固收+及纯债基金月度跟踪（2025年8月）：固收+整体调降转债配置，优选组合持续贡献超额收益-20250808

Quantitative Models and Construction Methods 1. Model Name: Risk Factor Exposure Regression for Fixed Income+ Funds - Model Construction Idea: The model regresses the net asset value (NAV) of Fixed Income+ funds against pure style factor returns of various asset classes to determine the risk exposure of these funds to different asset categories [4][19] - Model Construction Process: 1. The NAV of Fixed Income+ funds is regressed on the pure style factor returns of various asset classes, including bonds, stocks, and convertible bonds [4][19] 2. For bond assets, the model identifies changes in duration and credit strategy exposure by comparing the average risk factor exposure between the current and previous months [4][19] 3. For convertible bonds, the model calculates the exposure to parity risk factors and tracks changes in overall positioning [4][23] - Model Evaluation: The model effectively captures the dynamic risk exposure of Fixed Income+ funds across different asset classes, providing insights into their allocation strategies [4][19] 2. Model Name: Fixed Income+ Fund Optimal Portfolio Construction - Model Construction Idea: The model selects 10 funds quarterly based on multiple dimensions such as win rate and odds, and allocates them equally to construct an optimal portfolio [5][26] - Model Construction Process: 1. Quarterly selection of 10 funds based on criteria such as win rate and odds [5][26] 2. Equal-weight allocation of the selected funds to form the optimal portfolio [5][26] - Model Evaluation: The constructed portfolio demonstrates more stable performance compared to the secondary bond fund index, outperforming it by 0.36% in the current month [5][27] 3. Model Name: Pure Bond Fund Optimal Portfolio Construction - Model Construction Idea: The model selects funds with alpha characteristics within one standard deviation of market averages across various style exposures and allocates them equally to construct an optimal portfolio [7][45] - Model Construction Process: 1. Identify funds with style exposures within one standard deviation of market averages [7][45] 2. Select 10 funds with high alpha characteristics on a quarterly basis [7][45] 3. Allocate the selected funds equally to form the optimal portfolio [7][45] - Model Evaluation: The portfolio outperformed the mid-to-long-term pure bond fund index by 0.16% year-to-date, demonstrating its effectiveness in generating excess returns [7][46] --- Quantitative Factors and Construction Methods 1. Factor Name: Bond Risk Factors - Factor Construction Idea: The factors include duration, slope, convexity, credit, and default, which are used to measure the risk exposure of pure bond funds [6][42] - Factor Construction Process: 1. Regress the NAV of pure bond funds on the five factors: duration, slope, convexity, credit, and default [6][42] 2. Analyze the mean changes in factor exposures between the current and previous months [6][42] 3. Assess the dispersion of factor exposures to evaluate consistency in fund strategies [6][42] - Factor Evaluation: The analysis reveals a significant increase in credit exposure and a decrease in convexity exposure, with low dispersion in credit exposure, indicating consistent adoption of credit strategies among funds [6][42] --- Backtesting Results of Models 1. Risk Factor Exposure Regression for Fixed Income+ Funds - No specific numerical backtesting results provided 2. Fixed Income+ Fund Optimal Portfolio Construction - Outperformed the secondary bond fund index by 0.36% in the current month [5][27] 3. Pure Bond Fund Optimal Portfolio Construction - Outperformed the mid-to-long-term pure bond fund index by 0.16% year-to-date [7][46] --- Backtesting Results of Factors 1. Bond Risk Factors - Credit exposure increased significantly, while convexity exposure decreased [6][42] - Low standard deviation in credit exposure indicates consistent strategy adoption [6][42]