What moves commodity prices

Three quantities, and nothing else 

Every explanation of a commodity price move eventually reduces to one of three quantities: how much is being produced, how much is being consumed, and how much is already held in storage. A geopolitical event matters because it threatens production. A manufacturing survey matters because it implies consumption. An inventory report matters because it measures the buffer between them. This is a useful discipline when reading commentary, because a story that cannot be connected to one of the three is describing sentiment rather than the market's mechanism.

What makes commodity prices behave differently from financial ones is that all three quantities are difficult to change quickly. A mine takes years to permit and build. A refinery cannot be enlarged in a quarter. A field is planted once a season. On the demand side, a smelter that needs metal cannot run without it, a household that needs heat buys it at almost any price, and a food staple is not readily given up. Economists call this inelasticity, and it is the single most important structural fact about the whole asset class.

Why a small change in quantity moves price a long way 

When neither side of a market can adjust quantity, an imbalance has to be resolved entirely by price. The price must move far enough to persuade the marginal consumer to go without, because no other adjustment mechanism is available on the timescale required. That is why a disruption removing a modest share of world supply can produce a price change of a very different order of magnitude, and why the same headline produces a large move in a tight market and almost none in a comfortable one.

The last two rows are the whole reason inventory data is watched more closely than production data. An identical physical event is absorbed by the buffer in one case and threatens to exhaust it in the other, and the market prices the second far more violently than the first. This is also why a market can rise on news of a disruption that never actually removes a single unit of supply: what is being priced is the probability that the buffer is tested.

The dollar channel 

Almost every internationally traded commodity is quoted in US dollars, which means the dollar is an input to the price every buyer outside the dollar area actually faces. When the dollar strengthens against other currencies, the same dollar price is a higher local currency cost, which weighs on demand from those buyers, and it simultaneously raises the local currency revenue of producers whose costs are in a weaker currency, which encourages supply. Both effects push in the same direction, which is why the observed relationship between the dollar and commodity prices is generally inverse.

The relationship is a tendency rather than an identity, and treating it as mechanical is a common error. Both the dollar and commodity prices respond to the same underlying conditions, so the correlation between them varies with what is driving the cycle, and there are extended periods in which they rise together. A currency maintained at a fixed rate against the dollar removes this channel entirely for buyers in that currency, which is the situation across much of the Gulf.

Interest rates, storage and carry 

Holding a physical commodity costs money in three ways: the warehouse or tank has to be paid for, the goods have to be insured, and the capital tied up in them has to be funded. Together those form the cost of carry, and it is the main reason a price for delivery later is normally above a price for delivery now. Interest rates enter directly through the funding component, so a higher rate environment raises the cost of holding inventory and gives every holder an incentive to carry less of it.

There is a second and more contested rate channel that applies to commodities held as stores of value rather than as inputs. An asset producing no income is compared against a safe asset that does, so its relative attractiveness is commonly discussed in terms of the inflation adjusted yield being forgone. The relationship is loose, holds over long horizons rather than short ones, and applies far more clearly to precious metals than to anything consumed industrially.

Energy as an input to everything else 

Energy prices propagate through the whole class because energy is a cost of producing every other commodity. Smelting metal is electricity intensive, fertiliser is manufactured from natural gas, farm machinery and freight run on diesel, and mining and pumping consume fuel directly. A sustained rise in energy costs therefore raises the marginal cost of production across metals and agriculture alike, and it can take high cost capacity offline outright, converting a cost effect into a supply effect.

Some links run in the other direction and are specific rather than general. Sugarcane can be processed into either sugar or ethanol, so a fuel price change reallocates sugar supply. Corn is fermented into fuel ethanol, so a fuel policy change alters food and feed demand. Those connections mean a commodity can move sharply on news from a market it has no obvious relationship with.

Weather, geography and disruption 

Production of most commodities is concentrated in a small number of places, and that concentration turns local events into global prices. A drought in one growing region, a frost in one state, an outage at one smelter, a labour dispute at one producer or a disruption to one shipping strait can each remove a large share of world supply with no substitute available on the same timescale. The concentration also applies to routes: a significant share of seaborne crude and of grain exports passes through a handful of narrow waterways, and the availability of those routes is a supply input in its own right.

Two properties of disruption driven moves are worth stating because they are frequently misread. They price a probability rather than a realised loss, so a move can reverse fully when the feared disruption does not occur, without anything having been wrong about the original move. And they are asymmetric in speed: supply is lost in an instant and restored over months, so the rise is typically faster than the subsequent decline.

Policy, which is not a residual 

Government action is a first order driver in this asset class rather than a footnote. Production quotas agreed between exporting states set supply directly. Export bans imposed by a food producing country during a domestic shortage remove supply from the world market while increasing it at home. Tariffs redirect trade flows and create regional price differences that persist. Blending mandates create demand for a fuel by statute. Strategic reserves can be released to add supply or refilled to add demand, on a decision rather than on a price.

Environmental and permitting regimes act more slowly and more powerfully, because they determine which mines and fields are developed at all, and therefore the supply available years later. A market that appears comfortable can be carrying a structural shortfall created by investment decisions not taken a decade earlier, and that is a condition rather than an event, so it does not show up on any calendar.

Flows that are not fundamental at all 

A meaningful share of commodity market activity has no view about the commodity. Broad commodity index products hold a basket of contracts and roll them on a published schedule, so a mechanical flow arrives at a known time regardless of conditions. Multi asset portfolios rebalance to fixed weights, buying what has fallen and selling what has risen for reasons internal to the portfolio. Producers hedge on programmes set by a board rather than by a price view. These flows are real, observable and unrelated to production or consumption in the week they occur.

Positioning data is the standard way this is observed rather than guessed at. Regulators in the major jurisdictions publish a weekly breakdown of open positions classified by participant type, and the convention is to read extremes of concentration as a description of how crowded a position has become, not as a signal. The data is published with a lag and the classifications are imperfect, so it is a bounded convention with known limits rather than a measure of anything precise.

Seasonality, and its limits 

Many commodities have genuine seasonal patterns, because the underlying physical activity is seasonal: heating demand in winter, driving in summer, planting and harvest at fixed points, construction activity in the warmer months. Those patterns are real in the physical market and they are also known to everyone, which means the futures curve already reflects them. A pattern that is visible in a chart of historical prices is not necessarily available in a curve that has priced it, and that distinction is the standard caution attached to any seasonal statistic.