Why does a stronger dollar usually weigh on commodity prices?

Almost every globally traded commodity is quoted in US dollars, and one of the most quoted regularities in markets is that a stronger dollar tends to come with weaker commodity prices. The relationship is real. It is also smaller, less reliable and more tangled with causation than the folklore suggests — so it is worth taking apart properly.

Three channels, only one of which is about pricing

The denomination channel. If a commodity's dollar price is unchanged while the dollar strengthens 10% against a buyer's currency, that buyer's cost in their own money has risen 10%. Some demand is lost, and the dollar price must fall a little to clear the market.
The common-driver channel. A strong dollar usually arrives with something — tighter US monetary policy, weaker growth elsewhere, or a flight to safety. Those forces weigh on commodity demand directly. Here the dollar is a symptom sharing a cause, not the cause.
The producer-cost channel. Many producers earn dollars but pay wages, power and taxes in local currency. A stronger dollar lowers their costs in dollar terms, so they can profitably supply at a lower dollar price. The cost floor itself moves down.

Sizing channel 1 honestly

The naive version — "the dollar rose 10%, so commodities should fall 10%" — collapses immediately once you put numbers on it.

Rounded, illustrative: the dollar strengthens 10% against the euro. Oil is at $80.

For a euro-area buyer, an unchanged $80 now costs about 10% more in euros.
At a short-run demand elasticity of −0.2 for that buyer, its consumption falls about 2%.
That buyer is roughly 15% of world demand, so world demand falls 15% × 2% = 0.3%.
At a global elasticity of −0.1, the implied dollar-price fall is 0.3% ÷ 0.1 = ≈ 3%.

A 10% currency move, a 3% price effect through the pricing channel. The mechanism is real and the magnitude is modest — which tells you that when you do observe a large dollar-commodity co-movement, most of it is probably channel 2, the shared driver, not the currency arithmetic.

The reverse-causality caveat

This is the part most commentary skips. Causation runs both ways.

For commodity-exporting economies, the commodity price drives the currency, not the other way around. A rising oil price strengthens an exporter's terms of trade and its exchange rate. Any correlation you measure between "commodities" and "a currency basket" partly contains this.
The United States is itself a very large energy producer, so its own terms of trade — and therefore the dollar — respond to energy prices. Regressing one on the other and calling the coefficient causal is exactly the error your data-literacy training warned about.

And the relationship is unstable

In 2022 the dollar was exceptionally strong and commodity prices were exceptionally high, because a supply shock dominated everything the currency channel could contribute. The dollar link is a conditional tendency that supply shocks routinely overwhelm — not an identity, and not something you can lean on.

The useful posture: use the dollar as one input in a decomposition ("how much of this move is currency, how much is the physical balance?"), never as a predictor.

Try it now

Five years of crude and five years of the euro against the dollar are below. Read the second one upside down: EURUSD is dollars per euro, so the dollar is strong when that line is low. Switch both to Monthly and put your finger on the stretches where crude falls while the FX line falls too — those are the periods the lesson is about.

Interactive line chart: CL.COMM (5Y)

Interactive line chart: EURUSD.FOREX (5Y)

Now find one long stretch where the relationship plainly failed — both rising, or both falling. It will not take long. Write down what else was going on, and note that you have just explained a correlation by leaving it.
Redo the channel-1 arithmetic with your own elasticity assumptions: how large would elasticities have to be for a 10% currency move to justify a 10% price move? Then write the caveat in your own words: a correlation between the dollar and commodities does not tell you which one moved first.