Key Takeaways
- AI ETFs aim to provide exposure to companies involved in artificial intelligence but vary widely in their focus and holdings.
- Investors should evaluate an ETF’s methodology, holdings, fees, and overlap with existing investments before buying.
- AI ETFs reduce company-specific risk but may still carry concentration risk within the AI sector.
- Not every company that says it is an AI play is actually involved in a meaningful AI business.
Artificial intelligence is one of the hottest investment themes right now, but betting on AI is not that simple. There isn’t a standard definition for what it encompasses, largely because it cuts across industries. It’s not a sector in the way that utilities or consumer discretionary are.
An AI company can be in chipmaking, software, power generation, construction, healthcare, and on and on. And as with all crazes, there are companies that make flashy pronouncements about their AI focus when, in fact, they might just be a run-of-the-mill retailer with an internet connection.
That matters when it comes to selecting an exchange-traded fund that has the AI label. One AI ETF may target semiconductor makers and computing infrastructure. Another might focus on software developers and cloud computing platforms. But both might be marketed simply as AI.
In short, not all AI ETFs are created equal, and investors should understand what they are buying into by looking at their methodology, holdings, and expenses.
What Is an AI ETF?
An artificial intelligence ETF is designed to provide investors with exposure to companies whose business is largely related to artificial intelligence. They are generally considered thematic ETFs. Like all ETFs, AI-focused ETFs trade on exchanges throughout the day while owning a portfolio of underlying stocks.
Some AI ETFs track an index and select their holdings based on predetermined rules created by the index overseer. The index might screen stocks based on revenues, business activities, or involvement with certain technologies. The Nasdaq CTA Artificial Intelligence Index (NQINTEL), for example, follows publicly traded companies engaged in AI across technology, industrial, medical, and other economic sectors. There are two WisdomTree ETFs that track it.
Other AI ETFs are actively managed and give a portfolio manager discretion to decide which companies are best positioned to benefit from AI. Active ETFs usually have higher fees and costs than index ETFs.
Tip
Most AI ETFs pick companies active in AI, but some actually use AI bots to pick stocks. In the latter, AI refers to the investment process rather than the theme. For example, Amplify’s AI-Powered Equity ETF (AIEQ) uses AI to select securities and can own just about any company in the U.S. equity market.
Not All AI ETFs Invest in the Same Way
The easiest way to begin comparing AI ETFs is to consider where along the AI value chain they invest. On one end are pure-play companies building AI models, software, and applications. Alongside them are the makers of the proverbial “picks and shovels”: computer chips, servers, networking equipment, and cloud platforms needed to train and run AI models.
Some funds also include robotics and automation providers who represent AI moving into factories, warehouses, autonomous vehicles, and medical devices. Other funds look further downstream at companies using AI to transform existing lines of business. You can see how various ETFs interpret those choices by looking at products already on the market.
Global X’s Robotics & Artificial Intelligence ETF (BOTZ) targets industrial and non-industrial robot companies as well as autonomous vehicle makers. Amplify’s Bloomberg AI Equal Weight ETF (AIVC) weighs semiconductors, cloud providers, software companies, and hardware makers evenly. Roundhill’s Generative AI & Technology ETF (CHAT) focuses specifically on companies in the generative AI ecosystem.
The distinction between making AI and using AI will likely become even more important over time. If AI continues to diffuse throughout the economy, some of the eventual winners might be companies that didn’t invent AI but found particularly effective ways to deploy it. Think of Cisco Systems Inc. and Apple Inc. at the dawn of the internet age. Cisco made a lot of the equipment that forms the backbone of the World Wide Web and seemed indispensable, but Apple became the bigger winner by putting the internet in everyone’s pocket.
Should You Invest in an AI ETF?
When it comes to thematic ETFs, many planners say they shouldn’t be a core part of anyone’s portfolio, but rather a limited, targeted bet. AI is likely to rewire the global economy, but figuring out when that will pay off and which companies will be the ultimate winners is nearly impossible to get right at the right time.
Potential Benefits of AI ETFs
The big benefit to an AI ETF is the broad-based approach. Instead of picking the exact AI winner, you spread your bet among a number of AI companies. They also reduce company-specific risk. A software company can lose market share. A chipmaker can run into problems during a product cycle. A robotics company might face soft industrial demand. Spreading investments across different companies and parts of the value chain makes returns less reliant on any single business.
An AI ETF lets you bet on the AI “market,” rather than on one headline stock. Growth from AI could benefit model and software developers, chipmakers, cloud computing companies, networking firms, automation businesses, and companies that figure out how to use AI effectively. An ETF that holds some of each of those stands to benefit.
Risks of AI ETFs
While company-specific risk may be less of a concern using an ETF, sector-wide risk is still something to watch. Owning a dozen AI stocks might seem like a diversified bet, but if each of them responds similarly to the same external factors, then the wager is actually quite concentrated.
An AI ETF may also have outsized weighting on just a handful of megacap tech companies, which is also a form of concentration risk. Given tech’s predominance in U.S. financial markets, any cap-weighted AI ETF that has the biggest tech companies faces overlap risk with broader-market ETFs.
Important
If you already own a U.S. stock market ETF or a large-cap growth ETF, you likely already have exposure to some of the biggest AI-related companies. Buying an AI ETF could simply increase your exposure to those stocks rather than diversify your portfolio.
Thematic ETFs tend to have higher expense ratios than index-tracking broad-market funds, especially if they are actively managed. It is important to check fees before investing.
Valuation is also an important consideration. Thematic ETFs tend to hit the market after the trend has taken off, which means that the companies involved likely have already seen heady share-price gains—and possibly have sky-high valuations.
Important
AI ETFs face the usual risks associated with new technologies. Competitive advantages can erode. Pricey infrastructure can become outdated. Different parts of the AI ecosystem can face regulations, copyright concerns, data restrictions, and energy prices.
What to Evaluate Before Investing
- Evaluate methodology. For index-tracking ETFs, learn how companies qualify for inclusion, how stocks are weighted, and how frequently the index rebalances. For actively managed ETFs, check what the manager is looking for and how much discretion the fund has to buy and sell stocks.
- Look at holdings. Examine the fund’s largest positions, sector allocations, geographic exposure, and how much weighting its largest holdings have. Consider concentration and overlap risks.
- Don’t forget about fees. Check the expense ratio, assets under management, and liquidity metrics like volume and the bid-ask spread.
The Bottom Line
An AI ETF does not give you standardized exposure to the AI industry—a somewhat arbitrary moniker since AI companies can do business across a range of sectors. Instead, each AI ETF will have its own focus and characteristics. This requires an investor to do some homework before buying in. Check the fund’s methodology, holdings, and fees.
The goal before investing in an AI ETF is to understand which part of the AI ecosystem it owns, determine how concentrated the portfolio is, how much it costs to own, and whether the fund overlaps with any of your existing investments.