The $650 Billion Question: Why Betting on US Big Tech May Be Riskier Than You Think
The best investment stories are compelling right up until they aren’t. Artificial intelligence captured the imagination of global markets, drove valuations to historic highs, and quietly embedded itself at the core of millions of supposedly diversified portfolios. Now, with $650 billion in capital expenditure chasing $45 billion in revenues, investors are beginning to ask questions they perhaps should have asked sooner.
The Hidden Concentration in “Diversified” Portfolios
For many investors, a global index fund represents the golden standard of diversification: broad exposure to hundreds or thousands of companies across multiple markets. However, the reality of how these indices are constructed adds nuance to this story.
Major indices like the S&P500 and the MSCI World are weighted by market capitalisation, meaning the largest companies automatically receive the greatest allocations. As a result, a small number of US technology firms have come to represent an outsized share of what many investors consider to be a diversified portfolio.
As US Big Tech valuations have grown over the past decade, it becomes increasingly difficult to turn a blind eye to this issue. An investor holding a typical global tracker today may find five to seven companies accounting for a disproportionate share of their portfolio. These companies are all in the same sector, country, and increasingly, place the same thematic bet on artificial intelligence. The diversification investors were promised might just be a mere illusion. When the bet on AI unravels, where does that truly leave your portfolio?
From AI Awe to AI Poor: How the Narrative Is Shifting
For much of the past several years, artificial intelligence was the trade people were optimistic about. Capital flooded into anything connected to the theme: semiconductors, cloud infrastructure, software platforms and more.
However, sentiment in financial markets is rarely static, and the AI narrative is no exception. What was once characterised by unbridled enthusiasm has given way to more sober scrutiny. Investors are increasingly asking not just whether AI will transform industries, but when, and more critically, at whose expense. The mood has shifted, and with it, the calculus of risk for portfolios heavily exposed to the theme.
This shift in sentiment has not occurred in isolation. A growing unease has spread across industries once considered insulated from the AI story: from technology software to logistics, manufacturing, and beyond. The reality is that the AI narrative has two sides: for every industry positioned to benefit, there is another whose business model is being fundamentally reconsidered. In this environment, simply holding an index may no longer be sufficient. What is required is an active reassessment of portfolio positioning, and the investors who navigate this period successfully will likely be those who got ahead of that question, not those who discovered it too late.
The Capex Time Bomb: $650 Billion Spent, $45 Billion Earned
At the heart of the AI investment debate lies a number that deserves far more attention than it currently receives. The world’s largest technology companies, commonly referred to as the hyperscalers, are projecting combined AI-related capital expenditure of $650 billion in 2026 alone, against projected AI software revenues of just $45 billion. That is a price-to-sales ratio of 1,400%, extraordinary by any measure.
That is also not to say that the AI revolution will not materialise. Transformative technologies have always required significant upfront investment before generating meaningful returns, and it would be premature to dismiss the long-term potential of artificial intelligence entirely. However, the scale of this gap raises legitimate questions about timing, execution risk, and whether current valuations already reflect a best-case outcome.
The dot-com bubble is a case in point. When internet valuations collapsed in 2000, the NASDAQ lost nearly 80% of its value, and the damage was not contained to technology stocks alone. Portfolios that appeared diversified proved anything but. For investors navigating today’s AI-driven markets, the question is not whether history will repeat itself. It is whether their portfolio is built to withstand it if it does.
How to Position Your Portfolio for What Comes Next
For investors who have spent the past decade riding the US Big Tech wave, the question now is not whether to act, but how. The answer lies in genuine diversification across asset classes, geographies, and investment themes.
In practice, this means looking beyond the indices. This can be considering exposure to markets that have been overlooked in the rush towards AI — Japan, Europe, the UK, and select emerging markets — all of which have demonstrated renewed momentum as capital begins its rotation away from US technology leadership. It means allocating to real assets such as gold, commodities, and infrastructure, which have historically provided resilience precisely when financial assets are under pressure.
This is the philosophy underpinning the TEAM Multi-Asset UCITS Fund. With just 25 carefully selected positions spanning equities, fixed interest, alternatives, and real assets, the fund is deliberately constructed to stand in contrast to the index-heavy concentration that characterises much of the market today.
In an age of heightened volatility, the line between a winning portfolio and a losing one has never been thinner. Passive investors, anchored to indices increasingly dominated by a single theme, may find that the diversification they relied upon offers less protection than anticipated.
Active management cannot eliminate risk, but it can, at the very least, ensure that the risks being taken are deliberate ones.
Written by Finna Ng
