Active or passive in bonds and commodities: why the rules differ from equities

In depth12 min read · 19 sources

The question of active or passive is almost always discussed in terms of equities. For bonds and commodities, the situation is different: bond indices weight by debt and swap out holdings every month, while commodity indices are rulebooks for rolling futures. Active bond funds beat their benchmark more often than active equity funds, but often by taking extra credit risk rather than through better selection. Here you can see why, how the return of a commodity index is made up and how to decompose fund returns.

Why the equity arithmetic does not carry over one to one

The active or passive debate rests on an argument by Sharpe (1991): before costs, all active investors together earn the market return, so after costs they earn less than passive investors. This holds for every market, but only for the market as a whole. In bonds and commodities, the gap between market, benchmark and fund is wider:

• The benchmark is not the market: depending on its investment terms, a government bond fund may also add corporate bonds, and an investment grade fund some high yield. If you buy risk outside the benchmark, you beat it in good years without selecting any better.
• Other participants have other goals: central banks, banks and insurers hold bonds for reasons of monetary policy, liquidity or regulation, not just for the return. In commodity futures, producers and consumers hedge prices; Bessembinder (1992) found that this hedging pressure is related to futures returns.
• Rules force trading: bond indices change their composition every month, and commodity indices roll every month on a fixed calendar. If you know the rules, you can trade against them or avoid them.
• No market portfolio in commodities: futures are contracts; every long position is matched by a short position. A commodity index therefore does not represent a market portfolio but a strategy with weighting and roll rules.

Bond indices: weighted by debt, rebalanced monthly, rating thresholds

Broad bond indices weight by the market value of outstanding bonds, require a minimum issue size and a rating (for the Bloomberg indices, the middle of three ratings counts) and are reconstituted monthly (Bloomberg Methodology 2026). This leads to three peculiarities:

• The bums problem: whoever takes on more debt gets more weight. Such an index steers the money of the funds tracking it towards the most heavily indebted governments and companies, not the most solid ones. Arnott et al. (2010) found that bond portfolios with weights that do not depend on market prices beat their market-value-weighted benchmarks in historical tests, namely for US corporate bonds, high yield and emerging market bonds in hard currency, and all the more clearly the more inefficient and volatile the market.
• Many holdings, little trading: a broad bond index comprises thousands of bonds, many of them rarely traded. Index funds therefore track it by sampling (see index replication) and pay the bid-ask spread when they reallocate. Every month, new bonds are added and those with too short a maturity drop out.
• Rating thresholds: if a bond loses its investment grade rating, it no longer meets the rules of an investment grade index and drops out at the next rebalancing (fallen angel). Index funds have to sell, and so do insurers under capital pressure: Ellul et al. (2011) show that the prices of downgraded bonds then fall and later recover; Ambrose et al. (2008) find considerably more selling by insurers around the downgrade. An active fund can wait, or buy precisely these bonds.

Active bond funds: ahead more often, but why?

Measured against their benchmarks, active bond funds do better on average than active equity funds. In Europe, over the ten years to June 2026, 76% of bond funds lagged their benchmark, compared with 94% of equity funds; over the six months to June 2026, the figure for bond funds was 63% (SPIVA Europe 2026). The US report for year-end 2025 gives an average across categories of 70% for bond funds, compared with 62% for equity funds; for investment grade bond funds it was 82%, for high yield 76%, and for emerging market bonds only 31%, presumably helped by the weaker dollar (SPIVA U.S. 2026). So the results vary greatly with the period and the category.

Brooks et al. (2020) investigated where the lead of the successful funds comes from: active bond managers often beat their benchmarks over the past 20 years, but passive exposures to traditional risk premiums, above all credit risk, explain most of their active returns. Little genuine alpha remains, and adding credit risk can erode the diversification benefit that bonds are in the portfolio for. A fund that permanently carries somewhat more credit risk than its benchmark carries an additional credit beta.

Worked example: credit beta instead of skill

A fund is measured against an index of investment grade corporate bonds. It adds high yield and so carries an additional credit beta of 0.2 relative to the excess return of high yield over the benchmark. Suppose this excess return is +4% in calm years and −20% in a crisis year; the fund costs 0.50% a year, an index ETF 0.10%. All the figures are illustrative.

Commodities: what a futures index really measures

If you invest in a broad commodity index, you are not buying barrels and bushels but futures, which are replaced by later contracts before they expire (rolling). The money that is not needed as margin earns interest as collateral. The total return of a fully collateralised futures index therefore consists of three parts:

The total return indices calculate exactly like this: futures return plus interest on the hypothetically deposited funds (S&P GSCI methodology 2026), using the rate on three-month US Treasury bills for the Bloomberg Commodity Index (BCOM methodology 2026). The roll yield is the carry of a future; Koijen et al. (2018) found across many asset classes, including commodities, that high carry predicted higher returns.

Worked example: spot, roll and collateral

A commodity costs 80 on the spot market and 84 a year later. The index always holds the front-month future, which converges to the spot price at expiry, and rolls every month into the next one, which trades 0.5% above the spot price (contango). The collateral earns 2% a year. Everything is illustrative.

Index construction: production, liquidity and roll calendar

Sources for the table: (S&P GSCI methodology 2026; BCOM methodology 2026), methodology from August and February 2026 respectively.

Weighting by production gives the most weight to the commodities with the highest production value; the BCOM, by contrast, explicitly caps individual commodities and groups. Both roll on a fixed, public calendar and are therefore predictable. Mouakhar and Roberge (2009) argue that most index providers roll inefficiently because they always switch into the next contract to expire; depending on the curvature of the futures curve, this leads to automatic losses. Tang and Xiong (2012) show that as index investment grew from the early 2000s, the prices of non-energy US commodity futures became increasingly correlated with the oil price, considerably more so for commodities in two widely used indices. That is what is meant by the financialisation of commodity markets.

Active variants: optimised rolling, enhanced indices, physical gold

• Optimised rolling: instead of the next contract, the index or fund picks the one with the least contango or the strongest backwardation, often further out on the curve, or rolls on different dates from the large indices. Mouakhar and Roberge (2009) show for individual commodities and for portfolios that an optimised approach could improve performance.
• Enhanced and curve strategies: they focus on characteristics that, according to Erb and Harvey (2006), were historically rewarded, such as the shape of the futures curve. This is rule-based like smart beta: active relative to the standard index, but transparent.
• Discretionary commodity funds and CTAs: managers trade on judgement or trend signals. For CTAs, Bhardwaj et al. (2014) found no significant excess return over money market instruments after fees (more under CTAs and managed futures).
• Physical gold: a physically backed ETC has neither roll yield nor collateral yield, but instead ongoing costs for custody and issuer risk. It tracks the spot price. More in the article gold in the portfolio.

Equities, bonds and commodities compared

Last row: SPIVA Europe (2026). The other rows summarise the sections above.

For advanced readers: decompose returns instead of reading league tables

• Non-linear risks: credit risk behaves like a sold option: small, steady gains, rare large losses. A regression over calm years only underestimates the beta; crisis years belong in the sample.
• Decomposing commodity funds: separate the spot, roll and collateral parts. Only a comparison with an index with the same weighting but standard rolling shows what a roll strategy delivers.
• Surviving funds: closed funds are missing from many databases, so league tables look too good (survivorship bias).
• Pricing and spreads: indices priced at bid and funds priced at mid value differently; that alone produces deviations (see index providers and pricing).

What the research shows

• Blake et al. (1993) examined US bond funds, including in a sample free of survivorship bias: after costs, they lagged suitable indices, one percentage point more in costs cost one percentage point of return on average, and past results did not predict future ones.
• Brooks et al. (2020) explain most of the active returns of bond funds with passive risk premiums, above all credit risk; Arnott et al. (2010) find advantages for weights that do not depend on market prices.
• Gorton and Rouwenhorst (2006) show for an equally weighted index of fully collateralised commodity futures from 1959 to 2004 a return and Sharpe ratio similar to those of US equities, with negative correlation to equities and bonds and positive correlation to inflation. Ten years later, the results largely hold out of sample, including the relationship between average return and basis (Bhardwaj et al. 2015).
• Erb and Harvey (2006) find hardly any excess return on average for individual commodity futures; a rebalanced portfolio, by contrast, could deliver equity-like excess returns, and the term structure was historically rewarded. They warn against naively extrapolating historical returns.
• Tang and Xiong (2012) document financialisation: with index investment, commodity prices became more closely linked to the oil price, which helps explain the high volatility around 2008.

Limits and common mistakes

• League tables from calm years: if you buy the best bond fund after three good years, you often buy the biggest credit beta.
• Wrong benchmark: a fund with a high-yield component should be measured against a blended benchmark.
• Confusing the spot price with the index return: oil price charts show the spot price; an investor in futures gets spot plus roll plus interest.
• Extrapolating roll yields: whether a curve is in contango or backwardation changes with inventories and demand.
• Passive does not mean neutral: a bond index weights by debt, a commodity index by production or liquidity. Both are decisions.
• Taxes and product type: ETFs, ETCs and ETNs on commodities are taxed differently (see gold in the portfolio); this is not tax advice.

How Portfolio Lab calculates it

Under Asset classes you will find bonds and commodities as long series in euros. Bond classes chain a tradable fund with index data and, before that, often with synthetic bonds: each month, a bond is bought at par from yields of a fixed maturity and revalued a month later. Interest rate and duration risk are thus captured, defaults are not. Commodity classes consist of series of funds that hold futures, so roll and collateral yield are contained in the fund price. Where a series is extended before the fund's launch with the price of the front-month future, and for some individual commodities entirely, roll and collateral yield are missing; the page shows for each class which sources it consists of. There are no separate building blocks for spot, roll and collateral yield.

To decompose a bond fund, use the Style analysis with the bond template (money market, German federal bonds by maturity, euro government bonds, US Treasuries) or with your own style indices such as corporate and high-yield bonds, or the factor regression with the bond factors duration and credit, built from total returns of US bond funds. A clear credit beta with an alpha close to zero is exactly the pattern from the worked example. The asset classes methodology describes how the series are built.

This article explains a method; it is not investment or tax advice. Worked examples are simplified and rounded. All terms in the glossary.