Should You Index Your Bonds? BND vs. Active Funds vs. DFA and Avantis
Over 10 years, 43% of active core bond funds beat index peers vs 10.5% of large-blend stock funds (Morningstar). Research ties much of that edge to credit risk.
The case for a stock index fund rests on two facts. A market-cap index holds every company in proportion to its market value, which is the portfolio all investors hold in aggregate, so the fund owns what the average dollar owns and charges almost nothing to do it. And most active stock funds trail those indexes: over the 10 years to June 30, 2026, 88% of all U.S. domestic stock funds and 83% of large-cap funds underperformed their benchmarks in S&P’s SPIVA data.2
Bonds keep the cost argument and weaken the rest. BND and AGG charge 0.03% a year, the same as a total stock market fund.10,11 But a bond index is a rule-defined slice of the bond market that has to rebuild itself every month, and many of the largest bondholders are not trying to maximize return. Active bond funds have beaten their index-fund peers far more often than active stock funds have. Most of that record traces to one persistent bet: holding more corporate credit than the index does, which an investor can also buy with index funds for about 0.03%.
My recommendation, in short: decide what the bond sleeve is for, index the Treasury part or own Treasuries directly, and use BND or AGG as the default core. Dimensional’s and Avantis’s core bond ETFs are credit tilts with a systematic overlay that has not yet produced a measurable payoff. In our test, AVIG returned less than a fixed mix of BND and a corporate bond index fund with the same starting corporate share, and DFCF roughly matched its mix.1
What a bond index fund owns
AGG tracks the Bloomberg U.S. Aggregate Index and BND tracks its float-adjusted version. In AGG’s August 31, 2026 SEC filing, U.S. Treasuries were 46.0% of net assets, agency mortgage-backed pools 23.7%, and corporate bonds 24.5%. BND’s June 30 filing was similar: 48.1% Treasuries, 19.8% agency mortgages and 25.6% corporate bonds.1
So the common complaint that bond index funds lend the most to the biggest borrowers is accurate, and the biggest borrower is the U.S. government. The corporate quarter of AGG is spread across 1,160 issuers. The largest, JPMorgan Chase, was 0.57% of the fund; the 10 largest together were 3.9% and the 25 largest 6.6%. No single corporate bond was more than 0.03% of the fund. In BND the largest corporate issuer, again JPMorgan, was 0.61%.1 Banks top the list because selling bonds is one of the ways they fund themselves. A large issuer is not necessarily a riskier one.
The weighting does mean the fund’s Treasury share rises when Treasury borrowing grows faster than other borrowing, and its corporate exposure grows toward the companies borrowing the most. The figures above show that individual corporate positions are small. They do not measure concentration by sector (seven of the ten largest corporate issuers in AGG are banks) or the credit risk the whole corporate quarter shares, and grouping by legal entity counts a parent and its subsidiaries separately.
“Total bond market” also covers less than the name suggests. The Aggregate holds taxable, investment-grade, fixed-rate, U.S. dollar bonds with at least one year to maturity and a minimum issue size ($300 million for Treasuries and corporates). It excludes inflation-protected Treasuries (TIPS), tax-exempt municipal bonds, high-yield bonds, floating-rate notes and bonds about to mature. Bloomberg also deducts the Federal Reserve’s Treasury holdings from the index, and the float-adjusted version that BND tracks also removes the Fed’s mortgage holdings.9 The practical gap between the two funds is small. In 2022, BND returned -13.11% and AGG -13.02% at market prices.1
How debt weights differ from market-cap weights
A stock index weights each company by price times shares outstanding. When a stock rises, its weight rises by the same amount in the index and in every index fund, so the fund has no trades to make. New shares and index additions are the main reasons it trades at all. VTI’s portfolio turnover was 3% in its latest fiscal year.12
A bond index weights each bond by the market value of the amount outstanding, and the list of bonds changes constantly. Bonds mature, drop out when they fall under one year to maturity, leave when they are downgraded below investment grade, and get replaced by new issues every month. BND’s turnover was 38% and AGG’s 62% in their latest fiscal years.10,11 Lasse Pedersen estimates that a passive bond investor needs about 20% annual turnover just to buy new issues, against 7.6% for staying passive in the U.S. stock market.8
Weighting by debt outstanding also means the index buys more of a sector as that sector borrows more. The BBB share of a main U.S. investment-grade index rose from about 25% in 1990 to about half by 2022.17 Robin Greenwood and Samuel Hanson found that when the mix of corporate borrowing shifts toward lower-quality issuers, later excess returns on corporate bonds tend to be low.18 A debt-weighted index adds weight to whichever borrowers are issuing the most, including during credit booms. In the Aggregate the effect is diluted, because corporate bonds are about a quarter of the fund.
The usual alternative is to weight by something other than debt, such as an issuer’s sales or a country’s GDP. Rob Arnott and coauthors proposed this for bonds in 2010, and the backtests published by the firms that sell those indexes look better than market weights.19 An independent test on government bond markets found that fundamental weights did beat market weights, and that the extra risk exposure they took on explained the outperformance.20 I found no independent out-of-sample evidence that reweighting the Aggregate improves its risk-adjusted return.
Sharpe’s arithmetic in the bond market
William Sharpe’s 1991 argument applies to any market. If passive investors hold the market portfolio, they earn the market return before costs, and so the remaining active dollars, which together hold the rest of the market, must earn the market return too:7
After costs, the average active dollar trails. That logic holds in bonds, but four of its premises fit bonds less well than stocks.
- The index is a subset of the market. An active “core” bond fund can hold high-yield bonds, non-dollar bonds, floating-rate notes and securitized debt that the Aggregate leaves out. It can beat the index by owning different risks, and Morningstar reports that active funds in all three of its bond categories “tend to take more credit risk than indexed peers.”4
- The market portfolio keeps changing. Pedersen shows that once new issues and index changes are allowed for, passive investors have to trade, and the active and passive returns need not net to the market return.8 Bond indexes change far more than stock indexes.
- Many large holders are not maximizing return. The Federal Reserve buys Treasuries for monetary policy. Insurers face capital rules tied to ratings. Bo Becker and Victoria Ivashina found that insurers tilt toward the highest-yielding bonds within each rating, and Andrew Ellul and coauthors found that insurers under regulatory pressure sell downgraded bonds together, pushing prices down temporarily.21,22
- Index rules create forced trades. When a bond is scheduled to leave the index, index trackers sell it together, and dealers charge for taking the other side. Jens Dick-Nielsen and Marco Rossi measured that cost around index exclusions and found it more than doubled after the 2008 financial crisis. In the accepted version of their paper they estimate that trading on exclusion dates costs investors in corporate bond index trackers roughly 34 basis points a year, and call the figure only indicative. For Vanguard’s total bond index fund, where corporate bonds are about a third of the index, they estimate that trading away from those dates would have improved tracking by 3 to 9 basis points a year.23
Each of these leaves room for an active manager in theory. None of them tells you whether a given fund captures that room after its fees and trading costs. The scorecards are the next test.
The scorecards
S&P’s SPIVA compares each active fund with an index. Bond results vary more by category than stock results do. Over the 10 years to June 2026, 96% of active general government funds (measured against an iBoxx index of U.S. sovereign and sub-sovereign debt) and 87% of high-yield funds trailed their indexes, worse than the stock categories. Only 49% of short and intermediate investment-grade funds did.2 Over the first half of 2026 alone, the average underperformance rate across bond categories was 38%, against 60% for stock categories.3
The pattern fits the credit explanation. The category where active funds did best is measured against a broad index of all U.S. dollar bonds maturing in one to five years (iBoxx USD Overall 1-5Y), which includes government debt and leaves room to add corporate and securitized credit. The worst results came against an all-government index, where there is little credit to add, and against all-corporate indexes, which already hold the credit.
Morningstar’s Active/Passive Barometer measures active funds against the actual index funds and ETFs in the same category, after those funds’ own fees, and counts closed or merged funds as failures. Over the 10 years to June 2026, 45% of active bond funds survived and beat their passive peers, the highest of any group in the study, against 25% for all active funds.4
| Morningstar category | 1 yr | 10 yr | 20 yr | 10 yr, cheapest fifth | 10 yr, priciest fifth |
|---|---|---|---|---|---|
| Intermediate core bond | 66.0% | 43.1% | 18.1% | 70.8% | 21.7% |
| Corporate bond | 34.1% | 55.6% | n/a | 44.4% | 66.7% |
| High-yield bond | 44.7% | 42.9% | n/a | 47.5% | 32.5% |
| U.S. large blend (stocks) | 31.6% | 10.5% | 4.9% | 23.1% | 9.1% |
Success rate: share of active funds that existed at the start of the period, survived, and beat the average passive fund in their category. Periods end June 30, 2026. Source: Morningstar, U.S. Active/Passive Barometer, Mid-Year 2026, Exhibit 1.4
Three details matter. First, cost predicts results: in intermediate core bonds, 70.8% of the cheapest fifth of active funds succeeded over 10 years, against 21.7% of the most expensive fifth. That cheapest-fifth figure is the strongest evidence in these data for active bond management. (The corporate bond quintiles hold nine funds each, too few to read much into.) Second, the record fades with time. Of the 221 intermediate core funds that existed 20 years ago, 18.1% survived and beat their index peers, and 28.1% survived at all.4
Third, the two scorecards disagree most about high yield. SPIVA finds that 87% of high-yield funds trailed their index over 10 years, while Morningstar finds that 43% beat their passive peers over the same period. Part of the gap shows up in the passive funds. HYG, an index ETF that tracks the same iBoxx index SPIVA uses, returned 5.51% a year over the 10 years to December 2025, against 6.01% for the index.5 The two scorecards also cover different funds and use different methods, so that lag does not explain the whole difference. In less liquid markets, “passive” is harder to deliver.2,4
Where the extra return comes from
Morningstar itself explains the recent bond results by credit: active funds hold more of it than index funds, and credit spreads stayed tight. It concludes that “the value proposition for going active in fixed income remains strong.”4 The year-to-year pattern points the same way. The one-year success rate for active corporate bond funds was 4.1% in the year to June 2025 and 34.1% a year later; for intermediate core funds it rose from 49.3% to 66.0%.4 Manager skill does not swing that much from one year to the next. Credit spreads do.
AQR’s researchers tested this directly on institutional bond managers from 1997 to 2019, before fees. The average active return of U.S. Aggregate-benchmarked managers had a correlation of 0.94 with corporate credit excess returns. After allowing for persistent credit exposure and other standard risk premiums, the managers showed “virtually no alpha,” and in the 10 worst quarters for stocks the active bond strategies trailed their benchmarks.6 AQR sells its own systematic strategies, so read this as a rival manager’s analysis; its data and method are laid out in the paper.
Academic studies of bond mutual funds reach the same place by other routes. Blake, Elton and Gruber found that bond funds trailed their indexes after expenses, with performance falling roughly one for one with fees.25 Using fund holdings, Cici and Gibson found no ability to pick corporate bonds that beat bonds with similar characteristics.26 Chen, Ferson and Peters found performance that is positive before costs and often negative after them.27 A 2021 study of U.S. fixed-income funds found that fund management companies extract most of whatever abnormal performance their managers produce.28 Choi and Kronlund found that corporate bond funds that reach for yield earn higher returns and attract more money, but that those returns “tend to be negative on a risk-adjusted basis.”24
Fees and credit are linked. Vanguard’s researchers found that across 842 taxable and 502 municipal active bond funds, higher expense ratios went with more sensitivity to credit risk.29 Active managers make the opposite case. PIMCO reports that active bond funds beat their median passive peer, after fees, in 64% of rolling 10-year periods, against 42% for active stock funds, and argues that new-issue discounts, index rebalancing and holders who are not price-sensitive give bond managers structural advantages that stock managers lack.30 Those sources are the same ones listed in the previous section, and they are real. The evidence above says that, on average, little of their value survives fees once credit exposure is counted.
Three claims are worth keeping apart. Active bond funds have had higher success rates than active stock funds; the scorecards establish that. Much of the average excess return comes from credit exposure; AQR shows this for institutional products before fees through 2019, and Morningstar says the same about recent years. Whether a cheap active bond fund adds value that lasts, after allowing for its credit risk, is a separate question these data do not settle, and the 70.8% success rate of the cheapest fifth is the best sign that some do.
The credit tilt also costs something in the situation bonds are usually held for. Corporate credit tends to fall when stocks fall, so a core bond fund that leans on it gives up part of its protection in a recession, as the AQR study found in the worst quarters for stocks.6 For an investor whose portfolio is mostly stocks, extra credit inside the bond sleeve adds risk that overlaps with the stocks already owned. The guide to what a 60/40 portfolio owns and the guide to bonds as diversifiers cover that trade-off in detail.
DFA and Avantis: systematic tilts at near-index fees
Dimensional and Avantis run rules-based core bond funds benchmarked to the Aggregate. Neither forecasts interest rates. Both use the current yield curve and credit spreads to decide where expected returns look higher, inside a duration range tied to the index.
| Fund | Expense ratio | Duration rule | Corporate bonds | Turnover |
|---|---|---|---|---|
| AGG (index) | 0.03% | Matches the Aggregate | 24.5% | 62% |
| BND (index) | 0.03% | Matches the float-adjusted Aggregate | 25.6% | 38% |
| DFCF (Dimensional) | 0.17% | Aggregate’s duration minus 1 year to plus 0.25 years | 48.5% | 23% |
| AVIG (Avantis) | 0.15% | Within 2 years of the Aggregate’s duration | 60.6% | 332% |
Corporate bonds as a percent of net assets in each fund’s latest public SEC N-PORT filing (AGG and AVIG August 31, 2026; DFCF July 31; BND June 30), excluding supranational and foreign-agency issuers. DFCF and AVIG own their agency mortgages through forward (TBA) contracts, which filings report at full value alongside the bonds that back them, and DFCF also holds about 6% in a Dimensional cash fund, so their holdings add up to more than 100% of net assets. The corporate figure is corporate bond holdings per dollar invested in the fund. DFCF’s 0.17% is after a fee waiver (0.18% before it). Expense ratios, duration rules and turnover from each fund’s summary prospectus.1,10,11,13,14
DFCF’s prospectus says it adds or reduces intermediate-term exposure depending on the expected term premium, and adds or reduces corporate and other non-government bonds depending on the expected credit premium, with an emphasis on bonds rated A+ to BBB−.13 AVIG sorts bonds into groups by industry sector, rating, duration, country and currency, estimates each group’s expected return from its yield curve, and keeps its duration within 2 years of the Aggregate’s.14 In plain terms, both funds hold about twice the index’s share of corporate bonds, or more, and choose maturities using the shape of the curve.
The idea has a long academic record. Eugene Fama showed in the 1980s that forward rates, which can be read from the yield curve, carry information about future bond returns.31,32 Dimensional’s own research reports that the average term premium was 9 basis points a month from 1976 to 2023 and 33 when the term spread was in its top quarter, with a similar pattern for credit, and that a simulated U.S. core strategy using these signals beat the Aggregate by 41 basis points a year. The results are hypothetical, and Dimensional sells the strategy.33 Out-of-sample tests are less encouraging: Daniel Thornton and Giorgio Valente found that forward-rate forecasts did not beat a no-predictability benchmark for an investor allocating in real time.34
The broader research on corporate bond “factors” is shakier than the stock version. One widely cited 2019 paper on corporate bond risk factors was retracted by the Journal of Financial Economics after errors were found in its factor data, and a 2023 study in the same journal found that the proposed bond factors add little beyond the overall corporate bond market.35,37 Dimensional and Avantis rely mainly on term and credit premiums, which have longer records than those factors, although Dimensional’s own figures above show how much their size varies. The episode is a reason to ask for live results.
Our test: systematic funds against index funds
We compared each systematic fund with the index funds over its full live record, using monthly total returns at market prices through October 2026. To separate the static credit tilt from everything else, we also built a replica for each fund from two index funds: BND and Vanguard’s Intermediate-Term Corporate Bond ETF (VCIT, 0.03%).16 The VCIT weight was set so the replica’s corporate share matched the fund’s in its first SEC holdings filing, and each replica comparison starts the month after that filing and the matching BND and VCIT filings were public (February 2021 for AVIG, April 2022 for DFCF). The mix was rebalanced each January. The replica matches the corporate share and nothing else: its duration, credit ratings and agency mortgage share can differ from each fund’s.1
| Comparison | Window | Fund, annualized | Comparison, annualized | Gap | Tracking error | Worst drawdown, fund vs comparison |
|---|---|---|---|---|---|---|
| AVIG vs AGG | Nov 2020 to Oct 2026 | -0.19% | -0.21% | +0.02 | 0.76% | -18.5% vs -17.0% |
| DFCF vs AGG | Dec 2021 to Oct 2026 | -0.10% | -0.23% | +0.13 | 1.08% | -18.0% vs -15.8% |
| DFAPX vs VBTLX | Apr 2011 to Oct 2026 | 2.60% | 2.25% | +0.36 | 0.89% | -17.5% vs -17.4% |
| AVIG vs BND/VCIT replica | Feb 2021 to Oct 2026 | -0.33% | 0.04% | −0.37 | 0.42% | -18.1% vs -17.4% |
| DFCF vs BND/VCIT replica | Apr 2022 to Oct 2026 | 1.59% | 1.62% | −0.03 | 0.83% | -11.4% vs -10.6% |
Annualized total returns from Yahoo Finance dividend-adjusted month-end closes (market price for ETFs, NAV for the mutual funds DFAPX and VBTLX). Gap in percentage points a year. Replica weights: 38% VCIT for AVIG and 33% VCIT for DFCF, the rest BND. Summitward calculation.1
Against AGG, both systematic ETFs finished close to even: +0.02 points a year for AVIG over 6 years and +0.13 for DFCF over about 5 years, with deeper drawdowns in 2022. A regression of each fund’s monthly return difference on a credit factor and a term factor shows the expected tilt: both load positively on credit (t-statistics of 3.3 and 2.6), and what is left over is close to zero, +0.02 and −0.10 points a year with t-statistics of 0.12 and -0.25.1
Against the replicas the results split. From February 2021, a 38/62 mix of VCIT and BND returned 0.04% a year against AVIG’s -0.33%, a gap of −0.37 points with a t-statistic of -2.11; roughly 0.12 points of that is the fee difference. DFCF and its replica finished about even from April 2022: −0.03 points a year (t of -0.09). After adjusting for the remaining credit and term differences, the gaps were −0.20 points for AVIG (t of -1.57) and −0.17 for DFCF (t of -0.55), neither statistically reliable.1
Start dates move these numbers a lot. Starting DFCF’s comparison at its launch in December 2021, four months earlier and before its first holdings report was public, changes its gap to −0.28 points a year; AVIG’s barely moves (−0.38 from November 2020).1 These are 5- and 6-year windows that include the 2022 rate shock. They show that a cheaper index-fund mix with the same starting corporate share did at least as well as either fund. They cannot show whether the funds’ processes add or subtract value over longer periods.
DFAPX, Dimensional’s Aggregate-benchmarked mutual fund, has the longest record of the approach. Its current prospectus uses the same term and credit language and the same duration range as DFCF. Its prospectus does not list mortgage-backed securities among its investments, which by itself makes its returns differ from the Aggregate’s, and it charges 0.19%.15 From April 2011 to October 2026 it returned 2.60% a year against 2.25% for Vanguard’s total bond index fund, a gap of +0.36 points with a t-statistic of 1.59. After the credit and term regression the remaining edge was +0.17 points (t of 1.02).1 That is the best evidence for the approach. A result this size is consistent with a small positive edge, no edge, or a small negative one.
Short records are hard to judge because returns are noisy relative to the gaps being measured. With tracking error near 0.8% a year, a true edge of 0.25 points a year needs about 36 years of returns to reach a t-statistic of 2 on average:
What I recommend
Start with the job the bonds are doing. That choice moves outcomes far more than the choice between BND and DFCF.
| Job of the bond sleeve | What to hold |
|---|---|
| Spending in the next few years | T-bills, a money market fund, or a ladder of Treasuries or TIPS maturing when the money is needed |
| Cushion against a stock crash | A Treasury index fund, or a total bond index fund if you accept its corporate quarter |
| One core bond holding | BND or AGG |
| Deliberately more credit | Add a corporate bond index fund to BND, or use DFCF or AVIG if you accept 0.12 to 0.14 points a year more in fees and some tracking error |
| Discretionary active management | Only cheap funds, judged against a benchmark with the same credit exposure, and not for the Treasury sleeve |
The government bond results are the clearest guidance in the data: 96% of active general government funds trailed their index over 10 years, so the Treasury part of a portfolio belongs in an index fund or in Treasuries bought directly. For the core, BND and AGG are cheap, diversified and predictable, and nothing in the evidence says you are leaving a reliable return on the table. If you want more credit, buy it on purpose and in the open: a corporate bond index fund next to BND costs less than DFCF or AVIG, and in our test it matched DFCF and beat AVIG.
Where a bond fund sits matters too. Interest on taxable bond funds is taxed as ordinary income, so these funds usually belong in IRAs and 401(k)s when there is room. In a taxable account, compare a Treasury-only fund (Treasury interest is exempt from state and local income tax) and a municipal bond fund on an after-tax basis before defaulting to BND.36
When an index fund is the weaker choice
Indexing has the least advantage where the index is expensive to replicate. High-yield bonds are the clearest case in the scorecards: passive high-yield funds trail their index by enough that 43% of active high-yield funds beat them over 10 years. In a market like that, a low-cost active fund with a long record can be a reasonable choice. Municipal bonds look different: 76% of active general municipal funds trailed their index over the same period.2,4
A systematic fund like DFCF or AVIG can also make sense for an investor who wants a credit tilt, prefers one fund to two, and values a disciplined process that adjusts with spreads. That is a defensible preference. So far DFCF has roughly matched a cheaper index-fund mix and AVIG has trailed one, over records too short to judge.
Key Takeaways
- The cost case for bond index funds is as strong as for stocks. BND and AGG charge 0.03% a year.
- The equilibrium case is weaker. A bond index is a rule-defined subset of the market that keeps changing: bond index funds turn over 38% (BND) to 62% (AGG) a year, against 3% for VTI, and many large bondholders are not maximizing return.
- The biggest borrower in a total bond fund is the U.S. Treasury. In AGG’s August 2026 filing, the largest corporate issuer was 0.57% of the fund and the top 10 were 3.9%.
- Active bond funds beat index funds more often than stock funds do, 43.1% of intermediate core funds over 10 years against 10.5% of large-blend stock funds in Morningstar’s data, and research ties much of that to holding more credit.
- Active general government funds are the exception: 96% trailed their index over 10 years.
- In our test, a cheap index-fund mix with the same starting corporate share beat AVIG by 0.37 points a year and roughly matched DFCF. The records are too short to settle whether either fund’s process adds value.
Frequently Asked Questions
Is BND or AGG better?
They are close substitutes. Both charge 0.03% and track versions of the Bloomberg U.S. Aggregate; BND’s version also removes the Federal Reserve’s mortgage holdings. Their returns differ by small amounts from year to year. Pick whichever your brokerage trades without commission.
Do bond index funds lend the most to the most indebted companies?
They lend the most to the largest issuers, and the largest by far is the U.S. Treasury, at about 46% of AGG. The corporate sleeve is spread across more than a thousand issuers, and the largest, JPMorgan Chase, was under 0.6% of the fund in its August 2026 filing. Debt weighting does add exposure to sectors as they borrow more.
Is AVIG or DFCF better than BND?
They hold more corporate bonds than BND and adjust maturities using the yield curve. Over their live records they returned about the same as AGG and BND with deeper drawdowns in 2022. In our test, a BND and corporate index fund mix with the same starting corporate share beat AVIG and roughly matched DFCF. They are reasonable funds; the evidence that they improve on indexing is not there yet.
Do active bond funds beat the index?
More often than active stock funds. Over the 10 years to June 2026, 43% of active intermediate core bond funds survived and beat their passive peers, against 10.5% of large-blend stock funds. Research ties much of the advantage to holding more credit than the index, which tends to pay in calm years and hurt when stocks fall.
Should I index government bonds?
Yes. Active general government funds have the worst record of the SPIVA bond categories shown above: 96% trailed their index over the 10 years to June 2026. A Treasury index fund, or Treasuries bought directly, is the sensible default.
Should I buy individual bonds instead of a bond fund?
Individual Treasuries or TIPS held to maturity work well for spending on a known date, because you know what you will receive. For general diversification a fund is simpler. Corporate bonds are harder to buy in small amounts at fair prices, so a fund is usually the better way to hold them.
Related Guides
- What a 60/40 Portfolio Really Owns How much credit risk sits inside an aggregate bond fund, and when it hurts.
- Are Bonds Still Good Diversifiers? Treasuries versus credit as a hedge for stocks.
- Liability-Driven Investing for DIY Investors Matching bonds to future spending.
- How to Build a TIPS Ladder Owning inflation-protected bonds directly.
- The Problem With “Just Invest in Index Funds” What an index fund does and does not decide for you.
- The Rise of Avantis The firm behind AVIG and its stock funds.
Sources
- Summitward calculations: holdings from each fund’s latest public SEC Form N-PORT filing, returns from Yahoo Finance dividend-adjusted month-end closes through October 2026, regressions and the static replica as described under Method. Code, data and printed output: summitward-research/should-you-index-bonds
- S&P Dow Jones Indices, SPIVA U.S. data, data as of June 30, 2026 (10-year column, percentage of funds that underperformed the comparison index). spglobal.com
- Anu R. Ganti, Liam Flaherty and Nick Didio, “SPIVA U.S. Mid-Year 2026,” S&P Dow Jones Indices, September 17, 2026. spglobal.com
- Morningstar Manager Research, “U.S. Active/Passive Barometer, Mid-Year 2026,” data through June 30, 2026: Exhibits 1 to 3, the fixed-income section (p. 31) and the Intermediate Core Bond tables (p. 32). morningstar.com
- iShares Trust, iShares iBoxx $ High Yield Corporate Bond ETF (HYG) summary prospectus (497K), June 29, 2026: average annual total returns for the 10 years ended December 31, 2025, 5.51% for the fund and 6.01% for the iBoxx USD Liquid High Yield Index. sec.gov
- Jordan Brooks, Tony Gould and Scott Richardson, “Active Fixed Income Illusions,” Journal of Fixed Income 29(4), Spring 2020. Institutional products in the eVestment database, gross of fees, November 1997 to September 2019. aqr.com
- William F. Sharpe, “The Arithmetic of Active Management,” Financial Analysts Journal 47(1), 1991. stanford.edu
- Lasse Heje Pedersen, “Sharpening the Arithmetic of Active Management,” Financial Analysts Journal 74(1), 2018, pp. 27-28 (7.6% average annual turnover for all CRSP equities, 1926 to 2015; 20% to buy new fixed-income issues). doi.org; open-access copy at cbs.dk
- Bloomberg, “Bloomberg US Aggregate Index” index rules (eligibility, exclusions, treatment of Federal Reserve holdings). assets.bbhub.io; and “Bloomberg U.S. Aggregate Float Adjusted Index,” April 2024. assets.bbhub.io
- iShares Trust, iShares Core U.S. Aggregate Bond ETF summary prospectus (497K), June 29, 2026: expense ratio 0.03% and 62% portfolio turnover (fiscal year ended February 28, 2026). sec.gov
- Vanguard Bond Index Funds, Vanguard Total Bond Market ETF (BND) summary prospectus (497K), April 28, 2026: expense ratio 0.03%, benchmark Bloomberg U.S. Aggregate Float Adjusted Index, 38% portfolio turnover. sec.gov; Admiral Shares (VBTLX) summary prospectus, expense ratio 0.04%. sec.gov
- Vanguard Index Funds, Vanguard Total Stock Market ETF (VTI) summary prospectus (497K), April 28, 2026: expense ratio 0.03%, 3% portfolio turnover. sec.gov
- Dimensional ETF Trust, Dimensional Core Fixed Income ETF summary prospectus (497K), February 28, 2026: expenses 0.18% before and 0.17% after a fee waiver, duration range, strategy, 23% turnover. sec.gov
- American Century ETF Trust, Avantis Core Fixed Income ETF summary prospectus (497K), January 1, 2026: expense ratio 0.15%, duration range, strategy, 332% turnover (fiscal year ended August 31, 2025). sec.gov
- DFA Investment Dimensions Group, DFA Investment Grade Portfolio summary prospectus (497K), February 28, 2026: expenses 0.19%, the same term, credit and duration language as DFCF. sec.gov
- Vanguard Scottsdale Funds, Vanguard Intermediate-Term Corporate Bond ETF (VCIT) summary prospectus (497K), December 19, 2025: expense ratio 0.03%. sec.gov
- Bank for International Settlements, BIS Quarterly Review, September 2022, box on the BBB share of U.S. investment-grade corporate bonds. bis.org
- Robin Greenwood and Samuel G. Hanson, “Issuer Quality and Corporate Bond Returns,” Review of Financial Studies 26(6), 2013. doi.org
- Robert D. Arnott, Jason C. Hsu, Feifei Li and Shane D. Shepherd, “Valuation-Indifferent Weighting for Bonds,” Journal of Portfolio Management 36(3), 2010. doi.org
- Lidia Bolla, “Fundamental Indexing in Global Bond Markets: The Risk Exposure Explains It All,” Financial Analysts Journal 73(1), 2017. doi.org
- Bo Becker and Victoria Ivashina, “Reaching for Yield in the Bond Market,” Journal of Finance 70(5), 2015. doi.org
- Andrew Ellul, Chotibhak Jotikasthira and Christian T. Lundblad, “Regulatory Pressure and Fire Sales in the Corporate Bond Market,” Journal of Financial Economics 101(3), 2011. doi.org
- Jens Dick-Nielsen and Marco Rossi, “The Cost of Immediacy for Corporate Bonds,” Review of Financial Studies 32(1), 2019. doi.org; the 34 basis point estimate (p. 3 and section 5.3) and the 3 to 9 basis point Vanguard Total Bond Market Index Fund illustration (section 5.3) are in the accepted manuscript. cbs.dk
- Jaewon Choi and Mathias Kronlund, “Reaching for Yield in Corporate Bond Mutual Funds,” Review of Financial Studies 31(5), 2018. doi.org
- Christopher R. Blake, Edwin J. Elton and Martin J. Gruber, “The Performance of Bond Mutual Funds,” Journal of Business 66(3), 1993. doi.org
- Gjergji Cici and Scott Gibson, “The Performance of Corporate Bond Mutual Funds: Evidence Based on Security-Level Holdings,” Journal of Financial and Quantitative Analysis 47(1), 2012. doi.org
- Yong Chen, Wayne Ferson and Helen Peters, “Measuring the Timing Ability and Performance of Bond Mutual Funds,” Journal of Financial Economics 98(1), 2010. doi.org
- Andrew Clare, Keith Cuthbertson, Dirk Nitzsche and Niall O’Sullivan, “How Skilful Are US Fixed-Income Fund Managers?” International Review of Financial Analysis 74, 2021. doi.org
- Vanguard (Patterson and Yan), “Are Investors Paying Too Much for Risk Exposure?” May 15, 2025: 842 taxable and 502 municipal active bond funds. vanguard.com
- PIMCO (Seidner and Mittal), “Calculating the Active Advantage in Fixed Income,” December 2025. Written by an active bond manager; net of fees, against the median passive fund in the same Morningstar category, over rolling 10-year periods (the page’s stated date ranges differ slightly between its chart and footnote). pimco.com
- Eugene F. Fama, “The Information in the Term Structure,” Journal of Financial Economics 13(4), 1984, pp. 509-528. doi.org
- Eugene F. Fama and Robert R. Bliss, “The Information in Long-Maturity Forward Rates,” American Economic Review 77(4), 1987. repec.org
- Dimensional (Wang), “Higher Expected Returns with Systematic Fixed Income,” June 5, 2024. Sponsor research; its strategy results are simulated, and the page does not state the simulation period or whether returns are after fees. dimensional.com
- Daniel L. Thornton and Giorgio Valente, “Out-of-Sample Predictions of Bond Excess Returns and Forward Rates: An Asset Allocation Perspective,” Review of Financial Studies 25(10), 2012. doi.org
- Retraction notice for Jennie Bai, Turan G. Bali and Quan Wen, “Common Risk Factors in the Cross-Section of Corporate Bond Returns,” Journal of Financial Economics 131(3), 2019; notice in volume 150, 2023. doi.org
- Internal Revenue Service, Publication 550, Investment Income and Expenses, “U.S. Treasury Bills, Notes, and Bonds”: Treasury interest is taxable federally and exempt from state and local income taxes. irs.gov
- Alexander Dickerson, Philippe Mueller and Cesare Robotti, “Priced Risk in Corporate Bonds,” Journal of Financial Economics 150(2), 2023. doi.org
Method
Holdings come from each fund’s latest public SEC Form N-PORT filing (AGG and AVIG August 31, 2026; DFCF July 31, 2026; BND June 30, 2026), as a percent of net assets. Corporate bonds are positions the filer classifies as corporate or REIT debt; supranational and foreign-agency issuers, which Vanguard and Avantis file as corporate and Dimensional files separately, are moved to government-related for every fund. Issuers are grouped by legal entity identifier, so a parent and its subsidiaries can count separately and the top-issuer weights are lower bounds.
Returns are month-end changes in Yahoo Finance dividend-adjusted closes through October 2026. Comparisons with index funds start with the systematic fund’s first full month (November 2020 for AVIG, December 2021 for DFCF, April 2011 for DFAPX). The regressions use the monthly return difference against the index fund as the dependent variable, IEF minus BIL as the term factor and SPIB minus IEI (two ETFs of similar duration) as the credit factor, with Newey-West standard errors; using LQD minus IEF as the credit factor gives the same conclusions. The replica holds BND and VCIT with the VCIT weight set so that its corporate share equals the fund’s in its first N-PORT filing (AVIG November 2020, DFCF January 2022), using the BND and VCIT filings on or before that date, and is rebalanced each January. Each replica comparison starts the month after the last of those three filings appeared on EDGAR (January 29, 2021 for AVIG; March 31, 2022 for DFCF). The replica does not match each fund’s duration, credit ratings, or Treasury and mortgage mix. SPIVA figures come from S&P’s SPIVA data page, which lists U.S. bond categories for periods ending June 30, 2026; it shows two high-yield rows, at 86.55% and 86.61% over 10 years, and the chart uses the first. Revised November 20, 2026: the replica comparisons now start after the holdings filings they use were public. Code, data and printed output are in the summitward-research repository.
Author disclosure
I own one of the index funds compared here (BND or AGG). I do not hold DFCF, AVIG or other Dimensional or Avantis bond funds. Summitward has no business relationship with Vanguard, iShares, Dimensional, Avantis, PIMCO or any firm mentioned here and receives no compensation from them. Figures labeled as Summitward calculations are historical measurements over the stated windows, not forecasts. Nothing here is investment advice.
More in Investing & Portfolio
Browse all investing & portfolio guidesGet new guides by email
Evidence-based, no jargon. At most two emails a month. Unsubscribe any time.
Try it in Summitward
See portfolio factor analysis in action with your own financial data. Free to start, no credit card required.