Two slopes, one question: is the barrel short today?
An oil future does not have one price, it has one per delivery month. Lined up from nearest to furthest, those prices form the curve. Its slope says more than its level.
- Contango: the deferred months trade above the front. The barrel is plentiful today, and the market pays someone to hold it.
- Backwardation: the front trades above the deferred months. The barrel is short now, and the market pays a premium to have it immediately.
A desk reads the slope as spreads: the first month minus the second (M1 − M2), the first minus the sixth (M1 − M6). Positive is backwardation. Negative is contango.
Contango has a ceiling, backwardation does not
Contango is capped by carry. If December trades $1.50 over November and storing a barrel for a month costs $0.60, financing included, anyone with a tank buys the barrel, stores it, sells December and locks in $0.90. That cash-and-carry fills the tanks until the spread falls back to the cost of storage. A contango pinned to that ceiling, full carry, says inventories are overflowing.
Backwardation has no bound. No arbitrage delivers today a barrel that will only exist in six months. When physical supply is short, the front rises as far as buyers will pay, and inventories draw to close the gap.
The extreme case has a date. On 20 April 2020, the May WTI contract settled at −$37.63: the tanks at Cushing were close to full, and taking delivery of a barrel cost more than the barrel itself.
What the slope costs a long: the roll
A future expires. Staying long means selling the expiring month and buying the next one, every month. That roll is paid at the spread.
In backwardation, the long sells high and buys back lower: the roll earns, with the curve unchanged. In contango, it sells low and buys back higher: the roll costs, even if spot does not move. A contango of $0.80 a month on a $70 barrel costs a rolling long close to 14% a year. That is why an oil index fund lags spot in a well-supplied market.
How the desk reads the curve
On the Oil Desk, the desk follows the WTI and Brent curves contract by contract, and records every spread at the close to place it in its own range. Five readings come out of it.
M1 − M2 for immediate tightness. WTI delivers at Cushing, Oklahoma. The prompt spread reacts before anything else when physical supply there tightens or loosens. It is read next to Cushing stocks against their five-year season, published every week by the EIA. An M1 − M2 that steepens while Cushing drops below its season is tightness confirming itself. The same spread rising on stocks above the season is fragile.
M1 − M6 for the regime. The first month against the sixth gives the regime without the noise of expiry. The useful question is not “backwardation or contango” but “since when, and which way”: the spread is compared with where it stood a week, a month and three months ago.
Front against long end: where the tightness sits. Backwardation packed into the first six months, with a flat curve beyond M12, says spot tightness: a delivery problem, not a lasting deficit. Backwardation that runs to M12 and beyond says the market is pricing scarcity for longer. The first often resolves fast. The second changes drilling budgets.
Dec − Dec for the year ahead. This December against next December. It is where producers hedge their output and funds express a one-year view, away from the accidents of the monthly expiry.
Brent against WTI: the world or the US. A Brent M1 − M2 in deep backwardation while WTI stays flat says the Atlantic basin is tight in a way US crude does not share. And a widening Brent − WTI spread makes US exports to Europe pay, which eventually closes it.
Three traps
The expiring contract. The nearest contract trades until its last day, but volume and open interest leave it a week or more before. On 19 October 2026, November WTI had one session left and 63,000 contracts of open interest, against 311,000 in December. An M1 − M2 computed on the expiring contract measures an empty market. The desk takes as its front the first contract whose open interest the next one has not overtaken.
The slope is not a forecast. A six-month future below spot does not say the price will fall. It says what the barrel delivered in six months is worth today against the barrel available now.
The curve and flat price do not always move together. A price falling while backwardation steepens is often a sell-off driven by financial flows, not by the barrels. The reverse, a price rising on a curve that eases, is a rally the physical market does not confirm. That is what the curve is for: it says whether the price move has the barrels behind it.