Why do industrial metals turn before the economy does?

Commodity demand is derived demand — nobody wants copper, they want wiring in a building that somebody wants. So a commodity's demand cycle is the cycle of whatever it goes into, amplified by the supply chain that carries it.

A hierarchy of cyclicality

The three commodity groups sit at very different points on the cyclical scale:

• Industrial metals — the most cyclical. Copper, aluminium, zinc, iron ore go into construction, machinery, vehicles and the electrical grid: the most deferrable, most credit-sensitive, most interest-rate-sensitive parts of GDP.
• Energy — moderately cyclical. Transport fuel and power demand fall in a recession, but heating, lighting and basic mobility are hard to defer, so the swings are smaller in percentage terms.
• Agriculture — least cyclical on the demand side. Calorie demand barely notices a recession; agricultural volatility comes overwhelmingly from the supply side, via weather. Diet composition shifts with income over decades, not quarters.

This maps directly onto the cyclical-versus-defensive distinction you met in Fundamental Analysis — the same logic, applied to physical inputs rather than to sectors.

The restocking amplifier

Here is why metals often move before and more than the economy they serve. Every step in a supply chain holds inventory, and when final demand wobbles, each step adjusts both its purchases and its stock — a phenomenon usually called the bullwhip effect.

Worked example, rounded illustrative numbers:

• Final demand for a product falls 5%.
• The distributor, expecting weaker sales, also decides to hold two weeks less inventory. Its orders to the factory fall 15%.
• The factory, seeing orders down 15%, cuts its own raw-material purchases and runs down its metal stock. Its metal buying falls 25%.

A 5% consumer wobble has become a 25% collapse in metal orders, and it happens early, because the destocking decision comes before the sales decline shows up in official statistics. Then the whole thing reverses on the way up: restocking makes the recovery in orders look far stronger than the recovery in end demand.

What to actually watch

The observable leading indicators are the ones tied to the deferrable, credit-sensitive activity above: manufacturing survey indices, new orders components, housing starts and permits, vehicle production, and the credit conditions that fund all of it.

A caution on the folklore. "Dr Copper has a PhD in economics" is a good story with a mixed record: copper is also driven by mine outages, one country's property cycle (next lesson) and inventory games that have nothing to do with global growth. Treat it as one noisy signal among several, and check what is actually moving before attributing it to the world economy.

Try it now

1. Copper futures are drawn below, and under them US economic growth, one figure per year. Take the ten years to August 2026 on the chart and set them beside the growth years. Which one turns first? Note that setting one number a year against a daily price is a choice you are making about how to line them up, not one the data made.

2. Repeat with an agricultural commodity, corn, below, against the same growth years. The relationship should be visibly weaker — that difference is the hierarchy above.

3. Find one period where the metal moved and the macro series did not. Before calling it a leading indicator, go and check whether a mine outage explains it.