Attributing Bond Fund Returns to Rates, Credit Spreads, and Carry

Summary

This research note presents a framework for explaining bond fund returns and identifying portfolio style. It decomposes returns into income, government yield-curve movements, credit spread effects, convexity, and bond roll-down. Income is further divided into coupon income and price convergence; curve and spread contributions are broken down into allocation and security-selection components using a Brinson-style approach. The analysis uses disclosed bond holdings and reports a study of 1,338 pure bond funds.

The note says coupon income was a persistent major return source, while rate-curve and spread contributions varied more with interest-rate movements and changes in credit spreads. It also reports that stronger past coupon, rate-risk, credit-risk, and roll-down management were associated with better next-quarter performance; convexity management showed no relationship. A screening score averaging coupon, curve, roll-down, and spread contributions reportedly outperformed both the weakest-ranked group and a long-term bond fund index. These are reported findings, not a full account of methodology: the supplied text gives limited detail on definitions, controls, transaction costs, or out-of-sample validation, so the predictive claims require caution.

Key ideas

• Bond fund returns can be attributed to income, government yield-curve changes, credit spreads, convexity, and roll-down.
• Reported analysis uses portfolio holdings disclosed in fund reports to infer return sources and style.
• Coupon income was described as a major and relatively consistent contributor across years.
• Several past management measures were associated with stronger following-quarter performance, while convexity management was not.
• An average score based on four attribution components reportedly separated stronger-performing funds from weaker ones and beat a long-term bond fund index.

Tags

This summary was written by Stratmill's research agent from the original; it is not a copy of the source.