Skip to main content
Roll yield

ETF Tracking Error from Rolling

Pomegra Learn

ETF Tracking Error from Rolling

A commodity ETF is supposed to track the price of a commodity. If gold rises 10%, a gold ETF should return approximately 10%. If crude oil falls 5%, an oil ETF should return approximately −5%. The reality is far messier. Commodity ETFs regularly diverge from the underlying spot price—sometimes by small amounts (1–3% annually), sometimes by large amounts (10–20% or more). The primary culprit is tracking error from rolling—the persistent performance gap caused by rolling futures contracts in contango or backwardation.

Understanding tracking error is essential for any investor holding commodity ETFs. A 2% annual tracking error compounded over 20 years erodes more than a third of expected returns. Yet many retail investors do not realize their commodity ETF is losing money to rolling costs until they examine their returns and discover a disappointing gap between expected and realized performance.

Defining Tracking Error​

Tracking error is the difference between an ETF's actual return and the return of its benchmark. For most commodity ETFs, the benchmark is the spot price of the commodity. If the spot price rises 15% and the ETF returns 12%, the tracking error is −3%.

Tracking error can be positive (the ETF outperforms the benchmark) or negative (the ETF underperforms). Over longer periods, tracking error compounds, amplifying its impact on wealth.

Sources of Tracking Error​

1. Roll Yield and Futures Curve Dynamics

The largest component of tracking error is typically roll yield. An ETF holding crude oil futures in a contangoed market loses money every month rolling forward. If the contango is 1.3% monthly, the annual tracking error from rolling alone approaches −15%. This is independent of any change in the spot price; the ETF is underperforming purely due to mechanical rolling costs.

Conversely, in a backwardated market, an ETF gains from rolling, creating positive tracking error. An ETF tracking gold in a slightly backwardated market might outperform the spot price by 2–3% annually, even if spot price moves are minimal.

2. Expense Ratio

All ETFs charge operating fees, expressed as the annual expense ratio (ER). A commodity ETF might charge 0.45–0.95% annually. These fees are deducted from fund assets daily, reducing returns below the benchmark by the full ER amount. Over time, a 0.75% ER reduces returns by that amount each year. Over 20 years, a 0.75% annual fee reduces wealth by approximately 14% compared to an unmanaged benchmark.

3. Cash Drag and Rebalancing Slippage

Commodity ETFs must maintain small cash buffers for operational needs, margin requirements, and rolling transaction costs. This cash (held in money market funds) typically earns very little, creating a drag on returns. The magnitude is typically modest (10–30 basis points annually) but compounds over time.

Additionally, the act of rebalancing—selling expiring contracts and buying new ones—incurs bid-ask spreads and commissions. As discussed in the rolling mechanics article, this slippage can total 20–50 basis points annually.

4. Imperfect Futures Pricing

The spot price and the front futures contract price are not identical, though they converge at expiration. During periods of supply stress or delivery constraints, the basis (the difference between spot and futures) can widen significantly. An ETF buying futures contracts at prices above the true spot price will perpetually underperform the spot benchmark.

This is most acute in metals. Gold spot is often quoted at prices slightly below gold futures, reflecting the cost and logistics of physical storage and delivery. An ETF using futures will never quite track the spot as closely as an ETF using physical gold.

Case Studies in Tracking Error​

Crude Oil ETFs in 2014–2016

The most striking tracking error example occurred in crude oil. From mid-2014 to early 2016, crude oil spot prices fell from $105 to below $30, a 70% decline. A pure spot-price investor would lose 70%. In reality, crude oil ETFs lost approximately 80–85%, significantly more than the spot decline.

The reason: extreme contango throughout the collapse. As oil inventories swelled and storage was constrained, the futures curve became severely contangoed. Rolling costs reached 30–50% annually—far more than the spot price decline. The ETF lost money both from spot depreciation and from rolling costs incurred while holding a depreciating asset. The combination was devastating.

Natural Gas ETFs in 2016

Natural gas exhibited even more extreme tracking error. Spot natural gas prices fell from $3 per million BTU to below $2 in early 2016, a 35% decline. However, some natural gas ETFs lost 60–70%. The cause: persistent contango throughout this period, as storage filled and the futures curve became extremely steep. Rolling costs created a drag that exceeded the spot price decline.

Gold ETFs: Stability and Minimal Tracking Error

In contrast, gold ETFs typically exhibit very low tracking error. Gold rarely experiences extreme contango or backwardation. Convenience yield is minimal (gold does not deteriorate, and demand is stable). Roll yield is therefore small. A gold ETF tracking the spot gold price typically diverges by less than 1% annually. The main drag is the expense ratio, not rolling.

This explains why gold ETFs are generally high-quality core holdings, while energy ETFs require more careful monitoring and strategic timing.

Measuring and Monitoring Tracking Error​

Investors can compute tracking error by comparing the ETF return to the spot price return over any period. Monthly, quarterly, and annual tracking error tell different stories:

  • Positive tracking error in one month might indicate the ETF benefited from backwardation or favorable rolling.
  • Negative tracking error building month after month indicates persistent contango or high expenses.
  • Highly variable tracking error month to month suggests the ETF is not following a consistent rolling schedule or the spot-futures basis is volatile.

The SEC requires commodity ETF prospectuses to disclose rolling procedures, the expected tracking error range, and historical tracking error. Investors should review these disclosures carefully before purchasing.

Third-party research firms (Morningstar, Seeking Alpha, and others) regularly publish tracking error analyses. Many commodity fund managers publish annual reports explaining tracking error in detail and justifying it as a structural feature of futures-based tracking.

When Tracking Error Becomes Intentional​

Some commodity funds intentionally allow (or create) tracking error to implement strategic objectives. For example:

Curve Rolling Strategies: Instead of rolling mechanically to the next monthly contract, some funds roll to three-month-forward contracts or use other rolling schedules designed to optimize for expected curve dynamics. This intentionally diverges from a simple spot-price benchmark but may provide better risk-adjusted returns during certain market regimes.

Leverage and Amplification: Some leveraged commodity ETFs (e.g., 2x or 3x oil ETFs) are designed to amplify spot price moves. Their "tracking error" relative to spot price is intentional—they are tracking an amplified benchmark, not the spot price itself.

Actively Managed Funds: Some commodity funds employ active managers who choose rolling schedules, timing, and curve positioning to outperform a simple spot-tracking benchmark. Their tracking error (hopefully positive) reflects the value of active management.

Disclosure and Transparency Issues​

The SEC has periodically scrutinized commodity ETF tracking error and the adequacy of disclosures. The issue is that many retail investors do not understand that a commodity ETF is not a perfect replica of owning the physical commodity or the spot price.

Consider an investor who sees a news headline stating "Oil rises 15%" and checks their oil ETF to find it returned only 10%. They may blame the fund manager or attribute it to poor market conditions, not realizing the 5% shortfall is likely due to contango and rolling costs—a structural feature they were not aware of when buying the fund.

To protect investors, the SEC has encouraged funds to:

  • Clearly disclose the rolling schedule and expected roll yield impact
  • Provide historical tracking error statistics
  • Explain the difference between spot price returns and futures-based returns
  • Highlight the limitations of futures-based commodity exposure

Major commodity ETF sponsors (iShares, Vanguard, Invesco) now provide detailed tracking error analysis in their fund fact sheets and marketing materials.

The Volatility Smile: Contango and Risk​

There is an indirect relationship between roll yield tracking error and portfolio volatility. When contango is widening (suggesting weaker commodity supply and demand), commodity volatility often rises simultaneously. An ETF experiencing negative roll yield from contango is also likely experiencing increased volatility. Conversely, backwardation often coincides with low volatility and positive roll yield.

This creates a subtle dynamic: the periods when you most need commodity ETFs to hold up (during supply crises and high volatility) are the same periods when tracking error is most likely to be negative. The reverse is true during calm, well-supplied periods—the ETF outperforms, but you do not need the diversification benefit as much.

Sophisticated investors account for this by avoiding commodity ETF exposure during periods of suspected contango widening and increasing exposure during backwardation periods.

Comparing ETF Structures​

Different commodity ETF structures have different tracking error profiles:

Futures-Based ETFs: Track spot via rolling futures contracts. Subject to roll yield, slippage, and basis risk. Typical tracking error: ±1% to ±5% annually, depending on commodity and market regime.

Physical Commodity ETFs: Hold actual physical inventory (especially common in gold and silver). Track spot closely, with tracking error primarily from storage costs and expenses. Typical tracking error: ±0.1% to ±0.5% annually.

Swap-Based ETFs: Use total return swaps with banks to replicate commodity returns. Can be efficient but introduce counterparty risk. Tracking error varies widely depending on swap pricing.

Commodity Producer Equities: Some investors use mining company stocks or commodity producer ETFs as a substitute for direct commodity exposure. These have very different return profiles, driven by company operations, leverage, and management rather than commodity spot prices. Tracking error can exceed ±10% annually.

For investors seeking pure commodity exposure, physical-based ETFs (where available and cost-effective) offer superior tracking to futures-based structures. However, for many commodities (oil, natural gas), physical holding is impractical, making futures-based ETFs the only viable option despite their tracking error challenges.

Strategic Implications​

Understanding tracking error allows investors to make informed decisions:

  1. Avoid passive commodity ETFs during expected contango periods: If you believe a commodity will trade in contango, avoid or reduce futures-based ETF holdings. The roll costs will offset any spot price gains.

  2. Overweight during expected backwardation: If you expect backwardation, commodity ETF positions offer attractive returns from both spot appreciation and positive roll yield.

  3. Use physical ETFs for core holdings: For long-term commodity allocation, prefer physically-backed ETFs (if available) to minimize rolling costs.

  4. Monitor tracking error actively: Check your ETF's actual returns versus spot price quarterly. If tracking error is accumulating faster than expected, investigate the cause (contango, high expenses, basis issues) and consider alternatives.

  5. Adjust expectations appropriately: If the commodity rises 10% and your ETF returns 7%, understand why before blaming the fund manager. The difference is likely a feature of the market structure, not fund mismanagement.

References​