- The report constructs an industry crowding monitoring model to monitor the crowding degree of Shenwan First-Level Industry Indexes on a daily basis[4] - The Z-score model is used to build a related ETF product screening signal model, providing potential arbitrage opportunities through rolling calculations[5] Model Construction and Process Industry Crowding Monitoring Model 1. **Model Name**: Industry Crowding Monitoring Model 2. **Model Construction Idea**: Monitor the crowding degree of various industries to identify potential investment opportunities and risks[4] 3. **Model Construction Process**: - Daily monitoring of the crowding degree of Shenwan First-Level Industry Indexes - Identify industries with high and low crowding levels - Track the main fund flows in and out of these industries over recent trading days[4] 4. **Model Evaluation**: Provides insights into industry crowding levels, helping investors to make informed decisions[4] Z-score Model for ETF Product Screening 1. **Model Name**: Z-score Model 2. **Model Construction Idea**: Identify potential arbitrage opportunities in ETF products by calculating the Z-score of their premium rates[5] 3. **Model Construction Process**: - Calculate the Z-score of the premium rates of various ETF products - Identify ETFs with significant deviations from their historical average premium rates - Provide signals for potential arbitrage opportunities and caution for potential pullback risks[5] 4. **Model Evaluation**: Helps in identifying ETFs with potential arbitrage opportunities, but also warns of possible pullback risks[5] Model Backtest Results 1. **Industry Crowding Monitoring Model**: - Daily monitoring results show that industries like basic chemicals, textiles and apparel, and light manufacturing have high crowding levels, while home appliances, real estate, and electronics have low crowding levels[4] - Significant changes in crowding levels were observed in industries like construction and decoration, and non-ferrous metals[4] 2. **Z-score Model for ETF Product Screening**: - Identified ETFs with potential arbitrage opportunities, such as the National Defense ETF and the Hang Seng Medical Index ETF[14] Factor Construction and Process Industry Crowding Factor 1. **Factor Name**: Industry Crowding Factor 2. **Factor Construction Idea**: Measure the crowding degree of industries to identify potential investment opportunities and risks[4] 3. **Factor Construction Process**: - Calculate the crowding degree of each industry based on fund flows and other relevant metrics - Identify industries with high and low crowding levels[4] 4. **Factor Evaluation**: Provides valuable insights into industry crowding levels, aiding in investment decision-making[4] Factor Backtest Results 1. **Industry Crowding Factor**: - High crowding levels were observed in industries like basic chemicals, textiles and apparel, and light manufacturing[4] - Low crowding levels were observed in industries like home appliances, real estate, and electronics[4] - Significant changes in crowding levels were noted in industries like construction and decoration, and non-ferrous metals[4]

Quantitative Models and Construction 1. Model Name: Industry Crowdedness Monitoring Model - **Model Construction Idea**: This model is designed to monitor the crowdedness levels of Shenwan First-Level Industry Indices on a daily basis, identifying industries with high or low crowdedness levels to provide insights for potential investment opportunities or risks [4] - **Model Construction Process**: 1. The model calculates the crowdedness levels of each industry index based on specific metrics (not explicitly detailed in the report) 2. Daily updates are performed to track changes in crowdedness levels across industries 3. Industries with significant changes in crowdedness levels are highlighted for further analysis [4] - **Model Evaluation**: The model effectively identifies industries with extreme crowdedness levels, providing actionable insights for investors [4] 2. Model Name: Premium Rate Z-Score Model - **Model Construction Idea**: This model identifies potential arbitrage opportunities in ETF products by calculating the Z-score of their premium rates over a rolling window [5] - **Model Construction Process**: 1. The premium rate of each ETF is calculated as the difference between its market price and net asset value (NAV) 2. A rolling window is applied to compute the Z-score of the premium rate for each ETF 3. ETFs with Z-scores exceeding a certain threshold are flagged as potential arbitrage opportunities [5] - **Model Evaluation**: The model provides a systematic approach to detect arbitrage opportunities while also warning of potential price corrections [5] --- Backtesting Results of Models 1. Industry Crowdedness Monitoring Model - **Top Crowded Industries**: Basic Chemicals, Textile & Apparel, Light Manufacturing [4] - **Least Crowded Industries**: Real Estate, Electronics, Social Services, Steel, Non-Banking Financials [4] - **Significant Daily Changes**: Petroleum & Petrochemicals experienced notable daily crowdedness changes [4] 2. Premium Rate Z-Score Model - **Highlighted ETFs for Arbitrage**: Specific ETFs flagged for potential arbitrage opportunities are not detailed in the report [5]