- The report discusses the construction of an out-of-sample effective index allocation portfolio, focusing on three key aspects: price judgment, tool expression, and risk control. Price judgment involves forming predictions on the price trends of major assets, industries, or individual stocks using macro, meso, and micro-level information through qualitative, quantitative, or mixed methods. Tool expression refers to selecting investable tools for portfolio implementation, while risk control manages potential losses in the portfolio[9] - The primary goal of the strategy is to reduce overfitting risks to enhance out-of-sample effectiveness. This is achieved through three measures: expanding the pool of targets, neutralizing factors to reduce style impact, and managing portfolio risks to mitigate the impact of tail risks on excess returns. Signal sustainability outside the sample is emphasized as a critical factor[2] - The report highlights the advantages of using equity indices as allocation tools. Indices, being a basket of stocks, can hedge individual stock-specific risks to some extent. They also serve as better tools for expressing investment views due to their distinct target attributes. Additionally, risk models at the index level are more effective, providing better risk management outcomes[11][12] - The construction of the index risk control model follows a process similar to stock risk control models but requires additional steps to synthesize index-level data. The process includes selecting indices published before the given trading day, ensuring all index components are A-shares, obtaining index component lists and weights, and calculating weighted scores for industry/style exposures based on real-time weights. The model's effectiveness is significantly higher than individual stock models, with industry contributions surpassing style contributions[22][23] - The report categorizes factors into four main types: fundamental, analyst, price-volume, and high-frequency. Each type is further divided into subcategories, such as growth, profitability, valuation, momentum reversal, volatility, liquidity, and fund flows. The factor library includes a total of 275 factors, with specific counts for each subcategory[26][27][30] - Historical performance analysis of sub-strategies shows varying correlations among them, emphasizing the necessity of multi-strategy approaches. For the period of January to August 2025, fundamental factors like profitability and growth, as well as price-volume sub-strategies, performed well. However, individual sub-strategies experienced periodic drawdowns, highlighting the importance of diversification[27][30] - Based on selected sub-strategies, the report constructs a composite index scoring signal for portfolio allocation. Anchored to the CSI All Share Index, the portfolio controls deviations in industry and major style exposures. The out-of-sample performance, including returns, drawdowns, and tracking errors, aligns closely with backtest results[32][33] - The report evaluates the use of existing products, including active and passive types, for tracking the target index portfolio. Combining active and passive products yields better out-of-sample tracking results compared to using ETFs alone. While ETFs perform well in certain months, the combined approach demonstrates superior consistency[37][38] - The report identifies the overall performance of factors in 2025, with fundamental factors like growth and profitability, as well as price-volume factors such as momentum reversal, volatility, and liquidity, showing strong results[36]