Commodities

A commodity has no cash flows to discount and no management to judge; a barrel of oil pays no dividend. So the discipline built different machinery: the futures curve as its central object, storage as its central theory, and physical balances as its fundamental research. Professionals do not trade "the oil price"; they trade a curve of prices for delivery at different dates, and the shape of that curve is most of the information.

The futures curve: the object itself

For each commodity there is a strip of futures contracts, one per delivery month, and plotting their prices gives the term structure. Two canonical shapes:

CONTANGO         later > sooner     spot 70, 6-month 74
                 "the market pays you to store it"
                 signature of a well-supplied market

BACKWARDATION    sooner > later     spot 80, 6-month 74
                 "the market pays a premium for barrels NOW"
                 signature of scarcity

The shape matters more than the level because of the roll: a futures position must be rolled from the expiring contract into the next one, and that roll has a yield. In backwardation you roll from an expensive expiring contract into a cheaper later one and harvest the difference; in contango you pay it. Over years, roll yield has dominated the returns of commodity indices, which is why an investor can be right on the direction of spot and still lose money in the future.

The theory of storage: why the curve has its shape

futures ≈ spot + financing cost + storage cost - convenience yield

contango is CAPPED at full carry (finance + storage): above that, buy spot,
store, sell the future, lock the profit. Backwardation has NO cap, because
you cannot short a barrel you cannot borrow; scarcity premia can go vertical.

The convenience yield is the value of physically holding the commodity when it is scarce: a refinery cannot run on a futures contract. When inventories are low, convenience yield spikes and the curve flips backwardated. This makes the curve an inventory gauge you can read without any inventory report, and professionals treat curve flips as regime changes, not noise.

Balances: the fundamental work

Commodity fundamental research is an accounting identity applied forward: supply plus inventory draw must equal demand. Analysts build balance sheets for each market, quarter by quarter: production (OPEC decisions, shale response, mine supply), demand (GDP sensitivity, substitution, seasonality), and inventories as the shock absorber that closes the identity. Price is the negotiator that keeps the identity true: when the balance says inventories head toward empty, price must rise until demand surrenders or supply appears.

• The data is public and slow. Weekly US oil inventories (EIA), gas storage, USDA crop reports, exchange warehouse stocks for metals. Each release is that market's earnings day.
• Seasonality is structural, not superstition. Gas is burned in winter, gasoline in summer, harvests come annually; the balances breathe on a calendar.
• Elasticities are asymmetric and slow. Neither a mine nor a demand pattern can change quickly, which is why commodity shocks overshoot in both directions before supply and demand can answer.

The cost curve: the long-run anchor

Line up every producer from cheapest to most expensive and you have the industry cost curve. In the long run, price gravitates toward the cost of the marginal producer, the one just needed to satisfy demand: above it, new supply gets built; below it, the expensive tail shuts. This is the closest thing commodities have to intrinsic value, and the professional questions are always about the curve's edges: what does the marginal tonne cost today, and what will it cost after the next wave of projects?

Positioning: who is on the other side

The weekly Commitments of Traders report splits the futures open interest into commercial hedgers (producers and consumers laying off risk) and speculators. Extreme speculative positioning is a contrarian flag: when every fund is already long, the marginal buyer is missing and the exit is crowded. Professionals read positioning not as a signal by itself but as a measure of how much of a fundamental story is already in the price.

The special cases: gold, and metals with order books

Gold breaks the framework: barely consumed, almost all of it ever mined still above ground, so balances and cost curves matter little. It trades as a monetary asset, priced principally off real interest rates (the opportunity cost of a yieldless asset) and currency debasement fear, with central bank purchases as the structural bid. Industrial metals sit at the other pole: warehouse inventories, smelter economics, and, for copper especially, a read on global construction and the energy transition.

How views get expressed