What Is the Copper Price Forecast for the Next 6-12 Months?
All figures and forecasts in this article are as of 16 November 2026.
If you are setting a 2027 copper budget or deciding how much of next year's volume to fix, you will be asked for "the forecast". The trouble is that there are many, they disagree, and their track record is weaker than their confidence suggests.
Here is the short answer. As of 13 November 2026, with LME three-month copper trading around US$13,350-13,550 per tonne (Business Recorder/Reuters, 13 Nov; ABC News, citing Reuters, 16 Nov), published views for the next 6-12 months span roughly US$11,000 to US$15,500 per tonne. At the high end, Goldman Sachs expects a 2027 average of US$15,400/t on supply deficits outside the US. At the low end, Macquarie has argued that prices will correct towards about US$11,000/t in the third quarter of 2027 as the market moves into surplus. The International Copper Study Group (ICSG) expects the global refined surplus to widen in 2027. The biggest single near-term swing factor is whether the US imposes a tariff on refined copper imports from January 2027.
For a buyer, the useful takeaway is not which bank is right. It is that a US$4,000+ spread between credible forecasts is itself the forecast: the next 12 months could plausibly land anywhere in a wide range. That is why MetalAlert works with probabilities of specified price changes over defined horizons rather than a single number, and why the second half of this article is about how to use forecasts rather than which one to believe.
Where the main forecasters stand
| Source (date) | View for the next 6-12 months | Main reasoning |
|---|---|---|
| Goldman Sachs (20 Oct 2026) | End-2026 US$14,700/t; 2027 average US$15,400/t (on 13 Nov it said it sees downside risks to the 2027 figure) | Deficits outside the US of about 950 kt in 2026 and 220 kt in 2027; weaker mine supply; China scrap shortage (Canadian Mining Report, citing Reuters; ABC News, citing Reuters, 16 Nov) |
| UBS (Sept 2026, as reported) | About US$15,500/t by mid-2027 | Cited in a September roundup (same source) |
| Morgan Stanley (2 Nov 2026) | Q4 2026 average US$14,800/t | US stockpiling as the swing factor (Copper Weekly Brief, 12 Nov) |
| Macquarie (Jul 2026) | 2026 average US$13,165/t; correction with a floor around US$11,000/t in Q3 2027 | Rally driven more by positioning and tariff flows than physical tightness; surplus above 700 kt a year in 2027-28 (MINING.COM) |
| BMI (Jul 2026, updated 5 Nov 2026) | 2026 average US$13,500/t (US$12,700/t in July); 2027 average US$12,900/t; near-term caution, structural deficit from 2027 | Current rally "beyond what the fundamentals alone would suggest"; long-run deficits from energy transition and data centres (MINING.COM; November figures: Copper Weekly Brief, 12 Nov) |
| ICSG (14 Oct 2026) | No price forecast; refined surplus of about 104 kt in 2026 widening to about 368 kt in 2027 | Refined output +0.4% in 2026 and +3.4% in 2027; usage +1.5% and +2.5%; mine output +1.3% and +3.1% (Mining Weekly/Reuters) |
Two observations stand out. First, the views do not split neatly into bulls and bears. Goldman and the ICSG are describing different things: Goldman focuses on the market outside the US, where metal has been drained by US tariff stockpiling, while the ICSG balance is global and includes the large US stocks. Second, several of the views are months old. Macquarie published in July, before the September record highs. Forecasts are revised often, and you should check the date on any number you use.
Why the forecasts disagree: five drivers
1. The US refined copper tariff decision
The US has applied a 50% Section 232 tariff to semi-finished copper products since August 2025, while refined cathode has so far been exempt. A phased tariff on refined copper (15% from 1 January 2027, 30% from 2028) has been proposed, but as of 12 November no decision had been announced (Reuters columnist Andy Home via The Star, 11 Nov; Copper Weekly Brief, 12 Nov; CRS). The tariff threat has pulled metal into the US: stocks in Comex-linked warehouses reached about 733,000 t in early November, more than double a year earlier (MINING.COM, 9 Nov).
This is why forecasts diverge so sharply. A tariff would likely keep drawing metal into the US and tighten the rest of the world. No tariff could release some of that pressure. JP Morgan's Patrick Jones said investors see "US tariff inaction" as the main downside risk (Business Recorder/Reuters), while Macquarie's Alice Fox called a January deadline passing without a decision the most likely scenario (Reuters via Investing.com). The market itself puts modest odds on the tariff: Société Générale estimated in August that the Comex premium implied about a 14.6% probability of a 15% tariff in January 2027 (CNBC). For how this transmits to European prices, see how US copper tariffs affect European copper prices.
2. Chinese demand
China accounts for about 58% of global refined copper consumption (SMM, citing ICSG). The signals are mixed. The Yangshan import premium, a gauge of Chinese appetite for imported metal, rose to a record US$285/t on 13 November (Business Recorder/Reuters, 13 Nov). On the other hand, Macquarie noted in July that Chinese buyers step back at high prices (MINING.COM), and the official manufacturing PMI fell back into contraction at 49.3 in October (NBS).
3. Mine supply
Disruptions at Grasberg and Kamoa-Kakula have delayed recovery schedules, and Macquarie assumed in July that Cobre Panama restarts only in the second quarter of 2027 (MINING.COM). Concentrate is so scarce that smelters are being paid to process it: Antofagasta reportedly secured negative treatment charges of more than US$300/t in recent tenders (Kitco/Reuters, 16 Oct). Yet the ICSG still expects mine output to grow 3.1% in 2027. Whether that recovery arrives on time is a large part of the bull-bear split.
4. AI and data-centre demand
Data-centre construction is a common bullish argument; BMI cites AI-related build-out as one source of a structural deficit. Macquarie is sceptical, arguing that "the copper impact may be smaller and slower than the market assumes" (MINING.COM). In late October, concern about heavy AI spending at large technology companies was weighing on copper sentiment (Business Recorder/Reuters). The narrative can move prices in the short term even where the physical tonnages are uncertain.
5. The dollar and US interest rates
Copper is priced in dollars, and research finds that commodity prices tend to rise when real interest rates fall and the dollar weakens (Akram, 2009). The Federal Reserve raised its policy rate to 3.75-4.00% in September 2026 and signalled further tightening (Honolulu Star-Advertiser/AP). On 28 October it held the range, but two policymakers voted for another increase, which keeps a December hike on the table (Federal Reserve, 28 Oct 2026). For a European buyer, part of any dollar move is offset in the euro or krona price. How much these indicators really predict is covered in which indicators predict copper price moves.
How accurate are copper price forecasts?
Less accurate than most buyers assume. The research is fairly consistent:
- Futures prices are hard to beat. An IMF study of ten commodity futures at horizons up to two years found futures prices difficult to beat as out-of-sample forecasts of spot prices (Reichsfeld and Roache, IMF WP/11/254).
- But they are not better than "no change" for copper. Chinn and Coibion found that copper futures performed slightly worse than a random walk (assuming the price stays where it is), with relative errors of 1.01, 1.03 and 1.12 at 3, 6 and 12 months, though the differences were not statistically significant (NBER WP 15830).
- Expert judgment did worse than market-based models. An earlier IMF paper covering 15 commodities found that models using futures prices outperformed both judgment-based forecasts and purely historical models, especially at longer horizons (Husain and Bowman, IMF WP/04/41).
- Forecasters deliberately disagree. A study of more than 20,000 forecasts of nine metal prices found evidence of "anti-herding": forecasters scatter their numbers around the consensus for strategic reasons (Pierdzioch, Rülke and Stadtmann, Journal of Banking & Finance, 2013). The spread between banks is partly a product of incentives, not only of different analysis.
The past year is a live example. In October 2025, the median of 30 analysts in a Reuters poll expected LME cash copper to average US$10,500/t in 2026 (Reuters via 979weve). By January 2026 the median had risen to US$11,975/t (Reuters via WHBL), and by July the Bloomberg consensus stood at US$13,007/t (MINING.COM). Meanwhile the price itself was above US$14,000/t for much of the autumn and set records in September. The forecasts followed the market up rather than anticipating it. That is not a criticism of the analysts; it is what the academic evidence predicts.
From point forecasts to probability ranges
A point forecast hides the information a buyer most needs: how wide the range of outcomes is. The table below shows what a simple, neutral model implies. It assumes no expected drift and an annualised volatility of 22%. Both are illustrative assumptions for this example, not a forecast; check current implied volatility with your bank or broker.
| Horizon | Central value | 80% range | Probability above US$15,400 | Probability below US$13,000 |
|---|---|---|---|---|
| 6 months | about US$14,180 | US$11,600-17,300 | about 30% | about 29% |
| 12 months | about US$14,000 | US$10,600-18,600 | about 33% | about 37% |
Starting price US$14,350/t; lognormal model; figures rounded.
Three things follow. The range is wide enough to contain every institutional forecast in the first table. Goldman's US$15,400 and Macquarie's US$11,000 are both well inside the plausible zone. And the questions that matter for purchasing ("what is the chance copper is 10% higher by the time we need the metal?") are answered by probabilities over a specific horizon, not by an annual average.
A naive model like this one treats every period as equally uncertain. Better models condition on the current state of the market, such as inventories, spreads, positioning and macro data, and should then be judged on calibration: when the model says 30%, the event should happen about 30% of the time over many forecasts. Calibration is also the yardstick MetalAlert emphasises for its probability estimates for LME copper.
What a buyer should do with copper forecasts
Use the forward curve as the budget baseline. It is the price at which you can actually lock in, and the evidence says it is as good a central estimate as any. Build scenarios around it rather than around a favourite bank's number. A worked approach is in how to build a copper budget with best, base and worst-case scenarios.
Turn the forecast range into scenarios with a cost attached. For example, a bear case near US$11,000/t (Macquarie-style correction), a base case at the forward, and a bull case at US$15,500/t (Goldman/UBS territory). Multiply each by your annual tonnage. The difference between bull and bear is your budget at risk.
Decide by policy, not by conviction. Whether to fix more now depends on how much of that budget at risk you can tolerate, not on which bank sounds most persuasive. See should I lock in copper prices for 2027 or buy at spot and how much of your annual copper volume to hedge.
Track the catalysts. In the next 6-12 months the main ones are the US refined tariff decision ahead of 1 January 2027, the pace of mine restarts, Chinese import premiums and stock movements between exchanges. When one of them moves, revisit your probabilities rather than waiting for the next round of bank forecasts.
Ask any forecast provider for a track record. How large were their errors at your horizon? Were their stated probabilities calibrated? A provider that cannot answer is offering an opinion, not a forecast.
Frequently asked questions
Will copper reach US$15,000 per tonne in 2027?
It is plausible but far from certain. Goldman and UBS expect prices around or above that level; Macquarie and the ICSG balance point the other way. Under the neutral illustrative model above, the probability of being above US$15,400/t in 12 months is about one in three.
Is copper going to crash in 2027?
A correction is a credible scenario if mine supply recovers and the US tariff is not imposed, which is Macquarie's case. A collapse to pre-2025 levels would need a much larger demand shock. Budget for a downside scenario, but do not assume it.
Which copper price forecast is most reliable?
No forecaster has a consistent edge over market prices at 6-12 month horizons in the published research. Treat the futures curve as the baseline and use forecasts to understand the arguments and the range, not to pick a number.
Sources
- MINING.COM: Copper price bounces after worst week since March amid scramble for metal in China (9 Nov 2026)
- Business Recorder/Reuters: Copper falls, heads for fifth straight weekly drop on AI rout (13 Nov 2026)
- ABC News: ASX markets live, 16 Nov 2026 (LME copper close and Goldman Sachs comment, citing Reuters)
- Business Recorder/Reuters: Copper slips on concerns over AI demand, high US inventories (29 Oct 2026)
- Canadian Mining Report: Goldman Sachs raises its copper price forecast to $15,400 for 2027 (22 Oct 2026)
- MINING.COM: Macquarie says copper price rally still running ahead of reality (9 Jul 2026)
- MINING.COM: BMI hikes copper forecasts (16 Jul 2026)
- Mining Weekly/Reuters: ICSG sees global refined copper surplus widening in 2027 (14 Oct 2026)
- Copper Weekly Brief, to 12 November 2026 (ICAA)
- The Star/Reuters (Andy Home): China buyers squeeze copper's tariff trade (11 Nov 2026)
- Reuters via Investing.com: Trump tariff threat built a US copper mountain (14 Oct 2026)
- Congressional Research Service: Section 232 copper tariffs
- CNBC: Copper and Trump tariffs (14 Aug 2026)
- SMM: ICSG outlook, China share of refined consumption (9 Oct 2025)
- National Bureau of Statistics of China: PMI for October 2026
- Kitco/Reuters: Antofagasta nets its highest ever copper treatment charges (16 Oct 2026)
- Honolulu Star-Advertiser/AP: Fed raises interest rates (16 Sep 2026)
- Federal Reserve: FOMC statement, 28 October 2026
- Akram (2009), Commodity prices, interest rates and the dollar, Energy Economics
- Reichsfeld and Roache (2011), Do Commodity Futures Help Forecast Spot Prices? IMF WP/11/254
- Chinn and Coibion (2010), The Predictive Content of Commodity Futures, NBER WP 15830
- Husain and Bowman (2004), Forecasting Commodity Prices: Futures Versus Judgment, IMF WP/04/41
- Pierdzioch, Rülke and Stadtmann (2013), Forecasting metal prices: Do forecasters herd? Journal of Banking & Finance
- Reuters poll, October 2025 (via 979weve)
- Reuters poll, January 2026 (via WHBL)
Plan around the range, not the headline
The forecasts for the next 6-12 months are informative mainly because they disagree. If you want to see how calibrated probability estimates for LME copper price changes over defined horizons could support your 2027 purchasing decisions, MetalAlert is built for exactly that question.