Exclusive: Global X sees AI ETFs as key to growth amid sector evolution

Global ETF provider Global X is navigating the fast-moving artificial intelligence landscape, as investors look for both clarity and opportunity amid the sector's rapid growth and speculation.

During an exclusive interview, Billy Leung, Senior Investment Strategist at the firm, broke down how the AI value chain is taking shape and what it means for global markets and investment strategies.

Global X manages a diverse portfolio of more than 400 ETF strategies worldwide - totalling almost USD $100 billion in assets - with 45 funds available on Australia stock exchanges alone.

"We try to be very innovative with our products," Leung explained to TechDay. "AI is one of our key thematic ETFs as well."

That diversity of product signals the depth-and complexity-of opportunities arising from AI-driven transformation across sectors.

Amid escalating interest in AI, Leung stresses the importance of separating substance from speculation. "When people talk about AI, it's like a buzzword. It's very hard to distinguish it from hype," he said.

For Global X, identifying investment-worthy AI opportunities revolves around three pillars: technological advancement, significant regulatory or environmental change, and robust demographic or structural shifts.

"When we look for actual investments, it's important to find quality companies: those with growth, those that are fairly valued, and those with a strong balance sheet."

Valuation, frequently debated as leading AI firms surge, remains a challenge in a frothy market. "Some companies related to AI could be trading maybe even close to 70 times price earnings," Leung noted.

He emphasises balancing current valuations with forward-looking growth. "When you look at Nvidia, for example, you're paying 30 to 40 times price earnings, but you're looking at 90-100% earnings growth for the next few years. That's probably a good measure for high-growth companies."

While headlines often trumpet AI as a meteoric force, Leung points to historical parallels to illustrate the sector's infancy.

"Whenever we see innovations-steam engines, electricity-they always provide support to the economy, improving productivity. Those innovations brought about anywhere from 40% to 80% productivity improvements over a few decades," he said.

"AI has only brought about 5% so far, which gives you an idea that we're still at a very early stage."

Leung dissects AI's ecosystem into three segments: hardware, infrastructure, and adopters. "We've had a strong run in hardware-semiconductors have driven a lot of the technology advances. But there's a lot of growth opportunities now in infrastructure and the adoption side," he explained. Cloud and data centre buildouts, along with vertical integration by established industry players, are the next growth horizons.

"We're going to see a lot of companies starting to monetise opportunities. The next 10 years will be about applications-in robotics, healthcare, logistics, warehousing-where a lot of value will be created."

Critical to unlocking the next phase of AI advancement, according to Leung, is affordable energy.

"What's missing currently is sustainable energy-so we do look at uranium as a potential renewable source that could be powering a lot of this data centre and AI," he said, noting that major tech firms are actively investing in their own energy assets. In tandem, industry-specific adopters are poised to benefit as AI applications mature.

"Companies with vertical integration in robotics and healthcare are using AI to enhance efficiency. That's where we see significant growth."

This broadening of the AI landscape is also reflected in geographic diversification, with Asia's technology leaders playing a pivotal role. "Chinese tech companies such as Alibaba and Tencent have made huge advancements in large language models, applying AI to social networking and eCommerce services. These firms are quietly driving the next wave of AI adoption," said Leung.

For investors, Leung advocates for exposure across the AI value chain.

"The AI chain is just so huge-hardware, infrastructure, adopters. It's hard for investors to find the winner between all these sectors, which is why ETFs encompassing all these sectors are an efficient way to gain exposure to the thematic structure," he said. Rather than attempting to time individual sectors, diversified products "allow investors to benefit from the entire value chain."

Amid periodic headlines of AI slowdowns or market fatigue, Leung remains bullish on the sector's multi-decade trajectory.

"In the innovation cycle, growth comes in phases. The next growth in AI will really depend on cheap, sustainable energy and on companies monetising it. Once that happens, growth will be fast. Think back to smartphone adoption: it was slow until 'killer apps' spurred mass uptake. That's what AI needs-the next killer app."

Balancing exposure within AI requires a core focus on the broad theme, with optional tilts to specific subsectors depending on investor convictions.

"For investors who believe that semiconductors are good value, there are semiconductor ETFs. For those focused on infrastructure or adopters-there are products for those too. But AI as a core thematic will remain at the centre."

Identifying companies that will endure-rather than those capitalising on the AI label-requires rigorous analysis, Leung cautioned.

"It's really about fundamental cash flow. Look at whether a company can generate free cash flow, and whether they are focused on one specific innovation and applying it effectively. Companies like Nvidia differentiated themselves by focusing and executing on clear applications."

Leung views the impact of macroeconomic and regulatory shifts as largely transient. "There's definitely going to be some impact from trade wars or sanctions, but innovation has always been borderless."

"If innovation has application, it will always advance, no matter what hurdles it faces."

As an example, he highlighted China's leap in semiconductors and robotics, despite years of trade restrictions. "It's very hard to restrain innovation, because innovation will always find its way as long as there is a use case."

"In the future, we believe there will be only two kinds of companies - ones that use AI, and ones that don't exist," Leung said.