Why can a commodity fund and the commodity itself tell completely different stories?
Because they are not measuring the same thing, and they never were. This lesson is mechanics — how a tracking gap is generated, arithmetically — and it is not a warning about any product or a suggestion to hold or avoid one.
What a futures-based fund actually owns
A fund offering "exposure to oil" without owning barrels holds futures contracts plus collateral. Its net asset value is therefore:
NAV = value of the contracts held + collateral − liabilities
Read that carefully and notice what is missing. The spot price does not appear. It influences NAV only indirectly, through convergence pulling the held contracts toward it.
So the fund tracks the contracts it holds, faithfully and by design. The commodity's spot price is a different series, and the accumulated difference between them is the roll return from the last lesson.
The decomposition, with numbers
Take the contangoed year: spot flat at $80.00, monthly roll return −1.84%, expenses 0.75% a year, collateral earning 5.0% a year.
- Spot return: 0.0%
- Roll return: (1 − 0.0184)¹² = 0.800 → −20.0%
- Collateral return: +5.0%
- Fees: −0.75%
Combined: 0.800 × 1.05 × 0.9925 = 0.834 → roughly −16.6% for the year.
The commodity did nothing. The fund is down about a sixth. And critically, the fee is the smallest term in the equation by an order of magnitude — 0.75% against a 20% roll effect. Anyone diagnosing a tracking gap by reading the expense ratio is looking at the wrong line.
The documented 2020 case
In the first half of 2020, the largest US crude-oil exchange-traded product held predominantly the front-month WTI contract, rolling it monthly, exactly as its prospectus described.
Then two things happened at once. Demand collapsed and storage filled, so the front of the curve went into extreme contango — at points the second month traded several dollars above the first, which on the arithmetic above is a roll cost of many percent in a single month. And simultaneously the fund's assets grew very rapidly as buyers arrived, to the point where the fund held a substantial fraction of the entire open interest in the contract it was holding.
Around the negative settlement of 20 April 2020, and under exchange position limits and broker restrictions, the manager changed what the fund held: first moving out of the front month, then spreading holdings across multiple contracts further out the curve. Each change was disclosed in regulatory filings as it was made.
The result over that year: front-month crude fell sharply and then recovered a large part of the fall, while the fund did not recover to a comparable degree — because the contracts it had held across that period were rolled repeatedly at very wide contango, and each of those rolls was paid by convergence.
Nothing here was a malfunction. The product held futures and rolled them, which is what it said it would do. The gap was the curve arriving on schedule.
The literacy point
"Exposure to oil" is not a well-defined phrase. The well-defined questions are:
- Which contracts? Front month, a fixed deferred month, or a ladder?
- Rolled how, and when?
- What is the collateral earning?
- How large is the fund relative to the contract it holds?
Answer those and the tracking behaviour is predictable. Skip them and it will look mysterious.
The metals contrast
Precious-metal products often work differently: they hold the physical metal in a vault and charge a storage fee. No futures, no roll, no roll return — just a small, steady drag.
That is not a judgement about which design is better. It is a structural fact about which commodities permit it. Gold is dense, non-perishable and cheap to store relative to its value. Crude oil, natural gas and wheat are none of those things, so a physically-backed structure is impractical and futures are the only available route — which means the roll is unavoidable, not a design flaw.
Try it now
- The commodity and a futures-based fund on it are below. Measure both from the same date to today, and subtract one total return from the other. The gap is roll return plus fees; decide for yourself which of the two is doing most of the work before you look anything up.
- Now find the fund's published roll methodology (prospectus or factsheet) and write down which contracts it holds and when it rolls. Everything you measured in step 1 is a consequence of those two sentences.
- Recall the physically backed metal wrapper from the foundations course, where the gap shrank to roughly the storage fee. Say in one sentence why that product has no roll return to lose, and this one cannot avoid having it.