走在债市曲线之前系列报告（十）：债券指数体系与应用解析（下）

1. Report Industry Investment Rating No relevant content provided. 2. Core Viewpoints of the Report The report focuses on the practical applications of bond indices in investment, systematically exploring their three core functions: serving as a tool for measuring asset performance, acting as a benchmark and tracking target for fund performance, and facilitating the construction of investment portfolios. It also introduces the Campisi and Brinson performance attribution models to analyze the logic of portfolio return decomposition and practical application steps [3]. 3. Summary by Related Catalogs Index Application: Analytical Tools in Investment Decision - Making Index Application One: Measuring Asset Performance - The interval price change of an index provides a reference for comparing the returns of different indices. By comparing the performance of wealth and net - price indices, bond returns can be split into coupon income and capital gains. Coupon income is relatively stable, with a slow downward trend in recent years, while capital gains are more volatile and significantly impact overall bond asset returns [17]. - Different - duration indices show distinct performance in various market conditions. Long - duration indices perform well in bull markets, and short - duration indices are more defensive in bear markets. Across different asset classes, there is a rotation relationship between bond indices and stocks, commodities, etc., which provides important references for asset allocation [7][24]. - Bond indices can be used as benchmarks for measuring the performance of fund portfolios. Different types of funds have different performance characteristics. Bond funds generally show better risk - dispersion and drawdown control capabilities compared to direct bond investments [27][34]. - Customized indices can be constructed to observe their risk and return performance [37]. Index Application Two: Performance Benchmark and Tracking Target - For actively managed funds, bond indices are performance benchmarks, and fund managers use strategies such as credit downgrading and duration management to outperform the benchmark. For passive index funds, bond indices are tracking targets, with requirements for daily average tracking deviation not exceeding 0.2% and annual tracking error not exceeding 2%. In practice, most index bond funds do not reach these upper limits [8][42]. - Fund performance comparison benchmarks can be divided into four types: index - based, interest - rate - based, numerical, and hybrid. The choice of performance benchmark affects the objective evaluation of fund managers' investment capabilities. Using wealth or total return indices as benchmarks can more comprehensively reflect the actual returns of investors [49][50]. - Index bond funds usually use full replication or sampling replication methods to construct portfolios. Sampling replication is more commonly used, and there are different specific methods. Tracking errors are mainly caused by subjective and objective factors [51][54]. - Index - enhanced bond funds are relatively scarce. They seek excess returns by adjusting the proportion of non - index component bonds and investing in other assets such as stocks [55]. Index Application Three: Flexible Construction of Investment Portfolios - Based on interest - rate forecasts, a duration - timing strategy can optimize portfolio performance. When yields are falling, the portfolio shifts to the ChinaBond 7 - 10 - year Treasury Bond Index to lengthen the duration; when yields are rising, it switches to the ChinaBond 1 - 3 - year Treasury Bond Index to shorten the duration [9]. - Incorporating bond indices into the stock - bond equilibrium framework can capture rebalancing signals through relative price changes between assets. A multi - asset timing strategy can continuously generate excess returns by switching between bond holdings and cash positions [9]. Performance Analysis Campisi Attribution Model - The Campisi model decomposes bond returns into holding income, price income, and bond residual income. Holding income includes coupon income and discount/premium convergence income; price income can be further split into Treasury curve riding income, Treasury curve movement income, and spread change income [70][71]. - Through the analysis of a simulated portfolio A, it is found that the portfolio failed to outperform the benchmark mainly due to insufficient exposure to interest - rate risks, resulting in lower price income [78][82]. Brinson Attribution Model - The Brinson model can be divided into the BHB model and the BF model, which differ in the calculation of industry allocation income. The model decomposes excess returns into industry allocation, individual security selection, and cross - income [83][84]. - Through the analysis of a simulated portfolio B, it is found that the total excess return of the portfolio in the second quarter of 2025 was 0.2995%. The main source was individual security selection income, followed by cross - income, which offset the small negative impact of industry allocation [90][93].