Futures Roll and Roll Yield

A stock can be held forever. A futures contract cannot — it expires, and to maintain the exposure you must close the expiring month and open the next one. That mechanical necessity is called the roll, and if you hold commodity exposure through futures or through a fund that does, the roll rather than the spot price will dominate your long-run return. This is the single most expensive thing most investors do not know about commodities.

The operational half. Expiration cycles run monthly or quarterly depending on the product, so learn the cycle for anything you trade. As expiration approaches you have three choices: close the position, let it expire (only sane if cash-settled), or roll it forward.

Physically settled products must be closed on most retail brokerages, which do not permit delivery. Two dates govern:

First Notice Day

The first day the exchange can assign delivery to accounts that are long futures contracts.

Last Trading Day

The last day a futures contract may trade or be closed before delivery.

Close two business days before First Notice Day or one business day before Last Trading Day, whichever comes first. Miss both and the brokerage closes the position for you, at its convenience rather than yours.

The expensive half. The contract you sell and the contract you buy trade at different prices, and the sign of that difference is not random — it is set by the shape of the futures curve described in section “Futures Pricing”. The resulting gain or loss is roll yield:

Contango: the roll costs you

The deferred contract is more expensive than the expiring one. You sell low and buy high, every single roll, and you do it whether or not spot prices moved. A market in persistent contango bleeds a long position by the carry — storage, insurance, and financing — because that is precisely what the curve is pricing.

Backwardation: the roll pays you

The deferred contract is cheaper. You sell high and buy low, harvesting a positive carry. Keynes called the tendency toward this shape normal backwardation and attributed it to producers paying speculators to take price risk off their hands. It is the return the long-commodity investor is theoretically compensated for, and it is not always present.

Compounded over years, this dwarfs spot moves. Energy markets have spent long stretches in contango, and funds mechanically rolling front-month contracts through those stretches have lost enormous sums while the underlying commodity went nowhere. USO is the cautionary case: in April 2020, front-month WTI settled below zero, the fund was forced out of the front month into far more expensive deferred contracts, and it amended its stated strategy mid-crisis to spread holdings across the curve. Holders who bought “oil” near the bottom owned an instrument that could not, by construction, capture the recovery in spot.

What to do about it. If you want commodity exposure, choose the roll methodology deliberately, because it is the product’s most important design decision:

Front-month indices

Simple, transparent, and maximally exposed to contango. Avoid for long-term holdings.

Optimized-roll indices

The fund selects the contract along the curve with the most favorable implied roll rather than mechanically taking the next month. DBC works this way.

Constant-maturity or laddered indices

Exposure spread across several maturities, diluting any single roll. CMCI works this way.

Then check the tax wrapper, which is a separate decision from the roll methodology — and the two common structures give you genuinely different tax outcomes, not merely different paperwork.

Partnership (K-1)

DBC and USO hold the futures directly and pass everything through to you. The contracts are § 1256 contracts under IRC §1256(b)(1)(A), so they are marked to market on the last business day of the year under IRC §1256(a)(1) and taxed 60/40 under IRC §1256(a)(3) regardless of holding period. You owe tax on unrealized gains, and the K-1 arrives late enough to force an extension in a bad year. You get the favorable 26.8% blended rate in exchange for the annual mark and the filing complexity.

RIC with a Cayman subsidiary (1099)

BCI runs the futures inside a controlled foreign subsidiary and says so in its name. Here the § 1256 treatment stops at the entity: the fund marks to market and computes 60/40 internally, but you hold ordinary fund shares. What reaches your return is distributions (ordinary or qualified dividends) and capital gain or loss when you sell the shares — no 60/40 passthrough, and no tax on unrealized gains you have not received. You trade the preferential rate for a 1099 and control over your own timing.

Neither is strictly better. Take the K-1 partnership in a taxable account if you value the 26.8% blended rate and can live with phantom income; take the 1099 wrapper in a taxable account if you want to control realization timing, and in a retirement account always, where the 60/40 rate is worth nothing anyway. Plan the location accordingly (section “Assigning Assets into Tax Buckets” and section “Section 1256 Contracts and the 60/40 Regime”).