Guosen Securities(SZ:002736) Guoxin Securities·2025-09-27 08:40Quantitative Models and Construction Methods - Model Name: Maximized Factor Exposure Portfolio (MFE) Model Construction Idea: The MFE portfolio is designed to maximize the exposure of a single factor while controlling for various constraints such as industry exposure, style exposure, stock weight deviation, and turnover rate. This approach ensures that the factor's predictive power is tested under realistic constraints, making it more applicable in actual portfolio construction [42][43][44] Model Construction Process: The MFE portfolio is constructed using the following optimization model: $ \begin{array}{ll} max & f^{T} w \ s.t. & s_{l} \leq X(w-w_{b}) \leq s_{h} \ & h_{l} \leq H(w-w_{b}) \leq h_{h} \ & w_{l} \leq w-w_{b} \leq w_{h} \ & b_{l} \leq B_{b}w \leq b_{h} \ & \mathbf{0} \leq w \leq l \ & \mathbf{1}^{T} w = 1 \end{array} $ - Objective Function: Maximize single-factor exposure, where represents factor values, is the weighted exposure of the portfolio to the factor, and is the stock weight vector - Constraints: 1. Style Exposure: is the factor exposure matrix, is the benchmark weight vector, and are the lower and upper bounds for style factor exposure 2. Industry Exposure: is the industry exposure matrix, and are the lower and upper bounds for industry deviation 3. Stock Weight Deviation: are the lower and upper bounds for stock weight deviation 4. Constituent Stock Weight: is a 0-1 vector indicating whether a stock is a benchmark constituent, and are the lower and upper bounds for constituent stock weights 5. No Short Selling: Ensures non-negative weights and limits individual stock weights to 6. Full Investment: Ensures the portfolio is fully invested with - Implementation: 1. Set constraints for style, industry, and stock weights 2. Construct MFE portfolios for each factor at the end of each month 3. Backtest the MFE portfolios, calculate historical returns, and adjust for transaction costs (0.3% on both sides) [42][43][46] Model Evaluation: The MFE portfolio approach is effective in testing factor validity under realistic constraints, ensuring that factors deemed "effective" can contribute to actual portfolio performance [42][43] Quantitative Factors and Construction Methods - Factor Name: Single-Quarter ROE Factor Construction Idea: Measures the return on equity for a single quarter to capture profitability trends [17] Factor Construction Process: $ \text{Single-Quarter ROE} = \frac{\text{Net Income (Quarterly)} \times 2}{\text{Average Shareholders' Equity}} $ - Net Income (Quarterly): Quarterly net income attributable to shareholders - Average Shareholders' Equity: Average of beginning and ending equity for the quarter [17] - Factor Name: Single-Quarter Revenue Growth (YoY) Factor Construction Idea: Tracks the year-over-year growth in quarterly revenue to identify growth trends [17] Factor Construction Process: $ \text{Single-Quarter Revenue Growth (YoY)} = \frac{\text{Revenue (Current Quarter)} - \text{Revenue (Same Quarter Last Year)}}{\text{Revenue (Same Quarter Last Year)}} $ [17] - Factor Name: Analyst Coverage (3-Month) Factor Construction Idea: Measures the number of analysts covering a stock over the past three months to gauge market attention [17] Factor Construction Process: $ \text{3-Month Analyst Coverage} = \text{Number of Analysts Covering the Stock in the Last 3 Months} $ [17] Factor Backtesting Results - Single-Quarter ROE: - CSI 300: Weekly excess return: 0.42%, monthly: 2.94%, YTD: 15.41%, historical annualized: 4.92% [19] - CSI 500: Weekly excess return: 0.47%, monthly: 0.89%, YTD: 4.43%, historical annualized: 5.85% [21] - CSI 1000: Weekly excess return: 1.20%, monthly: 1.70%, YTD: -0.61%, historical annualized: 7.62% [23] - CSI A500: Weekly excess return: 0.30%, monthly: 1.68%, YTD: 13.78%, historical annualized: 3.35% [25] - Single-Quarter Revenue Growth (YoY): - CSI 300: Weekly excess return: 0.48%, monthly: 2.34%, YTD: 17.35%, historical annualized: 4.94% [19] - CSI 500: Weekly excess return: 1.28%, monthly: 2.58%, YTD: 15.18%, historical annualized: 3.81% [21] - CSI 1000: Weekly excess return: 0.69%, monthly: 2.73%, YTD: 15.73%, historical annualized: 5.17% [23] - CSI A500: Weekly excess return: 0.47%, monthly: 1.15%, YTD: 15.65%, historical annualized: 2.96% [25] - 3-Month Analyst Coverage: - CSI 300: Weekly excess return: 0.17%, monthly: 0.90%, YTD: 10.33%, historical annualized: 3.07% [19] - CSI 500: Weekly excess return: 0.29%, monthly: 0.07%, YTD: 4.10%, historical annualized: 5.56% [21] - CSI 1000: Weekly excess return: 1.30%, monthly: 0.52%, YTD: 5.98%, historical annualized: 7.22% [23] - CSI A500: Weekly excess return: -0.21%, monthly: 0.97%, YTD: 8.12%, historical annualized: 3.93% [25]