A crude oil futures ETF's twelve-month result is not decided by the direction of oil alone. On July 29, 2026, WTI September futures settled at $84.46 a barrel while December settled at $77.74 (Newspim). The near contract is 8.64% more expensive than the one three months out — backwardation — and if that slope simply held, the fund would earn roughly 2.80% a month without oil moving at all. Flip the curve into contango and the same mechanism cuts the other way. That is why the complaint "oil rallied but my account didn't" shows up in oil ETFs and almost nowhere else.
How the roll cost is calculated
Futures expire. An oil futures ETF sells the contract it holds before expiry and buys the next month — that swap is the roll. When the two prices differ, the same money buys a different number of contracts. In contango, where the deferred month is dearer, the fund buys fewer contracts; in backwardation it buys more. Net asset value moves even when the spot price does not.
The arithmetic is simple: take the ratio of the two prices, spread it over the number of months between them, then compound. September and December are three months apart, so the monthly roll return is (84.46 ÷ 77.74) raised to the one-third power, minus one — plus 2.80%. Compounded over twelve months that is 39.3% a year. Reverse the sign and you have the cost of contango.
| Curve shape | Monthly roll return (%) | Annualized (%) | 100 invested, one year later |
|---|---|---|---|
| Backwardation (curve observed 2026-07-29) | +2.80 | +39.3 | 139.3 |
| Mild contango (1% a month) | -1.00 | -11.4 | 88.6 |
| Steep contango (3% a month) | -3.00 | -30.6 | 69.4 |
| 2x leverage × 1% monthly contango | -2.00 | -21.5 | 78.5 |
These figures strip out everything but the curve, assuming oil sits perfectly still for a year. Real returns add price moves, management fees and currency hedging costs on top. In May 2020 Financial News pointed out that leveraged products hold twice the futures exposure, so the roll cost doubles as well. That is the last row of the table: even a mild 1% monthly contango becomes a 21.5% annual hole in a 2x product. It accumulates separately from the negative compounding of leveraged ETFs. The same report noted that the premium of WTI crude ETNs to their net asset value stretched as wide as 61% at the time.
Why the ETF lagged a rising oil price
The long-run scoreboard answers this directly. As compiled by Newspim, USO — the largest US oil futures ETF — returned about 5% over ten years against roughly 18% for spot WTI. Over five years the gap is 65% versus 20%. Annualizing makes the size of the leak clear.
| Period | Spot WTI cumulative (%) | USO cumulative (%) | Spot annualized (%) | USO annualized (%) | Annual gap (%p) |
|---|---|---|---|---|---|
| 10 years | 18 | 5 | 1.67 | 0.49 | 1.18 |
| 5 years | 65 | 20 | 10.53 | 3.71 | 6.82 |
The ten-year gap looks smaller than the five-year gap because of what falls inside each window, not because the drag eased. The worst contango episodes cluster in the last five years, and the same report put the monthly roll cost during contango at roughly 1%. It also noted USO's structure: front-month heavy, with 30% in the second month and 15% in the third. The closer a fund sits to the front of the curve, the more the roll matters.
The extreme case is 2020. On April 20 the WTI May contract printed minus $37.63 a barrel, and USO was already down 75% for the year at that point. Storage had run out, so only the front contract collapsed — and front-heavy products took the worst of it.
Which way is the curve pointing now
2026 has run the other direction. Since the Middle East conflict began in late February, WTI was up about 26% through July 29 while USO gained about 58%. Strip out the spot move and the excess is 1.58 ÷ 1.26 = 1.254, or 25.4 percentage points. Spread over five months that works out to roughly 4.6% a month of roll gain. The curve measured on July 29, by contrast, yields 2.80%. Backwardation is still intact, but the slope is flatter than it was in spring.
April showed the same pattern. As of April 10, USO was up 80.5% year to date against 66.2% for spot WTI, and oil peaked at $112.95 a barrel on April 7. In both snapshots, an inverted curve left the futures ETF ahead of the physical commodity.
Backwardation is a premium paid for a supply accident. When the accident is cleaned up the premium goes with it, and the same fund starts eating returns from the same position.
When this arithmetic breaks
Every number above assumes the curve keeps its shape. It gets redrawn at every roll, so three things break the calculation. First, if the geopolitical risk clears, the front-month premium deflates first and backwardation flips to contango — oil can drift higher while the ETF goes sideways or backward. Second, a surge in inventories raises storage costs and steepens contango, which is what 2020 was. Third, Korea-listed products differ in currency hedging and in how they track the curve, so identical oil prices produce different results. A fund that holds only the front month behaves differently from one spread across maturities, which is the same place tracking error and premium/discount open up.
What to watch
- The spread between the front month and the contract three months out — the moment the front month becomes the cheaper one is the flip into contango. Take the cube root of the ratio and you have the monthly roll return on the spot
- How fast the slope is changing — even with backwardation intact, moving from 4.6% to 2.80% is the premium draining away
- Each fund's maturity mix — front-month concentration versus spread across maturities produces different outcomes on the same curve
- Premium to net asset value — market price versus NAV. It reached 61% in 2020
- Leverage multiple — the roll cost is multiplied too. In a 2x product, 1% monthly contango is 21.5% a year
- The hedging label — whether a Korean listing carries the (H) suffix decides whether the won-dollar rate lands on top of your return