Posted inIndustry viewsAsset Class in FocusNews

The limits of bond fund tracking error

Individual bonds within a particular sector do not necessarily have similar risk-return characteristics, argues Colin Purdie, London-based chief investment officer for credit at Aviva Investors.
Colin Purdie, Aviva Investors

Calculating a bond fund’s tracking error has been a traditional risk management tool among corporate bond portfolio managers. The measure can be useful to explain potential deviations of performance versus the benchmark, said Purdie, who spoke to FSA on a recent visit to Hong Kong.

However, the tool has limitations.

“It’s not perfect and it’s flawed because tracking error ultimately looks at where you are in a sector, [and not all securities in a sector move the same way],” said Purdie, who is also the manager of the firm’s global corporate bond fund.

He used Dell and Microsoft bonds as an example. Although both companies belong to the IT sector, they are different in terms of their risk-return profiles.

Performance of key same-sector bonds 

Source: Aviva Investors

“They have different business models, very different credit ratings and outlooks, yet tracking error would suggest that if you are overweight one and underweight the other, from a sector perspective, you become neutral and as a result of that your tracking error might be low,” he said.

The same goes for Walmart and Macy’s, which belong to the same sector, he added.

Purdie argues that grouping bonds in terms of their volatility profile is more efficient for credit selection rather than just grouping them by sectors. By doing so, managers can identify what securities they would want to increase or reduce exposure to.

“Don’t group Dell and Microsoft together, group Dell and Macy’s, which have high volatility, and Microsoft and Walmart, which have low volatility. Dell and Macy’s should be competing with each other for a space in your portfolio, and so can Microsoft and Walmart.”

Source: Aviva Investors

Beware of sell-side bias

Purdie also cautioned against relying heavily on analyst reports.

“In the investment grade space alone, on both a coverage and a recommendation perspective, BBBs are always a preferred flavour.”

Of the 700 BBB-rated issuers, around 42% of them had outperform recommendations, while only 17% had underperform recommendations.

On the flipside, out of the 400 issuers  with an A rating, only 12% had outperform recommendations and 18% underperform recommendations.

Purdie explained that there is a bias for BBB-rated bonds because they have the widest spreads in the investment grade universe.

“You buy the thing with the widest spread, and in this case, it’s BBB, because you think you’ll make more money when spreads tighten.

“If portfolio managers follow analyst recommendations, their portfolios will have an inherent bias toward the riskier bonds. I am not saying that [being overweight in BBB] is necessarily wrong at all points in a cycle. The question is, are you aware of this bias and the impact it has on your portfolio?”

The firm’s Global Investment Grade Fund has 49% in BBB, which is overweight by 0.72% relative to its benchmark, the Bloomberg Barclays Global Aggregate Corporate Total Return Index, according to the fund factsheet.

A- and AA-rated bonds comprise around 36% of the portfolio. The biggest overweight is in AAA-rated bonds (9.75% of the portfolio), which is 8.43% higher than the benchmark.


The Aviva Investors Global Investment Grade Corporate Bond Fund versus its benchmark index.

Source: FE Analytics. Note: Fund and sector NAVs have been converted to US dollars

Part of the Mark Allen Group.