Almost 100 thematic ETFs launched since end of Q1 2026, many targeting AI

Almost 100 thematic ETFs have launched since the end of the first quarter alone, a count that Strategas Asset Management chief ETF strategist Todd Sohn shared and that ETF Trends describes as a huge leap over prior years. Many of those new funds target AI. The number lands differently depending on which side of the shelf you sit on: for an issuer it is inventory, and for an advisor it is a diligence queue that grew by roughly a hundred line items in one stretch of the year.

The backdrop is a decade in which issuers have moved into core equity and fixed income categories the ETF wrapper was once thought too narrow for, and done it to what the source calls great success. Thematics sit at the other end of the same shelf: narrative-friendly, headline-prone, and dependent on a story that has to keep working after the launch. That is why the industry keeps shipping them, and why the survival math on any large cohort is what it is.

Sohn's account of the cascade is that issuers are seeking to provide exposure to an AI-centric world and, in his phrasing, arguably doing it as a complement to an S&P 500 core that has grown increasingly tech-heavy. What he leaves the funds themselves is a standard rather than a forecast: what matters going forward, he told ETF Trends, is the funds that offer the most thoughtful constructions and solutions.

Thematics have long been the part of the ETF market that draws headlines for narrative-friendly strategies, and the segment's delivery has been mixed enough that the same coverage raises an uncomfortable possibility for this year's crop — many of the launches may fold within a few years, as many earlier thematic funds already have.

One wrinkle cuts the other way. With a launch base this wide, the absolute number of funds still standing a few years out will likely be higher than in past cycles even if the survival rate is no better; more entrants means more survivors and more failures at once. ETF Trends frames the investor's task accordingly: in an environment with more idiosyncratic thematic products, identifying the long-term standouts matters twice over, both for the returns a fund can deliver if it rides a powerful trend and for the concentration risk a narrow theme can add or offset.

The practical cost of a cohort this size lands on the advisor's calendar. A fund whose membership is set by a disclosed revenue threshold can be checked against a model portfolio in an afternoon. A fund whose theme is asserted rather than measured pushes that work back onto whoever has to reconstruct the exposure from a prospectus and a fact sheet, and that queue is the one an advisor is actually managing.

THNQ, the ROBO Global Artificial Intelligence ETF, launched in 2020 — six years ago, a distance the piece calls shocking to write — and charges 68 basis points to track the ROBO Global Artificial Intelligence index, built to measure the performance of companies deriving revenue from AI advancements. The index sorts constituents into infrastructure or applications and services, and the fund breaks them down further into network and security, semiconductors, big data and other areas. Holdings are scored for AI revenue from 1 to 100, and only those at 50 or above are admitted. The screen arrives with a record: 55.5% year to date and, per ETF Database data, 42.7% over three years, beating the ETF Database Tech Equities category average in both periods.

AIQ, the Global X Artificial Intelligence & Technology ETF, is older still, a 2018 launch carrying the same 68 basis point fee to track the Indxx Artificial Intelligence Big Data index. Identical pricing is its own kind of information: when two products charge the same, whatever differentiation exists has to live in the index, and that is where the disclosed detail thins. Past AIQ's vintage, fee and index name, the coverage does not set out its inclusion rules — the gap an advisor would have to close on every one of this year's launches.

For a firm building model portfolios, the question is not whether AI exposure belongs in a client account. On Sohn's framing, the core index is already delivering a good deal of it, which suggests the thematic sleeve has to justify itself on something other than the theme, an overlap question and a fee question, and the overlap is the harder of the two to answer from a fact sheet.

What THNQ demonstrates is that a theme can be expressed as a rule, with membership decided by a measurable revenue contribution and a floor an advisor can recite to a client. It does not demonstrate that the rule is what produced the return. A single year of performance cannot separate the screen from a six-year head start, and the coverage makes no attempt to isolate the two. What both funds do establish is narrower and more durable: the index rule, the fee and the vintage each fit on a page, and the page is where the diligence happens.

Sohn's standard of thoughtful constructions and solutions reads less as a slogan than as an instruction to open the index methodology. The count worth running as this year's cohort seasons is the number of funds that can name an inclusion rule as specific as a revenue score of 50.