What is contango and why does it erode commodity ETF returns?

The short answer

Imagine buying a barrel of oil today for $70, knowing storage costs $1 per month, and offering to deliver in three months. To make that contract worthwhile, the buyer pays roughly $73. That gap — futures above spot — is contango. It reflects the carrying cost of holding a physical commodity over time.

Futures-based ETFs do not own physical barrels. They hold contracts that expire each month. Before expiration, the ETF sells the cheaper near-month contract and buys the more expensive next-month one. Repeat this monthly, and even when spot prices stay flat, the fund loses money on each roll.

Contango is not a market accident. It is the default condition for most commodities under normal supply conditions, which means buy-and-hold investors in futures-tracking commodity ETFs face a persistent structural drag.

→ New to commodities? Commodity regimes hub

What the data shows

The historical evidence on contango drag is unambiguous. Research by Erb and Harvey (2006) found that roll returns explain approximately 91.6% of the cross-sectional variation in commodity futures excess returns — the curve shape, not the spot price, is what drives long-run performance differences across commodity ETFs.

The empirical record (Bloomberg, USCF Investments, 2012–2022):

• USO ETF returned -14.6% annualized over the ten years ending January 2022, while spot WTI rose materially over the same window
• An oil ETF facing persistent 2% monthly contango loses roughly 24% annually from rolls alone, even with flat spot prices
• VIXY ETF (volatility futures) returned -50.3% annualized over the same decade — same mechanism, more extreme curve
• Gold ETFs holding physical bullion (rather than futures) avoided this drag entirely over identical periods

The exception worth noting: during periods of acute supply tightness, futures curves can flip into backwardation, generating positive roll yield. Such episodes tend to be temporary.

→ Dataset: WTI crude oil price dataset

Why it happens — the macro mechanism

Contango emerges from a simple no-arbitrage condition. Future delivery has to compensate the seller for storage, insurance, and the opportunity cost of capital tied up in inventory. When these carrying costs are positive — which they usually are — the futures curve slopes upward.

The carrying cost channel. Storage tanks, warehouses, and tankers cost money. Financing inventory at prevailing interest rates costs money. The further out a delivery date, the more accumulated carry, and the higher the futures price. This is why the curve is upward-sloping by default and why higher rates have historically widened contango.

The crowding channel — the underappreciated mechanism. Research by Chincarini and Moneta (2021) shows that the rapid growth of commodity ETF assets under management has itself contributed to wider contango in oil markets. When ETFs systematically buy the front-month contract and roll forward, their predictable demand pushes near-month prices up relative to later contracts, then the reverse on the roll date — a self-reinforcing drag. ETF flows shape commodity futures pricing in observable ways.

The financialization channel. Pension funds and index investors treat commodity allocations as a portfolio asset class. Their flows are largely insensitive to curve shape. The structural buyer base for futures-tracking products has expanded faster than the underlying physical market.

Synthesis by regime: in persistent contango (the default state observed throughout most of the 2010s), futures-based ETFs underperform spot by 5-15% annually depending on curve steepness. In super-contango episodes — most dramatically the April 2020 oil collapse when WTI front-month briefly traded at -$37 — the roll cost can be catastrophic for funds forced to exit positions. In backwardation phases, such as oil markets through much of 2022 following Russia’s invasion of Ukraine, the same mechanism reverses and rolls become accretive. The pivot between regimes hinges on whether front-month supply is constrained: when buyers pay a premium for immediate delivery, the curve inverts.

Contango is not a market failure. It is the default condition of futures markets — and the structural tax that buy-and-hold ETF investors quietly pay.

→ Framework: Commodity price formation

What it means for different economic actors

Long-horizon investors. Futures-based commodity ETFs were not designed for buy-and-hold strategies. The structural drag from rolling contracts compounds over time, often turning positive spot returns into negative ETF returns over multi-year periods.

Short-term traders. For tactical exposure of a few days to weeks, the roll mechanism matters less. The fund tracks daily spot moves reasonably well in the short term. The drag is a function of holding period.

Producers and hedgers. Commercial users of futures benefit from the same curve shape that hurts ETF investors. A producer can lock in higher future delivery prices today, capturing positive roll yield on short positions.

A common error is reading commodity ETF underperformance as evidence that the underlying commodity is a poor investment. The structural drag is a property of the wrapper, not the commodity itself.

Practical observation

Go deeper

Related questions

• How does backwardation affect commodity investing?
• How do oil inventories affect crude prices?
• Why are leveraged ETFs dangerous long-term?
• What determines oil refining margins?

Frequently asked questions

Is contango always bad for commodity ETF holders?

The drag is structural for futures-tracking products held over multi-month periods, but contango itself reflects rational pricing of carrying costs and is not a market failure. For short tactical trades of a few days, the impact is minimal. For multi-year buy-and-hold, the cumulative cost can exceed the underlying commodity’s spot return. Physical-backed funds (gold, silver) and producer equities offer alternative wrappers without roll exposure, though they carry their own basis risks against pure spot prices.

Why is contango considered the default state rather than backwardation?

Most physical commodities can be stored at positive cost — tanks, silos, warehouses, financing. No-arbitrage logic implies that holding inventory and selling forward must be at least as profitable as immediate sale, which forces the curve upward unless near-term supply is unusually tight. Backwardation requires a specific inversion — buyers willing to pay a premium for immediate delivery — that occurs only during supply stress. Statistically, oil futures curves have been in contango more often than backwardation since 2010.

How does ETF size affect contango?

Research from 2021 documents that the growth of oil ETF assets under management has been associated with wider contango spreads in WTI futures. The mechanism is that predictable monthly rolling by large funds creates systematic buying pressure on the next-month contract, lifting it relative to the front-month. This crowding effect is one reason that the structural drag on funds like USO has been larger than the carrying-cost model alone would predict.

Last updated — 29 May 2026