Copper, aluminum, lithium, nickel.... "AI investment trends" in commodities beyond AI computing power trading theme.

During the global COVID-19 pandemic, Wall Street veteran strategist Jeff Currie accurately predicted a "commodity supercycle." In a recent media interview, Currie stated that the world is in the early stages of a new commodity supercycle, driven by a boom in artificial intelligence infrastructure construction and long-term underinvestment in energy and metal production capacity. This cycle could last another decade or longer. Currie's views align with those of Michael Hartnett, a strategist at Bank of America, who believes that the geopolitical rivalry between China and the US, ongoing conflicts in the Middle East, and the global race for artificial intelligence have brought increased attention to core supply chain issues related to traditional energy sources. Currie, known as the "commodity bull market frontman," mentioned in an interview on the "Surveillance" program on Tuesday that the global energy sector is experiencing the "largest asymmetrical trade in modern financial markets." He pointed out that oil companies are currently providing up to a 15.5% free cash flow yield, far exceeding that of large-scale cloud computing service providers. Despite supply shocks caused by geopolitical conflicts in the Middle East spreading globally, Currie believes that this supercycle could last for another 10 to 12 years. Currie's views on the commodity market carry significant weight on Wall Street. Having worked at Goldman Sachs for over 20 years and later as Chief Strategic Officer for Energy Pathways at Carlyle Group, he is now a senior advisor at the group. As the co-chairman of Abaxx Markets, Currie draws on recent historical trends to suggest that from the 1990s to the early 21st century, technology stocks dominated the market, followed by energy in the lead until around 2014. He believes that technology stocks are once again driving demand for electricity and metal raw materials. Currie's latest perspective on commodities indicates that there is a significant gap between oil supply and demand. Despite a sharp increase in oil prices, there is still pressure in the system due to the consumption of inventories during the weakest period of the year for demand. Once inventories are depleted, prices are expected to rise to drive down demand to available supply levels, replicating the 2020-2021 historical supply-demand dynamic when US inflation skyrocketed to its highest level since 1981. As of Monday, Brent crude oil has surged by 84% this year to around $112 per barrel. Currie believes that the real pain in oil prices is yet to come, as the difference between spot prices and long-term prices in the oil curve indicates a fundamental mispricing of the oil trading market's long-term cost structure. Currie also highlighted the shortage of sulfuric acid pushing copper prices to historic highs, as this chemical, derived from oil refining, is crucial for the copper production process. In modern industry, approximately 90% of sulfuric acid globally is recycled from the oil refining and natural gas processing processes. In the interview, Currie stated, "Every policymaker, macroeconomic forecaster, senior central bank official, technology innovator is telling you that there is no problem. But every commodity CEO and anyone who truly trades in physical commodity markets is telling you that you do indeed face significant problems." Currie suggests that investing in what he calls the "Munificent Seven," a play on the US tech giants known as the "Magnificent Seven" (including NVIDIA, Amazon, Meta, Google, Microsoft, Tesla, and Apple), is a way to participate in the record energy demand. The "Munificent Seven" includes ExxonMobil, Chevron, ConocoPhillips, Shell, TotalEnergies, BP, and Equinor. He said, "From an economic perspective, the trade here is that you hold these large oil companies." In the era of artificial intelligence, not only chips are needed but also active acquisition of energy and metals. With resource-exporting countries like Indonesia tightening commodity exports and industrial metal prices such as copper, aluminum, and nickel continuing to rise this year, there is a growing demand for key metals and minerals consumed in the construction of AI data centers. Commodities are becoming a core investment focus beyond the AI hardware and cloud computing services, emphasizing the underlying energy, metal, chemical, and resource security premiums that support AI infrastructure expansion. As Currie's core judgment implies, the boom in AI data center construction intersecting with long-term underinvestment in energy and material production capacities could lead to another decade or longer of a commodity supercycle. A recent study report by Barclays Bank suggests that emerging markets like Chile, Peru, Brazil, Indonesia, and China, among others, will benefit from the commodity price transmission resulting from the construction of AI infrastructure. Chile and Peru are major copper exporters, which are essential for AI-related electrification trends due to their extensive use in power grids, electrical infrastructure, and data center power management wiring. Chile also holds a crucial position in lithium, a key material for global energy storage technology - when it comes to battery storage, the vast majority of new energy storage projects use lithium-ion batteries, firmly placing lithium at the forefront of global energy storage technology. Indonesia is the largest nickel producer and Nickel is a critical input for batteries and storage systems to support the increasing demand for electricity driven by artificial intelligence. Meanwhile, China controls rare earth magnets that are essential for semiconductor manufacturing equipment, data center infrastructure, Siasun Robot&Automation technology, and many other areas. From the perspective of AI engineering, AI data centers are not abstract "clouds" but highly physical capital expenditure systems: GPUs/ASICs require near-limitless high-efficiency power supply, HBM/SSD capacity expansion requires large amounts of semiconductor materials and chemicals, and the construction of AI data center machine rooms requires copper, aluminum, steel, transformers, storage energy, cooling systems, and natural gas/electricity grid support. The price expressions of the sulfuric acid shortage driving up copper prices, the constraints on aviation fuel/diesel/gasoline inventories, and the underestimation of the long-term cost structure of oil prices essentially represent the pricing bottlenecks of the AI era. If Indonesia proceeds with the centralized management of key commodity exports such as coal, palm oil, ferroalloys, or strengthens resource control, this will further amplify the political risk and supply chain security premiums. Michael Hartnett, a strategist at Bank of America known as the "most accurate strategist on Wall Street," and their team recently released a research report indicating that investors will continue to flock to commodity trading markets in the next few years. Even if a new round of Middle East wars temporarily comes to an end, the bullish trend in the global commodity market is expected to continue for many years until the end of 2030. According to Hartnett and his team at Bank of America, commodities are the most logical and highest-level post-war trading theme, with investors betting that commodities will replace stocks as the biggest winners in the coming years. The core reason is that investors urgently need to hedge against risk, inflation, and a weakening dollar, and geopolitical tensions and the global AI race are fundamentally strengthening the competition for energy, rare earths, minerals, and key commodity resources. He even summarizes the core logic as: whoever controls the chips, rare earths, minerals, and high-efficiency energy will win this global AI war. This implies that the pricing core of the post-war world, in Bank of America's view, is no longer just interest rates and profits, but the security of commodity supply systems, control over supply chains, and expanding fiscal expenditures.