[Practical Finance] Entering the Supercycle: 'Jeollyeok Infrastructure' Is Hard to Pick Stocks, Are There Any ETFs?

AI Data Center-Driven Surge in Power Consumption Expected
Increased Demand for Nuclear Power Plants, Power Facilities, Components, and Cables
Outstanding Returns from US Energy-Related ETFs
Long-Term Upward Trend Expected at the Start of a Growth Cycle

The ‘Trump Trade’ refers to the flow of investment funds based on the policy direction of Donald Trump, the next President-elect of the United States. After the U.S. presidential election, investors focused on Trump’s policies, causing virtual currencies including Bitcoin to surge and related industry theme stocks to fluctuate wildly. Investors are also paying close attention to the words and actions of Elon Musk, Tesla CEO, who has been a strong supporter of Trump. This is to ride the massive flow of funds moving from investment destinations that were popular under the Joe Biden administration to places expected to grow during Trump’s second term.

However, the difficulty of investing has increased. It may seem like you just need to put money into promising investment destinations included in the Trump theme and wait, but choosing investment products is not as easy as it sounds. Even before the new administration takes office, there are controversies over overvaluation of investment destinations that surged in the short term due to the Trump theme. Some themes have already started to undergo corrections, increasing volatility. There is a risk of suffering significant losses if you buy a large number of stocks that seem likely to keep rising.

Investment experts unanimously recommend investing in artificial intelligence (AI), a field expected to grow steadily over the next 10 to 20 years. AI is analyzed as a revolutionary technology that will change the next 100 years regardless of which government takes office. Semiconductor companies such as Nvidia and SK Hynix have continued their rally on expectations that the demand for numerous semiconductors used in AI will increase. As semiconductor companies have shown a lull, investors’ attention has recently shifted to nuclear power and power infrastructure. This is due to the expectation that power consumption will increase exponentially as AI’s role expands.

Rapid Increase in Power Consumption from AI and Data Centers... Power Infrastructure and Nuclear Power Companies on the Rise

Earlier this year, the International Energy Agency (IEA) projected that global power consumption related to AI, data centers, and virtual currencies will surge from 460 terawatt-hours (TWh) in 2022 to a maximum of 1,050 TWh by 2026. According to a data center power outlook report published in 2021 by the U.S. Energy Information Administration (EIA), power consumption by U.S. data centers is expected to reach 88 TWh by 2030. This is 1.6 times the annual power consumption of the entire city of New York. This projection suggests a surge in demand for power generation facilities such as small modular reactors (SMRs), power infrastructure, and grids including cables.

Jung-han Kim, a researcher at Samsung Securities, said, "Despite repeated bubble controversies, the AI theme is becoming stronger along with Trump and Musk," diagnosing, "While the protagonists of the AI cycle were previously graphics processing units (GPUs) and high-bandwidth memory (HBM), now nuclear power, optical communication, and high-speed networks are taking over." Researcher Kim advised, "Until next year, the main focus of investment should still be found in AI."

Song-cheol Kang, a researcher at Eugene Investment & Securities, also said, "Last year, data center power consumption was estimated at an average of 20 to 25 gigawatts (GW) per hour, accounting for 4 to 5% of U.S. power consumption, and power demand due to data center expansion is expected to increase by 10 to 20% annually until 2030," forecasting, "This implies the possibility of competition to secure power or energy." He added, "The increase in data centers, power consumption, and power grid investment is not a short-term phenomenon but could become an important trend."

Moreover, the dominant analysis is that nuclear power is essential to meet power demand that cannot be handled by renewable energy alone. In fact, among recent U.S. big tech companies, Microsoft has decided to receive power from the restart of the Three Mile Island nuclear power plant starting in 2028. Google and Amazon are also showing interest in nuclear power generation. Jun-seop Kim, a researcher at KB Securities, interpreted this as "an attempt to compensate for the intermittency of renewable energy and delays in power grid connections with nuclear power." He added, "The U.S. government recently announced a roadmap to triple nuclear power generation capacity by 2050 to meet the power demand of AI data centers," pointing out, "Expanding the share of nuclear power is inevitable as renewable energy alone has limitations in achieving carbon neutrality."

Amid this trend, stock prices of power infrastructure and nuclear power-related companies continue to rally in domestic and international stock markets. In the U.S. market, companies such as GE Vernova (energy), Emerson Electric (power equipment), Vertiv Holdings (power management), NuScale Power (SMR), and Eaton (power devices) are showing upward trends. In Korea, investors are focusing on companies such as HD Hyundai Electric (power facilities), Doosan Enerbility (nuclear power facilities), LS Electric (power efficiency), Hyosung Heavy Industries (power equipment), and Daehan Cable (wires and cables).

Many Variables in Individual Power Infrastructure Stocks... ETF Investment Recommended

However, even though power infrastructure and nuclear power are promising, it is difficult for individuals to select and invest in individual stocks. Even within the same industry, competitiveness varies by company, and due to the nature of the order-based industry, there is inherent uncertainty in orders. Also, since the power industry is a business-to-business (B2B) market among experts, it is difficult for individuals to predict the stock price direction of individual companies based on their own insight. For this reason, experts recommend investing in exchange-traded funds (ETFs) related to power infrastructure and nuclear power.

Representative domestic ETFs include ‘KODEX AI Power Core Facilities’ and ‘KODEX U.S. AI Power Core Infrastructure’ listed by Samsung Asset Management, ‘SOL U.S. AI Power Infrastructure’ managed by Shinhan Asset Management, ‘RISE Global Nuclear Power’ listed by KB Asset Management, ‘PLUS Global AI Infrastructure’ by Hanwha Asset Management, and ‘TIGER Global AI Infrastructure Active’ by Mirae Asset Global Investments.

From October to the 15th of last month, over about one and a half months, ‘SOL U.S. AI Power Infrastructure’ posted the highest return of 22%. Shinhan Asset Management introduces ‘SOL U.S. AI Power Infrastructure’ as a product that can evenly invest in related sectors such as △nuclear power value chain (43.4%) △power grid system facilities (32.9%) △data center infrastructure (23%). ETFs such as ‘KODEX U.S. AI Power Core Infrastructure (16.54%)’ and ‘RISE Global Nuclear Power’ (14.69%) followed with double-digit returns during the same period.

Recently, top-performing ETFs mainly hold portfolios of U.S. energy and nuclear power companies. Nuclear power companies invested in include Constellation Energy, the No. 1 U.S. nuclear power company; NuScale Power, a leading SMR company certified by the U.S. Nuclear Regulatory Commission; Cameco, the world’s largest uranium mining company; and Oklo, an SMR company invested in by Sam Altman, CEO of OpenAI.

They also hold data center-related companies such as NextEra Energy, the largest renewable energy utility company in the U.S.; GE Vernova, a power grid operator spun off from GE; Vertiv Holdings, which provides power and cooling solutions for data center server systems; and Eaton, which produces transformers and distribution panels for data centers.

An asset management company official said, "Since Trump’s election, U.S. energy companies’ stock prices have rallied, and ETFs holding U.S. companies as major portfolios have performed relatively well," forecasting, "ETFs holding domestic power infrastructure companies, which have been neglected for a while, are also likely to show a long-term upward trend."

Hyun-bin Kim, head of ETF Investment at NH-Amundi Asset Management, said, "Investment in power facilities to operate AI data centers that consume massive amounts of power is currently ongoing," adding, "North American power infrastructure is over 25 years old, and a large-scale replacement cycle has arrived." He continued, "Large-scale transmission and distribution infrastructure investment is also necessary for the transition to renewable energy," explaining, "All these situations indicate that investment in power facilities has entered a super cycle." He predicted, "Investment related to nuclear power construction and facilities, power cables, transformers, and other power infrastructure will continue for the next several years."