By Editorial Team, AdvisorAnalyst
Why the conventional wisdom on active management doesn't apply to bonds — and what advisors need to know
The conventional wisdom has been clear for decades: active managers can't beat the index, so buy the market and go home. But a landmark April 2026 academic paper by Jaewon Choi (Seoul National University), K.J. Martijn Cremers (Notre Dame), and Timothy B. Riley (University of Arkansas) draws a precise and important boundary around that wisdom. It applies to equities. For bonds, it does not — and may even invert.
The paper's thesis is stated without reservation: "passive bond management is, in practice, something of an oxymoron." The reason isn't semantic. It's structural.
The Replication Problem
Bond indexes are not like equity indexes. The S&P 500 has 500 liquid, publicly traded securities. The Bloomberg US Aggregate Bond Index — the most common benchmark in the study's sample — holds thousands of bonds, many of which don't trade once in a given day. The median benchmark in the passive bond fund sample contains approximately 7,900 bonds, with an average fraction of non-trading days of 49%. Full replication is not just costly. It is operationally impossible.
Passive bond funds respond to this constraint by holding a fraction of their benchmarks. The average passive bond fund holds 2,344 bonds against a benchmark of 6,601. To measure the resulting divergence, the authors apply Active Share — the portfolio overlap metric introduced by Cremers and Petajisto (2009). The findings are striking: the median bond-level active share for passive bond funds is 59%. For passive equity funds, the comparable figure is 2%.
Put directly: a fund marketed as passive is, in practice, making active security selection decisions across more than half its portfolio.
The Performance Drag Is Real — And Pre-Fee
The implications for performance are severe. Passive bond funds don't just underperform after fees. They underperform before them. The study's portfolio analysis finds an average annualized gross alpha of −0.07% for passive bond funds relative to their stated benchmarks — meaning the underperformance cannot be explained by expense ratios alone.
The culprit is rebalancing cost. Bond indexes rebalance constantly: new issuances enter, bonds mature or get downgraded, and inclusion rules force turnover. The average turnover rate for passive bond funds in the sample is approximately 70%, compared to 42% for passive equity funds. And trading bonds costs orders of magnitude more than trading equities. The authors estimate average annualized rebalancing transaction costs of approximately 25 basis points for passive bond funds, compared to as little as 0.13 to 1.09 basis points for passive equity ETFs. During the COVID disruption of 2020, those bond rebalancing costs spiked above 60 basis points per year.
The Skewness Argument: Why "Buying the Haystack" Works in Equities But Not Bonds
Jack Bogle's famous logic — "don't look for the needle in the haystack; just buy the haystack" — rests on a specific empirical premise: that equity returns are so positively skewed that missing a handful of star stocks destroys long-run performance. Bessembinder (2018) demonstrated this conclusively for equities.
Choi, Cremers, and Riley show the premise doesn't transfer to bonds. Annual equity returns carry a cross-sectional skewness of 1.76. Corporate bonds: 0.47. The practical consequence is direct. A randomly selected portfolio of 100 equities will underperform the broad equity market by approximately 21 basis points per year simply due to skewness — what the authors call "the skewness cost of underdiversification." For 100 randomly selected bonds, that figure drops to just 5 basis points.
In equities, passive funds justify their existence by capturing stars that active managers miss. In bonds, those stars don't exist in the same form. Instead, broad bond index strategies may increase the likelihood of including large defaults.
Active Bond Funds: The Numbers
Against this backdrop, active bond fund performance looks materially different from the equity narrative. Using their five-factor CCR5 model constructed from investable passive ETFs, the authors find active bond funds generate an average annualized net alpha of 0.30% across the full sample, rising to 0.55% among high yield funds. Approximately 71% of active bond funds exhibit positive net alpha.
Over longer horizons, the advantage compounds. Where active equity fund outperformance declines over time — from 31.1% beating their benchmark over two years to just 19.0% over ten — active bond funds move in the opposite direction: 56.1% outperform over two years, rising to 68.6% over ten.
Among the most active bond funds — those with the highest firm-level active share combined with strong past performance — the annualized alpha reaches 1.76%, with statistical significance. These funds also show lower maximum drawdown and reduced financial fragility, with a linear rather than concave flow-performance relationship.
Key Takeaways for Advisors
1. Passive bond fund labels are misleading. With median active share of 59%, these funds are making significant active decisions without the mandate or the track record to justify them.
2. Rebalancing costs are a hidden drag. The gross underperformance of passive bond funds is a structural feature, not a fee artifact. Advisors should scrutinize total cost of ownership, not just expense ratios.
3. Active bond management has a credible, durable edge. The longer the horizon, the stronger the case. This is the inverse of the equity story.
4. Active share and past alpha are actionable screening tools. High firm-level active share combined with strong prior performance identifies a subset of active bond managers with persistent, economically significant outperformance.
5. The Bogle argument has limits. Buying the haystack works when stars drive returns. In bonds, the haystack may include the defaults.
Endnote:
1 Choi, Jaewon, et al. "Active versus Passive Management of Bonds (and why passive bond management is an oxymoron)." 26 Apr. 2026.
Copyright © AdvisorAnalyst
