债券投资组合业绩归因的多期实证研究

目前国内对投资组合的业绩归因研究主要从管理者层面着手，将超额收益的来源归结为择时能力和选股能力，但这并不适用于债券投资。本文基于Campisi模型，对债券定价公式进行分解，从债券自身的特性来研究组合的超额收益来源，并结合GRAP跨期处理方法，形成多期业绩归因模型，对长期债券投资组合进行归因分析。相对于单期的归因模型，多期归因模型可以对任意一段时间内投资组合的超额收益进行归因，而不是单期归因项的简单加总。本文以中证全债指数为基准组合，对32只债券构成的投资组合进行实证研究，结果表明模型符合市场情况和实际操作情况。因此本文提出的多期业绩归因研究具有实用性。

The Multi-period Empirical Research on Performance Attribution of Bond Portfolio

Currently, the domestic researches on performance attribution mainly start from management level, and the source of excess return would be attributed to the ability of timing and selection. However, this method isn't suitable for bond portfolio. Based on the Campisi model, this paper analyzes the bond pricing formula, studies the excess return of the portfolio from the characteristics of the bond itself, and combines with the GRAP inter-temporal approach to form multi-period attribution model. Then we use this new model to analyze the performance attribution of long-term bond portfolio. Compared to a single-period attribution model, the inter-temporal attribution model can be used for any time during one period. The excess returns of the portfolio during the time are not the simple sum of the single-period attributions. This paper uses the CSI Aggregate Bond Index as a benchmark to conduct an empirical study of the portfolio of 32 bonds. The results show that the model is in line with market conditions and actual operating conditions. Therefore, the multi-period performance attribution research proposed in this paper is practical.