Why does it take a decade to answer a price signal?
Every market economy runs on one loop: price rises, supply responds, price falls. In commodities that loop still works — it just runs on geological time. The lead time between the signal and the barrels is the most underappreciated number in the entire asset class.
The response-time ladder
Rough, rounded orders of magnitude from decision to first output:
- Grains and softs — one season. Farmers respond to price at planting. The fastest supply response in agriculture, and the reason agricultural prices are dominated by weather rather than capacity.
- US shale oil — weeks to months. A well can be drilled and completed quickly and produces almost immediately. This is genuinely short-cycle supply — and it declines steeply, often 60-70% in the first year, so it must be constantly redrilled just to stand still.
- Conventional and offshore oil — five to ten years. Appraisal, engineering, platforms, pipelines.
- LNG export terminals — four to six years, and enormous fixed cost.
- Copper and other hard-rock mines — a decade or more. S&P Global's 2023 study of 127 mines put the average from discovery to first production at 15.7 years, and its 2024 update found 17.9 years for mines that started up in 2020–2023, counting exploration, permitting, financing and construction.
A market with only long-cycle supply is a market where nothing can arrive in time to fix today's problem.
The sentence that explains everything downstream
The price that justifies an investment is never the price the project sells into.
A copper project approved in a shortage will produce its first metal into a market a decade later, whose balance nobody can know. The company is not being reckless; it is being asked to make a fifteen-year bet using today's price as the only evidence available.
A worked example
A metal spikes in year 0. Three companies read the signal correctly and sanction expansions.
- Year 0: price at $10,000/t against a marginal cost near $6,000. Investment looks obviously profitable.
- Years 1-7: permitting, financing, construction. No new metal reaches the market. The shortage is unrelieved, so the price stays high, which attracts more projects.
- Year 8: all three expansions start up within eighteen months of each other, adding 9% to global supply over a window in which demand grows about 3%.
- Year 9: the price is $5,500 — below the cost that was supposed to be the floor.
Every individual decision was defensible. The aggregate was an overshoot, because none of the three could see the other two, and all three were reading the same price signal at the same moment. That collective failure has a name, and it is the next lesson.
Try it now
- Rank five commodities you can name by how fast their supply can respond — season, months, years, decade. The ranking tells you which ones will show the sharpest spikes.
- Corn and copper over their full histories are below, both best read on Monthly bars. One shows frequent, short-lived spikes; the other shows long multi-year waves. Decide which is which before you read on, then check your answer against the lead times in the table above.
- Measure the longest single up-move on each chart and compare the bar counts. Then write the sentence: the price that sanctions a project is not the price it sells into.