What 150+ Large Mining projects Reveal About CAPEX Overruns, Part III: When is it Cheaper to Buy Mines Than to Build Mines?
At what valuation threshold does it make more sense to build a mine rather than buy one?
In collaboration with Mining Valuation Assistant. This analysis follows on from Part 1 and Part 2.
Summary
With capital intensity rising, most copper, lithium and western nickel projects sit in the buy camp, while only a small set of brownfield projects still justify building. The analysis points to backing expansions and a few exceptional deposits, and directing most growth capital toward acquiring existing tonnes rather than building from scratch.
The central question…
At current capital intensity and price levels, one question sits above most growth plans and M&A presentations:
When is it cheaper to build a mine and when is it cheaper to buy one?
Answering that needs a way to compare the cost of securing an incremental tonne of metal organically (new projects) versus inorganically (acquiring existing production or reserves).
That comparison can be reduced to a single synthetic metric:
Parity Ratio = Inorganic price (market EV per tonne) / Organic cost (capex per tonne)
Inorganic price is the market value per tonne of existing production or reserves.1
Organic cost is the capital intensity of a project, the capex per tonne of capacity.
Ratio is >1.2 it is cheaper to build than to buy.
Ratio is <0.8 it is cheaper to buy than to build.
Ratio is between 0.8 and 1.2 the answer depends on jurisdiction, ramp up risk and portfolio fit.
This shifts the focus from “Is this project attractive?” to “What is the cheapest risk-adjusted way to add one more tonne?”
Capital intensity drifting up for 4 key energy transition commodities
The 2×2 capital-intensity chart is the organic side of the build-vs-buy equation. Each point is a real project from the dataset, plotting capital intensity against actual start-up year; the trend lines and ±1σ bands show how the cost of new tonnes has shifted over time (Illustration 1).
Copper: capital intensity drifts up at around 2–3% CAGR from 2007 to 2027. Early projects cluster below $20,000/t; more recent builds sit closer to $25,000–30,000/kt and above.
Nickel: a similar upward slope, with capital intensity growing at roughly 4% CAGR, and sitting at a much higher absolute level than copper.
Gold: a slower but steady rise of around 2–3% CAGR, with the band lifting from roughly $2–3m/koz to $3–4m/koz.
Lithium: the steepest line by far, with capital intensity increasing sharply. (CAGR >20% but off a short timeframe and dataset of projects).
New projects now cost more per tonne to build than earlier ones. That rising capital intensity pushes Organic Cost (capex per tonne) higher in the Parity Ratio. When Organic Cost rises but the market value per tonne of existing producers does not rise the same way, the ratio moves toward the Buy zone, it becomes cheaper to buy existing capacity than to fund the next set of projects in these plots.
What the 150+ project samples show
The project bar charts (Illustration 2) apply the Parity Ratio of 150+ mining projects across copper, nickel, gold and lithium completed or due to complete between 2007 and 2027.
Two clusters stand out.
“Kill zone”
A wide run of red bars, especially in nickel and in many copper and gold builds, sits below 0.8. These projects are firmly in the buy zone. It is cheaper to buy similar tonnes in the market than to build them. In several nickel cases the ratio is only 0.2 to 0.4, which means the capital per tonne is five to seven times higher than the implied M&A cost.
“Alpha zone”
At the other end a short group of green bars marks the real build candidates. A small set of copper and gold projects, plus two standout lithium projects, sit well above 1.2. They add capacity at a clear discount to market value per tonne and improve portfolio cost positions.
The remaining grey bars sit in the neutral band between 0.8 and 1.2. In these cases the call is driven more by schedule risk, geology, jurisdiction and strategic fit than by price alone.
Under base price assumptions, more than half of copper projects fall in the buy zone and only about a quarter land in the build zone. Gold is more evenly spread, but close to forty per cent of projects are still cheaper to buy than to build. In lithium about two thirds of projects are buy cases. In nickel every project in the sample is a buy rather than a build.
A simple sensitivity test shifts prices 20% up or down (Illustration 3). Even in a bull case the nickel projects never reach the build zone. Copper and gold gain a few build candidates, but a large share of projects remain value destructive compared with buying existing tonnes.
Greenfield vs brownfield choices
One of the more useful checks is to ask whether the chosen project type lines up with the ratio. The greenfield versus brownfield mismatch chart does that (Illustration 4).
In copper there is a split. About 14% of cases look like they should have been greenfield but were built as brownfield expansions. Around 12% look the other way, where a brownfield expansion would have been cheaper than the greenfield route that was chosen.
In gold and lithium the pattern is clearer. A quarter of gold projects and about 18% of lithium projects appear to have been built as greenfield when the ratio suggests a brownfield expansion would have been cheaper. In nickel a smaller share shows the opposite, where the economics would have favoured a new greenfield development over the chosen brownfield path.
The time value of lost tonnes
Time further widens the gap between building and buying. A modern greenfield mine typically requires 10–15 years from concept to steady-state production.
In a high price environment (for example, copper >$12,000/t, those missing tonnes during construction carry significant value. If the apparent capital saving from building is modest, say a $1,000/t edge versus buying a two-year delay can eliminate that edge once cash flows are discounted at an 8–10% WACC.
Given that large projects commonly run 2–3 years late, many projects that look acceptable on a simple dollars per tonne chart move back into the buy camp.
This will be another topic for an article coming soon, stay tuned…
What this means for capital allocation
Taken together, the commodity benchmarks, mining project data and schedule risk point to a simple rule of thumb:
Organic growth makes sense when the Parity Ratio is clearly above 1.2 and delivery risk is tightly controlled. That is most often in brownfield expansions or truly exceptional deposits.
In many other cases the build versus buy gap now favours acquiring existing tonnes rather than funding more concrete and steel. As incentive prices drift higher and capital intensity rises, the replacement cost of today’s low cost assets keeps stepping up.
As incentive prices drift higher and capital intensity rises, the cost of replacing today’s low cost assets keeps stepping up. The edge over the next few years is likely to sit with teams that treat the Parity Ratio as a live build versus buy dashboard, keep it updated as markets move and are ready to move capital out of projects and into M&A when the numbers say that buying is cheaper than building.
Copper:
Goldman Sachs (Nov 20, 2025): Raised their long-term copper price forecast (the “Incentive Price” required to trigger new mines) to $15,000/t by 2035. They cite that “long-stalled mine projects” require this price level to be viable.
UBS (Nov 25, 2025): Forecasts prices to hit $12,000 – $13,000/t by late 2026, creating the arbitrage window you asked about.
Link: Goldman Sachs Raises Copper Forecast to $15k Incentive Price (Metal.com) | UBS Raises Forecast to $13,000 (Mining.com)
Gold
Metric: Market Buy Price ~ $3,000/oz (EV/Production) vs. Price Forecast ~ $3,750+ (Upside Case).
Source:
Discovery Alert (Nov 4, 2025): In an analysis of the “M&A Fever,” they explicitly list the EV/Production Ounce valuation metric for operating mines as $2,000 – $3,500 per ounce. This confirms that “Buying” an ounce of production capacity on the M&A market costs roughly $3,000.
Harmony Gold (Nov 24, 2025): Their feasibility study for the Eva project uses an “Upside Case” gold price of $3,750/oz, validating the bullish forecast buffer.
Link: Gold Giants Circle as M&A Fever Grips Sector (Discovery Alert)
Lithium
Metric: Market Buy Price $\approx$ $10,000/t (Spot/Distressed M&A).
Source:
Fastmarkets (Oct 2024): Assessed Lithium Carbonate spot prices at $10,700 – $11,300 per tonne (CIF China). This represents the “floor” price where distressed assets are trading.
Rio Tinto / Arcadium (Oct 2024): Rio’s acquisition of Arcadium was a counter-cyclical bet at the bottom of this cycle. While the deal value was higher (premium), the market spot price remains the baseline for “Buying” un-premiumed tonnes.
Link: Rio Tinto Acquires Arcadium at Market Lows (Fastmarkets)
Nickel
Metric: Market Buy Price $\approx$ $20,000/t (Western Incentive).
Source:
Crux Investor (Dec 1, 2025): Highlights the “Supply Transformation” where Indonesian supply dominates at lower costs, but Class 1 Sulphide projects (Western Supply) require a premium. The standard industry “Incentive Price” for Western nickel sulphide projects is widely cited at $20,000 – $22,000/t to cover the higher capital intensity compared to Indonesian NPI.
Link: Battery Metals Investment Case: Nickel Supply Dynamics (Crux Investor)
Im just starting out in my journey understanding the economics of resources and I really enjoy your work.
Keep it up!!
I really love your detailed analysis about mine industry production.
Also attaching the source make it more legit.
Thanks for sharing your article and keep writing 💫