Why the loudest AI trade might not be the best one

The hype around the upcoming US technology initial public offerings (IPOs) is hard to ignore. Anthropic, OpenAI and SpaceX are being positioned as once-in-a-generation investment opportunities.

Emanuel Datt, chief investment officer at Melbourne-based Datt Capital, has heard that pitch before.

“When the American railroads were being built in the 19th century, investors rushed to own the rail companies, but the real wealth was created by the businesses that shipped goods across those rails. AI is not that different. Anthropic, SpaceX, and OpenAI are building the infrastructure. The question investors should be asking is who benefits from using it?”

The valuation problem

The mechanics of these listings are worth understanding. Nasdaq’s new ‘fast entry’ rule lets companies that would rank among the Nasdaq-100’s top 40 constituents by market cap, roughly US$100 billion and above, join the index within 15 trading days of listing, bypassing the usual three-month seasoning period applied to most new listings.

Datt is candid about what that means for investors. “These IPOs are being marketed as once-in-a-generation opportunities, but the valuations defy conventional logic. When you see liquidity draining from other asset classes the moment these listings are announced, that tells you retail money is being repositioned. For example, money is moving from crypto towards these mega stock IPOs.”

The capital flows tell a story of their own. Retail investors chasing the next big thing are moving money out of one speculative asset class and into another, often without fully understanding what they are buying or at what price.

“It’s a sign of the times,” Datt says. “You have a lot of speculation, you have hot new technology, a confluence of different factors clearly designed to suck in the retail dollar. The promoters of these companies are trying to capture the zeitgeist and basically sell the top.”

Who actually benefits from AI

Datt’s railroad analogy carries a deeper point. The companies building AI infrastructure are not necessarily the ones that will generate the most durable returns. Businesses that adopt these tools and sharpen their competitive position with them tend to come out ahead.

“Technology companies who adopt or use these highly subsidised services at the present are best equipped to benefit ultimately,” he says.

That insight reframes the investment question entirely. Instead of asking whether OpenAI or Anthropic will justify their valuations, investors should be asking which businesses are quietly embedding AI into their operations, reducing costs and building competitive advantages not yet reflected in their share prices.

In many cases, those businesses are Australian small-cap equities, operating well below the radar of investors transfixed by Silicon Valley.

Where Datt is looking instead

The answer, according to Datt, is closer to home. Investors are overlooking Australian small-cap companies precisely because the spotlight is pointed elsewhere. And that, he argues, is exactly when they are worth paying attention to.

“That’s often the best time to invest,” he says. “These companies are quietly adopting AI tools that reduce their cost base and sharpen their competitive edge and they’re available at a fraction of the valuation multiples being asked of the US mega-caps. Australian small-cap equities are a safer place to invest than the US stocks, where markets have run really hot in comparison over the last six months or so.”

Few comparisons in global equities are as striking right now. US mega-cap IPOs are priced for perfection in a market that has already run hard. Meanwhile, Australian small-cap equities sit largely unnoticed, priced for indifference by investors looking the other way.

The opportunity in being contrarian

The loudest investment opportunities are rarely the best ones. The capital flooding towards US technology IPOs is creating a vacuum elsewhere. Australian small caps are sitting in that vacuum right now.

For investors willing to look past the noise, the companies quietly using AI to cut costs and strengthen their businesses could deliver far more than the infrastructure builders attracting all the attention.

The rails are being built. The real question is who ships the goods.