Pages

Tuesday, March 10, 2009

Fixed income attribution falls short

Attribution analysis can help investment managers keep their clients, even in down-markets, said David Spaulding, president of The Spaulding Group, Inc. in his presentation on "Fixed Income Attribution: An Introduction" to the Boston Security Analysts Society (BSAS) on March 5. But good attribution analysis has been hard for fixed income managers to find. While equity managers have long enjoyed good models and software, the fixed income world is only catching up now, according to Spaulding. The Campisi model for fixed income attribution offers a solution. 

Explanation of underperformance can save the day
Some managers underperform their benchmarks, but keep their clients because of attribution. How's that? Attribution helps them to explain what's working--and what's not. With that information, managers can reassure clients with their strategies for fixing things. This is a technique I talked about in "How can you report underperformance in your client letters?" 

Equity-based models don't cut it
But many fixed income managers create their performance attribution with the equivalent of one hand tied behind their back, based on what I learned from Spaulding. They're using attribution models developed for equities, which look only at security selection and sector allocation. That's a poor match for fixed income, where decisions about duration, sectors, and risk levels (ratings) are most important and security selection typically doesn't count for much.

"If you're not looking at duration, you don't have fixed income attribution," said Spaulding. That's because the duration decision typically has the greatest impact on fixed income performance. 

Campisi model fixes problems 
The Campisi model, developed by Stephen Campisi, CFA, may help. It is an attribution model with the potential to  play the role for fixed income that two Brinson models play for equities, said Spaulding. The model views bond returns as coming from income in addition to price change. Spaulding ran through the steps in applying the model, including gathering the data, calculating the contribution effect for the benchmark and the portfolio, and calculating the attribution effect.

The BSAS audience seemed receptive to the Campisi model. But some expressed concern about handling derivatives in a fixed income portfolio. Spaulding said that assets that aren't in a portfolio's benchmark should be isolated and only their contribution should be discussed. However, I got the sense that managers who invest heavily in derivatives aren't satisfied with that solution.

It looks as if challenges still remain until fixed income attribution achieves the usefulness of its equity counterpart.

If you'd like a copy of Spaulding's PowerPoint presentation, e-mail your request to The Spaulding Group.


 _________________
Susan B. Weiner, CFA
Check out my website at www.InvestmentWriting.com or sign up for my free monthly e-newsletter.
Copyright 2009 by Susan B. Weiner All rights reserved

2 comments:

  1. Two comments -- don't know if I'd characterize the talk as fixed income attribution "...falls short". Dave's point was that when done *well* -- that is, by NOT using an equity model and rather using one that rigorously takes into account all aspects of fixed income management (duration, spread, yield curve, etc.) that fixed income attribution provides a treasure-trove of information.

    Second point is to emphasize a KEY benefit, especially in THESE times: using attribution as a communication tool. We work with many clients on exactly this (see various articles at http://www.the-collaborative.com/html/pubs/newsletter.cfm and http://www.advisorstrustedadvisor.com/); the power of attribution to enhance communication -- even if the news is bad -- cannot be understated. It can make the difference between retaining and losing a client.

    ReplyDelete
  2. Mike,

    I agree that fixed income attribution can pay off when it's done well.

    ReplyDelete

Note: Only a member of this blog may post a comment.