The copper deficit everyone’s talking about is real. It’s also nowhere near the apocalyptic narrative some sell-side reports want you to believe. And that gap: between what’s actually happening and what your model might be pricing in: is where valuations break.
Copper hit $13,300 per metric ton on January 6, 2026. That’s an all-time high. It’s also not a sustainable run-rate for your NPV calculations. Goldman Sachs said it plainly: “We do not expect the price above $13,000 to be sustained.” They’re forecasting an 18% decline from those early-year highs. Meanwhile, BlackRock pegs the incentive price for new mines at $12,000/mt: the floor where project economics actually work.
Your P/NAV model needs to account for this disconnect. Fast.
The Deficit Reality: Modest, Not Massive
Start with the numbers that matter. The International Copper Study Group projects a 150,000-ton refined copper deficit for 2026. J.P. Morgan goes higher at 330,000 tons. That’s a 2.2x variance. Both are material. Neither is the 800,000-ton supply gap some headlines screamed about last quarter.
The structural challenge shows up later. Demand is climbing from 25 million metric tons today to a projected 33 million metric tons by decade-end. Annual deficits could hit 6 million metric tons by 2030 if no new supply comes online. That’s the long game. But 2026? The deficit is tight, not catastrophic.
This distinction matters because it underpins every price scenario in your NPV model. If you’re running a base case on sustained $13,000+ copper, you’re overvaluing development-stage assets by a significant margin.
The Price Forecast Table You Should Be Using
Stop anchoring to spot. Start weighting probabilities across the institutional consensus. The divergence tells you more than any single number.
| Source | 2026 Average Price ($/mt) | Q2 Peak Estimate | Year-End Forecast |
|---|---|---|---|
| J.P. Morgan | $12,075 | $12,500 | : |
| Goldman Sachs | $11,500 | : | $11,200 |
| UBS | $11,000 (Sept target) | : | : |
| Citibank | $12,000 | : | : |
| Goldman (H1) | $10,710 | : | : |
Source: Skillings Mining Review (Data as of February 13, 2026)
Goldman’s first-half forecast of $10,710 is the outlier low. J.P. Morgan’s $12,075 average sits at the top. The median lands around $11,500. That’s your base case. Not $13,300. Not $15,000 (which remains an outlier call despite the macro tailwinds from AI infrastructure build-out).
What This Does to Your Project Economics
Development-stage copper projects: greenfield assets, brownfield expansions with fresh capex: live and die on forward price assumptions. Marimaca Copper explicitly noted their definitive feasibility study relies on “higher copper price assumptions.” That’s code for: our economics compress fast if copper reverts to the $10,500–$11,000 range.
BlackRock’s $12,000/mt incentive price is the floor for new mine economics. Below that, projects don’t get funded. Above that, they compete for capital against established producers who are already profitable at lower prices. The spread between your base-case copper assumption and that $12,000 threshold is your real valuation risk.
Run the sensitivity. A 10% price swing from $12,000 to $10,800 doesn’t just knock 10% off NPV. It cascades through your model: lower revenue, compressed margins, extended payback, reduced optionality on expansions. For pre-development assets trading on P/NAV multiples, that’s a 20–30% haircut to market cap in a bad quarter.
The Supply-Constraint Optionality (And Its Limits)
The structural deficit gives copper a pricing floor that didn’t exist five years ago. Electrification, AI data centers, grid infrastructure: all demand copper. All are non-discretionary over the medium term. That’s the bull case, and it’s valid.
But it doesn’t mean every copper project gets a free pass on valuation discipline. Supply constraints create optionality for producers, not for every speculative explorer with a resource estimate. Freeport-McMoRan’s 500,000-ton production loss over the next 12–15 months tightens the market. It also reminds you that operational risk: weather, permitting, geotechnical failures: matters as much as macro demand drivers.
The projects that capture upside are the ones with near-term production visibility, jurisdictional stability, and permitting timelines that don’t stretch into 2029. Everything else trades at a discount to NAV because the market prices in execution risk and price-deck uncertainty.
Policy Wildcards That Belong in Your Model
Two policy dynamics deserve scenario weighting in 2026 valuations:
1. Strategic Stockpiling Programs
The U.S., EU, and Japan are all building critical minerals reserves. Copper isn’t officially on most “critical” lists, but government buying programs for adjacent supply chains (rare earths, lithium) create precedent. If a Western government decides copper qualifies for strategic procurement, that’s a demand tailwind. It also introduces non-market pricing floors that distort your supply-demand models.
2. Resource Nationalism and Royalty Creep
Chile, Peru, Zambia: copper-producing jurisdictions are raising royalties and tightening tax regimes. That’s a margin headwind for existing operations and a hurdle-rate problem for new projects. If your model assumes 2023-era fiscal terms, you’re underpricing sovereign risk.
The policy uncertainty doesn’t just affect project economics. It also compresses valuation multiples. Investors apply a jurisdiction discount to assets in politically volatile regions, even if the copper’s in the ground and the resource estimate is solid.
The Three-Scenario Framework That Actually Works
Single-point estimates are dangerous in 2026. The variance in deficit forecasts, the spread in price targets, the lag between permitting and production: it all argues for probabilistic modeling.
Weight your scenarios like this:
Base Case (60% probability): Goldman Sachs-style reversion to $11,000–$11,500 average for full-year 2026. Deficit stays modest (150,000–200,000 tons). Supply disruptions are temporary. AI infrastructure demand is real but offset by deferrable consumption in construction and consumer electronics.
Bull Case (25% probability): J.P. Morgan’s $12,075 average holds. Deficit widens to 300,000+ tons. Additional supply disruptions (labor strikes, weather, permitting delays) tighten physical market. Speculative positioning from institutional rotation (precious metals → base metals) provides support above $12,000.
Bear Case (15% probability): Macro slowdown (China real estate, U.S. recession fears) crimps demand faster than supply adjusts. Price reverts to $9,500–$10,500 range. Deficit narrows or flips to small surplus. Projects with all-in sustaining costs above $9,000 face margin compression.
That weighted average: roughly $11,200: should be your NPV anchor. Not spot. Not the January spike.
Market Dynamics: 2026 Copper Snapshot
| Metric | Value | Change vs. 2025 | Implication |
|---|---|---|---|
| Global Demand (mt) | 25.4 million | +2.8% | Steady, not explosive |
| Refined Deficit (ICSG) | 150,000 tons | -20kt vs. 2025 | Tight but moderating |
| Incentive Price (BlackRock) | $12,000/mt | +8% vs. 2024 | New mine floor rising |
| Spot Price (Feb 13) | $11,850/mt | -11% vs. Jan peak | Reversion underway |
| Freeport Production Loss | 500,000 tons (12–15mo) | New disruption | Supply-side risk elevated |
Source: Skillings Mining Review (Data as of February 13, 2026)
The table crystallizes the tension: demand is growing, deficits are real, but prices are already pulling back from unsustainable highs. That’s your 2026 reality.
Where Valuations Break (and How to Fix Them)
P/NAV models fail when they treat copper as a static input. The metal trades on macro sentiment, supply shocks, and policy whims as much as fundamentals. Your model needs to flex.
Three fixes:
1. Dynamic Price Decks: Use escalating assumptions tied to deficit scenarios, not flat-line forecasts. If your base case assumes $11,500 in 2026, model $12,000 in 2027–2028 as new supply fails to materialize, then revert to $11,000 in 2029–2030 as delayed projects finally hit production.
2. Jurisdiction-Adjusted Discount Rates: Apply a 200–300 basis point premium to WACC for assets in jurisdictions with rising royalty risk or permitting uncertainty. That’s not pessimism. That’s pricing reality.
3. Capex Inflation Buffers: Construction costs are running 15–20% above pre-2024 estimates. Labor, equipment, energy: it all costs more. If your DFS was completed in 2023, add 20% to the capex line and rerun the model.
The copper deficit is real. The price volatility is real. The valuation risk from anchoring to spot instead of probability-weighted forecasts? Also real. 2026 marks the inflection point where modeling discipline separates winners from write-downs.
Your P/NAV model needs to reflect that. Now.
