Let’s address the elephant in the room: that 800kt deficit figure everyone’s been throwing around? It’s not real.
At least not according to the major banks and research houses tracking this market. The actual 2026 copper deficit estimates from credible sources range between 150,000 and 330,000 metric tons. That’s a material difference when you’re building financial models that will determine whether a project gets funded or dies in due diligence.
J.P. Morgan pegs the 2026 refined copper deficit at approximately 330,000 metric tons. The International Copper Study Group (ICSG) forecasts 150,000 tons. Goldman Sachs takes the contrarian position entirely, projecting a 300,000-ton surplus for 2026.
Three major institutions. Three wildly different conclusions. Same data universe.
Why the Disconnect Matters More Than the Number
The spread between these forecasts tells you everything you need to know about copper modeling in 2026: the assumptions you make about demand elasticity and secondary supply response will determine whether your P/NAV looks brilliant or catastrophically wrong.
Goldman’s surplus forecast hinges on two critical assumptions. First, elevated prices will dampen demand growth across price-sensitive sectors. Second, high copper prices will unlock substantial scrap supply that’s currently sitting idle. They’ve already walked back their U.S. stockpiling forecast from 750,000 to 600,000 tons because import arbitrage opportunities aren’t materializing as expected.
Meanwhile, J.P. Morgan’s 330kt deficit scenario assumes demand holds firmer than Goldman expects. The ICSG’s 150kt deficit splits the difference.
But nobody credible is modeling 800kt.
What Your P/NAV Model Actually Needs to Account For
If you’re underwriting a copper project in 2026, the deficit size matters less than understanding which scenario you’re implicitly betting on. Every copper price deck in your financial model contains embedded assumptions about:
Supply-side responsiveness. How fast can marginal mines ramp? How much scrap gets mobilized at $4.50/lb versus $5.00/lb? What’s the lag between price signal and production response?
Demand elasticity. Which end-use sectors are truly price-inelastic (data centers, EV charging infrastructure) versus deferrable (construction, consumer electronics)?
Chinese consumption trends. Goldman notes that China’s refined copper consumption has weakened materially. That’s not a rounding error. China represents roughly 55% of global copper demand.
Most P/NAV models in circulation right now are running price decks that assume structural tightness persists through 2026 and beyond. They’re pricing in deficit scenarios closer to the 330kt range, not the Goldman surplus case.
The uncomfortable question: what happens to your project’s economics if Goldman is right?
The Structural Story Still Holds (Eventually)
Strip away the 2026 noise, and the longer-term copper thesis remains intact. Copper production is projected to peak around 33 million metric tons in 2030. Demand could hit 42 million metric tons by 2040.
That’s a 9-million-ton gap that has to get filled somehow. New mines. Expanded operations. Aggressive brownfield exploration. Technology improvements. Demand destruction. Some combination of all five.
But 2040 is 14 years away. Your equity committee wants to know what happens in the next 24 months.
The strategic challenge for project developers: you’re raising capital in a market where near-term fundamentals look ambiguous (150kt deficit versus 300kt surplus versus 330kt deficit), but the long-term structural case appears bulletproof.
That’s a needle that’s almost impossible to thread cleanly.
Scenario Planning: What Changes Under Each Case
Deficit Case (150-330kt): Copper prices trade in a $4.25-$4.75/lb range through 2026. Projects with all-in sustaining costs below $3.50/lb clear investment hurdles comfortably. Marginal projects in the $3.50-$4.00 AISC range remain viable but face tighter financing terms.
Surplus Case (300kt): Prices drift toward $3.75-$4.25/lb. Only sub-$3.00 AISC assets pencil out at conservative price decks. Development-stage projects face brutal capital-raising environments. M&A stalls as buyers wait for forced sellers to emerge.
Wildcard: Technology Breakthrough or Substitution. What if battery chemistry shifts away from copper-intensive designs? What if aluminum becomes genuinely competitive in more applications? This tail risk almost never appears in base-case P/NAV models, but it exists.
The probability-weighted outcome sits somewhere in the middle: modest deficit, prices in the low $4s, development discipline required but capital still available for quality assets.
What Investors Are Actually Modeling
Talk to 10 mining equity analysts. You’ll get 10 different price decks for 2026.
The consensus clusters around $4.25-$4.50/lb for 2026, which implies belief in a modest deficit scenario rather than Goldman’s surplus or the aggressive 330kt case. That pricing reflects a market that’s structurally tight but not in crisis mode.
Most P/NAV models being presented to boards right now assume:
- 2026 price: $4.30-$4.50/lb
- 2027-2030 price: $4.50-$5.00/lb
- Long-term price: $4.00-$4.25/lb (real)
Those assumptions embed a deficit narrative without going full bull case. They reflect optimism about electrification and data center demand while acknowledging that scrap supply and demand elasticity provide safety valves.
But notice what’s missing: explicit sensitivity analysis around the Goldman surplus scenario. Very few presentations honestly stress-test what happens if copper trades down to $3.80/lb for 18 months.
The China Problem Nobody Wants to Talk About
Goldman’s reduced consumption forecast for China isn’t a minor technical adjustment. It’s a fundamental reassessment of the world’s dominant copper consumer.
China’s refined copper consumption growth has decelerated materially as the country’s infrastructure build-out matures and its real estate sector remains in structural decline. Electric vehicle and renewable energy demand growth is real, but it’s not offsetting the construction slowdown at the pace many models assumed.
If China’s copper intensity continues declining faster than EV/renewable penetration rises, the global deficit story gets pushed further into the future. Not canceled. Just delayed.
That delay matters enormously for 2026-2028 project economics. A developer counting on tight markets in 2027 could find themselves ramping production into softer conditions than modeled.
What This Means for Your Model
If you’re presenting a copper project to an investment committee in 2026, the credible approach includes:
Scenario analysis that brackets the range. Base case deficit of 200kt. Bear case surplus of 200kt. Bull case deficit of 400kt. Show what each means for project NPV.
Explicit assumptions about demand elasticity. Don’t just assume consumption growth continues unabated at elevated prices. Model what happens if price-sensitive demand gets deferred.
Scrap supply assumptions. At what price does meaningful secondary supply get activated? How much? How fast?
China-specific consumption modeling. Don’t treat Chinese demand as monolithic. Break it out by sector and show how each component responds to domestic economic conditions.
Time-to-market sensitivity. A project coming online in Q2 2026 faces different market conditions than one starting production in Q4 2027. That timing risk needs to be visible.
The models that survive scrutiny in 2026 won’t be the ones projecting the tightest markets or highest prices. They’ll be the ones that honestly stress-test the downside while making a credible structural case for the upside.
The Bottom Line
There is no 800kt deficit in 2026. The range is 150-330kt deficit, or possibly a 300kt surplus depending on whose analysis you trust.
That uncertainty should make you uncomfortable. It should force harder questions about the assumptions buried in your copper price deck.
The structural copper thesis: supply constraint meeting secular demand growth from electrification: remains intact for the next decade. But the path from here to there looks messier than the bull case presentations suggest. Price-sensitive demand will get deferred. Scrap supply will respond. Chinese consumption growth will moderate.
Projects will still get funded. Mines will still get built. But the winners will be the operators and developers who modeled the volatility honestly rather than extrapolating the tightest possible supply story into their base case.
2026 copper markets will be tight enough to support development of quality assets. They probably won’t be tight enough to save marginal projects built on optimistic assumptions about deficit severity. Know which category you’re in before you commit the capital.
