Insights
What Is Roll Yield?
Alphanume Team · May 28, 2026
The hidden return (or drag) from rolling contracts.
Roll yield is the return component a futures investor earns — or loses — purely from rolling an expiring contract into the next one, independent of any move in the underlying spot price. It is not theoretical. Over a decade of holding a commodity ETF, roll yield often swamps the contribution of spot price changes entirely, silently compounding into a gap of tens of percentage points between what the commodity did and what the fund returned. Understanding roll yield is as fundamental to futures investing as understanding the futures pricing calculator you would use to value the contracts themselves. This post builds the concept from first principles, works through a concrete roll example, and explains how the curve shape — contango and backwardation — determines whether roll yield is an asset or a liability.
What roll yield actually measures
A futures contract has a fixed expiration date. A fund or trader that wants continuous exposure cannot hold it to delivery — they must sell the expiring contract and buy the next one. That transaction is the roll. If the two contracts are priced identically, the roll is neutral. They almost never are.
Roll yield captures the price difference between the contract being sold and the contract being bought, expressed as a return on the position. The formal definition is:
Roll yield ≈ (Fnear − Ffar) / Fnear
where Fnear is the price of the expiring (near-month) contract and Ffar is the price of the next contract. A positive number means you are selling high and buying low — a gain. A negative number means you are selling low and buying high — a drag. That drag is sometimes called roll cost or negative roll yield, and it is the default condition in most commodity markets most of the time.
The curve shape determines the sign
Whether roll yield is positive or negative depends entirely on the shape of the futures curve, which is driven by cost of carry — storage costs, convenience yield, financing, and seasonal supply-demand dynamics.
Contango describes a market where the futures curve slopes upward: the December contract is more expensive than the October contract, which is more expensive than spot. When you roll forward in contango, you sell the cheaper near contract and buy the more expensive far contract. You receive fewer barrels' worth of exposure per dollar spent. Roll yield is negative.
Backwardation describes the opposite: the curve slopes downward, and near-month contracts trade at a premium to deferred ones. Rolling forward means selling high and buying low. Roll yield is positive — you accumulate more exposure for the same notional, effectively being paid to maintain the position.
A few reference points for how large these effects can be:
- Crude oil spent much of 2009–2016 in steep contango, imposing roll costs of 5–10% per year on long-only ETFs.
- Copper in backwardation has at times generated roll yields exceeding 8% annualised for long holders.
- Natural gas markets routinely exhibit seasonal contango in summer and backwardation heading into winter, cycling through both regimes within a single calendar year.
A worked roll example
Suppose you hold 100 WTI crude oil futures contracts (each representing 1,000 barrels) through expiration. The current prices are:
| Contract | Price ($/bbl) |
|---|---|
| Oct (near, expiring) | $78.00 |
| Nov (far, next month) | $79.56 |
The roll: sell 100 Oct contracts at $78.00, buy 100 Nov contracts at $79.56. The roll cost per barrel is $79.56 − $78.00 = $1.56. Across 100,000 barrels, the immediate dollar cost of rolling is $156,000 — paid without the spot price moving a cent. As a percentage of the near-month value: $1.56 / $78.00 = 2.0% in a single monthly roll. Annualised across twelve rolls, even if the contango premium remained constant, that would compound to roughly 21.5% of drag — (1 − 0.02)12 ≈ 0.785, meaning the position retains only 78.5 cents of every dollar in futures exposure.
Now run the same example in backwardation. If Nov trades at $76.44 instead, you sell Oct at $78.00 and buy Nov at $76.44. The roll gain per barrel is $1.56, the percentage gain is +2.0%, and the annualised tailwind compounds to roughly +26.8% — (1.02)12 ≈ 1.268.
How roll yield compounds across many rolls
Roll yield is not additive — it compounds. Each time you roll in a contango market, you receive slightly fewer contracts' worth of exposure than you started with (or equivalently, the same number of contracts represents a larger paper loss). That attrition multiplies across rolls:
Terminal exposure multiple = (1 + rroll)n
where rroll is the per-roll yield (positive in backwardation, negative in contango) and n is the number of rolls. Over three years of monthly rolls with a −1.5% per-roll drag — not unusual in energy markets — the compounded loss is (0.985)36 ≈ 0.578: the fund has shed 42% of its real exposure even if spot is unchanged. This is the mechanism behind the persistent underperformance of many commodity ETFs versus naive spot benchmarks. It is also why the S&P GSCI and Bloomberg Commodity Index total-return series diverge so dramatically from their excess-return (spot-only) counterparts over long horizons.
Practical implications: contract selection and constant-maturity strategies
Recognising roll yield as a distinct return driver changes how you approach futures construction:
- Contract month selection. Rolling into a contract two or three months out rather than one month out can reduce roll cost if the curve steepens in the near months — a common feature in energy markets during inventory builds. The trade-off is less liquidity and a wider bid-ask spread.
- Curve-optimised rolls. Some commodity indices select the contract with the most favourable roll yield across the available maturities each month rather than mechanically rolling to the prompt contract. This is the logic behind enhanced or optimised-roll indices.
- Constant-maturity strategies. A constant-maturity approach maintains a fixed weighted-average duration — say, three months — by holding a blend of the two surrounding contracts. The roll is continuous rather than discrete, which smooths the yield but does not eliminate it; the average roll yield is still a function of the curve shape over the two-to-three month segment.
- Regime awareness. Roll yield is not static. Monitoring the spread between near and far contracts — and whether it is widening or compressing — gives real-time signal about whether the roll environment is improving or deteriorating.
The bottom line is that roll yield is not a rounding error. In commodity futures, it is often the single largest determinant of long-run returns, outweighing spot price movements over holding periods of two years or more. Treating it as incidental — focusing only on where spot is headed — is the most common and most expensive mistake in commodity investing.