When the price of copper follows the AI computing power industry chain to dance, the fate of copper and the global AI boom are already bound together.

Copper has always been known for tracking global industrial economic changes in depth and is often referred to as the "Copper Doctor." In the eyes of some economists, it is considered to be a barometer of the global economy. The recent surge in copper prices to historic highs is reminiscent of a popular technology stock closely tied to AI computing power infrastructure. This is mainly because investors are heavily betting on the unprecedented global AI infrastructure construction frenzy and the surge in electricity consumption brought about by AI training/inference will drive a sharp increase in copper demand. It can be said that the fate of this industrial metal is tightly bound to the AI boom. The copper super cycle is likely entering a startup zone, but this is not the old "real estate boom driving cyclical copper bulls," but a new type of copper super cycle shaped by the combination of "massive expansion of demand for AI computing power infrastructure + accelerated expansion of power grid construction + global energy transformation + hard asset inflation + mining supply constraints." In the AI computing power industry chain, AI GPU/AI ASIC is the core computing power, HBM is the core data throughput, electricity is the prerequisite for data center operation, and copper is the underlying conductor that connects unprecedented electricity and computing power systems. Without enough copper, there will not be enough high-speed copper cables, copper foils, and other copper components inside cloud computing server racks, as well as AI data center basic infrastructure such as transmission, transformers, distribution, and liquid cooling systems; without these, the empire of AI computing power can only remain at the level of chip orders and capital expenditure planning without being able to successfully build a complete AI data center. This crucial metal with high conductivity, which is vital for data center construction and power grid, has been mirroring the stock prices of AI computing industry leaders like NVIDIA and ASML in recent times. The LME copper price hit a record high last week, but fell with the volatility of AI computing-related stocks. On Friday, as investors returned to this sector, copper prices soared again. Since the second half of 2025, the relatively low volatility LME copper price has soared by as much as 50% and reached historic highs. This intense fluctuation reflects that copper has become a key pillar for bullish prospects in the market, as it is crucial for all the transmission lines, transformers, and key electrical equipment needed to power AI data centers. Currently, this factor is overshadowing concerns about traditional industrial infrastructure related to the impact of high oil prices brought about by the Iran war and copper demand. A rare surge in the stock market is driving copper prices up - the correlation between copper and US technology stocks has reached the highest level since 2012. Note: Shows a 40-day rolling correlation. DH Fund Management Co., one of China's largest hedge funds, head of basic metals trading, said, "This round of rally is mainly driven by structural AI computing power theme trades." She holds a neutral to bullish outlook on the future of copper and believes that further rise may require new inflows of capital into the technology sector, while mining supply tightens. From being an economic barometer to a computing power pipeline Copper is transforming from a traditional "global economic barometer" to an "AI infrastructure hard asset." Copper prices have been closely mirroring core stocks in the AI computing industry chain like NVIDIA, ASML, Broadcom, indicating that global capital has considered copper as part of "AI computing power chain overflow or AI derivative trading": as AI data centers expand massively, they are in greater need of high capacity transformers, low-loss distribution, liquid cooling pumps, backup energy storage, and stronger grid access. This makes copper a bottleneck asset that is crucial to the underlying infrastructure of AI hardware. AI data centers cannot operate solely on AI GPU/AI ASIC, HBM/NAND, data center CPU, and larger AI server clusters; they need stable electricity delivery from the power generator to the chip racks and modules. Global grid expansion is driving new demand for copper, with demand for copper in power grid infrastructure expected to rise from 12.52 million tons in 2025 to 14.87 million tons in 2030, further fueled by AI data centers and electrification. Additionally, the bull market for copper is not just driven by AI demand but also by inflation hedging and insufficient mining supplies. Investors are using copper and other hard assets as inflation hedging tools, while a shortage of new mining investments is creating major supply shortages. S&P Global's related research warns that global copper demand could increase from about 28 million tons in 2025 to about 42 million tons in 2040, driven by AI infrastructure, energy transition, and grid upgrades, while mining supply growth struggles to match demand. There is significant divergence in analysts' predictions of copper demand driven by the global AI wave - analysts have varying predictions for the impact of AI on copper demand. Marex data reflects a midpoint estimate of 100,000 to 150,000 tons of demand, while S&P Global data predicts that AI-related copper demand will increase from 1.1 million tons in 2025 to 2.5 million tons in 2040. However, some analysts insist that the copper market has prematurely priced in the demand increase. Commodity brokerage Marex Group and a recent study from Oxford University point out that the US domestic data center sector is facing serious bottlenecks in labor, power, equipment, and permitting, despite commitments from Microsoft, Alphabet's Google, and Amazon to invest around $580 billion in US data center projects. Marex's global market analysis director, Guy Wolf, says, "For copper, what's really important is not the data center construction frenzy itself, but the power and transmission networks that support these data centers." "While the AI narrative is indeed positive for copper, the actual realization of metal demand may be further away than many investors assume." Nevertheless, Wolf points out that given the massive influx of funds into tech stocks recently, traders looking to ride the wave of rise and fall face the risk of shorting. With copper prices rising, Wall Street fund management institutions have added approximately $14 billion net long positions on the London and New York futures markets this quarter so far. Concurrently, according to institutional calculations, the total market value of Nasdaq 100 index component companies has increased by $7.8 trillion. Wolf states, "Since the China economic supercycle, we have not seen this level of institutional interest before." He is referring to the period earlier this century when China's rapid economic growth led to a boom in commodities. "Even a small shift in allocation towards hard assets in the portfolio can potentially dominate this market." As shown in the figure above, the inflow of funds into tech stocks could potentially overwhelm the copper commodity futures trading market - in the second quarter, the Nasdaq 100 index saw a daily increase in market value of over $200 billion. Bearish investors also face another profitable opportunity associated with the risk of US imposing tariffs on copper imports. In recent weeks, the price of copper on the New York Comex exchange has risen above the London Metal Exchange price, enabling traders to capture arbitrage opportunities by shipping goods to the United States. On Friday, the London Metal Exchange copper futures price was around $13,630 per ton, while the Comex price was about $400 higher. The London Metal Exchange saw the largest surge in copper delivery orders in warehouses since 2013, indicating that traders may be seeking to move inventory to Comex warehouses to realize higher prices. This dynamic could ultimately trigger a new bidding war among buyers to secure available metal, even in a weak demand environment. Earlier this year, a rebound in Chinese demand supported copper prices, with strong spending on power grid and renewable energy projects offsetting losses in the real estate sector; the real estate industry had been the engine driving copper demand for much of the 21st century. However, prices close to record highs are now starting to dampen demand from manufacturers. In the past, investors have often been caught off guard as Chinese spot demand rapidly declines and overturns bullish bets on future growth. But since Matthew Fine joined Third Avenue Management as a portfolio manager in 2017 and began firmly betting on copper miners, he still believes the long-term trajectory will see significant growth. From a relatively conservative standpoint, even considering the new demand brought by the AI data center industry, copper demand may only increase by 2.7% annually by 2040. However, the key point is that mining companies are increasingly unable to keep up with the pace of demand expansion, as new projects become scarcer and development costs rise. Copper needs more than just GPU and HBM! Copper becoming the "physical blood vessel" of AI infrastructure, as the hard asset super cycle receives a catalyst from computing power Copper is essential in AI infrastructure because it is a fundamental metal for electricity transmission and power distribution systems. The pull of AI on copper is not so much within the construction of data centers but in the restructuring of the power systems behind the data centers. For copper, the most important aspect is not the AI data centers themselves, but the power generation and transmission systems needed to support them. The higher the power density of AI server cabinets, the more data centers need high-capacity transformers, low-loss distribution, liquid cooling pumps, backup energy storage, and stronger grid access. This makes copper the "physical blood vessel" that converts electricity into computing power. AI data centers cannot operate solely on AI GPU/AI ASIC, HBM/NAND, data center CPU, and a wider range of AI server clusters; they need stable electricity delivery from the power generator to the chip racks and modules. According to forecasts, global grid expansion is driving new demand for copper, with copper demand in power grid infrastructure expected to rise from 12.52 million tons in 2025 to 14.87 million tons in 2030, further fueled by AI data centers and electrification. Additionally, the bull market for copper is not just driven by AI demand but also by inflation hedging and insufficient mining supplies. Investors are using copper and other hard assets as inflation hedging tools, while a shortage of new mining investments is creating major supply shortages. S&P Global's related research warns that global copper demand could increase from about 28 million tons in 2025 to about 42 million tons in 2040, driven by AI infrastructure, energy transition, and grid upgrades, while mining supply growth struggles to match demand. Recently, one of Wall Street's senior strategists, Jeff Currie, accurately predicted the "super commodity bull market" during the global COVID-19 pandemic, stating that the world is in the early stages of a new super cycle of commodity demand. As the AI computing infrastructure construction frenzy collides with long-term investments in energy and raw materials production, this cycle could last another ten years or even longer. Currie's latest views are consistent with those of Michael Hartnett, a strategist at Bank of America, who believes that the complex dynamics between China and the US, ongoing geopolitical tensions in the Middle East, and the global AI competition are now drawing higher attention to core supply chain issues related to traditional energy sources. With major exporter Indonesia tightening commodity exports and prices for industrial metals like copper, aluminum, and nickel continuing to rise, driven mainly by the increased demand for key metals and minerals required for constructing artificial intelligence data centers, commodities are becoming a cornerstone that covers beyond the AI GPU/ASIC, data center CPU, HBM/NAND/HDD storage, 2.5D/3D advanced packaging, liquid cooling systems, optical interconnection supply chains, and data center power supply chains. Investors are not just buying models, chips, and high-performance AI cloud computing servers, but are investing heavily in underlying energy, metals, chemicals, and resource security premiums that support the expansion of AI computing infrastructure. Currie stated in an interview that US oil inventories are particularly tight, and at the same time, a shortage of sulfuric acid has pushed copper prices to historic highs, as this chemical derivative from oil refining is crucial to the copper production process. The primary raw material for producing sulfuric acid is sulfur. In modern industry, about 90% of the sulfur is recovered on a large scale from oil refining and natural gas processing. According to Currie's core analysis - the AI data center construction frenzy colliding with the long-term underinvestment in energy and materials capacity - the world may be entering a super cycle of commodities that could last another decade or longer. A research report from Barclays Bank recently showed that as the AI construction boom spills over into commodity prices, developing countries like Chile, Peru, Brazil, Indonesia, and China are expected to benefit. Chile and Peru are large copper exporting countries, crucial for the electrification trend related to artificial intelligence and the power management wiring of data centers, as copper is vital in power grid and data center core power management. Chile also holds a crucial position in the lithium supply chain, which is essential for global energy storage technology routes - when we mention "battery storage," the majority of new storage projects are based on lithium-ion batteries, and the global energy storage route has firmly locked lithium into the position of the "biggest beneficiary." Indonesia, on the other hand, is the largest producer of nickel, a key input for batteries and energy storage systems supporting the increasing electricity demand driven by artificial intelligence. Meanwhile, China controls the semiconductor manufacturing equipment, data center infrastructure, Siasun Robot & Automation technology, and many other areas crucial to rare earth magnets. AI