Bridge over ocean
1 November 2014 CFA Institute Journal Review

A Simplified Fixed Income Performance Attribution Model (Digest Summary)

  1. Sandra Krueger, CFA

The challenge of fixed-income performance attribution is to create a model that evaluates the investment process objectively without being excessively complex. The authors present a performance attribution model that uses returns rather than yields to simplify implementation of the model and provide meaningful output.

What’s Inside?

The authors create an attribution model that follows a top-down portfolio management approach but is built from the bottom up. As a result, the attribution output can be viewed at various levels of granularity. The attribution system groups the output into five sections—duration management, spread management, issue selection, yield management, and currency management.

How Is This Research Useful to Practitioners?

The need for fixed-income attribution has increased partly as a result of the global financial crisis in 2007, when collateralized debt obligations (CDOs) dropped dramatically in value. Attribution provides a proof statement of what happened in the portfolio.

Although sector-based models can be customized to any portfolio segmentation to show outperformance, they are not consistent with how fixed-income managers actually make investment decisions. Fixed-income managers analyze sources of return and risk that will have a positive impact on the portfolio. Key drivers of bond fund returns include interest rate levels, credit spread movements, yield curves, and so on. The authors incorporate these important drivers into the attribution modeling process.

They create an attribution model that provides results that can be looked at in aggregate or traced back to an individual security. The output can be viewed through their interactive web front end, which gives the viewer the ability to investigate the attribution and allows for easy interpretation.

How Did the Authors Conduct This Research?

The authors want their model to have the ability to be implemented easily, so it is based on returns rather than yields. Assets are assigned a market based on their trade currency and assigned a sector based on the sector definitions. The authors analyze the portfolio return by breaking each security return into three parts. One part of the security return is related to changes in government yields, another portion is related to the changes in the yield spread over governments, and a third part is related to a currency element.

Duration management is captured by three terms:

  • Active duration captures whether the overall portfolio was correctly positioned for government yield changes.
  • Market allocation captures how the portfolio was positioned in each market for portfolios that invest across multiple currency blocks.
  • Duration curve captures the point on the curve at which active duration bets were placed.

Sector or spread management is defined as the measurement of the portfolio’s sector deviation from the benchmark. By sector, the authors mean a user-defined group (e.g., country, sector rating, or financial subordination). Spread management follows the same type of procedure and interpretation as duration management does.

The attribution results are made available through web reporting tools. The web output has a drop-down menu that allows for alternative groupings of terms. Because attribution is calculated at the security level, the results can be regrouped into any set of sector or duration buckets, which allows the user to drill down into more detail.

Abstractor’s Viewpoint

From an investor’s perspective, a transparent approach to pinpointing sources of risk and return is crucial. Often, fixed-income attribution is readily available and difficult to use. An attribution model must be sophisticated enough to accommodate the range of market instruments as well as the changing nature of the market. Additionally, it must be able to convey the results in an easily understood manner to facilitate informed investment decisions. The authors have created a web-based tool that is usable by participants at various levels ranging from analysts to clients. It appears to be a model that warrants further examination.

We’re using cookies, but you can turn them off in Privacy Settings.  Otherwise, you are agreeing to our use of cookies.  Accepting cookies does not mean that we are collecting personal data. Learn more in our Privacy Policy.