TMX VettaFi writers Ben Hernandez and DJ Shaw delved into the world of crypto ETFs in the last Bull vs. Bear. In this edition, writers Nick Peters-Golden and Elle Fitzgerald focus on artificial intelligence — specifically, AI ETFs.
It’s been over two years since the launch of ChatGPT sparked a global frenzy for generative AI, sending hyperscalers like Nvidia (NVDA) and Microsoft (MSFT) to historic valuations. In 2025, the narrative shifted from simple software speculation to a massive industrial build-out, with data center demand and energy needs reaching unprecedented levels.
Increased institutional adoption and the proliferation of specialized thematic funds are wins that will help further the entire ETF marketplace. But as valuation fatigue sets in and circular capex concerns grow, are these AI gains just a passing fancy? Or do they have staying power?
AI: More Than Software Stocks
Peters-Golden: It’s hard to have a conversation about AI in investing because we’re often talking about so many different trends, companies, and technologies at once. The dominance of the so-called Magnificent Seven has kind of morphed into an ongoing conversation about the AI Hyperscalers, given the overlap there.
Those companies of course play an outsize role in the stock market and in AI advancements but they’re not the whole story. I like to think about AI investing less as a story about individual models or companies, but more as a world historical opportunity to increase productivity.
T. Rowe Price portfolio manager Dom Rizzo has captured this succinctly in past interviews. At our Exchange conference last year, Rizzo, who manages the firm’s tech ETF, the T. Rowe Price Technology ETF (TTEQ), told VettaFi’s Roxanna Islam that he believes AI has the potential to be the “biggest productivity enhancer for the global economy since electricity.”
Yes, those big software companies, the MSFTs and Metas (META) of the world, are pushing AI adoption and model advancement further. They play a crucial role in doing so. Many of us already own those companies. Rather than doubling or tripling down on them, I would suggest that the way to embrace AI’s massive potential is to invest in the companies that are best able to harness AI to boost productivity.
To put a pin in this point, let’s think about the internet. The so-called “digital economy” drove almost 20% of U.S. GDP, according to a report by the Interactive Advertising Bureau (IAB). Some of the biggest companies in the stock market, like Amazon (AMZN), META, and Alphabet (GOOGL) basically could not exist without the internet.
AI has the potential to have an even greater impact than the advent of the internet by bringing the internet into the physical world. In some ways, it’s hard to wrap our heads around it. However, amid headlines about speculation and a bubble, we shouldn’t forget what exactly is driving this excitement. It’s exciting for very good reasons.
Valuation Fatigue and Circular Capex
Fitzgerald: The enthusiasm surrounding generative AI has created a feedback loop in which tech giants are spending record amounts on hardware to sell AI services back to the same ecosystem. This “circular capex” raises concerns about long-term ROI, especially as the S&P 500’s price-to-earnings ratio remains near cycle highs. Recent data show that while 91% of organizations are increasing their AI spend, only 10% are realizing significant returns today.
Funds like the Global X Artificial Intelligence & Technology ETF (AIQ) and the iShares Future AI & Tech ETF (ARTY) have seen incredible runs, but the current high valuations suggest that much of the future growth is already baked in. If enterprise adoption slows or margins compress due to rising energy costs, these concentrated portfolios face significant repricing risk.
In early February, AI infrastructure names like CoreWeave saw 20% intra-week swings. After that mini crash, it is clear that even minor guidance misses can trigger aggressive repricing.
The ETF Wrapper Is Perfect for This AI Investing Moment
Peters-Golden: This AI investing moment is a broad, transformational productivity enhancer. Yes, there is risk for those big firms, and clients rightly have questions.
The ETF wrapper has answers. Since the 2019 ETF Rule dropped, asset managers have been able to launch new ETF strategies much more easily. That has allowed a wide variety of ETFs to proliferate, targeting themes ranging from KPop to Jim Cramer.
Among that massive class of recent ETFs are a variety of funds poised to benefit from the AI revolution. Crucially, they may do so without doubling or tripling down on the AI hyperscalers. Let’s be clear — it’s probably worth strongly considering having names like MSFT, NVDA and Oracle (ORCL) in a core equity allocation.
Outside of that, however, are myriad subcategories that can really ride the AI wave to new heights. Data centers are, of course, one step away from these hyperscalers, who are desperate for computing power.
Take the Global X Data Center and Digital Infrastructure ETF (DTCR), for example. The fund leads with data center real estate names like Digital Realty Trust (DLR), a key data center company. DTCR charges 50 basis points (bps) to track a market cap-weighted index of firms in that digital infrastructure space. It holds data center names with a dash of NVDA and other hyperscaler exposure in its lower-weighted holdings.
Robotics, too, represents a strong investment area poised to explode because of AI. The ROBO Global Robotics & Automation Index ETF (ROBO) leans into that potential-filled world with an exciting approach.
ROBO charges a 95-bps fee to track the ROBO Global Robotics and Automation TR Index. It uses a tiered weighting strategy, applying qualitative and quantitative criteria to target companies deriving revenue directly from the industry or a related subsector.
Specifically, that includes firms active in areas like sensors for any type of robot, unmanned vehicles, 3D printers and navigation systems, and more. In the future, AI could empower robots to do countless jobs humans do today, for better or worse. Either way, it promises remarkable productivity gains.
Those are just the obvious areas. Drug discovery, financial modeling, scientific research, and more offer big opportunities empowered by AI productivity enhancement. Together, the ETF wrapper has allowed such specific, targeted ETFs to come to market for investors. Tax efficient, tradable, and easily used as building blocks to craft an allocation — what’s not to like?
The Energy Bottleneck and Infrastructure Limits
Fitzgerald: AI requires a tremendous amount of electricity, and we are hitting the physical limits of the power grid. Data center energy consumption is projected to double by 2030, putting immense pressure on the aging power grid. Instead of chasing high-multiple software plays in AI ETFs, advisors might find better risk-adjusted growth in the “pick and shovel” plays of the energy transition.
The Virtus Reaves Utilities ETF (UTES) or the Range Nuclear Renaissance ETF (NUKZ) offer exposure to the essential infrastructure that AI literally cannot function without, often at more attractive valuations.
Furthermore, the real value may lie in pipelines. Midstream leaders like those found in the Alerian MLP ETF (AMLP) and the Alerian Energy Infrastructure ETF (ENFR) are seeing a structural tailwind as natural gas becomes the primary bridge to meet data center demand. These funds also offer attractive yields, providing a defensive income cushion that pure tech plays lack.
The Big Question: Is It a Bubble?
Peters-Golden: We know the potential of AI and ETFs combined. However, we still must speak directly to the big story or question looming over AI. Does what we’re seeing constitute a bubble? You mentioned this above, Elle, and I think it’s fair to be concerned. I have my own concerns about AI hyperscalers and companies like OpenAI. Can they deliver on the investments they’re getting?
Let’s break it down. Tech AI may spend up to $700 billion this year — a huge number. But yes, it’s eating into free cash flow for those big firms. Big tech has lost value in the last few weeks as fear took a sell-off bite out of their valuations. HSBC estimated that OpenAI won’t be profitable by 2030.
Look at how asset managers are thinking about this, however, and a different picture emerges. They’re not just buying the AI story without thinking it through. There’s plenty of evidence that these firms are earning their returns.
JP Morgan Private Bank shared an analysis in December that, while noting credit availability could add more air to an AI bubble-shaped trend, the so-called AI hyperscalers are not as leveraged as we may commonly think.
I’ll return to our past interviews with T. Rowe Price’s Dom Rizzo. Rizzo told our Active ETF Content Hub that TTEQ, the firm’s active tech ETF, is in “AI on” mode, but could swap to “AI off” mode as needed. AI isn’t going away, and its impact is only going to grow. However, that doesn’t mean portfolios can’t adapt. ETFs are powerful tools to craft exposure to AI and get the best out of an era of profound innovation.
Diversification Beyond the “Hype Cycle”
Fitzgerald: Market leadership has been concerningly narrow in recent history. Just a handful of names are driving the bulk of broad market returns. For clients worried about over-concentration in “broad” tech ETFs like the Vanguard Information Technology ETF (VGT), which holds over 43% of its value in just three stocks, it may be time to intentionally lean into concentration and tactically rotate into high-growth themes.
We are seeing a supercycle in robotics and healthcare where AI is a component — not the entire thesis. ROBO and the ROBO Global Healthcare Technology and Innovation ETF (HTEC) provide a way to capture technological disruption. These funds offer global diversification across industrials and diagnostics, sectors benefiting from AI implementation rather than just AI speculation.
Additionally, for those seeking high-growth opportunities beyond AI ETFs, the Amplify Blockchain Technology ETF (BLOK) and the Amplify Video Game Tech ETF (GAMR) provide exposure to cutting-edge technology development. BLOK and GAMR are positioned to capitalize on blockchain-enabled applications and the global video game industry. Those sectors are developing alongside AI but trade outside the immediate hyperscaler bubble.
For more news, information, and analysis visit the Thematic Investing Content Hub.
vettafi.com is owned by VettaFi LLC (“VettaFi”). VettaFi is the index provider for NUKZ, HTEC, BLOK, AMLP, ENFR, and GAMR, for which it receives an index licensing fee. However, NUKZ, HTEC, BLOK, AMLP, ENFR, and GAMR are not issued, sponsored, endorsed, or sold by VettaFi. VettaFi and its affiliates have no obligation or liability in connection with the issuance, administration, marketing, or trading of NUKZ, HTEC, BLOK, AMLP, ENFR, and GAMR.