micro-cap stocks have risen but still have a significant distance to historical peaks, suggesting that the small-cap style may reflect the innovative exploration characteristics during the economic transformation and upgrading period [3]

证券研究报告：金融工程报告 研究所 分析师:肖承志 SAC 登记编号:S1340524090001 Email:xiaochengzhi@cnpsec.com 近期研究报告 《微盘股持续创新高背后的历史意义 有何不同？——微盘股指数周报 20250727》 - 2025.07.28 《微盘股的流动性风险在哪？——微 盘股指数周报 20250720》 - 2025.07.21 《上证站上 3500 点，后续关注小盘+ 科技成长——微盘股指数周报 20250713》 - 2025.07.14 《"量化新规"或将平稳落地，双均线 法再现买点——微盘股指数周报 20250706》 - 2025.07.07 《现阶段主要矛盾是交易范式之争— —微盘股指数周报 20250629》 - 2025.06.30 《调整仍不充分——微盘股指数周报 20250622》 - 2025.06.23 《短期上涨动能枯竭，控制仓位做好 防御——微盘股指数周报 20250615》 - 2025.06.16 《为何微盘股基金仓位下降指数却不 断新高？——微盘股指数周报 20250608》 - 2025.06.09 《小盘股成交占比高意味着拥挤 ...

Quantitative Models and Construction Methods Model Name: Black Swan Index - **Construction Idea**: Measure the extremity of market transactions based on the deviation of style factor returns[24][25] - **Construction Process**: 1. Calculate the daily return deviation of style factors: $$ \sigma_{s,t}=\frac{\bar{r}_{s,t}-\bar{r}_{s}}{\sigma_{s}} $$ where $\bar{r}_{s,t}$ is the daily return of style factor $s$ on day $t$, $\bar{r}_{s}$ is the average daily return of style factor $s$ over the entire sample period, and $\sigma_{s}$ is the standard deviation of daily returns of style factor $s$ over the entire sample period[25] 2. Calculate the Black Swan Index: $$ BlackSwan_{t}=\frac{1}{N}\times\sum_{s\in S}\left|\sigma_{s,t}\right| $$ where $BlackSwan_{t}$ is the Black Swan Index on day $t$, $S$ is the set of all style factors, and $N$ is the number of style factors[25] - **Evaluation**: The Black Swan Index effectively captures the extremity of market transactions, indicating higher probabilities of extreme tail risks[24][25] Model Name: Heuristic Style Classification for Cognitive Risk Control - **Construction Idea**: Address the discrepancy between individual and collective cognition in style classification to control cognitive risk[80][81] - **Construction Process**: 1. Calculate the value and growth factors for each stock based on predefined metrics[85] 2. Construct value and growth portfolios by selecting the top 10% and bottom 10% stocks based on factor scores[82] 3. Perform time-series regression to classify stocks into value, growth, or balanced styles: $$ r_{t,t}\sim\beta_{\mathit{Value}}\cdot r_{\mathit{Value},t}+\beta_{\mathit{Growth}}\cdot r_{\mathit{Growth},t}+\varepsilon_{t} $$ subject to $0\leq\beta_{\mathit{Value}}\leq1$, $0\leq\beta_{\mathit{Growth}}\leq1$, and $\beta_{\mathit{Value}}+\beta_{\mathit{Growth}}=1$[97] 4. Use weighted least squares (WLS) to estimate regression coefficients based on the most differentiated trading days[98] - **Evaluation**: The heuristic style classification method captures market consensus more accurately than traditional factor scoring methods, reducing cognitive risk[80][81] Model Name: Louvain Community Detection for Hidden Risk Control - **Construction Idea**: Cluster stocks based on excess return correlations to identify hidden risks[116][117] - **Construction Process**: 1. Calculate weighted correlation of excess returns between stocks: $$ Corr_{w}(X,Y)=\frac{Cov_{w}(X,Y)}{\sigma_{w,X}\cdot\sigma_{w,Y}}=\frac{\sum_{i=1}^{n}w_{i}(x_{i}-\overline{X_{w}})(y_{i}-\overline{Y_{w}})}{\sqrt{\sum_{i=1}^{n}w_{i}(x_{i}-\overline{X_{w}})^{2}}\cdot\sqrt{\sum_{i=1}^{n}w_{i}(y_{i}-\overline{Y_{w}})^{2}}} $$ where $w_{i}$ is the weight for day $i$, reflecting market volatility[118] 2. Use Louvain algorithm to cluster stocks based on weighted correlation matrix[117] 3. Ensure clusters have at least 20 stocks and remove clusters with fewer stocks[121] - **Evaluation**: The Louvain community detection method effectively identifies hidden risks by clustering stocks with similar return patterns, which traditional risk models may overlook[116][117] Model Name: Dynamic Style Factor Control - **Construction Idea**: Control style factors dynamically based on their volatility clustering effect[128][129] - **Construction Process**: 1. Identify style factors with high volatility or significant volatility increase: $$ \text{High volatility: Rolling 3-month volatility in top 3} $$ $$ \text{Volatility increase: Rolling 3-month volatility > historical mean + 1 standard deviation} $$ 2. Set the exposure of these style factors to zero in the portfolio[136] - **Evaluation**: Dynamic style factor control captures major market risks without significantly affecting portfolio returns, leveraging the predictability of volatility clustering[128][129] Model Name: Adaptive Stock Deviation Control under Target Tracking Error - **Construction Idea**: Adjust stock deviation based on tracking error to control portfolio risk[146][147] - **Construction Process**: 1. Calculate rolling 3-month tracking error for different stock deviation levels[153] 2. Set the maximum stock deviation that keeps tracking error within the target range[153] - **Evaluation**: Adaptive stock deviation control effectively reduces tracking error during high market volatility, maintaining portfolio stability[146][147] Model Backtest Results Traditional CSI 500 Enhanced Index - **Annualized Excess Return**: 18.77%[5][162] - **Maximum Drawdown**: 9.68%[5][162] - **Information Ratio (IR)**: 3.56[5][162] - **Return-to-Drawdown Ratio**: 1.94[5][162] - **Annualized Tracking Error**: 4.88%[5][162] CSI 500 Enhanced Index with Full-Process Risk Control - **Annualized Excess Return**: 16.51%[5][169] - **Maximum Drawdown**: 4.90%[5][169] - **Information Ratio (IR)**: 3.94[5][169] - **Return-to-Drawdown Ratio**: 3.37[5][169] - **Annualized Tracking Error**: 3.98%[5][169]

Quantitative Models and Construction Methods - **Model Name**: Diffusion Index Model **Construction Idea**: The model monitors the future critical points of diffusion index changes to predict market trends[40][41] **Construction Process**: 1. Define the horizontal axis as the relative price change of stocks in the index, ranging from +10% to -10% 2. Define the vertical axis as the review period length (T days) or future period length (N days), where T ranges from 20 to 10 days, and N = 20 - T 3. Calculate the diffusion index value for each combination of horizontal and vertical axis values 4. Example: For N=5 days and T=15 days, if all stocks drop by 5%, the diffusion index value is 0.51[40] **Evaluation**: The model indicates that current trends are driven by leading stocks rather than bottom stock rotation, suggesting the weakening of reversal factors and strengthening of fundamental factors[41] - **Model Name**: First Threshold Method (Left-Side Trading) **Construction Idea**: This method triggers signals based on predefined threshold values[44] **Construction Process**: 1. Monitor the diffusion index value daily 2. Trigger a signal when the index reaches a specific threshold 3. Example: On May 8, 2025, the diffusion index value of 0.9850 triggered a sell signal[44] **Evaluation**: Provides early warning signals for market reversals[44] - **Model Name**: Delayed Threshold Method (Right-Side Trading) **Construction Idea**: Similar to the first threshold method but with delayed signal generation[46][48] **Construction Process**: 1. Monitor the diffusion index value daily 2. Trigger a signal when the index reaches a delayed threshold 3. Example: On May 15, 2025, the diffusion index value of 0.8975 triggered a sell signal[48] **Evaluation**: Offers a more conservative approach compared to the first threshold method[48] - **Model Name**: Dual Moving Average Method (Adaptive Trading) **Construction Idea**: Uses moving averages to adapt to market trends[49] **Construction Process**: 1. Calculate short-term and long-term moving averages of the diffusion index 2. Generate buy or sell signals based on the crossover of these averages 3. Example: On July 3, 2025, the method issued a buy signal[49] **Evaluation**: Effective in capturing trend reversals during adaptive market conditions[49] Model Backtesting Results - **Diffusion Index Model**: Current value is 0.89, showing an increasing trend at the bottom level[40][41] - **First Threshold Method**: Triggered a sell signal at 0.9850 on May 8, 2025[44] - **Delayed Threshold Method**: Triggered a sell signal at 0.8975 on May 15, 2025[48] - **Dual Moving Average Method**: Triggered a buy signal on July 3, 2025[49] Quantitative Factors and Construction Methods - **Factor Name**: Illiquidity Factor **Construction Idea**: Measures the impact of stock liquidity on returns[4][35] **Construction Process**: 1. Rank stocks based on their liquidity metrics 2. Calculate the rank IC (information coefficient) for the factor 3. Example: Weekly rank IC is 0.268, historical average is 0.04[4][35] **Evaluation**: Strong performance in the current week, significantly above historical averages[4][35] - **Factor Name**: Single-Quarter Net Profit Growth Factor **Construction Idea**: Evaluates the growth rate of net profit over a single quarter[4][35] **Construction Process**: 1. Calculate the quarterly net profit growth rate for each stock 2. Rank stocks based on growth rates 3. Example: Weekly rank IC is 0.062, historical average is 0.02[4][35] **Evaluation**: Moderate performance, slightly above historical averages[4][35] - **Factor Name**: Unadjusted Stock Price Factor **Construction Idea**: Uses raw stock prices without adjustments for splits or dividends[4][35] **Construction Process**: 1. Rank stocks based on their unadjusted prices 2. Calculate the rank IC for the factor 3. Example: Weekly rank IC is 0.055, historical average is -0.016[4][35] **Evaluation**: Positive performance, reversing historical negative trends[4][35] - **Factor Name**: Dividend Yield Factor **Construction Idea**: Measures the dividend yield of stocks[4][35] **Construction Process**: 1. Calculate the dividend yield for each stock 2. Rank stocks based on yield values 3. Example: Weekly rank IC is 0.046, historical average is 0.021[4][35] **Evaluation**: Consistent performance, slightly above historical averages[4][35] - **Factor Name**: PB Reciprocal Factor **Construction Idea**: Uses the reciprocal of the price-to-book ratio[4][35] **Construction Process**: 1. Calculate the reciprocal of PB for each stock 2. Rank stocks based on reciprocal values 3. Example: Weekly rank IC is 0.042, historical average is 0.033[4][35] **Evaluation**: Stable performance, aligned with historical averages[4][35] Factor Backtesting Results - **Illiquidity Factor**: Weekly rank IC is 0.268, historical average is 0.04[4][35] - **Single-Quarter Net Profit Growth Factor**: Weekly rank IC is 0.062, historical average is 0.02[4][35] - **Unadjusted Stock Price Factor**: Weekly rank IC is 0.055, historical average is -0.016[4][35] - **Dividend Yield Factor**: Weekly rank IC is 0.046, historical average is 0.021[4][35] - **PB Reciprocal Factor**: Weekly rank IC is 0.042, historical average is 0.033[4][35]

Quantitative Models and Construction Methods Diffusion Index Model - **Model Name**: Diffusion Index Model - **Construction Idea**: The model monitors the relative performance of stocks within the micro-cap index over different time windows to identify potential turning points in market trends [41][42] - **Construction Process**: - The horizontal axis represents the percentage change in stock prices from +10% to -10% (1.1 to 0.9) - The vertical axis represents the length of the review window, ranging from 20 days to 10 days - For example, at horizontal axis 0.95 and vertical axis 15 days, the value of 0.37 indicates that if all stocks in the micro-cap index drop by 5% after 5 days, the diffusion index value is 0.37 - Formula: Diffusion Index = $\frac{\text{Number of stocks outperforming the benchmark}}{\text{Total number of stocks}}$ [41][42] - **Evaluation**: The model effectively identifies market trends but faces challenges when bottom-performing stocks are abandoned during strong upward trends [42] - **Testing Results**: Current diffusion index value is 0.94, indicating a strong upward trend [41][42] Threshold Methods - **Model Name**: Threshold Methods (First Threshold Method and Delayed Threshold Method) - **Construction Idea**: These methods use predefined thresholds to generate trading signals based on the diffusion index [45][49] - **Construction Process**: - First Threshold Method: Triggered a sell signal on May 8, 2025, when the diffusion index reached 0.9850 [45] - Delayed Threshold Method: Triggered a sell signal on May 15, 2025, when the diffusion index reached 0.8975 [49] - **Evaluation**: These methods provide clear trading signals but may lag during rapid market changes [45][49] - **Testing Results**: First Threshold Method value: 0.9850; Delayed Threshold Method value: 0.8975 [45][49] Dual Moving Average Method - **Model Name**: Dual Moving Average Method - **Construction Idea**: This method uses adaptive moving averages to generate trading signals based on market trends [50] - **Construction Process**: - The method compares short-term and long-term moving averages to identify buy or sell signals - On July 3, 2025, the method generated a buy signal [50] - **Evaluation**: The method adapts well to changing market conditions and provides timely signals [50] - **Testing Results**: Buy signal generated on July 3, 2025 [50] --- Quantitative Factors and Construction Methods Top Performing Factors - **Factor Names**: Non-liquidity factor, Unadjusted stock price factor, Beta factor, Standardized expected earnings factor, PE_TTM reciprocal factor [4][19][36] - **Construction Idea**: These factors are derived from stock characteristics and financial metrics to predict future returns [4][19][36] - **Construction Process**: - Non-liquidity factor: Measures the illiquidity of stocks - Unadjusted stock price factor: Uses raw stock prices without adjustments - Beta factor: Captures the sensitivity of stock returns to market movements - Standardized expected earnings factor: Standardizes analysts' earnings forecasts - PE_TTM reciprocal factor: Calculates the reciprocal of the trailing twelve-month price-to-earnings ratio - **Evaluation**: These factors show strong predictive power for stock returns [4][19][36] - **Testing Results**: - Non-liquidity factor IC: 0.353 (historical average: 0.04) - Unadjusted stock price factor IC: 0.348 (historical average: -0.016) - Beta factor IC: 0.247 (historical average: 0.005) - Standardized expected earnings factor IC: 0.141 (historical average: 0.014) - PE_TTM reciprocal factor IC: 0.092 (historical average: 0.017) [4][19][36] Underperforming Factors - **Factor Names**: Turnover factor, 10-day total market capitalization turnover rate factor, Liquidity factor, 10-day free float market capitalization turnover rate factor, Leverage factor [4][19][36] - **Construction Idea**: These factors are derived from trading activity and financial leverage metrics [4][19][36] - **Construction Process**: - Turnover factor: Measures trading volume relative to market capitalization - 10-day total market capitalization turnover rate factor: Calculates turnover rate over a 10-day window - Liquidity factor: Assesses the ease of trading stocks - 10-day free float market capitalization turnover rate factor: Similar to the total turnover rate but focuses on free float shares - Leverage factor: Measures financial leverage of companies - **Evaluation**: These factors exhibit weak predictive power and negative correlations with returns [4][19][36] - **Testing Results**: - Turnover factor IC: -0.336 (historical average: -0.082) - 10-day total market capitalization turnover rate factor IC: -0.286 (historical average: -0.06) - Liquidity factor IC: -0.278 (historical average: -0.041) - 10-day free float market capitalization turnover rate factor IC: -0.276 (historical average: -0.062) - Leverage factor IC: -0.225 (historical average: -0.006) [4][19][36] --- Strategy Performance Small-Cap Low-Volatility 50 Strategy - **Strategy Name**: Small-Cap Low-Volatility 50 Strategy - **Construction Idea**: Selects 50 stocks with small market capitalization and low volatility from the micro-cap index [7][19][37] - **Construction Process**: - Stocks are selected bi-weekly based on market capitalization and volatility criteria - Benchmark: Wind Micro-Cap Index (8841431.WI) - Transaction cost: 0.3% on both sides [7][19][37] - **Evaluation**: The strategy demonstrates strong performance but occasionally underperforms the benchmark [7][19][37] - **Testing Results**: - 2024 return: 7.07% (excess return: -2.93%) - 2025 YTD return: 62.07% (weekly excess return: -2.44%) [7][19][37]

Core Viewpoint - The article discusses the effectiveness and risks associated with the "micro-cap stock strategy," highlighting its significant returns while also addressing the inherent volatility and potential for substantial losses [2][40]. Group 1: Micro-Cap Stock Strategy Effectiveness - The micro-cap stock strategy has shown long-term effectiveness, with the Wind Micro-Cap Stock Index achieving a remarkable 51.43% increase in 2025, contrasting with a mere 15.39% rise in bank stocks [2][40]. - The Wind Micro-Cap Stock Index is composed of the smallest 400 stocks from the A-share market, excluding certain categories, and is rebalanced monthly, which allows for high-frequency trading that capitalizes on market volatility [4][11]. - The strategy's success is attributed to its ability to generate excess returns through high-frequency trading rather than relying on fundamental company performance [12][19]. Group 2: Impact of Major Shareholder Actions - Major shareholder sell-offs have limited impact on micro-cap stocks due to regulatory constraints, making the perceived risks of such actions largely unfounded [14][40]. - The article emphasizes that the micro-cap stock strategy's returns are primarily driven by the index's trading rules rather than the underlying fundamentals of the companies involved [39][40]. Group 3: Risks and Considerations - The micro-cap stock strategy carries significant tail risks, with historical data showing potential declines of 40%-50% during market downturns, which investors must be prepared to endure [40][41]. - The article warns that while the micro-cap index can yield high returns, it is essential to recognize the associated risks and not to concentrate investments solely in this strategy [41][42]. - Diversification is recommended to mitigate risks, suggesting that investors should not allocate all resources to micro-cap stocks but rather create a balanced portfolio [42][43].

Quantitative Models and Construction 1. Model Name: Diffusion Index Model - **Model Construction Idea**: The model monitors the breadth of market movements and identifies turning points in stock price diffusion[5][38]. - **Model Construction Process**: The diffusion index is calculated based on the relative price movements of constituent stocks over a specific time window. For example, the current diffusion index value of 0.72 is derived from the relative price changes of stocks in the Wind Micro-Cap Index. The model uses thresholds to signal trading actions: - Left-side threshold method triggered a sell signal on May 8, 2025, at a value of 0.9850[43]. - Right-side threshold method triggered a sell signal on May 15, 2025, at a value of 0.8975[47]. - Dual moving average method triggered a buy signal on July 3, 2025[48]. - **Model Evaluation**: The model effectively identifies market turning points and provides actionable signals for trading strategies[39]. 2. Model Name: Small-Cap Low-Volatility 50 Strategy - **Model Construction Idea**: This strategy selects stocks with small market capitalization and low volatility to construct a portfolio[16][35]. - **Model Construction Process**: - Select 50 stocks from the Wind Micro-Cap Index based on small market capitalization and low volatility. - Rebalance the portfolio bi-weekly. - Transaction costs are set at 0.3% for both sides. - Benchmark: Wind Micro-Cap Index (8841431.WI)[16][35]. - **Model Evaluation**: The strategy demonstrates strong performance in 2025, with a year-to-date return of 56.90% and a weekly excess return of 0.04%[16][35]. --- Model Backtesting Results 1. Diffusion Index Model - Left-side threshold method: Sell signal at 0.9850 on May 8, 2025[43]. - Right-side threshold method: Sell signal at 0.8975 on May 15, 2025[47]. - Dual moving average method: Buy signal on July 3, 2025[48]. 2. Small-Cap Low-Volatility 50 Strategy - 2024 return: 7.07%, excess return: -2.93%[16][35]. - 2025 YTD return: 56.90%, weekly excess return: 0.04%[16][35]. --- Quantitative Factors and Construction 1. Factor Name: PB Inverse Factor - **Factor Construction Idea**: Measures the inverse of the price-to-book ratio to identify undervalued stocks[4][33]. - **Factor Construction Process**: - Calculate the inverse of the PB ratio for each stock in the Wind Micro-Cap Index. - Rank the stocks based on this value. - **Factor Evaluation**: This factor shows strong performance with a weekly rank IC of 0.152, significantly above its historical average of 0.034[4][33]. 2. Factor Name: Illiquidity Factor - **Factor Construction Idea**: Captures the illiquidity of stocks to identify those with higher potential returns[4][33]. - **Factor Construction Process**: - Measure the average daily turnover over a specific period. - Rank stocks inversely based on their turnover values. - **Factor Evaluation**: The factor has a weekly rank IC of 0.107, outperforming its historical average of 0.039[4][33]. 3. Factor Name: Profitability Factor - **Factor Construction Idea**: Identifies stocks with strong profitability metrics[4][33]. - **Factor Construction Process**: - Use metrics such as ROE or net profit margin to rank stocks. - **Factor Evaluation**: The factor has a weekly rank IC of 0.085, well above its historical average of 0.022[4][33]. 4. Factor Name: Momentum Factor - **Factor Construction Idea**: Tracks the momentum of stock prices to identify trends[4][33]. - **Factor Construction Process**: - Calculate the cumulative return over a specific period. - Rank stocks based on their momentum scores. - **Factor Evaluation**: The factor has a weekly rank IC of 0.069, improving from its historical average of -0.005[4][33]. 5. Factor Name: Leverage Factor - **Factor Construction Idea**: Measures the financial leverage of companies to identify risk-adjusted opportunities[4][33]. - **Factor Construction Process**: - Calculate the debt-to-equity ratio for each stock. - Rank stocks based on their leverage levels. - **Factor Evaluation**: The factor has a weekly rank IC of 0.064, outperforming its historical average of -0.005[4][33]. --- Factor Backtesting Results Top 5 Factors (Weekly Rank IC) 1. PB Inverse Factor: 0.152 (Historical Average: 0.034)[4][33]. 2. Illiquidity Factor: 0.107 (Historical Average: 0.039)[4][33]. 3. Profitability Factor: 0.085 (Historical Average: 0.022)[4][33]. 4. Momentum Factor: 0.069 (Historical Average: -0.005)[4][33]. 5. Leverage Factor: 0.064 (Historical Average: -0.005)[4][33]. Bottom 5 Factors (Weekly Rank IC) 1. Turnover Factor: -0.186 (Historical Average: -0.081)[4][33]. 2. Residual Volatility Factor: -0.154 (Historical Average: -0.040)[4][33]. 3. 10-Day Return Factor: -0.153 (Historical Average: -0.062)[4][33]. 4. 1-Year Volatility Factor: -0.153 (Historical Average: -0.033)[4][33]. 5. 10-Day Free Float Turnover Factor: -0.132 (Historical Average: -0.061)[4][33].