R14 · free executive summary
Copper and the Grid Metals Underneath R1–R4
Executive summary · free edition · September 2026
The full report runs to thirteen sections with nine markers we re-score every quarter. This summary is complete in itself and contains the finding. Nothing is held back to make you buy the report.
We commissioned this report as Fifteen Years From a Hole in the Ground: measure the interval from discovery to first production for copper projects, and show it runs to about fifteen years.
It cannot be measured from filings. Twenty-four projects here carry at least one verified dated milestone. Four state a discovery year. Two of those four have produced copper. A mining company's filed clock starts at a feasibility study or a board vote, not at the drillhole.
The number everyone quotes — roughly 29 years from discovery to production in the United States — is S&P Global's, a commercial estimate. We name it and use it nowhere. The US Geological Survey publishes no lead-time figure of its own: we swept the copper pages of Mineral Commodity Summaries 2024, 2025 and 2026, the full 2026 volume of 14,766 lines, and the Minerals Yearbook copper chapters for 2003 through 2022, across nine phrasings. Zero hits.
So we dropped the headline rather than sourcing it from a consultancy. What replaced it is narrower, filed, and better.
Board sanction to first production is month-dated at both ends for six mines. That distribution is 15 to 82 months — median 40 months, mean 43.2. One and a third to under seven years, not fifteen.
Greenfield and new underground mines run 29, 51 and 82 months. Brownfield expansions at operating mines run 15, 40 and 42.
The fifteen years are not in the construction. They are in the waiting before the board ever votes.
Rio Tinto's Form 20-F for FY2021 states that in 1996 the Resolution deposit was discovered by an underground drillhole. Its 20-F for FY2025, filed February 2026, states in full: "Resolution is in the permitting and study stage of the project."
Thirty years, a published environmental impact statement, an act of Congress, and no mine. The same corpus that cannot source a fifteen-year figure can source a thirty-year one — for a project that has not started. And that number can only grow.
The guided start-up date absorbs the bad news. The capital estimate does not move until money is being spent.
Southern Copper's Michiquillay project guided first production by 2025 in its FY2018 filing, by 2026 in FY2019, by 2028 in FY2021, and by 2032 in FY2025. Across all seven of those annual filings the capital estimate was "approximately $2.5 billion", unchanged. The most recent one concedes that mineral reserve estimation will begin in 2026.
A reader watching only the capital line would have seen a project that never changed.
Usually not. Fifteen projects here filed a capital estimate. Two also filed a delivered total. Both overran: Oyu Tolgoi's underground by 33%, like for like, and Quebrada Blanca Phase 2 by 81% to 86% on bases that are not symmetric — its sanction figure excluded escalation, working capital and interest during construction, and the delivered range excluded none of that.
The other thirteen close with the words "on budget", or with silence. That is the practical answer: for most copper projects, the filings do not let you find out.
Southern Copper's Tia Maria received its construction permit on 8 July 2019 and had it ratified on 30 October 2019. At its FY2025 filing, five lawsuits are pending and nothing is being built — fourteen years past the start-up year it first guided.
We also read what the funds named for this theme actually hold, from their own regulatory filings.
There is no price forecast in this report. Where we print a futures curve, it is what two SEC-filed schedules priced on 31 December 2025 and 30 June 2026, and nothing more. Both were upward-sloping, which means a fund that rolls futures forward paid a spread — a present-tense cost that appears in no expense ratio. It is not a prediction that copper will rise.
The full report names nine markers with thresholds and re-scores them every quarter in public. The one that matters most: if two or more projects now in construction file a delivered capital total against their sanction estimate on the same basis, our claim that the filed record almost never lets you check what a mine cost is wrong. We will print that in the quarter it happens.
The two research passes behind this report corrected fifteen errors in their own work and found twenty-four holes in their own checking scripts. Both lists are printed in the report rather than summarised as a clean run — including the largest one, which is that its founding premise was wrong. The inspectability is the product.
Finnacl Research publishes impersonal investment research, identical for every reader and tailored to no one. It is not individualised investment advice, and Finnacl Research is not an investment adviser. The author holds no position in any company covered in this report. Nothing here is an offer to buy or sell any security. As-of dates are printed per figure in the full report; markets move.