净值回归法
Search documents
走在债市曲线之前系列报告(七):基金久期测算方法全解
Changjiang Securities· 2025-11-21 01:34
1. Report Industry Investment Rating No relevant content provided. 2. Core Viewpoints of the Report - The report systematically sorts out mainstream fund duration measurement methods, focusing on the supplementary bond method and the net - value regression method. The supplementary bond method can optimize the heavy - position weighted method, but it is difficult to capture the dynamic changes of fund positions and duration. The net - value regression method has advantages in fitting stability and dynamic tracking ability. The measurement results show that since 2025, the duration of bond funds has fluctuated, with the central position remaining at a relatively high level. The duration management ability is an important factor affecting the differentiation of fund performance [5][82]. 3. Summary According to the Directory 3.1 Fund Duration: Significance and Measurement Methods - **Significance of Fund Duration**: Fund duration is an important indicator to measure interest - rate risk and investment style. A longer duration means greater price sensitivity to interest - rate fluctuations. Adjusting the duration reflects investors' risk preferences and operational ideas, and the duration level represents managers' judgments on interest - rate trends [22]. - **Mainstream Measurement Methods and Their Advantages and Disadvantages**: Common measurement methods include the interest - rate sensitivity method, the recursive method, the heavy - position weighted method, the supplementary bond method, and the net - value regression method. Low - frequency methods are simple to calculate and relatively accurate, while high - frequency methods are continuous and suitable for dynamic tracking. In practice, a trade - off is needed between frequency and accuracy, and multiple methods should be used for cross - verification [23][25]. 3.2 Supplementary Bond Method: Constructing a Simulated Portfolio to Optimize the Heavy - Position Weighted Method - **Basic Principle**: The supplementary bond method optimizes the heavy - position weighted method by filling in specific bonds based on the scale information of various bond holdings in the fund's investment portfolio, constructing a new simulated portfolio with the same duration as the fund's real duration [35]. - **Effectiveness and Limitations**: The supplementary bond method can effectively improve the accuracy of the simulated portfolio's duration at the time of the semi - annual disclosure of the fund's real duration. However, during the period without disclosure of heavy - position holdings and real duration, it cannot reflect the dynamic adjustment of the duration center caused by market changes, and its applicability in high - frequency real - time tracking scenarios is limited [39]. 3.3 Net - Value Regression Method: High - Frequency Dynamic Duration Tracking - **Basic Model**: The net - value regression method takes the daily fund return as the explained variable and the index return as the explanatory variable. The regression coefficient reflects the fund's sensitivity and allocation preference for various bond indexes, and the fund's duration is the weighted result of its holdings on different - term bond indexes [44]. - **Problems and Solutions**: The explanatory variables of the model have a multi - collinearity problem. The PLS method is finally adopted to compress multi - dimensional and highly correlated independent variables, avoid problems caused by multi - collinearity, and improve the model's prediction performance [49][53]. - **Duration Results Measured by the Net - Value Regression Method**: The overall fitting effect of the measured duration is good, with about 41% of the samples having a measurement error within 0.5 years. Since 2025, the duration of the entire market's bond funds has been oscillating, and the central position of the duration reached its highest point in July and August, then declined rapidly, and rebounded at the end of September [58]. - **Using the Net - Value Regression Method to Guide Investment**: Two indicators, the correlation coefficient between duration and Treasury bond yield and the timing accuracy rate, are constructed to test the fund's timing ability. The results show that high - performing funds have a higher timing accuracy rate and a stronger negative correlation between duration and interest - rate trends [72][77].