Commodity Seasonality and the Annual Curve Shape
Many commodities are produced or consumed on an annual clock — one harvest a year, one winter heating season — and that rhythm gets baked directly into the futures curve as a repeating seasonal pattern, not just into the spot price.
Prerequisites: Commodity Futures Basics, Contango and Backwardation
Wheat is harvested once a year, not continuously. Natural gas is burned much more heavily in January than in June. These are not subtle statistical patterns buried in years of data — they are calendar facts everyone in the market already knows, and because everyone knows them, they show up directly in the shape of the futures curve today, not just in how the spot price later turns out to move.
Commodity seasonality is not something you have to forecast from historical spot prices — it is already priced into the relative levels of different futures contract months today, because storage costs and known production or consumption cycles pin specific calendar months to predictably higher or lower prices relative to their neighbors.
Why the curve, not just the price, has a season
Take natural gas. Demand spikes every winter for heating, and supply is roughly steady year-round, so storage has to be built up over the summer to be drawn down each winter. That means winter-month futures contracts (say, January and February) trade at a persistent premium to shoulder-season contracts (April, October) every single year, regardless of what the overall price level is. The curve does not just drift up gradually into winter — it has visible peaks at the winter contract months and troughs at the shoulder months, a shape that repeats year after year on top of whatever the general price trend is doing.
Grains show the mirror image: prices are typically lowest right after a harvest, when storage is fullest and supply is most abundant, and rise into the following spring as stocks are drawn down before the next harvest arrives — the "new crop" contract just after harvest trading at a seasonal discount to the "old crop" contract from the prior cycle.
Worked example
Suppose October natural gas futures trade at $2.80 and the following January contract trades at $3.40 — a $0.60 seasonal premium reflecting expected winter heating demand, priced in months in advance. A trader who believes that seasonal spread is wider than it should be given current storage inventories (perhaps storage is unusually full heading into autumn) can sell the January contract and buy the October contract, betting the spread narrows, without taking any view at all on whether natural gas prices overall rise or fall — a pure seasonal spread trade isolated from the outright price level.
What this means in practice
Because the seasonal shape is already known and priced in, trading seasonality profitably is not about noticing "gas is expensive in winter" — everyone knows that — but about having a view on whether the size of this year's seasonal premium is too large or too small relative to current fundamentals like storage levels, weather forecasts, or crop conditions compared to the historical seasonal pattern.
A recurring historical seasonal pattern is not a guarantee — an unusual event (a mild winter, a drought-shortened crop, a pipeline outage) can override or invert the typical seasonal shape in any given year. Treat historical seasonality as the market's well-known starting assumption, not as a standalone trading signal that ignores the current year's actual fundamentals.
Related concepts
Practice in interviews
Further reading
- Geman, Commodities and Commodity Derivatives (ch. on seasonality)