Base Case
Copper Market Intelligence Workbench
Decision Question
--Is refined copper likely to tighten by the selected year?
----
Scenario Read-Through
----
- Demand growth
- --
- Supply growth
- --
Commercial Implications
NarrativeScenario Comparison
--| Scenario | Demand | Refined supply | Balance | Cumulative balance |
|---|
What Moves The Balance
----
| Driver | Base | Bull | Bear | Why it matters |
|---|
Demand Breakdown
--Demand is split by model region. End-use segments are not included in this public-data version.
Supply Breakdown
----
Demand Growth Drivers
Use the scope selector to switch between world-weighted demand growth and individual regions. Growth is built from fixed 10-year historical CAGRs for industry activity, GDP per capita, and population, then adjusted for energy transition, China property downturn, substitution, and scenario shock.
Supply Growth Drivers
--Primary refined
Secondary refined
Primary refined supply now comes from the Supply tab's mine-level forecast and conversion bridge. Secondary refined supply still follows the scrap/recycling response rule.
Global Balance
-- ktSupply Mix
Scenario assumptionPrimary refined copper is linked to the Supply tab's mine-supply and conversion bridge. Secondary refined copper comes from scrap or recycled copper.
Regional Demand
-- ktMine Supply By Country
-- ktHow Scenarios Are Built
Assumptions, not sourced forecastsThe same public data baseline is used in every scenario: ICSG 2024 refined usage, ICSG 2024 world mine/refined production, and World Bank macro drivers.
The World Bank macro growth rates are fixed historical lookback rates applied across the forecast horizon; they are not rolling future forecasts.
The demand-driver weights are scenario assumptions. A regression diagnostic is shown below, but it is not used as the forecast weighting because fit is weak and the macro predictors overlap mechanically.
The scenarios change only the assumption knobs: demand growth, transition-demand bonus, China property drag, substitution, mine/refinery growth, disruption loss, conversion constraints, and scrap/recycling response.
Bull means stronger demand and tighter supply. Bear means weaker demand and looser supply. Base sits between them.
| Assumption | Source / Rationale | Base | Bull | Bear |
|---|
Data Sources & Assumptions
Public data + scenario inputsMacro Drivers
World Bank indicators for World, China, United States, and European Union. Rest of World is calculated as World minus those regions.
GDP, constant 2015 USD Industry value added, % of GDP Population, totalICSG Market Context
Used for the exact 2024 mine production, refined production, and refined usage anchors; China/regional usage shares; primary/secondary definitions; capacity trends; supply constraints; scrap/recycling context; and historical world refined usage for the demand-driver and world-GDP regressions.
ICSG World Copper Factbook Local historical balance seedEnergy Transition Context
Used only to justify scenario framing around stronger or weaker transition-related mineral demand, not the exact model percentages.
IEA Global Critical Minerals Outlook 2025World Bank API Calls
Exact API endpoints used to build data/raw/world_bank_indicators.csv.
GDP-Copper Diagnostic
Diagnostic tests of ICSG world refined usage against World Bank macro data. The rule-of-thumb test uses total world real GDP growth, not GDP per capita. It is still not used for forecast weights because annual fit is moderate and copper-specific shocks matter.
Regression weight summary Regression fit diagnostics GDP relationship diagnostics Regression dataset World GDP regression dataset ICSG monthly publications and database contextMine & Refinery Cross-Check
USGS Mineral Commodity Summaries 2026 is retained as a public cross-check, but the model baseline now uses the transcribed ICSG 2024 world mine production and refined production figures exactly. The country table remains the local transcribed MCS 2025 seed.
USGS MCS 2026 PDF USGS MCS 2025 Copper PDF Local transcribed country table Local global seed tableScenario Assumptions
Demand driver weights, supply-growth rules, and scenario stress knobs. Driver weights are scenario assumptions, with the regression retained only as a diagnostic. The base-case 83%/17% primary/secondary split follows the ICSG 2024 refined-production split after grouping SX-EW with primary supply.
Base-case assumptions JSON Bull-case assumptions JSON Bear-case assumptions JSON Regression weight summary Regression fit diagnostics Macro regression dataset World GDP regression dataset GDP relationship diagnostics Forecast model code Regression code Data fetcher codeDashboard Data Files
The dashboard reads these generated CSV files directly.
Global forecast CSV Regional demand CSV Mine supply CSV Supply assets CSV Supply summary CSV Supply conversion bridge CSV Supply asset seed CSV Supply disruption seed CSV Bull forecast CSV Bear forecast CSV Reference model market balance CSV Reference model demand components CSV Reference model mine supply by country CSV Reference model major mines CSV ICSG forecast CSV ICSG regional forecast CSVSupply
--This tab builds upstream mine supply and then converts it into primary refined supply. Mine supply is mined copper content; the Model tab's refined supply is primary refined supply plus secondary refined supply.
--
Supply Method
selected yearcontained copper kt = ore processed Mt * grade % * recovery % * 10
achievable mine supply = gross output - maintenance - disruptions - project probability discount
mine supply is mined copper content, not refined output
total refined supply = primary refined after conversion constraints + secondary refined
Mine Supply To Refined Supply Bridge
kt copperSupply Buckets
mine supplyRisk & Disruption
largest expected losses| Asset | Disruption | Probability discount | Maintenance | Risk tags |
|---|
Asset-Level Forecast
top contributors| Asset | Country | Bucket | Gross | Risk-adjusted | Probability | Grade / recovery | Source and assumption |
|---|
Project Pipeline
nameplate vs risk-adjusted| Project / expansion | Status | Gross nameplate | Risk-adjusted | Discount | Probability |
|---|
The attached ICSG project chart is used as a pipeline anchor for large projects above roughly 100 ktpa Cu. Company sources and explicit model probabilities determine how much enters the risk-adjusted forecast.
Supply Sources & Assumptions
audit trail| Asset | Source | Assumption note |
|---|
Reference Model Market Balance
--Yearly copper supply, demand, price, scrap, and electrification rows extracted from the reference model's Market Balance sheet.
--
Reference Model Narrative
2024 to 2030Refined Market Balance
'000 tonnesDemand Mix
--Copper Price Path
$/tDemand Component Trends
index, 2024 = 100| Component | 2024 | 2024 share | 2030 | Change | 2030 share |
|---|
China Split
--Scrap System
--Mine Supply By Country
2030 rankMajor Mine Movers
2024 to 2030| Mine | 2024 | 2030 | Change |
|---|
Reference Model Rows
latest six years| Year | Refined production | Refined consumption | Market balance | Copper price | Electrification share |
|---|
ICSG Model
23 Apr 2026ICSG's April 2026 world refined copper forecast covers 2025-2027. The balance uses the published China apparent usage basis and the adjusted world supply rows.
--
ICSG Refined Market Balance
2025-2027Published Growth Rates
% change2027 Refined Production By Region
--2027 Refined Consumption By Region
--ICSG Regional Forecast
'000 tonnes Cu| Region | Year | Mine production | Refined production | Refined consumption |
|---|
Model Comparison
common years: 2025-2027This compares the selected dashboard scenario, the reference model, and the ICSG press-release model on total refined consumption and total refined production.
--
Total Refined Consumption
'000 tonnes CuTotal Refined Production
'000 tonnes CuComparison Table
Model/reference gaps vs ICSG| Year | Metric | Model | Reference Model | ICSG Model | Model vs ICSG | Reference vs ICSG |
|---|
GDP-Copper Usage Relationship
Diagnostic onlyThis page answers one question: does world GDP explain refined copper usage? The answer is yes over the long run, and partly year to year. The annual rule-of-thumb slope is useful, but GDP alone does not explain most short-term copper demand moves.
refined_usage_growth = beta * world_real_gdp_growth + error
refined_usage_growth = beta * world_real_gdp_growth + error
log_refined_usage = intercept + beta * log_world_real_gdp + error
What The Evidence Says
Main read-throughLong-run: world GDP and copper usage move together strongly because both grow with industrial scale. The log-level test is high R², but that mostly reflects the long-term upward trend.
Annual growth: the no-intercept slope is about 0.93. That supports the rule that a 1 pp GDP-growth move maps to roughly a 0.9 pp copper-usage-growth move.
Short-term noise: annual R² is low because copper usage also moves with China cycles, inventories, prices, substitution, policy, scrap availability, and sector mix.
| Check | Sample | Equation | Slope | R-squared | What it means |
|---|
The exact Goldman sentence is treated here as a rule to test against public data. I could not verify the original note from a public Goldman source.
Model Scatterplots
Fitted lineOnly the highlighted evidence rows are plotted here.
How To Read This
Interview notes- Log levels
- Use this to say copper usage scales with the size of the world economy. Do not present it as a clean annual forecast.
- No-intercept annual growth
- Use this for the rule-of-thumb: if GDP growth changes by 1 percentage point, copper usage growth changes by about 0.93 percentage points.
- Low annual R²
- Use this to explain humility: GDP matters, but copper has many commodity-specific short-term drivers.
Data & Method
Reproducible- Transcribe annual world mine production, refined production, and refined usage from the ICSG Factbook annex for 1960-2024.
- Fetch World Bank world real GDP, industry value-added share of GDP, and population.
- Calculate industry activity as real GDP multiplied by industry share of GDP.
- Calculate GDP per capita as real GDP divided by population.
- Convert usage and macro variables into growth rates where each test has overlapping data. The world-GDP sample runs from 1961-2024. The multivariate industry/GDP-per-capita sample starts in 1992 because World Bank world industry-share data begins in 1991.