EBSCN·2025-06-28 08:44Quantitative Models and Construction Methods 1. Model Name: PB-ROE-50 - Model Construction Idea: The PB-ROE-50 model selects stocks based on a combination of Price-to-Book (PB) ratio and Return on Equity (ROE), aiming to capture value and profitability factors[25][26]. - Model Construction Process: The model ranks stocks within the China Securities 800 Index (CSI 800) and the broader market based on PB and ROE metrics. Stocks with the best combined scores are selected to form a portfolio of 50 stocks. The portfolio is rebalanced periodically to maintain its factor exposure[25][26]. - Model Evaluation: The model demonstrates the ability to generate excess returns over benchmarks, particularly in the CSI 800 and broader market stock pools[25][26]. 2. Model Name: Bulk Transaction Portfolio - Model Construction Idea: This model leverages the information embedded in bulk transactions, focusing on stocks with high transaction amounts and low volatility[31]. - Model Construction Process: Stocks are selected based on two key metrics: "bulk transaction amount ratio" (higher is better) and "6-day transaction amount volatility" (lower is better). The portfolio is rebalanced monthly to align with these criteria[31]. - Model Evaluation: The model effectively captures excess returns by exploiting the "high transaction, low volatility" principle[31]. 3. Model Name: Directed Issuance Portfolio - Model Construction Idea: This model identifies investment opportunities in stocks involved in directed issuance events, considering factors like market capitalization and rebalancing cycles[36]. - Model Construction Process: Stocks are selected based on their involvement in directed issuance events, with adjustments for market capitalization and portfolio constraints. The portfolio is rebalanced periodically to reflect updated event data[36]. - Model Evaluation: The model demonstrates consistent excess returns, indicating the effectiveness of event-driven strategies in the directed issuance space[36]. --- Model Backtesting Results 1. PB-ROE-50 Model - CSI 500: Weekly excess return -1.38%, YTD excess return 2.37%, weekly absolute return 2.54%, YTD absolute return 4.84%[26]. - CSI 800: Weekly excess return 0.50%, YTD excess return 5.53%, weekly absolute return 2.99%, YTD absolute return 5.92%[26]. - Broad Market: Weekly excess return 0.09%, YTD excess return 6.83%, weekly absolute return 3.43%, YTD absolute return 10.50%[26]. 2. Bulk Transaction Portfolio - Weekly Excess Return: 1.16% - YTD Excess Return: 24.68% - Weekly Absolute Return: 4.53% - YTD Absolute Return: 28.95%[32]. 3. Directed Issuance Portfolio - Weekly Excess Return: 1.05% - YTD Excess Return: 9.32% - Weekly Absolute Return: 4.42% - YTD Absolute Return: 13.07%[37]. --- Quantitative Factors and Construction Methods 1. Factor Name: Beta Factor - Factor Construction Idea: Measures a stock's sensitivity to market movements, capturing systematic risk[20]. - Factor Construction Process: Calculated as the covariance of a stock's returns with market returns, divided by the variance of market returns. $ \beta = \frac{\text{Cov}(R_i, R_m)}{\text{Var}(R_m)} $ Where is the stock return and is the market return[20]. - Factor Evaluation: Demonstrates positive returns in the current week, indicating favorable market conditions for high-beta stocks[20]. 2. Factor Name: Liquidity Factor - Factor Construction Idea: Captures the ease of trading a stock, with higher liquidity stocks expected to perform better in certain market conditions[20]. - Factor Construction Process: Measured using metrics like average daily trading volume or bid-ask spread over a specific period[20]. - Factor Evaluation: Generated positive returns this week, reflecting a preference for liquid stocks in the market[20]. 3. Factor Name: Residual Volatility Factor - Factor Construction Idea: Measures the idiosyncratic risk of a stock, with lower residual volatility stocks often preferred in risk-averse environments[20]. - Factor Construction Process: Calculated as the standard deviation of the residuals from a stock's regression on market returns[20]. - Factor Evaluation: Negative returns this week, indicating underperformance of low-volatility stocks[20]. --- Factor Backtesting Results 1. Beta Factor - Weekly return: 1.06%[20]. 2. Liquidity Factor - Weekly return: 0.37%[20]. 3. Residual Volatility Factor - Weekly return: -0.31%[20].