The name’s Bond, AI Bond.

Earlier this month, interest on bonds issued by neocloud company CoreWeave climbed past 15%, pushing the company’s share price down more than 30% and bringing concerns over AI-related debt back into the spotlight. AI capital expenditure is projected to top $1 trillion a year by 2030, with more and more of this spending paid for by issuing debt. Against this backdrop, at least five issuers have recently filed for ETFs focused on AI-related borrowing.

“If you wanted to take AI exposure, it does seem like you’re getting better pricing to take risk in the credit market compared to the equity market,” said Stephen Coltman, head of macro at 21shares. But given the wide variety of debt issued by AI companies, due diligence is critical. “Wrapping it in an ETF and saying, ‘This is AI debt’ doesn’t tell you the full story of what the risks really are and where you stand in the capital structure.”

Plus, during a crisis, credit markets’ liquidity tends to freeze, which means that investors might be forced to trade at a discount. “If you have redemptions and the underlying market is just basically frozen up, it’s difficult to get people in and out,” Coltman said.

Leverage the Moment

Even before the CoreWeave news, AI debt has been a contentious topic, and while not every product is likely to succeed, demand is strong enough that there may be one or two winners, said James Seyffart, ETF analyst at Bloomberg Intelligence. “There’s plenty of people with bullish and bearish views on these things right now, and anything along those lines tends to do well in an ETF wrapper,” he said.

There are some differences in the funds’ makeup that could dramatically affect performance. “Your everyday investor is probably going to just see the names of these and not really understand how different they all are,” Seyffart said. “Much like anything we’re looking at with an ETF, you really need to look under the hood and try to understand what the strategy is doing and what it’s holding.”

There are a few differences already clear from the preliminary filings:

• The actively managed BondBloxx AI Infrastructure Debt ETF, the Allspring AI Infrastructure Bond ETF and the Bitwise AI Bond ETF will invest in a mix of debt products issued by AI infrastructure companies, including noninvestment grade debt, securitized debt obligations and debt issued by special purpose vehicles.
• The Defiance AI Infrastructure Bond ETF will similarly invest in a mix of debt, but is index-tracking instead of actively managed.
• The Proshares AI Hyperscaler Bond ETF is index-tracking and appears to be the only fund that will limit its holdings to investment-grade bonds issued by hyperscalers.

Bond Appetit: These funds open up a path to investing in corporate debt that isn’t usually possible outside the ETF wrapper, and as bond funds, they’re inherently a less risky way to play the AI theme. “This, theoretically, should be a lot less risky than buying the equity, because a lot of these are going to be asset-backed or backed by large companies,” Seyffart said, although he noted they’re likely to be more volatile than a typical bond product.

But how much AI debt, if any, is appropriate for retail investors? “There’s so much debt coming out. There’s a lot out there already, and there’s so much more in the works that that market is already challenged,” said Jay Goldberg, senior analyst of semiconductors and electronics at Seaport Research Partners. “There’s still appetite, but I definitely think that yields are going to start going up.”

Some investors are taking a gap year from the S&P to explore Asia.

Net flows into US-listed Asia-Pacific ETFs surpassed $15 billion through the end of September, a sharp increase from the $5 billion they took in during the whole of 2025, according to Morningstar data. Emerging markets have outperformed the S&P 500 for the past two years and have returned nearly 17% to the S&P’s 11% so far this year, and investors are thinking hard about diversifying.

“There’s two layers of the diversification story,” said Anthony Caruso, head of ETFs at Nomura Asset Management. The outperformance of emerging markets signals an attractive opportunity to diversify away from the Magnificent 7, he said, but APAC ETFs additionally offer investors the ability to spread their AI exposure across the global supply chain.

Travel Advisory

Asia’s AI supply chain and the subsequent explosion of ETFs like the Roundhill Memory ETF (DRAM) are creating significantly more opportunities for issuers to explore specialized ETFs in this market, said Chris Mullen, senior vice president of derivatives trading at Tidal Financial Group. “Everyone wants to catch lightning, and so they’re seeing other opportunities, specifically in Asia,” Mullen said. “We’ve seen an explosion of ETF baskets and ETF funds specifically created around Asian names.”

A recent survey from William Blair Investment Management sheds more light on the trend:

• Active ETFs seem to be advisors’ preferred vehicle for gaining exposure to international and emerging markets, as 56% of RIAs said they expect to increase their allocations in the next year.
• When asked which markets offer the strongest investment opportunities, 77% of RIAs choose at least one Asian market, led by China (28%), Japan (24%) and India (17%).

Eat, Pray, Allocate: Despite the growing popularity of this ETF category, managing the funds themselves can be challenging for issuers. “You need to make sure you have a tight bid-ask spread, but also you need to find a lead market maker to be able to host it for you,” which isn’t always possible, Mullen said. Plus, the time difference means that everything is trading overnight. “No one’s on the desk to help there, so we have to fair value a lot,” or estimate what holdings are worth without live pricing, Mullen said. “From A to B, there is a very large barrier of entry. However, I think with each fund we launch, it opens up more doors, and it continues to make the process easier.”