金工ETF点评：跨境ETF近3交易日净流入47.02亿元，传媒、通信拥挤幅度收窄

• The industry crowding monitoring model is constructed to monitor the crowding level of Shenwan first-level industry indices daily[3] - The Z-score model is used to build a signal screening model for related ETF products, providing potential arbitrage opportunities[4] Model Construction Process - Industry crowding monitoring model: The model monitors the crowding level of Shenwan first-level industry indices daily, identifying industries with high and low crowding levels[3] - Z-score model: The model calculates the Z-score of the premium rate for ETF products, identifying potential arbitrage opportunities and warning of potential pullback risks[4] Model Evaluation - Industry crowding monitoring model: The model effectively identifies industries with significant crowding levels, providing valuable insights for investment decisions[3] - Z-score model: The model is useful for identifying potential arbitrage opportunities in ETF products, but investors should be cautious of potential pullback risks[4] Model Testing Results - Industry crowding monitoring model: The model identified high crowding levels in the pharmaceutical, environmental protection, and steel industries, while communication and media industries had low crowding levels[3] - Z-score model: The model provided signals for potential arbitrage opportunities in various ETF products, including the Real Estate ETF, Gold ETF, Green Power ETF, and Biopharmaceutical ETF[15] Factor Construction Process - Industry crowding factor: The factor is constructed based on the daily monitoring of the crowding level of Shenwan first-level industry indices[3] - Premium rate Z-score factor: The factor is constructed by calculating the Z-score of the premium rate for ETF products[4] Factor Evaluation - Industry crowding factor: The factor is effective in identifying industries with significant crowding levels, providing valuable insights for investment decisions[3] - Premium rate Z-score factor: The factor is useful for identifying potential arbitrage opportunities in ETF products, but investors should be cautious of potential pullback risks[4] Factor Testing Results - Industry crowding factor: The factor identified high crowding levels in the pharmaceutical, environmental protection, and steel industries, while communication and media industries had low crowding levels[3] - Premium rate Z-score factor: The factor provided signals for potential arbitrage opportunities in various ETF products, including the Real Estate ETF, Gold ETF, Green Power ETF, and Biopharmaceutical ETF[15]