Are your active bond managers earning their fees?

March 2, 2021

A/AText size:AAA
 

With the proliferation of low-cost passive and factor-based products, how can institutional investors evaluate whether a particular active bond manager is worth the extra cost?

Vanguard's latest research paper answers this question with a systematic five-step process.

"We tried to bridge the gap between theory and practical application," said Inna Zorina, senior investment strategist in Vanguard Investment Strategy Group and coauthor of Can Active Fixed Income Managers Be Cloned Using Style Funds? A Practical Testing Method. "Our model used combinations of existing bond ETFs to simulate the performance of 625 actively managed U.S. fixed income funds that existed during the five years through 2019."

The results indicated that while there are some active fixed income managers providing true alpha, there is a sizable portion who could be replicated with lower-fee passive style (i.e., factor) products. We asked the authors to apply the same methodology to one of Vanguard's newer and popular active funds to see where it falls within this framework.

But first, here are the findings of the overall study:

More than one-fifth of active managers could have been cloned

The chart below shows where the active managers landed in terms of net alpha relative to a proxy portfolio of ETFs that simulated their performance (horizontal axis) and tracking error relative to that proxy (vertical axis).

Those managers in the upper two quadrants had a tracking error greater than 1.5%, meaning that they were not easy to replicate with a proxy portfolio, while those in the upper- and lower-right quadrants added alpha that exceeded the average weighted expense ratio for the proxy portfolios. The left two quadrants are where active managers do not want to land, as these managers underperformed the proxies.

Alpha and replicability of active bond funds

Alpha and replicability of active bond funds chart

Note: The chart illustrates individual U.S.-domiciled funds investing in U.S. fixed income. The tracking error and alpha values are calculated versus the given manager's factor-mimicking portfolio. The factors used in the analysis included the total U.S. market, high-yield credit, and ultrashort Treasuries acting as a cash term. Average cost of cloning is estimated to be 20 basis points (0.20%), which represents the weighted expense ratio of the factor proxy products of an average factor-mimicking portfolio. Alpha could result from model misspecification (such as a suboptimal choice of factor proxies or omitted factors). Past performance is no guarantee of future returns.
Source: Can Active Fixed Income Managers Be Cloned Using Style Funds? A Practical Testing Method. Vanguard calculations using monthly gross return data from Morningstar for the five years through December 31, 2019.

"Roughly 22% of the funds fell in the lower-left quadrant," said Zorina. "We deemed those to be 'likely cloneable.' Their performance can be mostly replicated with passive products, and they didn't deliver enough alpha to offset the cost. On the other hand, some funds fell in the lower-right quadrant, offering similar factor profile and greater alpha relative to their respective factor-mimicking portfolios."

Investors should also consider other factors that will contribute to the fund's ability to achieve better long-term results, including lower fees, long-term perspective, a keen focus on risk optimization and downside protection, and exposure to diversified sources of alpha.

Vanguard Core Bond Fund: True-to-label alpha

The paper's authors applied the same methodology to gauge the "cloneability" of Vanguard Core Bond Fund. (The fund was launched March 28, 2016, so it was not part of the original study, which counted only funds with five-year track records through 2019.)

A combination of two bond ETFs—one a total market product and the other tracking an intermediate-term corporate credit benchmark—came close to simulating the fund's return profile since inception though year-end 2020.

The chart below shows that the weighting between the two proxy products had to change over time to mimic the fund's performance. The weighting averaged around 91% for the market and 9% for intermediate-term credit.

Proportional changes in factors to mimic Vanguard Core Bond Fund's returns

Proportional changes in factors to mimic Vanguard Core Bond Fund's returns chart

Note: The graph illustrates a 36-month rolling window of the factor exposure of Vanguard Core Bond Investor Shares. The varying proportion represents the optimal combination of market and corporate intermediate benchmarks to replicate Vanguard Core Bond Fund's returns as closely as possible. The factor portfolios assumed monthly rebalancing and no transaction costs for rebalancing.
Source: Can Active Fixed Income Managers Be Cloned Using Style Funds? A Practical Testing Method. Vanguard calculations using data from Morningstar Direct from April 1, 2016, to December 31, 2020.

The regression analysis revealed that the tracking error was 0.50% and the adjusted R-squared was 0.98%, meaning that the factor-mimicking portfolio tracked the Core Bond Fund's returns very closely. Portfolio weights of both market and intermediate-term credit were statistically significant at 1%.

However, the Core Bond Fund delivered a gross alpha of 63 basis points relative to the proxy portfolio—63 bps generated from security selection and other alpha sources that could not be delivered by the proxies. This amount of alpha is significant for fixed income, particularly for a fund whose mandate is to keep a risk profile similar to the broad investment-grade bond market, putting it in the lower-right quadrant of the chart shown earlier.

Taking into account the 10 bps expense ratio of the fund's Admiral Shares, the fund delivered 53 bps of net alpha (or 57 bps if you factor in the 4 bps expense ratio of the proxy portfolio).

"During the research and design process before the fund's launch, the Vanguard Fixed Income Group said it would expect the fund to have a slight credit tilt on average over time," Zorina said.

"As per our analysis, the fund delivered returns equivalent to a mix of roughly 90% market and 10% intermediate-term credit while still providing alpha. This is exactly what we would expect. It's showcasing that the fund is acting true to label and has generated a net excess return relative to what could have been earned in a low-cost index fund portfolio with a similar style profile."

Find out how our active fixed income funds accomplish this on our investment approach overview page.

To learn more about Vanguard's research on factor replacement and the methodology used, see our paper.

Read PDF  

Notes:

  • All investing is subject to risk, including the possible loss of the money you invest.
  • Past performance is no guarantee of future returns.
  • Investments in bonds are subject to interest rate, credit, and inflation risk.