Why do traders watch the front spread more closely than the price?

Because the flat price is contaminated and the spread is not.

A commodity's outright price absorbs everything: the dollar, interest rates, risk appetite, index flows, macro positioning, a headline about a central bank. Very little of that has anything to do with whether there are enough barrels in the right place this month.

A calendar spread — the difference between two contract months on the same commodity — cancels most of it. Both legs are the same thing, exposed to the same macro. What survives the subtraction is close to a pure read on the physical balance between those two dates.

The same price, two different worlds

Crude at $78.00 in both scenarios:

• Scenario A. M1 $78.00, M2 $78.55 → prompt spread −$0.55 (contango). Annualised carry ≈ 8.5%, comfortably above financing. Storage is being paid for; the market is adequately supplied.
• Scenario B. M1 $78.00, M2 $77.20 → prompt spread +$0.80 (backwardation). Nobody is being paid to store anything; someone is paying up for prompt barrels. The market is tight.

The flat price is identical. Everything that differs is the spread, and only the spread said anything at all about the physical market. This is why a trading desk's screen leads with spreads and a news bulletin leads with the outright.

Reading the prompt spread

Using the tools from Units 1 and 2, in order of tightness:

• At or near full carry → inventories comfortable, storage economics binding, cash-and-carry active.
• Shallow contango, well under full carry → normal; convenience yield is real but modest.
• Flat → transitional. Convenience yield roughly equals carry.
• Backwardated → inventories drawing; the physical market is bidding for prompt supply.
• Steeply backwardated → operational tightness. Expect physical differentials to widen and inventories to keep falling.

Because convenience yield is a convex function of inventory, that last category arrives faster than intuition expects. Shallow backwardation can become steep on a modest further draw.

Reading the shape across the whole curve

The front tells you about now. The back tells you about structure.

• Backwardated front, flat back — tight today, expected to normalise. The most common tight-market shape.
• Contangoed front, contangoed back at full carry — well supplied throughout, storage priced everywhere.
• Steep contango at the front, flattening quickly — an acute storage problem with a short expected life. This is the super-contango signature.
• Contangoed front, backwardated back — unusual. Ample supply now, and something in the long-dated part of the curve pricing a structural constraint later.

Two cautions on the back end, both important. Deferred contracts are thinly traded, so a single quote two years out carries less information than a front-month print. And the back of the curve is where the risk premium from Unit 2 lives most heavily — the part you cannot observe or separate.

The discipline

A spread tells you what the market is currently paying for storage and for prompt delivery. It tells you the physical balance as the people transacting in it see it right now.

It does not tell you what the price will do. The mixing problem never goes away: carry, convenience yield and risk premium remain three unknowns in one equation, and the last two remain unobservable no matter how carefully you read the first. Description is available; prediction is not on offer, here or anywhere in this Academy.

Clean as information is not the same as safe as a position. Exchanges margin a calendar spread at a fraction of two outrights because the legs usually move together, so a spread book is large in barrels for little margin. In the delivery month the front leg detaches from the rest of the curve, because only the front leg carries the delivery obligation. On 20 April 2020 the expiring May WTI contract settled at −$37.63 and the June contract at $20.43 (CME Group settlements, as recorded in the CFTC staff report of November 2020), a prompt spread of −$58.06 in a market whose spread had been a few dollars wide weeks earlier. The spread is the best read of inventories on the board and the worst place to be large near expiry.

Try it now

1. For one commodity, record the prompt spread (M1 − M2) once a week for a month from the exchange's settlement page, alongside the published inventory statistic for the same weeks — for crude, the US Energy Information Administration's Weekly Petroleum Status Report, free on eia.gov every Wednesday.
2. The flat price for those same weeks is below. Measure it week by week and write the four numbers in a third column beside your spread and your inventory readings.

3. Answer two questions: which of the two market series was steadier, and which of them changed sign or direction when inventories changed? Then write your conclusion as a description — never as a forecast.