Australian Investors' U.S. Stock Rush Raises Red Flags

Australian investors are flocking to U.S. stocks, chasing the artificial intelligence (AI) boom. Over the past three years, Wall Street's S&P 500 index has delivered a 78% return, dwarfing the Australian S&P/ASX 200's 24%. However, in 2025, with AI speculation and a handful of giant companies driving market growth, warnings emerge that a fund composed of 500 stocks may not be as diversified as investors think.

Tony Sycamore, analyst at IG Markets, stated that "the so-called 'Magnificent Seven' companies—Apple, Nvidia, Alphabet (Google's parent), Amazon, Tesla, Meta, and Microsoft—account for more than one-third of the entire S&P 500 index market capitalization." He added, "When you include Netflix and Broadcom, these 9 companies represent nearly 45% of the index."

This is an extreme figure even compared to the dot-com bubble peak in 2000, when the top 6 tech companies accounted for only about 15% of the index.

Annual $520 Billion AI Investment, Returns Uncertain

What differs from the dot-com era is that most of the Magnificent Seven have existing businesses unrelated to AI. However, the scale of capital being poured into AI research and development (R&D) is staggering.

Microsoft, Amazon, Meta, and Alphabet alone invested approximately $350 billion (about 520 trillion won) in AI research and infrastructure in 2025. This includes the cost of building data centers to run large language models (LLMs).

Sycamore noted, "Concerns about the Magnificent Seven are warranted. They are not generating enough revenue to offset their massive debt burden. The cash burn rate is tremendous, and earnings growth has begun to slow."

The Trap of Circular Trading Structures

A bigger problem is the circular flow of capital between big tech companies and unprofitable startups. Notably, OpenAI, developer of ChatGPT, is estimated to be the world's largest private company with a $700 billion valuation, yet it does not expect to turn a profit until 2030 while planning to invest $1.4 trillion in AI infrastructure.

As questions grow about where this massive funding will come from, mutual investment deals between big tech companies continue to emerge. Nvidia invested $100 billion in OpenAI, and Broadcom, AMD, and others have struck similar deals.

Chris Weston, Head of Research at Pepperstone, warned, "These circular transactions are happening everywhere. Everything is based on the premise that OpenAI will someday be able to fully monetize its services. I've never seen anything like this."

Sycamore commented, "While these announcements look great on the surface, when you look inside, they are deeply concerning, artificial, and could collapse quickly. This is one of the reasons tech stocks have declined recently."

Differences from the Dot-com Bubble

Of course, the current situation is not entirely identical to the 2000 dot-com bubble. While many companies back then relied on unproven business models, today's Magnificent Seven already have profitable core businesses.

• Apple: iPhone and Mac hardware sales
• Microsoft: Windows, Office 365, and enterprise software
• Amazon: E-commerce and AWS cloud
• Alphabet: Search and YouTube advertising
• Meta: Facebook and Instagram advertising
• Nvidia: GPU chip sales
• Tesla: Electric vehicle manufacturing

These companies differ from dot-com era firms in that they could survive on existing businesses even if AI investments fail.

[AI Analysis] Concentration Risk Likely to Intensify

Current big tech concentration structurally contains three risks.

First, weakened diversification benefits. Even investing in an S&P 500 index fund effectively means betting nearly half on 9 companies. Poor performance from one or two companies could deliver a fatal blow to the entire portfolio.

Second, AI monetization uncertainty. OpenAI's 2030 profitability target means trillions of dollars will be invested without returns over the next five years. If investor confidence wavers during this period, cascading shocks are highly likely.

Third, risk of circular transaction chain collapse. An ecosystem has formed where Big Tech A invests in Startup B, B purchases chips from Big Tech C, and C uses A's cloud services. If cracks appear in one area, it could collapse like dominoes.

However, regulatory antitrust enforcement, emergence of competing AI models, and increased real-world AI use cases could act as variables to ease concentration. Investors need to move beyond the illusion that "500 stocks = safe diversification" and re-examine their actual portfolio composition.