Artificial intelligence is no longer a narrow technology trade. This is restructuring energy marketsinfrastructure spending and portfolio construction. According to investing experts on this week’s episode of CNBC’s “ETF Edge,” investors who focus solely on chips and software risk missing out on where the next round of value is coming.
Some of the trends and innovations that drive the market and the rapid scaling of companies are related AI physical requirements. Power, cooling, network stability and data center efficiency have become mandatory constraints. Just look at the stock price Bloom energyit has struggled to generate returns above its IPO price for years after its 2018 IPO. Starting last year, its local fuel cells began to be ordered for data centers. Bloom has seen its stock rise more than 500%. and the company has reached a market cap of more than $30 billion.
Many opportunities are being created for investors in small and medium-sized companies. Firms that were once overlooked by the market are now “moving up the charts very quickly,” TCW Group head of global distribution Jennifer Grancio said Monday on “ETF Edge.” In many cases, these companies operate in narrow segments with limited competition, allowing their fundamentals to improve faster than investor awareness.
Energy reliability is the central issue. In recent years, as renewables have come down in cost and become more competitive with fossil fuels, the market has been asking, “How regularly can we get off wind or off solar?” There was a discussion. said Grancio. But AI has changed the conversation because data centers cannot tolerate interruptions, needing constant power supply to avoid unexpected outages.
According to Grancio, this reality has led to a “major shift to nuclear,” including renewed investment in existing plants and maintenance of existing plants. development of small modular reactors. These projects will create new suppliers and accelerate the growth of specialized players upstream of utilities and hyperscalers.
Nuclear Energy ETFs
- First Trust Bloomberg Nuclear Power ETF (RCTR)
- VanEck Uranium and Nuclear ETF (NLR)
- Headlines Uranium and Nuclear ETF (CAREER)
- Range Nuclear Renaissance Index ETF (WRONG)
- Global X Uranium ETF (URA)
Efficiency inside the data center equally critical. As AI workloads expand, cooling and power management become bottlenecks. According to Grancio, investors are increasingly attracted to companies that are “one or two in their field” and “the best in a particular technology,” especially in places where alternatives are limited.
The structure of these markets is important. In some cases, there are “multiple providers” that border on oligopolies, Grancio said. This concentration creates operational leverage, but it also means that missteps can be costly.
As a result, actively managed ETFs have gained traction. While passive indices can capture broad market returns and add new companies as constituents as the indices scale, active strategies aim to identify them early and hold them through multiple phases of growth.
But the risks can be significant. Some parts of AI-powered ecosystems include “smaller, financially weaker companies” with electricity demand, VanEck CEO Jan van Eck said. “It also means you get a lot of volatility down the road,” he told ETF Edge.
As a result, he said no single AI theme should dominate an investor’s asset allocation. “You don’t want to be overweight them in your portfolio,” Van Eck said.
He described Van Eck’s nuclear ETF trading at “nosebleed levels” last year before reaching a more reasonable entry point for new investors.
ETF analysts said investors will incorporate AI into their portfolio construction A more targeted path in 2026active rebalancing and realistic risk expectations allow investors to stay invested without chasing peaks or panicking during downturns.