Copper's Plunge Could Make the AI Data-Center Buildout More Painful

Dow Jones11-19 05:55

The world might not be able to mine enough copper to meet AI's demand for years to come

Copper prices fell to a seven-month low on Wednesday. Here's what that suggests about the AI buildout.

In many ways, copper acts like the veins of the artificial-intelligence boom: It carries the electricity that powers data centers, connects power grids and serves as the backbone of the historic AI infrastructure buildout.

Yet copper prices have dropped alongside semiconductor stocks and other tech stocks in the past month, leaving investors concerned that enthusiasm for the AI trade might be coming up short.

Wednesday alone saw copper tumble 5.4%, settling at a seven-month low of $5.6595 per pound, according to FactSet data.

"A lot of steam is coming out of the AI trade," said Steve Sosnick, chief strategist at Interactive Brokers, on Wednesday. "In the same way that people were buying peripheral stocks to the AI buildout, you are now seeing those stocks get hit too."

A big concern is that the world will need less copper if the AI boom loses momentum. The reality might be the world can't produce enough copper to meet demand in the first place.

Copper prices (HG00) in November have already plunged 12.6%, putting the metal on pace for its worst month since July 2025, after it hit a record high just two months ago.

Copper's decline has coincided with a drop in the tech-centric Nasdaq Composite COMP and other AI benchmarks, as investors raise more questions about about the revenue needed to justify the rapid pace of infrastructure spending.

The tech-heavy Nasdaq has fallen over 4% in November, while the PHLX Semiconductor Index SOX was off 10.4% over the same period, according to FactSet data.

The selloff in copper and chip stocks at the same time has left "investors selling chip stocks and liquidating their portfolios, which is affecting copper prices," said Jacob White, director of ETF Product Management at Sprott Asset Management.

The tone around building data centers also has become more cautious since the U.S. midterm elections, though demand for data centers has yet to show signs of cooling. Goldman Sachs analysts said Wednesday they expect to see nearly $1 trillion in debt raised next year by hyperscalers for the financing of AI projects.

Data centers alone could account for approximately 3% of global copper demand by 2030, according to Ivan Bebek, CEO of Coppernico Metals, a mining company focused on copper and gold exploration in South America. When including the infrastructure to power those data centers, that number would go up to between 16% to 18% roughly over the next 10 years, he said.

"I think everyone needs to understand that just because copper prices went down, the demand from AI did not go away," Bebek told MarketWatch. "AI data centers are still going to be built, and we are still moving toward that clean, futuristic planet, which is going to require so much more copper."

Yet the decline in copper prices could end up exacerbating the looming supply shortage, and that supply constraint could eventually become a bottleneck for the AI buildout itself, he added.

Discoveries of new copper mines have fallen sharply since 2015, while existing copper deposits have become harder and more expensive to develop. Because the market needs sustained capital to finance exploration and mining capacity, falling copper prices and wild swings in the metal's price would threaten to undermine development projects, Bebek said.

Long-term demand for copper looks "very consistent," but short- and long-term supplies of the metal look increasingly at risk of not being met, he said. That could make the cost of building data centers even more expensive, at time when interest rates are increasing. You may not be able to predict a recession with copper, Bebek noted - but it's a good temperature check about the broader economy.

-Isabel Wang

 

At the request of the copyright holder, you need to log in to view this content

Disclaimer: Investing carries risk. This is not financial advice. The above content should not be regarded as an offer, recommendation, or solicitation on acquiring or disposing of any financial products, any associated discussions, comments, or posts by author or other users should not be considered as such either. It is solely for general information purpose only, which does not consider your own investment objectives, financial situations or needs. TTM assumes no responsibility or warranty for the accuracy and completeness of the information, investors should do their own research and may seek professional advice before investing.

Comments

We need your insight to fill this gap
Leave a comment